evidence why paroxetine dose escalation is not effective in major depressive disorder: a randomized...

Downloaded from UvA-DARE, the Institutional Repository of the University of Amsterdam (UvA)http://dare.uva.nl/document/124010

Description ThesisVersion Final published version (publisher's pdf)File ID 124010Filename thesis.pdf

SOURCE, OR PART OF THE FOLLOWING SOURCE:Type DissertationTitle Dose-escalation in the picture : pharmacological and imaging studies in depressionAuthor H.G. RuhéFaculty Faculty of MedicineYear 2008Pages 300

FULL BIBLIOGRAPHIC DETAILS: http://dare.uva.nl/record/283611

Copyrights It is not permitted to download or to forward/distribute the text or part of it without the consent of the copyright holder(usually the author), other then for strictly personal, individual use. UvA-DARE is a service provided by the Library of the University of Amsterdam (http://dare.uva.nl)

http://dare.uva.nl/document/124010

http://dare.uva.nl/record/283611

http://dare.uva.nl

Do

se-escalation

in th

e pictu

reH

.G. R

uh

éPharmacological and Imaging studies in depression

H.G. Ruhé

Dose-escalation in the picture

Dose-escalation in the picture

deprssion

Omslag Ruhe.indd 1 17-9-2008 15:56:58

DOSE-ESCALATION IN THE PICTURE

PHARMACOLOGICAL AND IMAGING STUDIES IN DEPRESSION

Eric Ruhé

The work described in this thesis was mostly performed at the Academic Medical Center, Programfor Mood Disorders in the Netherlands. The studies were conducted in close collaboration withthe departments of Nuclear Medicine, General Practice, Pharmacology and Pharmacotherapy,Clinical Epidemiology, Biostatistics and Bioinformatics and Radiology.

Most of the studies in this thesis were financially supported by grants from the Academic MedicalCenter (SFA.07.012), the Netherlands Organisation for Health Research and Development(ZonMw), program Mental Health, education of investigators in mental health (OOG; #100-002-002), and from the Dutch Brain Foundation (14F06.45).

Additional support for the completion of this thesis was provided by Amicale Facel Holland.

Dose-Escalation in the Picture. Pharmacological and Imaging studies in depressionThesis, University of Amsterdam, The Netherlands. With a summary in Dutch.

Copyright © 2008, Henricus G. Ruhé, Amstelveen, The Netherlands [email protected]

All rights reserved. No part of this thesis may be reproduced, stored or transmitted in any form orby any means, without prior permission of the author.

Layout: ProBook with FLE technology by GDS Cross Media Group BV, Nieuw Vennep, TheNetherlands

Cover: Inge E. Kos, Medical Photography and Illustration Academic Medical Center, TheNetherlands

Printed by : PrintPartners Ipskamp BV, Enschede, The Netherlands

ISBN: 978-90-9023559-2

Dose-Escalation in the PicturePharmacological and Imaging studies

in depression

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctoraan de Universiteit van Amsterdamop gezag van de Rector Magnificus

prof. dr. D.C. van den Boomten overstaan van een door het college voor promoties

ingesteldecommissie,in het openbaar te verdedigen in de Aula der Universiteit

op woensdag 5 november 2008, te 10:00 uur

door

Henricus Gerardus Ruhé

geboren te Amsterdam

Promotiecommissie

Promotor: Prof. dr. A.H. Schene

Co-promotor: Dr. J. Booij

Overige leden: Prof. dr. D.A.J.P. DenysProf. dr. R. DierckxProf. dr. B.L.F. van Eck-SmitProf. dr. C.G. KruseProf. dr. W.A. Nolen

Faculteit der Geneeskunde

‘The brain is not a chemical factory but an extremely complicated survival machine’

Arvid Carlsson, Nobel lecture 2000

Contents

Part I. General introduction 9

1. General introduction and outline of the thesis 11

2. Design and methods of the DELPHI-study 31

Part II. Studies to guide clinical treatment of Major Depressive Disorder 41

3. Clinical use of the Hamilton Depression Rating Scale: Is increased efficiency possible? A post hoc comparison of HDRS, Maier and Bech subscales, CGI and SCL-90 scores. 43Comprehensive Psychiatry 2005; 46: 417-427

4. Dose-escalation for insufficient response to a standard dose of a selective serotonin reuptake inhibitor in major depressive disorder: a systematic review. 59British Journal of Psychiatry 2006; 189: 309-316

5. Switching antidepressants after a first selective serotonin reuptake inhibitor in major depressive disorder: a systematic review 73Journal of Clinical Psychiatry 2006; 67: 1836-1855

Part III. Serotonin in the etiology of Major Depressive Disorder 99

6. Mood is indirectly related to serotonin, norepinephrine and dopamine levels in humans: a meta-analysis of monoamine depletion studies 101Molecular Psychiatry 2007; 12: 331-359

7. Serotonin transporter binding with [123I]β-CIT SPECT in major depressive disorder versus controls: effect of season and gender. 139Submitted

Part IV. Neurobiological effects of paroxetine in the treatment of Major Depressive Disorder 155

8. Serotonin transporter gene promoter polymorphisms modify the association between paroxetine serotonin transporter occupancy and clinical response in major depressive disorder 157Pharmacogenetics and Genomics 2008; in press

9. Evidence why paroxetine dose-escalation is not effective in major depressive disorder: a randomized-controlled trial with assessment of serotonin transporter occupancy 173Neuropsychopharmacology 2008; in press

10. Successful pharmacological treatment of major depressive disorder attenuates amygdala activation to negative facial expressions. An fMRI study 195Submitted

11. Effect of the selective serotonin reuptake inhibitor paroxetine on platelet function is modified by a SLC6A4 serotonin transporter polymorphism 217Journal of Thrombosis and Haemostasis 2008; in press

Part V. Summary and general discussion 231

12. Summary, conclusions and general discussion 233

13. Samenvatting en conclusies 253

Part VI. Abbreviations, color figures, publications, acknowledgement and curriculum vitae 269

Abbreviations 271

Color figures 275

Publications 285

Dankwoord 291

Curriculum Vitae 297

GENERAL INTRODUCTION

PART

General introduction

CHA

PTER

GENERAL INTRODUCTION AND OUTLINE OF THE THESIS

ChapterGeneral introduction and outline

1

13

General introduction

The present thesis concerns the pharmacological treatment of Major Depressive Disorder (MDD).It especially focuses on what to do when patients do not respond to a standard dose of anantidepressant.

In this thesis, the results of studies that were performed since 2001, while working at theProgram for Mood Disorders of the department of Psychiatry of the Academic Medical Center willbe reported. These studies first aimed at systematically reviewing the existing literature abouttreatment strategies for non-response to the first antidepressant (included in Part II). Thesereviews identified an important gap between equivocal evidence for dose-escalation and the firmrecommendation of this strategy in clinical guidelines. This was the starting point for theexperimental studies to further investigate the clinical efficacy and the mechanism behind dose-escalation, described in Part IV. Meanwhile, the findings, and further reading of conflicting resultsof others raised questions about the pathophysiological origin of MDD. Therefore, the generallyknown serotonin hypothesis for MDD, will be addressed in two additional studies (Part III). Part Vof the thesis comprises the general discussion and will address directions for future research.

In this introduction chapter, the reader will be introduced to some backgrounds of MDD, itspharmacological treatment, the monoamine hypothesis about the etiology of MDD, and briefly onthe mechanisms of action of antidepressant drugs. After a summary, the separate studies of thethesiswill be introduced. In chapter 2, the methods of the Dose-Escalation Legitimate?Pharmacology and Imaging studies in depression (DELPHI-study) will be described. However, let’sstart with some history.

MELANCHOLY, the subject of our present discourse, is either in disposition, or habit. In dispositionis that transitory melancholy which comes and goes upon every small occasion of sorrow, need,sickness, trouble, fear, grief, passion or perturbation of the mind, any manner of care, discontent orthought, which causeth anguish, dullness, heaviness and vexation of spirit, any wayes opposite topleasure, mirth, joy, delight, causing forwardness in us, or a dislike. In which equivocal and impropersense, we call him melancholy, that is dull, sad, sowr, lumpish, ill disposed, solitary, any way moved ordispleased. And from these melancholy dispositions, no man living is free. […] Melancholy, in thissense, is the character of mortality. […]

It falleth out oftentimes that these dispositions become habits, and many affects contemnedmake a disease. […]for that which is but a flea-biting to one, causeth unsufferable torment toanother; and which one by his singular moderation and well composed carriage can happilyovercome, a second is no whit able to sustain; but upon every small occasion of mis-conceived abuse,injury, grief, disgrace, loss, cross, rumour, etc. (if solitary, or idle) yields so far to passion, that hiscomplexion is altered, his digestion is hindered, his sleep gone, his spirits obscured, and his heartheavy, his hypocondries mis-affected; wind, crudity, on a sudden overtake him, and he himselfovercome with melancholy. […] If any discontent seise upon a patient, in an instant, all otherperturbations will set upon him; and then, like a lame dog or broken-winged goose, he droops, andpines away, and is brought at last to that ill habit or malady of melancholy it self. […]

But all these melancholy fits, […] displeasing, violent and tyrannizing over those whom theyseise on for the time–yet these fits, I say, or men affected, are but improperly so called, because theycontinue not, but come and go, as by some objects they are moved. This melancholy, of which we areto treat, is an habit, morbus sonticus, or chronicus, a chronick or continuate disease, a settledhumour, not errant, but fixed; and as it was long increasing, so, now being (pleasant or painful)grown to an habit, it will hardly be removed. […]

INVETERATE melancholy, howsoever it may seem to be a continuate, inexorable disease, hard tobe cured, accompanying them to their graves most part, yet many times it may be helped, even thatwhich is most violent, or at least it may be mitigated and much eased. Nil desperandum. It may behard to cure, but not impossible for him that is, most grieviously affected, if he be but willing to behelped.”1

Chap

ter

Chapter 1

1

14

Burton’s work The anatomy of melancholy* (first published in 1621) was a best-seller.1 Itdescribes many aspects of the disease that we currently recognize as depression:2 the sometimesdifficult distinction of the illness melancholy (depression) from ‘normal’ sadness, pain or sorrowafter misery in life, the relation between stressful life-events and the development of the illness,the interaction of life-events and coping styles to develop the illness or remain well, the clinicalsymptoms, the enormous impact of the illness on a depressed person’s life, social dysfunction andthe tendency of recurrence and chronicity.1

Nowadays, MDD is one of the most prevalent and disabling illnesses in psychiatry, afterischemic heart disease the second most common cause of disability worldwide and expected tobe the world’s second cause of disability by 2030, behind HIV/AIDS.3 MDD is diagnosed by theoccurrence of a cluster of different symptoms affecting mood, pleasure, attention, activities, vitalsomatic functions (eating, sleep), and (ruminative) thoughts over oneself, guilt or suicide over aprolonged period of time. For clinical and scientific uniformity, MDD is currently defined bycriteria specified in the Diagnostic and Statistical Manual (4th edition; DSM-IV),4 which is a non-theoretical classification-system.

Epidemiology of MDDThe 12-month prevalence of MDD is estimated to be 5.8% in the Netherlands,5 and 5.3% in theUnited States.6 Prevalence rates are two times higher in women,5;7-10 and higher in widowed,divorced, unemployed or disabled people, people with somatic disease and first degree relativesof patients with MDD.6;9 In the general Dutch population, the median duration of MDD is 3months, with 63% of new MDD episodes recovering within 6 months (regardless whethertreatment is offered or the natural course of the disease is awaited), however, this prognosis isless favourable for people who seek help.11 Estimations of the lifetime prevalence are 15.4% in theNetherlands,5 >16% in Europe,8 and 13.2%-16.2%6 in the USA, with the same 2:1 female-maledistribution. Most epidemiological studies indicate that nowadays 51-69% of depressed patientsseek or receive treatment,8;9 which has increased over the last decade relative to the nineties ofthe previous century.10 However, only 50% of patients who are treated for MDD meet diagnosticcriteria of MDD or meet indication criteria for treatment,10;12 and it is estimated that only 21.7% ofMDD-patients in the USA receive adequate treatment.9 Therefore, depending on one’s view, MDDis frequently over-treated, under-treated or mis-treated.

MDD is a recurrent and potentially chronic disease. After the first episode, 50-80% of patientshave a recurrence within 10 year.13;14 Of incident episodes of MDD 15-20% will become a chronicdepression (defined as a duration of ≥2 years), although higher percentages (46%) have beenreported.11;15;16

MDD is associated with high direct treatment costs as well as indirect costs of loss ofproductivity and quality of life.8;17 In the Netherlands, treatment costs of MDD (€250 million) in1994 were estimated to comprise 1% of the total healthcare budget, and 13.6% of the total mentalhealth budget.18 In 2003, MDD treatment costs in the Netherlands had increased to €660 million,still consuming 1.1% of the total healthcare budget.19 In 1999 10% of newly assigned disability-payments (WAO) in the Netherlands was due to MDD.20

Treatment of MDDIn order to treat major depressive disorder effectively, many national clinical guidelines weredeveloped.13;21-30 Besides various forms of psychotherapy, pharmacotherapy is often applied asmonotherapy or in combination with psychotherapy. Antidepressants are effective drugs forMDD when compared to placebo, although effect sizes are moderate,24;26;31-33 and some criticsclaim that effects are exaggerated by pharmaceutical companies.34-37 Antidepressants areprescribed for MDD, but also for other psychiatric disorders, and are of considerable financialinterest. In 2006, approximately 780.000 inhabitants (4.7%) of the Netherlands used

* Melancholy is derived from the Greek Μελαγχολια (melancholia), which refers to the black choler (bile) that was seen asthe material cause of the illness those days.


1

15

antidepressants, mostly Selective Serotonin Reuptake Inhibitors (SSRIs). Annual prescriptions of antidepressants increased over the last decade by ~6% every year. Thecosts of prescribed antidepressants were €156 million in 2006,38 again, most costs represent theprescription of SSRIs.

A range of antidepressants have been developed, all aiming at the stimulation of serotonergicand/or noradrenergic neurotransmission.39;40 Four classes of antidepressants exist: SSRIs, Tricyclicantidepressants (TCAs), Monoamine-oxidase inhibitors (MAOIs) and a miscellaneous group. TCAsand SSRIs, but also ‘dual action’ antidepressants block the transporters at the pre-synaptic nerveending, which causes inhibition of the reuptake of either serotonin (5-HT), noradrenaline (NA; =norepinephrine), or both. The classic MAOIs irreversibly inhibit the enzymes mono-amine oxidaseA and B, which degradeserotonin, noradrenaline and dopamine. The miscellaneous groupharbours ‘dual action’ antidepressants (e.g. venlafaxine or duloxetine), and antidepressants withpre- and post-synaptic targets different from serotonin or noradrenaline transporters.39;40

By nature of their largely similar, but also slightly distinct pharmacological affinities fortransporters and receptors, antidepressants have (almost) comparable efficacy,32;41-43 but aredifferent in their adverse effects. Because SSRIs are tolerated better than TCAs,32;44;45 SSRIs havebecome the antidepressants of first choicein most countries.38 Within their class, different SSRIsalso have comparable efficacy, but different side effect profiles.46;47

Achieving remission is the aim of treatment,because persistence of (residual) symptomsincreases the risk of a future relapse or recurrence and is associated with ongoing psychosocialdysfunction.48;49 If remission is achieved, patients are advised to continue their antidepressants(at least 6 months) in order to prevent relapse,21;50 although guidelines’ recommendations areinconsistent.13;21;23-25 Nevertheless, only 50% of the patients respond (defined as a 50% decrease indepression severity)40 to the initial antidepressant, and remission rates (defined as a depressionseverity score below a certain cut-off) are lower (28-35%).28;29;51 Therefore, clinicians who treatMDD are often confronted with non-response, for which evidence-based approaches arenecessary.52

Insufficient response to antidepressants: A challenge for the clinician.The clinician faces several dilemmas when a patient shows insufficient response to anantidepressant. First, how does one ascertain clinical improvement?

A variety of measurement scales is available to measure depression severity and improvementof therapy.53 Psychopathology rating scales with clear cut-off scores provide a more directrepresentation of the effects of treatment than quality of life scales or a functional outcome.Thus, the use of a symptom-rating scale is highly recommended to assess treatment effects,either clinician-rated (e.g. the Hamilton Depression Rating Scale (HDRS),54 Montgomery ÅsbergDepression Rating Scale (MADRS)55 or Inventory for Depressive Symptoms (IDS-C))56 or self-rated(Inventory for Depressive Symptoms (IDS-SR)56, Quick-IDS (QIDS-SR)57 or Beck DepressionInventory (BDI)).58 Despite criticism on internal validity, selection and rating of items, andsensitivity to change,59-64 most scales correlate with each other.57;65 Frank et al.66 proposedtentative definitions for response (≥50% decrease of pretreatment rating scale scores) andremission (e.g. HDRS ≤7, MADRS ≤7, IDS-SR ≤13, BDI ≤9). Therefore, measurements should beroutinely performed before the start of treatment and at consecutive critical decision points.67;68

However, scales are not routinely implemented in daily practice. Apparently, the currentchallenge of implementation is to persuade clinicians to measure MDD severity repeatedly withappropriate scales.

The next dilemma is when to measure the effects of the antidepressants. Although a recentmeta-analysis indicated that the effects of antidepressants on depressive symptoms can beobserved within the first weeks of treatment,69 patients may need 6-10 weeks to achievesubstantial improvement.51;70 Therefore, the initiated treatment should not be changed too early,neither too late. Changing a regimen early may forestall a potentially late effect of the currenttreatment, but may also offer an increased chance of response. Contrary, unchanged treatmentmay increase despair due to the failure of response,but may prevent the (possible premature)

Chap

ter

Chapter 1

1

16

start of a more aggressive treatment with increased side effects. Over the last decades, the timeto declare that an antidepressant trial failed, increased from 3 weeks71 to 6-8 weeks.70;72

Interestingly, recommendations on this lag-time were never derived from studies primarilydesigned to quantify the timing of the determination of response. Current guidelines fordepression differentially recommend a minimal duration of antidepressant therapy between 4and 6 weeks.13;21;23;25;26;28;29;67;73 At these time points (‘critical decision points’)67 a next treatment-strategy must be considered.

A final dilemma is to adequately re-evaluate the diagnosis, without the (counter-) transferenceof helplessness. Critical decision points are important for the reassessment of somatic orpsychiatric co-morbidity, re-evaluation of persistent psychosocial problems and treatmentadherence. However, when none of these issues apply, the patient may be accused of ‘a failure tocooperate’, or blamed for the non-response by ‘having a personality disorder’ or ‘not beingmotivated for treatment’. Although these reproaches occasionally may be justified, they couldalso indicate the patient’s and/or clinician’s helplessness, which is often transferred from adepressed person to the persons surrounding him or her. Most depressed patients want toimprove, but are indeed disabled in their coping-styles, their motivation and their problem solvingcapacities. As such, routinely administered behavioural interventions (registration, planning andgradual reactivation)74 are valuable in the pharmacological treatment of MDD, also deferringincreased helplessness.

Insufficient response to antidepressants: What to do when the miracle doesn’thappen?Five pharmacological strategies for insufficient response to an antidepressant can be distinguished: 1. prolongation, 2. dose-escalation, 3. switching, 4. augmentation and 5. combina tion, allregarding the antidepressant initially given. Of course, a sixth strategy could be the addition ofpsychotherapy to the monotherapy of pharmacotherapy,75;76 but as this is not a topic of thisthesis, only the first 5 strategies are described.

1 Prolongation of the trial for another 2-4 weeks. This strategy is applicable only in case of doubtof the nature of the non-response (e.g. amendable psychosocial stressors, non-adherance),but is not favourable given the relative importance assigned to critical decision points.

2 Dose-escalation. With this strategy the dose of the prescribed antidepressant is increased tothe maximal tolerable dose. Dose-escalation assumes a dose-response relationship, which isequivocal for SSRIs. The advantage of dose-escalation is that it is easy and quick to apply.

3 Switching of the antidepressant. With this strategy the prescribed antidepressant isdiscontinued (eventually after tapering and a wash-out period) and the antidepressant isswitched to another antidepressant either belonging to the same class, or to a different classcompared to the initial antidepressant. The obvious advantage of switching is that it facilitatesa new pharmacological approach, possibly targeting different neurotransmitter systems. Itsdisadvantage is that it requires some time, and that one may unwontedly discard achievedimprovement when the chosen antidepressant appears to be a less effective one for thispatient.

4 Augmentation of the prescribed antidepressant. With this strategy a drug is added that has nostrong antidepressant effects by itself, but is known to increase the effects of antidepressants(e.g. lithium). The advantage of augmentation is that the effect of the antidepressant initiallyprescribed is maintained. Furthermore augmentation might result in faster responses thanswitching (as tapering is not required, nor a wash-out). Disadvantage is the risk of poly-pharmacy and the related interactions.

5 Combination of antidepressants. With this strategy another antidepressant drug, mostly withdifferent pharmacological properties compared to the antidepressant already prescribed, isadded to the initial antidepressant. This strategy has the same advantages and disadvantagesas augmentation strategies, but the risks of severe interactions are higher.


1

17

Neurotransmission, the serotonergic system and neuronal networksThe brain passes information through electrical signals in neurons. Synapses form the functionalcontacts between neurons and mainly communicate through neurotransmitter secretion.Neurotransmitters are stored in synaptic vesicles. The vesicles are released into the synaptic cleftin response to presynaptic depolarisation. After the release of neurotransmitters in the synapse,they are either metabolized or transported back into the terminal to be used again.77

Neurotransmitters bind to specific receptor proteins on the membrane of the pre- andpostsynaptic neurons. Postsynaptically, receptors either induce a change in postsynaptic ionchannels (ionotropic receptors), causing a depolarisation, or they activate G-protein mediatedcomplexes (metabotropic receptors), which activate one or more metabolic steps (‘secondmessengers’). Activation of metabotropic receptors is generally responsible for forming enzymesthat regulate gene expression, neurotransmitter synthesis, receptors and neuroplasticity. Whichof the mechanisms is used depends on the particular postsynaptic receptor type.78 Three major‘monoamine’ neurotransmitters are associated with psychiatric disorders: serotonin,noradrenaline and dopamine, of which the latter is traditionally related to psychotic disorders.SSRIs target the serotonergic system and increase serotonergic neurotransmission. From here,the serotonergic system and the neuronal networks that are believed to be involved in MDD willbe described.79

SerotoninThe serotonin neurons of the brain start in the raphe nuclei in the midbrain and project to theneocortex, basal ganglia, temporolimbic zones, hypothalamus, cerebellum and the brain stem(Figure 1.1).80;81 Serotonin is involved in several functions: sleep and wakefulness, appetite, nausea,migraine, headaches and regulation of mood.77 The serotonin receptors comprise of 5-HT1A, 1B,

1C, 1D, 1E, 1F, 5-HT2A, 2B, 2C, 5-HT4, 5-HT5, 5-HT6 and 5-HT7 (metabotropic) and the 5HT3 (ionotropic)receptors.78 Serotonin plays an important role in brain development via regulation of neuriteoutgrowth, synaptogenesis and cell survival.82;83 Serotonin that is released into the synaptic cleftis either taken up back into the presynaptic nerve ending by the serotonin transporter (SERT), ordegraded by MAO-A.77;78

The serotonin (and noradrenaline) deficiency hypothesisIn the late fifties of the previous century, TCAs appeared to be effective in treating MDD byincreasing serotonergic and noradrenergic neurotransmission. This discovery led to themonoamine hypothesis: MDD might etiologically be explained by a deficiency in monoamineneurotransmitters: serotonin ornoradrenaline. Since then, the working mechanism of AD isbelieved to be by (1) increased neurotransmission by increased synaptic levels of serotonin,noradrenaline and/or (2) specific agonistic effects on serotonin or noradrenaline (sub-)receptors.

Figure 1.1. Serotonergic pathways through the human brain.

Serotonergic system. From the raphe nuclei in the midbrain, neurons project to the neocortex, basal ganglia, temporolimbiczones, hypothalamus, cerebellum and the brain stem.

��

��

��

�� !��

��!��

��

Chap

ter

Chapter 1

1

18

Depletion of the available serotonin andnoradrenaline is used as a model to test theinvolvement of monoaminergic systems in MDD. Serotonin depletion can be achieved by rapidlylowering the essential amino-acid tryptophan which cannot be synthesized by the body and mustbe ingested to enable formation of serotonin. To achieve depletion, a tryptophan free amino-acidmixture is administered (acute tryptophan depletion).84 Depletion ofnoradrenaline anddopamineoccurs simultaneously, and uses the same concept (acute depletion of the essentialamino-acids phenylalanine and tyrosine).85 As an alternative to induce a state of depletion,enzyme-blocking agents decrease the production of the monoamines. Para-chlorophenylalanineblocks serotonin synthesis,86 and Alpha-methyl-para-tyrosine blocks noradrenaline and dopaminesynthesis.87

Since 1975 an increasing number of depletion studies have been conducted, with differenteffects in different study-populations. In general in healthy controls no clear mood-effects werefound, unless they had relatives with MDD. In remitted MDD patients who used antidepressants(or shortly after tapering) approximately 50% of the patients experienced a relapse afterdepletion. However this occurred only after depletion of the monoamine that theirantidepressant targeted. In depressed patients no consistent deterioration of the mood effectswere found.88-93

Thus, the monoamine-deficiency theory, in its purest form, states that depression can becured by the increase of serotonergic and/or noradrenergic neurotransmission. However, thereverse train of thought, that depression is bio-etiologically caused by a deficiency of monoamines(e.g. serotonin and/or noradrenaline) has attractive face-validity, but probably is an untenable,superficial simplification.75;94 Therefore, the current, less pertinent view is that the monoaminehypothesis only partially explains MDD and the response to AD.95-98

The limbic-cortical dysregulation hypothesisIn a more multidimensional, systems-level model, MDD can be viewed as a disorder affectingdiscrete but functionally integrated pathways; neural networks, which can be identified byneuroimaging techniques.† In such a network, dysfunction in one or more of the elements (e.g.after cognitive or somatic stress), will initially be tried to be influenced (or compensated) byother, remaining parts of the network, that try to maintain homeostatic emotional control.Therefore, results from neuroimaging studies investigating differences between healthy controlsand MDD patients, must be considered as the identification of regions to be either etiologicallyabnormal or regions involved in (mal-)adaptive compensatory processes.99

Because MDD is an affective disorder, the neurobiology of emotion processes is likelyinvolved. For the processing of emotions two systems are important: a ventral system (consistingof amygdala, insula, ventral striatum, ventral anterior cingulate gyrus, and ventral prefrontalcortex) and a dorsal system (consisting of hippocampus, dorsal anterior cingulate gyrus, anddorsal prefrontal cortex). The ventral system serves to identify the emotional significance of astimulus, the production of mood states, and automatic regulation of emotional responses, whilethe dorsal system serves to effortfully regulate mood states and subsequent behaviour.100

Initial lesion-deficit studies, early Positron Emission Tomography (PET) studies (measuringregional resting state glucose metabolism or blood flow) and later functional MagneticResonance Imaging (fMRI)-studies identified several brain regions to be affected by MDD: thelimbic structures (amygdala, hippocampus, hypothalamus and brainstem), the subcortical (basalganglia and thalamus) and the cortical (dorsolateralprefrontal, ventrolateral prefrontal andorbitofrontalcortex (DLPFC, VLPFC, OFC respectively) structures. In these, the most consistentfinding is a hypoactivity of the (dorsal) frontal lobe, while often a hyperactivity in the (ventral)VLPFCand OFC is found.99;101 Additionally, an increased activity of the (rostral, subgenual) anteriorcingulated gyrus99;101 and the amygdala, anterior insula, and ventral striatum was found, althoughless consistently.101 Furthermore, fMRI studies point to a increased sensitivity of the amygdala for

† Single Photon Emission Computed Tomography (SPECT) and Positron Emission Tomography (PET) use a variety ofradioligands with various half-lives to quantify different targets (transporters, receptors, or blood-flow or metabolism).Functional Magnetic Resonance Imaging (fMRI) is a technique to image brain activity (hemodynamic response) related toa specific task or stimulus.


1

19

emotional stimuli, and a bias to interpret these stimuli in a negative context.101 Interestingly,induction of sad mood in healthy volunteers also produces increased blood flow in the insula,subgenual anterior cingulate gyrus and decreased blood flow in the dorsomedial prefrontalcortex (DMPFC).102

The above abnormalities were combined in the limbic-cortical dysregulation model, whichincludes a dorsal neocortical hypofunction, which results in ventral (para)limbic hyperactivity,with a reciprocity in this inverse relationship.103;104

Which effects in the brain cause antidepressants to be antidepressants? The various treatment forms available to relieve MDD and their moderate efficacy poses thequestion: How do antidepressants bring about their therapeutic effects? For this question, threelevels of action must be distinguished: 1. direct neurotransmission effects, 2. second messengereffects, and 3. change in neuronal networks. For levels 1. and 2. see Figure 1.2. At all three levels,there appears to be a time dependent differentiation of the effects that occur, e.g. not all effectsare the same in the consecutive weeks after the initiation of treatment. This may explain why itsometimes takes several weeks before the therapeutic effects become apparent. Because mostof the recent research was done with SSRIs and SNRIs, the effects of these drugs are describedhereafter, unless indicated differently.

Direct effects of antidepressants on neurotransmission When an SSRI is ingested, the blockade of the target transporter (i.e. the serotonin transporter(SERT)) occurs within several minutes.105;106 Therefore, the antidepressant effects cannot only bebased on increased neurotransmission by decreased serotonin reuptake, as these effects takemuch longer than this direct pharmacological effect. Further research revealed that after twodays of treatment, the firing-rate of SERT-containing neurons in rats decreased, but that thisfiring-rate was restored within 2 weeks of continued treatment. This was attributed tosomatodendritic 5-HT1A autoreceptors, which normally have a negative feedback on the neuron’sfiring rate. These 5-HT1A autoreceptors appeared to desensitize.105 Microdialysis-experiments inrats showed that the restored firing-rate after 14 days was responsible for a 6-fold increase inintrasynaptic serotonin level, while after the acute blockade of the SERT this increase of theserotonin level was only small and transient.107 Furthermore, in humans, serotonin autoreceptors(5-HT1B/1D) (and also noradrenergic α2 autoreceptors) in the synapse normally inhibit serotoninrelease by feedback-mechanisms as well. Prolonged treatment with SSRIs again desensitize thesereceptors, resulting in increased serotonin release in the synaptic cleft.105;108 As such,desensitisation of HT1A autoreceptors in the raphe nuclei in the midbrain may have effects onserotonergic neurotransmission in critical brain areas where these serotonergic neurons projectto.109 Finally, in rats, after prolonged administration of SSRIs the SERT itself is downregulated by80-90%.110-112 This is probably caused by trafficking and internalization of the SERTs instead ofaltered SERT-gene-regulation, because mRNA expression in the cells studiedwas unaltered by thetreatment.106

Other antidepressants have rather different effects. TCAs (except clomipramine) do notchange the pre-synaptic serotonin-containing neurons, but appear to sensitize the postsynaptic 5-HT receptors for serotonin.105 Along with serotonin, the responsiveness for noradrenaline wasalso found to be enhanced, likely due to enhanced α-adrenoreceptor-mediated transmission.MAOIs also increase serotonergic transmission by desensitisation of 5-HT1A autoreceptors, but donot desensitize other 5-HT autoreceptors, and desensitize noradrenergic α2 autoreceptors, whichindirectly enhances serotonergic transmission.

Chap

ter

Chapter 1

1

20

Figure 1.2. Transporters, receptors and second messenger systems involved in the effects of antidepressants.

Figure adapted from Belmaker and Agam.79 The left half of the presynaptic neuron represents a serotonergic neuron, theright half a norepinephrinergic neuron. For color figure see page 277.In the presynaptic neuron, serotonin is synthesized from tryptophan by tryptophan hydroxylase and stored in vesicles.Likewise, norepinephrine is synthesized from tyrosine by tyrosine hydroxylase. These vesicles merge with the cell membranewhen the neuron is depolarized, thereby releasing their contents into the synaptic cleft. After release, serotonin and norepinephrine are transported back into the presynaptic neuron by serotonine andnorerepinephrine transporters. Furthermore, serotonin and norepinephrine are catabolized by the monoamine-oxidase A(MAO-A) enzyme. In the synaptic cleft, serotonin and norepinephrine affect both the pre-and post-synaptic neuron. The pre-synaptic 5-HT1A and 5-HT1B auto-receptors decrease serotonin release by inhibitory feedback; the α2-adrenergic receptordoes the same for the release of norepinephrine. Post-synaptically, serotonin and norepinephrine bind to G-protein-coupled monoamine receptors (MARs): the cyclicAMP(cAMP)-coupled receptor, which activates protein kinase A (PKA), and the Phosphatidylinositol (PI)-coupled receptor, whichactivates phospholipase C (PLC) which thereafter form inositol triphosphate (IP3) and diacylglycerol (DAG). IP3 and DAGactivate protein kinase C (PKC). Both PKA and PKC finally activate cAMP responsive element binding (CREB) protein, whichstimulates DNA transcription. For example, this might result in the production of brain derived neurotrophic factor (BDNF).


1

21

Second messenge reffects of antidepressants The secondary effects of antidepressants can be divided in post-synaptic effects on thesensitivityand availability of receptors, and the effects caused by the activation of post-synaptic receptors.79

Effects on post-synaptic receptors include a decrease of 5-HT1A and 5-HT2Areceptor density,113 and5-HT1D

114 and 5-HT2C/2B responsivity.113;115 Because these effects are either investigated in animals orby using indirect measures (e.g. growth-hormone release), it is important that Meyer et al. onlydemonstrated 10% in-vivo 5-HT2A receptor downregulation in MDD patients between 20-30 years,but not in older patients.116 The desensitisation of post-synaptic receptors might explain thetolerance to adverse effects that occurs after some days of exposure to antidepressants.109

Interestingly, in rodents, the activation of post-synaptic 5-HT1A receptors in the hippocampusstimulates neurogenesis, which appears to be required for therapeutic effects ofantidepressants.117

After the activation of the G-protein metabotropic receptors (either by serotonin ornoradrenaline), second messenger systems mediate signals to the neuron’s nucleus, where cAMPresponsive element binding protein (CREB) regulates CREB-directed gene transcription.79;118 Mostevidence indicates that CREB is upregulated by chronic antidepressant use, but not for allantidepressants, and results of studies appear to be biased by the cell type investigated, the brainregion where these cells originate from and the timing after the first antidepressant exposure.118

One of the genes that is (positively) influenced by CREB is the brain derived neurotrophic factor(BDNF) gene.119 BDNF is a plethoric growth factor, which regulates neuronal survival, migration,differentiation, axonal and dendritic growth and synapse formation. The genomic structure ofBDNF is complex, which facilitates differential activation by diverse and variable stimuli, which canbe different in different brain regions and even in different parts of the cell.83

Changes in neuronal networksAt the neuro-anatomical level, the novel neuroimaging techniques have ‘opened’ the brain tostudy the in-vivo effects of psychiatric treatment.104 In MDD, early PET studies found timedependent changes during the treatment with fluoxetine.120 After 1 week, glucose metabolismincreased in the hippocampus and brainstem, and decreased in posterior cingulate gyrus, striatumand thalamus compared to the pre-treatment scan. After 6 weeks, patients who responded to thetreatment had a decrease in metabolism in the subgenual cingulate gyrus, the hippocampus, thepallidum and insula and an increase in the anterior and posterior cingulate gyrus, prefrontal andparietal cortex. However, the non-responders showed a persistent, unchanged pattern of changeas seen in week 1. The changes in prefrontal cortex and subgenual cingulate gyrus correlated bestwith symptomatic improvement.99;101;120 These findings were replicated in patients treated withparoxetine.121 Moreover, patients who responded while on placebo-treatment showed somehowsimilar but also distinct‡ changes in brain metabolism compared to fluoxetine responders.99;122 Thisindicates that in neuroimaging studies, response andtreatment effects may coincide, but bothmay also have their specific, distinguishable effects.

In recent fMRI studies, the increased activation of the amygdala to negative (sad, fearful,angry) and happy faces have been investigated after treatment of SSRIs (fluoxetine,123;124

sertraline125), venlafaxine126;127 and bupropion.128 These studies rather consistently founddecreased activation of the amygdala, insula and increased cortical activity after treatment.However,most patients in these studies were treatment responders at the end of the study. Incontrast with previous PET-studies, none of the fMRI studies used a placebo-comparisonmeasured twice, so no distinction between specific drug and response effects in fMRI can bemade yet.

‡ Similar changes included increased metabolism in frontal, parietal and posterior cingulate, and decreased metabolism insubgenual cingulate gyrus. Distinct changes included no changes in subcortical brainstem, hippocampal, and caudatemetabolism in placebo-responders.

Chap

ter

Chapter 1

1

22

(Pharmaco-)genetic effects relevant for antidepressantsSeveral genetic polymorphisms have been investigated in relation to treatment response toSSRIs. The polymorphism studied most is the SERT promoter gene (5-HTTLPR), for which a long(L) and a short (S) variant were identified, with a recently discovered functional tri-allelic variant(rs25531).129 The 5-HTTLPR is associated with the transcriptional activity of the SERT gene.130 Cellshomozygous for the L-allele produce higher concentrations of SERT mRNA, and the rate ofserotonin uptake by the transporter is >2-fold higher than in cells containing one or two copies ofthe S-allele. A meta-analysis of 15 studies showed a pooled association between the 5-HTTLPR-polymorphism and SSRI efficacy,131 with MDD patients with at least one L-allele having higherresponse rates to SSRIs. However, in a large sample of patients treated with citalopram (an SSRI),treatment response was not associated with the tri-allelic 5-HTTLPR-polymorphism.132;133

Furthermore, individuals carrying the S-allele experience increased adverse events after SSRItreatment,132;134 have elevated risk of depression in relation to life events,135 but alsoshowincreased amygdala reactivity to fearful stimuli.136;137 A large MRI-study in healthy controls showedassociations of the 5-HTTLPR S/S polymorphism with unfavourable alterations in anatomy andfunction of the amygdala-cingulate feedback circuit.138 Thesefindings strongly argue for animportant role of the 5-HTTLPR-polymorphism in the development and functioning of emotionalnetworks involved in MDD. Other pharmacogenetic associations with clinical response have beeninvestigated, but will not be further addressed in this thesis.

Summary and questions addressed in this thesisMajor Depressive Disorder is a prevalent and disabling illness, which potentially recurs and maybecome a chronic disease. It is the second most common cause of disability worldwide, and has a12-month prevalence of ±5.5%, with lifetime prevalences of 12-14% in males and 22-24% in females.3;5

MDD is associated with high direct treatment costs as well as indirect costs of loss of productivityand quality of life. Besides various forms of psychotherapy, pharmacotherapy with antidepressantdrugs is often applied.13;21-25

Antidepressants are grouped in four different classes: Selective serotonin reuptake inhibitors(SSRIs), tricyclic antidepressants (TCAs), monoamine oxidase inhibitors (MAOIs) and amiscellaneous group including so-called dual action antidepressants. SSRIs, dual-actionantidepressants and TCAs are used most often.39 Most antidepressants increase serotonergic and/or noradrenergic neurotransmission. When considered more precisely, different classes ofantidepressants appear to have distinct additional pre- and postsynaptic effects.105;106 Secondaryto increased serotonergic and/or noradrenergic neurotransmission, complex second messengerpathways are activated, which are only partly understood.79 Macroscopically, presumably relevantchanges in neuronal networks following antidepressant treatment have been identified.99;101;104

Nevertheless, which changes are required and specific for the improvement of symptomsremainsan enigma. Furthermore, the effects of increased neurotransmission might be affected by thegenetic make-up of patients (i.e. the 5-HTTLPR polymorphism129). Genetic polymorphisms appearto be associated with treatment effects,131 but might also influence the development andconnectivity of the neuronal networks.138

Despite comparable efficacy of antidepressants, only 50% of the MDD-patients respond to thefirst antidepressant trial given, while fewer achieve full remission of symptoms.48;51 Therefore,insufficient response to the first antidepressant is a relevant clinical problem that challengesclinicians.

The clinician’s first dilemma is how to adequately and efficiently measure the changes insymptom severity, for which the further development and implementation of short, easy to useand valid clinician rated questionnaires would improve clinical decision making. The seconddilemma is to time the change of the treatment initiallystarted: this should not be changed tooearly, neither too late.70 The third dilemma is to balance helplessness, impaired functioning by thedisease and (counter-) transference by ‘blaming the victim’, i.e. reproaching the patient of themoderate efficacy of antidepressants as ‘being unmotivated’ or ‘having a personality disorder’.


1

23

When ‘the miracle doesn’t happen’, 5 strategies for non-response are possible options:prolongation of the initial trial, dose-escalation, switching, augmentation and combination ofantidepressants. For all these options the current evidence is either equivocal or fragmentary. Inorder to develop evidence based treatment algorithms for the 50% of MDD-patients who turn outto be non-responsive to the first antidepressant trial, systematic overviews of the literature arerequired and gaps in the evidence-base need to be addressed by new research-projects.

More fundamentally, the perceived delayed onset of symptom improvement withantidepressants, their moderated efficacy, and their extensive effects in the brain beyond anincrease in serotonergic or noradrenergic neurotransmission, summons the question what is theetiopathogenetic model for MDD. Or, more specifically, what is the evidence that corroboratesthe corner-stone of most antidepressant’s action: the monoamine hypothesis and theinvolvement of the serotonergic system as a causal explanation for MDD.

Research projects underlying this thesisThis thesis incorporates three research projects that were initiated and conducted by the Programfor Mood Disorders of the Academic Medical Center (AMC). One additional project was initiatedby the department of vascular medicine of the AMC.

First, a guideline-project was initiated with a grant from the Academic Medical Center (NoSFA.07.012). Aim of this guideline project was to develop an evidence based guideline for non-response after a first SSRI.

Second, as a result of this guideline project, two grants were obtained from the NetherlandsOrganization for Health Research and Development (ZonMw), program Mental Health, educationof investigators in mental health (Geestkracht-OOG; projects #100-002-001 and #100-002-002).These grants were applied to initiate the DELPHI-study (Dose-Escalation Legitimate?Pharmacology and Imaging studies in depression). The DELPHI-study was set up as amethodologically sound trial to study clinical effectiveness of dose-escalation as a strategy fornon-response. As a novel extension to existent dose-escalation trials, we aimed to investigate themolecular neurobiological target of dose-escalation: the occupancy of the serotonin transporter(SERT) by paroxetine (a SSRI). This was done by the acquisition of two or three single photonemission computed tomography (SPECT) scans in a subgroup of drug-free patients of the totalDELPHI-cohort.

A third project was started as an extension to the DELPHI-SPECT study. This third project couldbe planned after a grant from the Dutch Brain Foundation (Hersenstichting) was obtained(project #14F06.45). The DELPHI-fMRI aimed to investigate the neurobiological changes in brainactivation by treatment with paroxetine in a functional Magnetic Resonance Imaging (fMRI)study. This study was superimposed on the SPECT-imaging of the last 22 drug-free patientsparticipating in DELPHI.

The final fourth research project in this thesis was initiated after an almost fatal bleedingcomplication, which occurred in a patient who underwent surgery, and lost 12 litres of blood. Thiscomplication was afterwards attributed to fluoxetine use. This project was funded internally bymutual contributions of both participating AMC-departments.

The questions addressed in this thesis

1 Is a short, easy to use clinician rated questionnaires as effective and precise as the routineHamilton depression rating scale (HDRS)?The HDRS54 is most frequently used as the golden-standardin clinical trials, but is probably tooextensive to use in clinical practice.61 To facilitate clinical measurements of depression severityand objectivity for critical decision points, we reanalyzed the treatment-outcomes of twoantidepressant-psychotherapy trials, which were performed by the Mentrum researchgroup.139;140 We therefore investigated whether the effect-sizes of two 6-item subscales of theHDRS (17-items) – the Maier141 and Bech59 subscales – were comparable to the original HDRS inthe measurement of depression severity and the sensitivity to measure changes.Furthermore, we investigated whether this comparability was stable across the full range of

Chap

ter

Chapter 1

1

24

response to treatment, and across different treatments and for different baseline severity ofdepression. We also determined cut-off points for remission for these subscales compared toconventional HDRS definitions.66;142;143 See chapter 3.

2 What is the evidence for dose-escalation as a strategy for non-response to a first SSRI?For this question, we performed a systematic review of the evidence for the dose responserelationship for SSRIs in MDD. See chapter 4.

3 What is the evidence for switching antidepressants as a strategy for non-response to a firstSSRI?For this question, we performed a systematic review of the evidence for switching afterfailure of a first SSRI in MDD. Part of this study was a meta-analysis of three switch-studies.See chapter 5.

4 Does the depletion of monoamine (5-HT and NA/DA) systems lower mood in humans, and isthis lowering of mood different across different populations?For this question, we performed a systematic review of monoamine depletion studiesreporting mood effects of depletion. As an extension of previous systematic reviews ofmonoamine depletion studies,88-93 we aimed to pool the results ofthe small-sized depletionstudies, because they might not have detected small differences by a lack of power, andpooling would quantify the balance of positive versus negative studies. Therefore, we applieda pooling technique (modified from conventional meta-analyses of randomized controlledtrial data and including an adjustment for small sample bias) to handle the statistically pairedcross-over designs of these studies in formal, stratified meta-analyses.144 See chapter 6.

5 Do MDD-patients and healthy controls differ in the number of central serotonin transporters,and is the amount of available SERTs correlated with depression severity?Despite the fact that the working-mechanism of antidepressants supports the monoaminedeficiency theory, the pathogenesis of MDD remains unclear.95 Therefore, differences in SERTavailability in patients and healthy controls have been studied previously, with conflictingresults. Additionally, significant effects on SERT availability have been reported for gender,145

smoking behavior,145 aging146;147 and season of scanning.148 Therefore, we analyzed thebaseline SPECT-scans of the DELPHI-SPECT participants versus age and sex-matched healthycontrols. Because our sample size was large, we were able to properly account for potentialconfounders and possible interactions, of which the multivariate effects are reported. Seechapter 7.

6 Does a common genetic polymorphism of the promoter region of the serotonin transportergene (SLC6A4) modify the association between the SERT occupancy by paroxetine and theclinical response?We performed this study because SSRI-response is likely associated with 5-HTTLPRpolymorphisms, 5-HTTLPR polymorphisms might influence SERT availability (the target forSSRIs), and it is unclear how occupancy of the available SERTs is related to clinical response.Thus, we aimed to investigate the paroxetine treatment by genotype interaction regardingclinical response on the molecular level of SERT occupancy. We quantified the relationbetween SERT occupancy and clinical response, and studied how the 5-HTTLPR -polymorphism affected this SERT occupancy-response relationship. We performed this studyin the open phase of the DELPHI-SPECT study, when patients were treated with paroxetine 20mg/day for 6 weeks. See chapter 8.

7 Is dose-escalation of paroxetine an effective clinical strategy for non-response in MDD?The systematic review of dose-escalation (chapter 4), identified methodological flaws inprevious dose-escalation trials. In this study we reevaluated the clinical efficacy of dose-escalation of paroxetine without these flaws, and, considering the molecular target of SSRIs,we also tested whether paroxetine dose-escalation increasedSERT occupancy more thanplacebo dose-escalation. We therefore performed a 6 week, multicenter, randomized study indepressed patients not responding to 6 weeks of paroxetine at 20 mg/day. As a novelextension to previous clinical trials, and in order to elucidate the neurobiological basis for an


1

25

expected lack of benefit of dose-escalation, we included a SPECT imaging approach.Herewith, we quantified whether paroxetine dose-escalation increased SERT occupancy morethan placebo dose-escalation. This enabled us to relate clinical findings to the neurobiologicalcorrelate of SERT occupancy. See chapter 9.

8 Does treatment with paroxetine normalize amygdala hyperactivation in MDD?We initiated this study after the first reports of attenuated amygdala activation aftertreatment with sertraline.125 Thereafter several groups replicated a baseline hyperactivation ofthe amygdala but less consistently reported the attenuation of amygdala-hyperactivationafter treatment.123;126;128 We therefore investigated whether: activation of the amygdala by(negative) facial expressions differed from healthy controls, this activation of the amygdalachanged after 6 and 12 weeks of treatment with paroxetine, the activationof the amygdalaand other brain areas merely changedby paroxetine treatment or in relation with clinicalresponse, and whether dose-escalation of paroxetine in week 6 non-responders affectedactivations, compared to placebo-dose-escalation. For this study, we performed an fMRI studyin 22MDD patients who participated in the DELPHI-fMRI study. Patients were treated withparoxetine (20 mg/day followed by a randomized dose-escalation for non-responders) andwere scanned at baseline, 6 weeks and 12 weeks of treatment. We obtained a baseline scanfor 21 matched controls, to contrast baseline amygdala activation in MDD-patients. Seechapter 10.

9 What are the changes in hemostasis and blood platelet parameters when patients are treatedwith paroxetine, and are these changes modified by dose-escalation or a geneticpolymorphism of the promoter region of the serotonin transporter gene?In this study, we evaluated the effects of standard and increasing dosages of paroxetine onthe bleeding tendency and hemostatic functions of platelets in patients who were drug-freebefore the start of paroxetine. In addition, we assessed whether these effects are modified bythe 5-HTTLPR polymorphism. See chapter 11.

References

1. Burton R. The anatomy of melancholy, what it is, with all the kinds, causes, symptoms, prognostics, and several cures of it. Inthree partitions. With their several sections, members, and subsections, philosophically, medecinally, historically opened andcut up. 6th ed. London: J. Cuthell; J. Nunn; Longman and Co.; J. Mawman; Baldwin and Co.; Lackington and Co.; Black andCo.; S. Bagster; Edwards and Knibb; Ogle, Duncan, and Co. R. Saunders; J. Sheldon; Wilsons, York; 1651:1-461.

2. Schene AH, Huyser J, Ruhe HG. Validiteit van de diagnose depressie. In: Koerselman F, Vergouwen AC, eds. Degecompliceerde kliniek 2005: over de geldigheid van diagnoses. Amsterdam: Benecke; 2005.

3. Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med. 2006; 3: e442.4. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, fourth edition. 4 ed. Washington

DC: American Psychiatric Association; 1994.5. Bijl RV, Zessen Gv, Ravelli A. Psychiatrische morbiditeit onder volwassenen in Nederland: het NEMESIS-onderzoek. II.

Prevalentie van psychiatrische stoornissen. NedTijdschrGeneesk. 1997; 141: 2453-2460.6. Hasin DS, Goodwin RD, Stinson FS, Grant BF. Epidemiology of major depressive disorder: results from the National

Epidemiologic Survey on Alcoholism and Related Conditions. Arch Gen Psychiatry. 2005; 62: 1097-1106.7. Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M, Eshleman S et al. Lifetime and 12-month prevalence of DSM-III-

R psychiatric disorders in the United States. Results from the National Comorbidity Survey. Arch Gen Psychiatry. 1994; 51:8-19.

8. Lepine JP, Gastpar M, Mendlewicz J, Tylee A. Depression in the community: the first pan-European study DEPRES(Depression Research in European Society). Int Clin Psychopharmacol. 1997; 12: 19-29.

9. Kessler RC, Berglund P, Demler O, Jin R, Koretz D, Merikangas KR et al. The epidemiology of major depressive disorder:results from the National Comorbidity Survey Replication (NCS-R). JAMA. 2003; 289: 3095-3105.

10. Kessler RC, Demler O, Frank RG, Olfson M, Pincus HA, Walters EE et al. Prevalence and treatment of mental disorders,1990 to 2003. N Engl J Med. 2005; 352: 2515-2523.

11. Spijker J, de Graaf R, Bijl RV, Beekman AT, Ormel J, Nolen WA.Duration of major depressive episodes in the generalpopulation: results from The Netherlands Mental Health Survey and Incidence Study (NEMESIS). Br J Psychiatry. 2002; 181:208-213.

12. Kendrick T, King F, Albertella L, Smith PW. GP treatment decisions for patients with depression: an observational study.Br J Gen Pract. 2005; 55: 280-286.

Chap

ter

Chapter 1

1

26

13. American Psychiatric Association. Practice guideline for the treatment of patients with major depressive disorder(revision). American Psychiatric Association. Am J Psychiatry. 2000; 157: 1-45.

14. Beekman AT, Ormel J. Depressie. In: Jong Ad, Brink Wvd, Ormel J, Wiersma D, eds. Handboek psychiatrischeepidemiologie. 1 ed. Maarssen: Elsevier/ De Tijdstroom; 1999:300-328.

15. Keller MB, Lavori PW, Mueller TI, Endicott J, Coryell W, Hirschfeld RM et al. Time to recovery, chronicity, and levels ofpsychopathology in major depression. A 5-year prospective follow-up of 431 subjects. Arch Gen Psychiatry. 1992; 49: 809-816.

16. Wells KB, Burnam MA, Rogers W, Hays R, Camp P. The course of depression in adult outpatients. Results from theMedical Outcomes Study. Arch Gen Psychiatry. 1992; 49: 788-794.

17. Greenberg PE, Stiglin LE, Finkelstein SN, Berndt ER. Depression: a neglected major illness. J Clin Psychiatry. 1993; 54: 419-424.

18. Ruysbroek JM. Aan welke ziekten en aandoeningen wordt het geld besteedt? 1.2 (25 juni 2001). 2001. Bilthoven, RIVM.Volksgezondheid Toekomst Verkenning, Nationaal Kompas Volksgezondheid.

19. Slobbe LC, Kommer GJ, Smit JM, Groen J, Meerding WJ, Polder JJ. Kosten van ziekten in Nederland 2003. Zorg vooreuro's. [RIVM rapport# 270751010]. 2008. Bilthoven, RIVM. Volksgezondheid Toekomst Verkenning, Nationaal KompasVolksgezondheid.

20. Ziektendiagnosen bij uitkeringen voor arbeidsongeschiktheid. Statistische informatie over medische classificaties inWAO, WAZ en Wajong 1999.2001. Amsterdam, Landelijk instituut sociale verzekeringen.

21. Richtlijn Depressie. Multidisciplinaire Richtlijn Depressie. Richtlijn voor de diagnostiek en behandeling van volwassencliënten met een depressie. Brandt-Dominicus JC.2005. Utrecht, Trimbos-instituut, in opdracht van de LandelijkeStuurgroep Multidisciplinaire Richtlijnontwikkeling in de GGZ.

22. Marwijk HWv, Grundmeijer HG, Bijl D, Gelderen MGv, Haan Md, Weel-Baumgarten EMv et al. NHG-Standaard Depressievestoornis (depressie) (eerste herziening). Huisarts Wet. 2003; 46: 614-623.

23. NICE. Clinical Guideline 23. Depression: management of depression in primary and secondary care. December 2004. 23, 1-63. 2004. London, National Institute for Clinical Excellence.

24. Kennedy SH, Lam RW, Cohen NL, Ravindran AV, CANMAT Depression Work Group. Clinical guidelines for the treatment ofdepressive disorders. IV. Medications and other biological treatments. Can J Psychiatry. 2001; 46 Suppl 1: 38S-58S.

25. Anderson IM, Nutt DJ, Deakin JFW. Evidence-based guidelines for treating depressive disorders with antidepressants: Arevision of the 1993 British Association for Psychopharmacology guidelines. J Psychopharmacol. 2000; 14: 3-20.

26. Mulrow CD, Williams JW, Jr., Trivedi M, Chiquette E, Aguilar C, Cornell JE et al. Evidence report on: Treatment ofdepression - Newer pharmacotherapies. Psychopharmacol Bull. 1999; 34: 409-795.

27. Centraal Begeleidingsorgaan voor de Intercollegiale Toetsing. Consensusbijeenkomst depressie bij volwassenen: vrijdag 16september 1994 Utrecht. Utrecht: Centraal Begeleidingsorgaan voor de Intercollegiale Toetsing; 1994.

28. Depression Guideline Panel. Depression in Primary Care: Volume 1. Detection and Diagnosis. Clinical Practice Guideline,Number 5. AHCPR Publication No. 93-0550. 1993. Rockville, MD, U.S. Department of Health and Human Services, PublicHealth Service, Agency for Health Care Policy and Research.

29. Depression Guideline Panel. Depression in Primary Care: Volume 2. Treatment of Major Depression. Clinical PracticeGuideline, Number 5. AHCPR Publication No. 93-0551. 1993. Rockville, MD, U.S. Department of Health and HumanServices, Public Health Service, Agency for Health Care Policy and Research.

30. Jonghe Fde, Swinkels JA, Vergouwen AC, Brink Gvan de. Antidepressiva 2008. Een leidraad voor het rationeel omgaan metantidepressiva bij de behandeling van patiënten met een depressie. Amsterdam: Benecke NL; 2008.

31. Anderson IM, Tomenson BM. The efficacy of selective serotonin re-uptake inhibitors in depression: A meta-analysis ofstudies against tricyclic antidepressants. J Psychopharmacol. 1994; 8: 238-249.

32. Anderson IM. Meta-analytical studies on new antidepressants. Br Med Bull. 2001; 57: 161-178.33. Anderson IM. Drug treatment of depression: Reflections on the evidence. AdvPsychiatrTreat. 2003; 9: 11-20.34. Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, Johnson BT. Initial severity and antidepressant benefits: a

meta-analysis of data submitted to the Food and Drug Administration. PLoS Med. 2008; 5: e45.35. Turner EH, Matthews AM, Linardatos E, Tell RA, Rosenthal R. Selective publication of antidepressant trials and its

influence on apparent efficacy. N Engl J Med. 2008; 358: 252-260.36. Barbui C, Furukawa TA, Cipriani A. Effectiveness of paroxetine in the treatment of acute major depression in adults: a

systematic re-examination of published and unpublished data from randomized trials. CMAJ. 2008; 178: 296-305.37. Moncrieff J, Cohen D. Do antidepressants cure or create abnormal brain states? PLoS Med. 2006; 3: e240.38. Stichting Farmaceutische Kengetallen. Data en feiten 2007.2007. Den Haag, Stichting Farmaceutische Kengetallen. 39. Ruhe HG, Rijswijk Evan, Verkes RJ. Biologische interventies bij depressie. In: Huyser J, Schene AH, Sabbe B, Spinhoven Ph,

eds. Handboek depressieve stoornissen. Utrecht: De Tijdstroom; 2008: 201-218.40. Thase ME, Rush AJ. Treatment-Resistant Depression. In: Bloom FE, Kupfer DJ, eds. Psychopharmacology: The Fourth

Generation of Progress. New York: Raven Press Ltd.; 1995:1081-1097.41. Freemantle N, Anderson IM, Young P. Predictive value of pharmacological activity for the relative efficacy of

antidepressant drugs. Meta-regression analysis. Br J Psychiatry. 2000; 177: 292-302.42. Smith D, Dempster C, Glanville J, Freemantle N, Anderson I. Efficacy and tolerability of venlafaxine compared with

selective serotonin reuptake inhibitors and other antidepressants: a meta-analysis. Br J Psychiatry. 2002; 180: 396-404.43. Papakostas GI, Thase ME, Fava M, Nelson JC, Shelton RC. Are antidepressant drugs that combine serotonergic and

noradrenergic mechanisms of action more effective than the selective serotonin reuptake inhibitors in treating majordepressive disorder? A meta-analysis of studies of newer agents. Biol Psychiatry. 2007; 62: 1217-1227.


1

27

44. Arroll B, MacGillivray S, Ogston S, Reid I, Sullivan F, Williams B et al. Efficacy and tolerability of tricyclic antidepressantsand SSRIs compared with placebo for treatment of depression in primary care: a meta-analysis. Ann Fam Med. 2005; 3:449-456.

45. Anderson IM. Selective serotonin reuptake inhibitors versus tricyclic antidepressants: a meta-analysis of efficacy andtolerability. J Affect Disord. 2000; 58: 19-36.

46. Kroenke K, West SL, Swindle R, Gilsenan A, Eckert GJ, Dolor R et al. Similar effectiveness of paroxetine, fluoxetine, andsertraline in primary care: a randomized trial. JAMA. 2001; 286: 2947-2955.

47. Simon G. Choosing a first-line antidepressant: equal on average does not mean equal for everyone. JAMA. 2001; 286:3003-3004.

48. Thase ME. Evaluating antidepressant therapies: remission as the optimal outcome. J Clin Psychiatry. 2003; 64 Suppl 13: 18-25.

49. McIntyre RS, O'Donovan C. The human cost of not achieving full remission in depression. Can J Psychiatry. 2004; 49: 10S-16S.

50. Geddes JR, Carney SM, Davies C, Furukawa TA, Kupfer DJ, Frank E et al. Relapse prevention with antidepressant drugtreatment in depressive disorders: a systematic review. Lancet. 2003; 361: 653-661.

51. Trivedi MH, Rush AJ, Wisniewski SR, Nierenberg AA, Warden D, Ritz L et al. Evaluation of outcomes with citalopram fordepression using measurement-based care in STAR*D: implications for clinical practice. Am J Psychiatry. 2006; 163: 28-40.

52. Rush AJ, Fava M, Wisniewski SR, Lavori PW, Trivedi MH, Sackeim HA et al. Sequenced treatment alternatives to relievedepression (STAR*D): rationale and design. Control Clin Trials. 2004; 25: 119-142.

53. Yonkers KA, Samson J. Mood disorders measures. In: Rush AJ, Pincus HA, First MB, et al., eds. Handbook of psychiatricmeasures. 1st ed. Washington DC: American Psychiatric Association; 2000:515-548.

54. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 23: 56-61.55. Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979; 134: 382-

389.56. Rush AJ, Giles DE, Schlesser MA, Fulton CL, Weissenburger J, Burns CT. The Inventory for Depression Symptomatology

(IDS): preliminary findings. Psychiatry Res. 1986; 18: 65-87.57. Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, Klein DN et al. The 16-Item Quick Inventory of Depressive

Symptomatology (QIDS), clinician rating (QIDS-C), and self-report (QIDS-SR): a psychometric evaluation in patients withchronic major depression. Biol Psychiatry. 2003; 54: 573-583.

58. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch GenPsychiatry. 1961; 4:561 571: 561.

59. Bech P, Gram LF, Dein E, Jacobsen O, Vitger J, Bolwig TG. Quantitative rating of depressive states. Acta Psychiatr Scand.1975; 51: 161-170.

60. Jonghe F de. Leidraad voor het scoren van de Hamilton Depression Rating Scale: HDRS leidraad. Amsterdam: BeneckeConsultants; 1994.

61. Faries D, Herrera J, Rayamajhi J, DeBrota D, Demitrack M, Potter WZ. The responsiveness of the Hamilton DepressionRating Scale. J Psychiatr Res. 2000; 34: 3-10.

62. Gibbons RD, Clark DC, Kupfer DJ. Exactly what does the Hamilton Depression Rating Scale measure? J Psychiatr Res. 1993;27: 259-273.

63. Moeller H-J, Mueller H, Volz HP. How to assess the onset of antidepressant effect: Comparison of global ratings andfindings based on depression scales. Pharmacopsychiatry. 1996; 29: 57-62.

64. Santor DA, Coyne JC. Examining symptom expression as a function of symptom severity: Item performance on theHamilton Rating Scale for Depression. Psychol Assess. 2001; 13: 127-139.

65. Vittengl JR, Clark LA, Kraft D, Jarrett RB. Multiple measures, methods, and moments: a factor-analytic investigation ofchange in depressive symptoms during acute-phase cognitive therapy for depression. Psychol Med. 2005; 35: 693-704.

66. Frank E, Prien RF, Jarrett RB, Keller MB, Kupfer DJ, Lavori PW et al. Conceptualization and rationale for consensusdefinitions of terms in major depressive disorder. Remission, recovery, relapse, and recurrence. Arch Gen Psychiatry. 1991;48: 851-855.

67. Crismon ML, Trivedi M, Pigott TA, Rush AJ, Hirschfeld RMA, Kahn DA et al. The Texas medication algorithm project:Report of the Texas consensus conference panel on medication treatment of major depressive disorder. J Clin Psychiatry.1999; 60: 142-156.

68. Trivedi MH. Using treatment algorithms to bring patients to remission. J Clin Psychiatry. 2003; 64 Suppl 2: 8-13.69. Taylor MJ, Freemantle N, Geddes JR, Bhagwagar Z. Early onset of selective serotonin reuptake inhibitor antidepressant

action: systematic review and meta-analysis. Arch Gen Psychiatry. 2006; 63: 1217-1223.70. Quitkin FM, Petkova E, McGrath PJ, Taylor B, Beasley C, Stewart J et al. When should a trial of fluoxetine for major

depression be declared failed? Am J Psychiatry. 2003; 160: 734-740.71. Katz MM, Koslow SH, Maas JW, Frazer A, Bowden CL, Casper R et al. The timing, specificity and clinical prediction of

tricyclic drug effects in depression. Psychol Med. 1987; 17: 297-309.72. Quitkin FM, McGrath PJ, Stewart JW, Ocepek-Welikson K, Taylor BP, Nunes E et al. Chronological milestones to guide

drug change. When should clinicians switch antidepressants? Arch Gen Psychiatry. 1996; 53: 785-792.73. Kennedy N, McDonough M. Pharmacological management of treatment resistant depression: A clinical review.

IrJPsycholMed. 2003; 20: 18-23.74. Boer NAde. Verpleegkundige interventies bij depressie. 1 ed. Amsterdam: Programma Stemmingstoornissen AMC/De

Meren; 2003.75. Friedman MA, Detweiler-Bedell JB, Leventhal HE, Horne R, Keitner GI, Miller IW. Combined psychotherapy and

pharmacotherapy for the treatment of major depressive disorder. Clin Psychol Sci Prac. 2004; 11: 47-68.

Chap

ter

Chapter 1

1

28

76. Pampallona S, Bollini P, Tibaldi G, Kupelnick B, Munizza C. Combined pharmacotherapy and psychological treatment fordepression: a systematic review. Arch Gen Psychiatry. 2004; 61: 714-719.

77. Siegel GJ, Agranoff BW, Albers RW, Fisher SK, Uhler MD. Basic neurochemistry, molecular, cellular, and medical aspects. 6ed. Philadelphia: Lippincott, Williams & Wilkins; 1999.

78. Purves D, Augustine G, Fitzpatrick D, Katz L, LaMantia A. Neuroscience. 2 ed. Sunderland: Sinauer Associates, Inc.; 2001.79. Belmaker RH, Agam G. Major depressive disorder. N Engl J Med. 2008; 358: 55-68.80. Andreasen NC. Brave new brain: conquering mental illness in the era of the genome [Schitterend nieuw brein: psychiatrie in

het tijdperk van het genoom]. 1 ed. Amsterdam: Nieuwezijds B.V.; 2002.81. Marieb E. Human anatomy & Physiology. 5 ed. Redwood City: The Benjamin/ Cummings Publishing Company, Inc.; 1999.82. Gaspar P, Cases O, Maroteaux L. The developmental role of serotonin: news from mouse molecular genetics. Nat Rev

Neurosci. 2003; 4: 1002-1012.83. Martinowich K, Lu B. Interaction between BDNF and serotonin: role in mood disorders. Neuropsychopharmacology. 2008;

33: 73-83.84. Young SN, Smith SE, Pihl RO, Ervin FR. Tryptophan depletion causes a rapid lowering of mood in normal males.

Psychopharmacology (Berl). 1985; 87: 173-177.85. Moja EA, Lucini V, Benedetti F, Lucca A. Decrease in plasma phenylalanine and tyrosine after phenylalanine-tyrosine free

amino acid solutions in man. Life Sci. 1996; 58: 2389-2395.86. Shopsin B, Gershon S, Goldstein M, Friedman E, Wilk S. Use of synthesis inhibitors in defining a role for biogenic amines

during imipramine treatment in depressed patients. Psychopharmacol Commun. 1975; 1: 239-249.87. Delgado PL, Miller HL, Salomon RM, Licinio J, Heninger GR, Gelenberg AJ et al. Monoamines and the mechanism of

antidepressant action: effects of catecholamine depletion on mood of patients treated with antidepressants.Psychopharmacol Bull. 1993; 29: 389-396.

88. Bell C, Abrams J, Nutt D. Tryptophan depletion and its implications for psychiatry. Br J Psychiatry. 2001; 178: 399-405.89. Booij L, Van der Does AJ, Riedel WJ. Monoamine depletion in psychiatric and healthy populations: review. Mol Psychiatry.

2003; 8: 951-973.90. Van der Does AJ. The effects of tryptophan depletion on mood and psychiatric symptoms. J Affect Disord. 2001; 64: 107-

119.91. Moore P, Landolt HP, Seifritz E, Clark C, Bhatti T, Kelsoe J et al. Clinical and physiological consequences of rapid

tryptophan depletion. Neuropsychopharmacology. 2000; 23: 601-622.92. Booij L, Van der DW, Benkelfat C, Bremner JD, Cowen PJ, Fava M et al. Predictors of mood response to acute tryptophan

depletion. A reanalysis. Neuropsychopharmacology. 2002; 27: 852-861.93. Van der Does AJ. The mood-lowering effect of tryptophan depletion: possible explanation for discrepant findings. Arch

Gen Psychiatry. 2001; 58: 200-202.94. Damasio AR. Descartes' error - emotion, reason and the human brain [De vergissing van Descartes - gevoel, verstand en het

menselijk brein]. 2nd ed. Amsterdam: Wereldbibliotheek; 1994.95. Owens MJ. Selectivity of antidepressants: from the monoamine hypothesis of depression to the SSRI revolution and

beyond. J Clin Psychiatry. 2004; 65 Suppl 4: 5-10.96. Burke WJ. Selective versus multi-transmitter antidepressants: are two mechanisms better than one? J Clin Psychiatry.

2004; 65 Suppl 4: 37-45.97. Fava M, Kendler KS. Major depressive disorder. Neuron. 2000; 28: 335-341.98. Delgado PL. How antidepressants help depression: mechanisms of action and clinical response. J Clin Psychiatry. 2004; 65

Suppl 4: 25-30.99. Mayberg HS. Modulating dysfunctional limbic-cortical circuits in depression: towards development of brain-based

algorithms for diagnosis and optimised treatment. Br Med Bull. 2003; 65: 193-207.100. Phillips ML, Drevets WC, Rauch SL, Lane R. Neurobiology of emotion perception I: The neural basis of normal emotion

perception. Biol Psychiatry. 2003; 54: 504-514.101. Phillips ML, Drevets WC, Rauch SL, Lane R. Neurobiology of emotion perception II: Implications for major psychiatric

disorders. Biol Psychiatry. 2003; 54: 515-528.102. Mayberg HS, Liotti M, Brannan SK, McGinnis S, Mahurin RK, Jerabek PA et al. Reciprocal limbic-cortical function and

negative mood: converging PET findings in depression and normal sadness. Am J Psychiatry. 1999; 156: 675-682.103. Mayberg HS. Limbic-cortical dysregulation: a proposed model of depression. J Neuropsychiatry Clin Neurosci. 1997; 9: 471-

481.104. Malhi GS, Lagopoulos J. Making sense of neuroimaging in psychiatry. Acta Psychiatr Scand. 2008; 117: 100-117.105. Blier P, De Montigny C. Current advances and trends in the treatment of depression. Trends Pharmacol Sci. 1994; 15: 220-

226.106. Nemeroff CB, Owens MJ. Neuropharmacology of paroxetine. Psychopharmacol Bull. 2003; 37 Suppl 1: 8-18.107. Bel N, Artigas F. Chronic treatment with fluvoxamine increases extracellular serotonin in frontal cortex but not in raphe

nuclei. Synapse. 1993; 15: 243-245.108. Davidson C, Stamford JA. Effect of chronic paroxetine treatment on 5-HT1B and 5-HT1D autoreceptors in rat dorsal raphe

nucleus. Neurochem Int. 2000; 36: 91-96.109. Vaswani M, Linda FK, Ramesh S. Role of selective serotonin reuptake inhibitors in psychiatric disorders: a comprehensive

review. Prog Neuropsychopharmacol Biol Psychiatry. 2003; 27: 85-102.110. Benmansour S, Cecchi M, Morilak DA, Gerhardt GA, Javors MA, Gould GG et al. Effects of chronic antidepressant

treatments on serotonin transporter function, density, and mRNA level. J Neurosci. 1999; 19: 10494-10501.


1

29

111. Benmansour S, Owens WA, Cecchi M, Morilak DA, Frazer A. Serotonin clearance in vivo is altered to a greater extent byantidepressant-induced downregulation of the serotonin transporter than by acute blockade of this transporter.JNeurosci. 2002; 22: 6766-6772.

112. Pineyro G, Blier P, Dennis T, De Montigny C. Desensitization of the neuronal 5-HT carrier following its long-term blockade.J Neurosci. 1994; 14: 3036-3047.

113. Maj J, Bijak M, Dziedzicka-Wasylewska M, Rogoz R, Rogoz Z, Skuza G et al. The effects of paroxetine given repeatedly onthe 5-HT receptor subpopulations in the rat brain. Psychopharmacology (Berl). 1996; 127: 73-82.

114. Whale R, Clifford EM, Bhagwagar Z, Cowen PJ. Decreased sensitivity of 5-HT1D receptors in melancholic depression. Br JPsychiatry. 2001; 178: 454-457.

115. Kennett GA, Lightowler S, de B, V, Stevens NC, Wood MD, Tulloch IF et al. Effect of chronic administration of selective 5-hydroxytryptamine and noradrenaline uptake inhibitors on a putative index of 5-HT2C/2B receptor function.Neuropharmacology. 1994; 33: 1581-1588.

116. Meyer JH, Kapur S, Eisfeld B, Brown GM, Houle S, DaSilva J et al. The effect of paroxetine on 5-HT(2A) receptors indepression: an [18F]setoperone PET imaging study. Am J Psychiatry. 2001; 158: 78-85.

117. Djavadian RL. Serotonin and neurogenesis in the hippocampal dentate gyrus of adult mammals. Acta Neurobiol Exp(Wars). 2004; 64: 189-200.

118. Blendy JA. The role of CREB in depression and antidepressant treatment. Biol Psychiatry. 2006; 59: 1144-1150.119. Angelucci F, Brene S, Mathe AA. BDNF in schizophrenia, depression and corresponding animal models. Mol Psychiatry.

2005; 10: 345-352.120. Mayberg HS, Brannan SK, Tekell JL, Silva JA, Mahurin RK, McGinnis S et al. Regional metabolic effects of fluoxetine in

major depression: serial changes and relationship to clinical response. Biol Psychiatry. 2000; 48: 830-843.121. Kennedy SH, Evans KR, Kruger S, Mayberg HS, Meyer JH, McCann S et al. Changes in regional brain glucose metabolism

measured with positron emission tomography after paroxetine treatment of major depression. Am J Psychiatry. 2001;158: 899-905.

122. Mayberg HS, Silva JA, Brannan SK, Tekell JL, Mahurin RK, McGinnis S et al. The functional neuroanatomy of the placeboeffect. Am J Psychiatry. 2002; 159: 728-737.

123. Fu CH, Williams SC, Cleare AJ, Brammer MJ, Walsh ND, Kim J et al. Attenuation of the neural response to sad faces inmajor depression by antidepressant treatment: a prospective, event-related functional magnetic resonance imagingstudy. Arch Gen Psychiatry. 2004; 61: 877-889.

124. Fu CH, Williams SC, Brammer MJ, Suckling J, Kim J, Cleare AJ et al. Neural responses to happy facial expressions in majordepression following antidepressant treatment. Am J Psychiatry. 2007; 164: 599-607.

125. Sheline YI, Barch DM, Donnelly JM, Ollinger JM, Snyder AZ, Mintun MA. Increased amygdala response to maskedemotional faces in depressed subjects resolves with antidepressant treatment: an fMRI study. Biol Psychiatry. 2001; 50:651-658.

126. Davidson RJ, Irwin W, Anderle MJ, Kalin NH. The neural substrates of affective processing in depressed patients treatedwith venlafaxine. Am J Psychiatry. 2003; 160: 64-75.

127. Schaefer HS, Putnam KM, Benca RM, Davidson RJ. Event-related functional magnetic resonance imaging measures ofneural activity to positive social stimuli in pre- and post-treatment depression. Biol Psychiatry. 2006; 60: 974-986.

128. Robertson B, Wang L, Diaz MT, Aiello M, Gersing K, Beyer J et al. Effect of bupropion extended release on negativeemotion processing in major depressive disorder: a pilot functional magnetic resonance imaging study. J Clin Psychiatry.2007; 68: 261-267.

129. Hu XZ, Lipsky RH, Zhu G, Akhtar LA, Taubman J, Greenberg BD et al. Serotonin transporter promoter gain-of-functiongenotypes are linked to obsessive-compulsive disorder. Am J Hum Genet. 2006; 78: 815-826.

130. Lesch KP, Gutknecht L. Pharmacogenetics of the serotonin transporter. Prog Neuropsychopharmacol Biol Psychiatry.2005; 29: 1062-1073.

131. Serretti A, Kato M, De Ronchi D, Kinoshita T. Meta-analysis of serotonin transporter gene promoter polymorphism (5-HTTLPR) association with selective serotonin reuptake inhibitor efficacy in depressed patients. Mol Psychiatry. 2007; 12:247-257.

132. Hu XZ, Rush AJ, Charney D, Wilson AF, Sorant AJ, Papanicolaou GJ et al. Association between a functional serotonintransporter promoter polymorphism and citalopram treatment in adult outpatients with major depression. Arch GenPsychiatry. 2007; 64: 783-792.

133. Kraft JB, Peters EJ, Slager SL, Jenkins GD, Reinalda MS, McGrath PJ et al. Analysis of association between the serotonintransporter and antidepressant response in a large clinical sample. Biol Psychiatry. 2007; 61: 734-742.

134. Murphy DL, Lerner A, Rudnick G, Lesch KP. Serotonin transporter: gene, genetic disorders, and pharmacogenetics. MolInterv. 2004; 4: 109-123.

135. Caspi A, Sugden K, Moffitt TE, Taylor A, Craig IW, Harrington H et al. Influence of life stress on depression: moderation bya polymorphism in the 5-HTT gene. Science. 2003; 301: 386-389.

136. Hariri AR, Mattay VS, Tessitore A, Kolachana B, Fera F, Goldman D et al. Serotonin transporter genetic variation and theresponse of the human amygdala. Science. 2002; 297: 400-403.

137. Hariri AR, Drabant EM, Munoz KE, Kolachana BS, Mattay VS, Egan MF et al. A susceptibility gene for affective disordersand the response of the human amygdala. Arch Gen Psychiatry. 2005; 62: 146-152.

138. Pezawas L, Meyer-Lindenberg A, Drabant EM, Verchinski BA, Munoz KE, Kolachana BS et al. 5-HTTLPR polymorphismimpacts human cingulate-amygdala interactions: a genetic susceptibility mechanism for depression. Nat Neurosci. 2005;8: 828-834.

139. Jonghe F de, Kool S, Aalst G van, Dekker J, Peen J. Combining psychotherapy and antidepressants in the treatment ofdepression. J Affect Disord. 2001; 64: 217-229.

Chap

ter

Chapter 1

1

30

140. Dekker J, Molenaar PJ, Kool S, Aalst G van, Peen J, Jonghe F de. Dose-effect relations in time-limited combined psycho-pharmacological treatment for depression. Psychol Med. 2005; 35: 47-58.

141. Maier W, Philipp M. Improving the assessment of severity of depressive states: a reduction of the Hamilton DepressionScale. Pharmacopsychiatry. 1985; 18: 114-115.

142. Prien RF, Carpenter LL, Kupfer DJ. The definition and operational criteria for treatment outcome of major depressivedisorder. A review of the current research literature. Arch Gen Psychiatry. 1991; 48: 796-800.

143. Ballenger JC. Clinical guidelines for establishing remission in patients with depression and anxiety. J Clin Psychiatry. 1999;60 Suppl 22: 29-34.

144. The Cochrane Collaboration. RevMan 4.2 User Guide, version 4.2 for Windows. 4.2 ed. Oxford, UK: The CochraneCollaboration; 2002.

145. Staley JK, Krishnan-Sarin S, Zoghbi S, Tamagnan G, Fujita M, Seibyl JP et al. Sex differences in [123I]beta-CIT SPECTmeasures of dopamine and serotonin transporter availability in healthy smokers and nonsmokers. Synapse. 2001; 41: 275-284.

146. Pirker W, Asenbaum S, Hauk M, Kandlhofer S, Tauscher J, Willeit M et al. Imaging serotonin and dopamine transporterswith 123I-beta-CIT SPECT: binding kinetics and effects of normal aging. J Nucl Med. 2000; 41: 36-44.

147. van Dyck CH, Malison RT, Seibyl JP, Laruelle M, Klumpp H, Zoghbi SS et al. Age-related decline in central serotonintransporter availability with [123I]beta-CIT SPECT. Neurobiol Aging. 2000; 21: 497-501.

148. Buchert R, Schulze O, Wilke F, Berding G, Thomasius R, Petersen K et al. Is correction for age necessary in SPECT or PET ofthe central serotonin transporter in young, healthy adults? J Nucl Med. 2006; 47: 38-42.

CHA

PTER

DESIGN AND METHODS OF THE DELPHI-STUDY

ChapterDesign of the DELPHI study

2

33

The DELPHI-study

DELPHI is an acronym for Dose-Escalation Legitimate? Pharmacology and Imaging studies indepression. The DELPHI-study is the backbone of the studies described in Part IV of this thesis.Because the DELPHI study in fact comprised three research projects, with slightly differentpatient populations and different numbers of participants, this chapter will describe the design ofthe DELPHI-study and indicate how the nested sub-studies relate to the major DELPHI-study.Furthermore the patient disposition over these projects will be given.

Objectives, rationale and research questionsThe major aims of the DELPHI-study were toinvestigate the clinical efficacy of and theneurobiological mechanism behind dose-escalation. In order to determine the neurobiologicaleffects of dose-escalation, we intended to study different biological measures during treatment:paroxetine serum concentrations (PSC), serotonin transporter (SERT) occupancy, functionalactivations in the cortico-limbic-network, awakening cortisol changes and changes in ω-3/ω-6 polyunsaturated fatty acids (PUFAs). Furthermore, we planned to study the effects of geneticpolymorphisms on outcomes (i.e. the serotonin transporter gene promoter region (5-HTTLPR)).Results related to some of these neurobiological measures are described in this thesis.

As a study drug, we choseparoxetine for three reasons. First, paroxetine is the selectiveserotonin reuptake inhibitor (SSRI) which is prescribed most frequently in the Netherlands.1

Second, paroxetine is a potent inhibitor of the cytochrome P450 2D6 sub-enzyme, which also isthe enzyme that is responsible for its metabolism. Therefore paroxetine inhibits its ownmetabolism, which causes an exponential rise in blood serum concentrations after dose-escalation.2 Third, when we initiated this study, Gilmor et al. reported noradrenergic reuptakeinhibition by paroxetine, suggesting that paroxetine, like venlafaxine and duloxetine, in fact was a‘dual action’ antidepressant, especially at higher doses.3

Our main research-questions were:1 Is a 6-week true dose-escalation of paroxetine (up to 30-50 mg/day) in patients non-responsive

to a 6-week trial with standard dose (20 mg/day) more effective than placebo dose-escalation?2 Does a 6 week true dose-escalation of paroxetine increase SERT occupancy more than a

placebo dose-escalation?3 What is the relation between SERT occupancy and clinical response to paroxetine (either at a

standard dose or after dose-escalation)? Is this relation modified by other neurobiologicalparameters (i.e. genetic polymorphisms)?

4 Which changes in the functional activation of the cortico-limbic brain regions correlate withclinical response to paroxetine? Is there a common and/or differential effect of paroxetineexposure and/or clinical response?

5 Does a 6 week true dose-escalation of paroxetine generate additional changes (e.g. innoradrenergic reuptake inhibition, cortisol awakening responses, or ω-3/ω-6 PUFAs) comparedto a placebo dose-escalation, and are these changes related to clinical response?

Only questions 1-4 will be discussed in this thesis, and the methods used for these questions willbriefly be discussed.However, for specific technical we refer to the method-sections in therelevant chapters.

Design and interventionsThe DELPHI-study was set up as a randomized clinical trial (ISRCTN register nr. ISRCTN44111488;http://www.trialregister.nl/trialreg/admin/rctview.asp?TC=193). The study was approved by theAcademic Medical Center (AMC) medical ethical committee (03/120, 03/287 and addendum to 03/120), and all participants provided written informed consent.

Chap

ter

Chapter 2

2

34

Between October 2003 and February 2007 patients were recruited from primary care, ourAMC Program for Mood Disorders, and public psychiatric settings. Patients were treated by theirreferring physician or were referred to our outpatient department. In DELPHI two nested sub-studies were embedded, the DELPHI-SPECT and the DELPHI-fMRI (Figure 2.1). These sub-studieshad slightly different in- and exclusion criteria (see below).

All eligible patients were treated open-label with paroxetine 20 mg/day for 6 weeks. Whensevere adverse effects occurred, dosages were reduced to 10 mg/day and again increased to 20mg/day after one week. After 6 weeks of treatment, patients who responded (defined as ≥50%reduction of the pretreatment Hamilton depression rating scale (HDRS)) continued paroxetine 20mg/day. Treatment non-responders were randomized to a true dose-escalation or a placebo dose-escalation, added to paroxetine 20 mg/day in a double blind design. Dose-escalation was providedin blue capsules containing 10 mg paroxetine or placebo. Randomization was stratified fortreatment setting (SPECT/fMRI-group, outpatient department AMC, primary care, and publicpsychiatry), gender and age. Within strata, we applied a minimization method to achieve abalanced distribution. We concealed allocation by using an independently operated computerprogram.

Dose-escalation consisted of incremental steps of one capsule every 5 days towards amaximum of 50 mg/day (20 mg + 3 capsules). Patients were allowed to increase at a slower pace(e.g. by 7 days) or stop further escalation (e.g. 20 mg + 2 capsules) according to adverse effects.4

No dosage adjustments were allowed during the last 3 weeks of the study. We checkedadherence by pill-counts and anamnesis.5

Figure 2.1. Design of the DELPHI-study.

DELPHI-SPECT and DELPHI-fMRI represent sub-studies nested within the total study. For these 2 sub-studies drug-freepatients were recruited and treated in the Academic Medical Center. For color figure see page 280.

��

��

��

�� !" ��

��

� �#�$��% #& #� ' &�� '��(")

��

��

�� *��(�+�,��(� ��"��"��(�

�� -.�� -.�� !" -.��


2

35

Patients Figure 2.2 summarizes patient disposition in the DELPHI-study and the DELPHI-SPECT/fMRIneuroimaging sub-study.

General in- and exclusion criteriaEligible patients met the following inclusion criteria: Age between 18 and 70 years,majordepressive disorder (MDD) determined by the structured clinical interview for DSM-IV (SCID),6

and a HDRS (17 items)7 score above 18. All participants were either drug-free and/or hadundergone no more than one antidepressant treatment (other than paroxetine) at an effectivedose for ≥ 6 weeks for the present MDD-episode. By the latter criterion, we avoided treatmentresistance as potential bias for inefficacy of dose-escalation. Exclusion criteria, apart frompregnancy (or wish to become pregnant), were bipolar disorder, psychotic features, neurologicalcognitive impairments (i.e. dementia), primary anxiety and/or substance abuse disorders andacute, severe suicidal ideation. Contrary, we allowed secondary co-morbid anxiety and/orsubstance abuse to increase applicability of our findings. In total, 278 patients were referred, ofwhom 107 patients started treatment with paroxetine 20 mg/day. Twenty patients withdrew fromthe study before week 6, and 27 patients were responder by then. Sixty non-responders wererandomized.

Figure 2.2. Patient enrolment and disposition over sub-studies.

The patients in DELPHI-SPECT and DELPHI-fMRI sub-studies are combined (left). For disposition of patients randomized seechapter 9.

��

�� !�� "�#� �$��$%�&��'�� (� %�#�( ��

'�� ) �#� �#��

*��#�� #�"�&�� $�# ��$� � ��

�!�� %�# #��&��& ��+�� *$��,%�&� ��

�� #�"�&�� -,�./ ��

*$��$%�&� -,�./�"01'�� *$��$%�&� 2��-,�./��

3#�$�� $��$%�&�

��#&� �""�� & ��

3#�$�� $��$%�&�

��""��) ��

��#&� �""�� & ��

� %�# #��&��& ��

(� %�#�(� �" �# �� #�4� ��

3#�$�� #��4�5� �� $%�&�

��""��) ��

��#&� �""�� & ��

3#�$�� #��4�5� �� $%�&�

��""��) ��

��#&� �""�� & ��

#�"�&�� " �# #��4�5 � ��

67 ��.�4$�� #�� #��4�5��

� ��)5��

� �!�� "�# ��)&�& ��

��

��

��

��

��

67 ��.�4$�� #�� #��4�5��

� ��)5��

� ��)5�� -,�./�"01' ��

�!�� "�# -,�./�"01'� ��)&�& ��

67 ��.�4$�� #�-,�./�"01'��

� 67 ��1�&$��# -,�./�"01'��

� 67 ��21 -,�./�"01'� #��4�5��

67 ��.�4$�� #�2��-,�./�"01'��

� 67 ��1�&$��# ��

� 67 ��21 #��4�5��

��

��

Chap

ter

Chapter 2

2

36

DELPHI-SPECT and fMRIPatients who were drug-free (for >4 weeks and ≥5 half-lives of a previous antidepressant) wereasked to additionally participate in the neuroimaging sub-studies. Patients were initially asked toparticipate in DELPHI-SPECT only, but from august 2005 onwards, when DELPHI-fMRI wasapproved, we asked them for both DELPHI-SPECT and DELPHI fMRI. Participation in only one ofthe neuroimaging studies was also possible. We limited age to 25-55 years to reduce variability inSERT-measurements by age.8 For participation in the fMRIs, patients had to be free of metalobjects in their body. None of the included patients reported past or present use of 3,4-methylenedioxymethamphetamine. We treated DELPHI-SPECT and DELPHI-fMRI patients attheoutpatient department of the Program for Mood Disorders. We supplied medication inpillboxes.

We included 51 patients in the DELPHI-SPECT study, of whom 33 non-responders wererandomized. For DELPHI-fMRI we included 22 patients, of whom 16 were randomized. Twentypatients participated both in the DELPHI-SPECT and DELPHI-fMRI. Unfortunately not all(repeated) scans were analyzable adequately, the reason why in chapters 8-10 different numbersof patients are described.

Healthy controlsWe recruited 53 healthy controls as reference for the study-entry scans. We individually matchedeach patient in DELPHI-SPECT and DELPHI-fMRI by gender and age (±2.5 years). Healthy controlswere in good physical health, and had never used psychotropic medication. Exclusion criteriawere current or lifetime psychiatric disorder(s) according to the SCID (including abuse oraddiction disorders), a Beck Depression Inventory (BDI) score >9, alcohol use >4 units per day (lastmonth) or a 1st-degree relative with psychiatric disorder(s). We allowed healthy controls to haveincidentally used illicit drugs, unless criteria for a DSM-IV disorder was met, but we prohibited illicitdrug use the month prior to scanning. Twenty healthy controls also participated both in theDELPHI-SPECT and DELPHI-fMRI.

Measurements Time points and questionnairesWe administered the HDRS17,7 Inventory for depressive symptoms (IDS-SR30),9 the occurrence ofadverse effects and health-related quality of life (MOS-SF36)10 at study-entry, randomization (T0),and 6 weeks after randomization (T1). Adverse effects and depressive symptoms were alsomonitored in the weeks 1, 2 and 4 after the initial start of treatment, and after randomization,using the Maier and Bech subscales11 and IDS-SR30

9 (Figure 2.1). Three trained investigatorsadministered clinician-rated questionnaires. Agreement between raters was good (intraclasscorrelation coefficient = 0.98). Raters and patients were blinded for treatment.

Definition of primary and secondary outcomesPrimary clinical outcomes were HDRS17-scores and the proportion of patients achieving response(≥50% decrease in HDRS17) or remission (HDRS17 ≤7). Secondary outcomes were total and specific(adverse effects / inefficacy) dropout rates, the Maier and Bech subscales and IDS-SR30-scores,the occurrence of adverse effects and health-related quality of life.

Neurobiological measurementsParoxetine serum concentrations: See chapters 8 and 9 for details.SERT-gene promoter polymorphism: See chapters 8 and 11 for details.Saliva and additional blood specimens

At study-entry, at T0, and T1, we collected two saliva specimens at awakening and 30 minutesthereafter, to determine salivary cortisol, dehydroepiandrosterone-sulphate, and α-amylase.12-15

Furthermore, we collected blood-specimens to quantify [3H]Noradrenaline and [3H]Serotonineuptake in ex-vivo models.3;16;17 Finally we collected blood-specimens to determine levels of plateletand plasma ω-3/ω-6 PUFAs.18 These neurobiological measurements remain to be analysed indifferent papers not in this thesis.


2

37

Neuroimaging Measurement of SERT occupancy We performed SPECT imaging at study-entry, T0 and T1 (Figure 2.1), between 2 to 10 pm accordingto previously described procedures.19 We made all scans 230 ±18 (SD) minutes after intravenousinjection of approximately 100 MBq [123I]β-CIT, when the radioligand is at equilibrium for SERTbinding in brain areas expressing high densities of SERTs.20 We performed SPECT imaging using a12-detector single slice brain-dedicated scanner (Neurofocus 810, Strichmann Medical Equipment;Cleveland, OH). After attenuation correction and reconstruction in 3D mode (http://www.neurophysics.com), we defined regions of interest (RoIs) for midbrain, diencephalon(regions rich of SERT) and cerebellum (as a reference) by using validated templates (Figure 2.3).19

For further details of SPECT-procedures see chapters 7-10.

Magnetic resonance imaging of functional activation of thecortico-limbic-network We acquired fMRI scans at study-entry, T0 and T1 (Figure 2.1), between 2 to 10 pm. FMRI-sessionslasted 50-60 minutes, each including a cognitive task (Tower of London),21 a structural scan and afacial expression task, reported in this thesis.22;23 We used a 3Tesla Intera MRI scanner (Philips,Eindhoven, NL), with a 6 channel head-coil for radiofrequency reception. Two magnet compatibleresponse boxes were used to record subject’s performance and reaction times. For further detailsof fMRI-settings and the parameters of the faces paradigm: see chapter 10.

Power and interim analysisFor the randomization-phase of the total DELPHI-study, we performed a-priori power-calculationsfor two co-primary endpoints. We planned an interim analysis after SPECT data had been collectedon at least 30 patients. Stopping criteriawere predetermined using the O'Brien and Fleming24

approach, and were p<0.0026 in case of superiority and p>0.50 for futility. See chapter 9 forfurther details.

Figure 2.3. Regions of Interest (RoI) for Midbrain, Cerebellum and Diencephalon.

Example of SPECT images after 3D reconstruction. Templates with fixed RoIs are shown. For color figure see page 281.A. Midbrain (circle) and cerebellum. B. Striatum (for demarcation midbrain-diencephalon) and diencephalon (circle). RoIswere positioned by hand and situated based on anatomy and maximum concentration of activity/ml in the RoI.

Chap

ter

Chapter 2

2

38

References

1. Stichting Farmaceutische Kengetallen. Data en feiten 2007.2007. Den Haag, Stichting Farmaceutische Kengetallen. 2. Anonymous. Paxil® Prescribing information and medication guide PXL:47PI (available at http://us.gsk.com/products/

assets/us_paxil.pdf). 1-44. 2008. Research Triangle Park, North Carolina, GlaxoSmithKline. 3. Gilmor ML, Owens MJ, Nemeroff CB. Inhibition of norepinephrine uptake in patients with major depression treated with

paroxetine. Am J Psychiatry. 2002; 159: 1702-1710.4. Baker CB, Woods SW. Is there a SSRI dose response in treating major depression? The case for re-analysis of current data

and for enhancing future study design. Depress Anxiety. 2003; 17: 10-18.5. Saunders K, Simon G, Bush T, Grothaus L. Assessing the feasibility of using computerized pharmacy refill data to monitor

antidepressant treatment on a population basis: a comparison of automated and self-report data. J Clin Epidemiol. 1998;51: 883-890.

6. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV Axis I Disorders Patient Edition(SCID-I/P version 2.0). Translated in Dutch by. Groenestijn MAC, Akkerhuis GW, Kupka RW, Schneider N, and NolenWA.1999. Lisse, the Netherlands, Swets & Zeitlinger B.V.

7. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 23: 56-61.8. van Dyck CH, Malison RT, Seibyl JP, Laruelle M, Klumpp H, Zoghbi SS et al. Age-related decline in central serotonin

transporter availability with [123I]β-CIT SPECT. Neurobiol Aging. 2000; 21: 497-501.9. Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH. The Inventory of Depressive Symptomatology (IDS):

psychometric properties. Psychol Med. 1996; 26: 477-486.10. Aaronson NK, Muller M, Cohen PD, Essink-Bot ML, Fekkes M, Sanderman R et al. Translation, validation, and norming of

the Dutch language version of the SF-36 Health Survey in community and chronic disease populations. J Clin Epidemiol.1998; 51: 1055-1068.

11. Ruhe HG, Dekker JJ, Peen J, Holman R, Jonghe F de. Clinical use of the Hamilton Depression Rating Scale: is increasedefficiency possible? A post hoc comparison of Hamilton Depression Rating Scale, Maier and Bech subscales, ClinicalGlobal Impression, and Symptom Checklist-90 scores.Compr Psychiatry. 2005; 46: 417-427.

12. Nater UM, Rohleder N, Schlotz W, Ehlert U, Kirschbaum C. Determinants of the diurnal course of salivary alpha-amylase.Psychoneuroendocrinology. 2007; 32: 392-401.

13. Pruessner JC, Wolf OT, Hellhammer DH, Buske-Kirschbaum A, von Auer K, Jobst S et al. Free cortisol levels afterawakening: a reliable biological marker for the assessment of adrenocortical activity. Life Sci. 1997; 61: 2539-2549.

14. Rohleder N, Nater UM, Wolf JM, Ehlert U, Kirschbaum C. Psychosocial stress-induced activation of salivary alpha-amylase:an indicator of sympathetic activity? Ann N Y Acad Sci. 2004; 1032: 258-263.

15. Schmidt-Reinwald A, Pruessner JC, Hellhammer DH, Federenko I, Rohleder N, Schurmeyer TH et al. The cortisol responseto awakening in relation to different challenge tests and a 12-hour cortisol rhythm. Life Sci. 1999; 64: 1653-1660.

16. Owens MJ, Morgan WN, Plott SJ, Nemeroff CB. Neurotransmitter receptor and transporter binding profile ofantidepressants and their metabolites. J Pharmacol Exp Ther. 1997; 283: 1305-1322.

17. Owens MJ, Krulewicz S, Simon JS, Sheehan DV, Thase ME, Carpenter DJ et al. Estimates of Serotonin and NorepinephrineTransporter Inhibition in Depressed Patients Treated with Paroxetine or Venlafaxine. Neuropsychopharmacology. 2008.

18. Dacremont G, Cocquyt G, Vincent G. Measurement of very long-chain fatty acids, phytanic and pristanic acid in plasmaand cultured fibroblasts by gas chromatography. J Inherit Metab Dis. 1995; 18 Suppl 1: 76-83.

19. Win MM de, Habraken JB, Reneman L, Brink W van den, Heeten GJ den, Booij J. Validation of [123I]β-CIT SPECT to assessserotonin transporters in vivo in humans: a double-blind, placebo-controlled, crossover study with the selective serotoninreuptake inhibitor citalopram. Neuropsychopharmacology. 2005; 30: 996-1005.

20. Pirker W, Asenbaum S, Hauk M, Kandlhofer S, Tauscher J, Willeit M et al. Imaging serotonin and dopamine transporterswith [123I]β-CIT SPECT: binding kinetics and effects of normal aging. J Nucl Med. 2000; 41: 36-44.

21. Heuvel OAvan den, Veltman DJ, Groenewegen HJ, Cath DC, Balkom AJ van, Hartskamp Jvan et al. Frontal-striataldysfunction during planning in obsessive-compulsive disorder. Arch Gen Psychiatry. 2005; 62: 301-309.

22. Hariri AR, Bookheimer SY, Mazziotta JC. Modulating emotional responses: effects of a neocortical network on the limbicsystem. Neuroreport. 2000; 11: 43-48.

23. Wolfensberger SPA, Veltman DJ, Hoogendijk WJG, Boomsma DI, Geus EJC de. Amygdala responses to emotional faces intwins discordant or concordant for the risk for anxiety and depression. Neuroimage. 2008; E-pub: doi: 10.1016/j.neuroimage.2008.01.053.

24. O'Brien PC, Fleming TR. A multiple testing procedure for clinical trials. Biometrics. 1979; 35: 549-556.


2

39

STUDIES TO GUIDE CLINICAL TREATMENT OF MAJOR DEPRESSIVE DISORDER

PART

Studies to guide clinical treatment of Major Depressive Disorder

CHA

PTER

CLINICAL USE OF THE HAMILTON DEPRESSION

RATING SCALE: IS INCREASED EFFICIENCY

POSSIBLE? A POST HOC COMPARISON OF HDRS, MAIER AND

BECH SUBSCALES, CGI AND SCL-90 SCORES.

Comprehensive Psychiatry 2005; 46: 417-427

Henricus G. Ruhé1,2

Jack J. Dekker1,3

Jaap Peen1

Rebecca Holman4

Frans de Jonghe1,2

1 Mentrum, Amsterdam, The Netherlands2 Department of Psychiatry, Academic Medical Center, University of Amsterdam, Amsterdam, TheNetherlands3 Department of Clinical Psychology, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands4 Department of Clinical Epidemiology and Biostatistics, Academic Medical Center, University ofAmsterdam, Amsterdam, The Netherlands

Chap

ter

Chapter 3

3

44

Abstract

BackgroundThe 17-item Hamilton Depression Scale (HDRS) is used as a semi gold standard in research. Intreatment guidelines, the HDRS-measurements serve to determine response, remission and guideclinical decision-making for non-responders. However, its use in clinical practice is limited, possiblybecause the HDRS is time-consuming. Additionally, the multidimensional HDRS is criticized for notmeasuring a unidimensional aspect as depression severity. The Maier and the Bech, two 6-itemseverity subscales extracted from the HDRS, are relatively unknown.

AimTo investigate whether the measurements obtained with Maier and Bech subscales arecomparable with the original HDRS-measurements.

MethodsData from two randomized controlled trials in 482 male and female patients, diagnosed with amajor depression (with or without dysthymia) according to DSM-III-R, of whom 219 participated inthe trials, were reanalyzed. A standardized stepwise psychopharmacological treatment wascompared with a combination of pharmacotherapy with Short Psychodynamic SupportivePsychotherapy in a psychiatric outpatient department. Outcome measures were internalconsistency and concurrent validity of HDRS, Maier, Bech, Clinical Global Impression (CGI) scalesand Symptom CheckList depression subscale. Effect sizes of HDRS, Maier and Bech were used tocompare measured treatment effects for the randomized subjects participating in the trials. ItemResponse Theory was used to obtain conversion tables for the HDRS, Maier, Bech and SCL-90depression subscale.

ResultsWe found moderate internal consistence (Cronbach α ≈ 0.6-0.7) and high correlations of the Maierand Bech subscales with overall HDRS-scores. Overall there were no clinically relevant differencesin effect sizes between Maier, Bech and HDRS, although some differences were statisticallysignificant. Receiver operating characteristic curves showed no difference between Maier andBech to define remission, but showed the CGI ratings to be unreliable. A cut-off ≤4 correspondedwith a HDRS ≤ 7 criterion in both subscales.

ConclusionsIn clinical practice, both Maier and Bech scales can be used as equivalents of the HDRS, but will bemore efficient.

ChapterClinical use of HDRS subscales

3

45

Introduction

Major depressive disorder is a severe, disabling illness, expected to be the world’s second health-problem in 2020.1 Depression is associated with high costs, regarding direct treatment andindirect costs of loss of productivity and quality of life.2 Several clinical guidelines were developedto guide the treatment of this disorder, both psychotherapy and pharmacotherapy (or incombination) appear effective.3-10

The use of self-report or clinician-rated symptom-scales is recommended to assess severityand response to treatment.8;11;12 Some experts claim clinician-rated symptom scales to have alarger validity and reliability than self-reporting scales, especially in patients with cognitiveimpairment, and more severe or psychotic depressions.11;13;14 Specific symptom-scales are morereliable than global rating scales.11;13;15 Especially, rating scales can be used to objectivelydetermine specific cut-off points for response and remission.12;16;17

In most clinical trials the Hamilton Depression Rating Scale (HDRS)18;19 – a clinician ratedsymptom-scale – is used as a standard to determine severity and response.5;8;11;15;20-23 Many versionsof the HDRS exist, with the number of items usually varying between 17 and 24,11;18;19;22 however upto 36 items have been described.23 Longer versions were especially developed to cover reverseneurovegetative (atypical) symptoms.23 The Clinical Global Impression (CGI)24 – a clinician-ratedglobal scale – is also frequently used.5;8;15;25 In clinical practice, although recommended, rating-scales are not used routinely. Explanations for this discrepancy could be ignorance of existingscales, a strong belief in one’s clinical judgment, an unsystematic approach of depression, but alsothe amount of time needed for rating-scales (e.g. 15-20 minutes for the HDRS11) and the necessityof training.20;26

The HDRS is criticized as being sensitive to somatic symptoms (e.g. somatic illness or side-effects of drugs),11;15;27;28 for not rating all 9 DSM-IV domains, its unequal weightings of differentsymptoms and for the multidimensionality of the HDRS total score.13;21;29-31 Multidimensionality isimportant to cover the maximum range of clinical features of major depressive disorder, but doesnot necessarily measures depression severity. Multidimensional scales can be misleading whenmeasurement of severity and treatment response is concerned,13;21;28 especially when themeasured depressive symptoms do not change proportionally with depression severity.Finally,some reports emphasize that the HDRS systematically favors (sedative) Tricyclic Antidepressants(TCAs) above Selective Serotonin Reuptake Inhibitors (SSRIs).27;32-35 Sleep and somatic items mayappear to be 'improved' by side-effects of TCAs, but worsened by side-effects (e.g. insomnia,gastrointestinal complaints, agitation) of SSRIs.

In order to overcome the problems of the multidimensional HDRS mentioned above, a moreunidimensional subscale from the HDRS covering core-symptoms of severity is desired. Also, froma clinical point of view, fewer items will be less time consuming for application by busy clinicians.However, for the purpose of reference, subscale scores must remain anchored to the originalHDRS. To identify shorter unidimensional subscales, Maier et al.28 used Rasch- and Mokken-analyses and Gibbons et al.29 used Factor-analysis. Bech and collegues developed another 6-itemsubscale. This scale initially emerged from an analysis with experienced psychiatrists as a validitycriterium,36 and was validated psychometrically thereafter using Rasch-analyses.37;38 This Bechsubscale was combined with four items of the Cronholm-Ottosson Depression Scale to form theBech-Rafaelsen Melancholia Scale.39 Santor and Coyne examined the score-performances ofindividual HDRS-items as a function of depression severity with a nonparametric Item ResponseTheory (IRT) approach, retaining 14 items.21 These 14 items included all 6 items of the Maiersubscale and all 8 items of the Gibbons subscale. However one item from the Bech subscale (13,somatic symptoms) was not included.

In a meta-analysis of individual patient-data, Faries et al. evaluated the responsiveness of totalHDRS and subscale scores in TCA and SSRI pharmacotherapy trials, finding a maximal sensitivityfor the Maier subscale.40 In a similar reanalysis, Entsuah et al. found larger effect sizes for theBech, Maier and Gibbons subscales compared to the HDRS in trials comparing SSRIs orvenlafaxine.41 O'Sullivan et al. found comparable sensitivity to detect changes for the six-item

Chap

ter

Chapter 3

3

46

Bech subscale compared to the 17-item HDRS.20 Hooper et al. found equal sensitivity to changeduring treatment for the 6-item Bech subscale compared to the HDRS 17 item version.26 Möllerand Bech et al. used the Bech subscale to reexamine treatment efficacy of SSRIs and mirtazapine(versus TCAs or placebo).32;42-44 The latter publications did not provide data for the Maier subscale.

In this paper we describe a secondary analysis of our trial data, in order to answer thefollowing questions:

1) Are the Maier, Bech and HDRS comparable in the measurement of depression severity andthe sensitivity to measure changes in severity? 2) Is this comparability stable across the full rangeof response to treatment (e.g. non-response, partial and full response), across differenttreatments and different baseline severity of depression? and 3) What are clinical cut-off pointsfor the subscales to determine remission compared to conventional definitions.12;16;17

We hypothesized that the differences between Maier, Bech and HDRS-scales would be smalland that there would be no apparent effect modification across neither treatments nor baselineseverity. In contrast, we hypothesized that for non-responders and partial responders the effectsizes would be smaller than for responders. This would additionally prove the hypothesis ofsensitivity to change.

Methods

Patient selectionIn the present analyses we use data from two published randomized controlled trials conductedbetween 1993 and 1998 which were published or accepted for publication.45;46 The first trial aimedat efficacy and effectiveness of pharmacotherapy versus the combination of pharmacotherapywith Short Psychodynamic Supportive Psychotherapy (SPSP)47-50 (16 sessions).45 The second trialinvestigated efficacy and effectiveness of a combination of pharmacotherapy with 8 versus 16sessions of SPSP.46 Pharmacotherapy in both trials consisted of three successive steps in case ofintolerance or inefficacy. Both trials started with fluoxetine (20 mg/ day), when this wasunsuccessful (CGI-I >2, only 'minimally improved' or worse) after 6 weeks amitriptyline (≥150 mg/day, dependent of plasma-levels) was initiated in trial 1 and nortriptyline (≥150 mg/day, dependentof plasma-levels) in trial 2. If again unsuccessful after 6 weeks, moclobemide (300-600 mg/day)was started in trial 1 and mirtazapine (30-45 mg/day) in trial 2.

Inclusion criteria for participation in the trials were age between 18 and 60 years, DSM-III-Rdefined Major Depression (with or without dysthymia) assessed in a structured clinical interview,a 17-item HDRS baseline score of at least 14 points and written informed consent. Patients wereexcluded in case of psycho-organic or psychotic or dissociative disorders, drug abuse, or when thepatient was considered to be too unreliable to participate in a clinical trial. Other axis 1 co-morbidity was not excluded. Further exclusion criteria were if there was a serious communicativeor practical problem (e.g. language barrier or the patient will soon leave the country), if there wasa contraindication for one of the antidepressants used, if the patient was adequately treated withantidepressants during the present depressive episode, if the patient used other psychotropicmedication, or if the patient was or planned to become pregnant. Additional exclusion criteriawere of the usual kind in drug research: “too ill” (e.g. antidepressants must be startedimmediately) and/or “too suicidal” (e.g. hospitalization is unavoidable) to participate in a clinicaltrial. The study was approved by the medical ethics committee. After complete description of thestudy to the subjects, written informed consent was obtained.

Of 3226 newly registered outpatients, 988 patients had a depressive disorder. By initialscreening 503 of these 988 patients were excluded by the above exclusion-criteria leaving 485subjects (including patients that later refused to participate or had a HDRS below 14; furtherreferred to as the diagnostic sample). To enter the trials, a second exclusion check was performedby a psychiatrist (excluding 73 patients), and 142 subjects with a HDRS-17 <14 were excluded,


3

47

leaving 270 patients for randomization. After randomization 51 patients refused participation,leaving 219 patients who started the proposed therapy (further referred to as the per protocolsample).45;46

In this manuscript we used the diagnostic sample for most cross-sectional analyses, and therandomized patients in the per protocol sample for analyses of sensitivity of response-data. Fornon-completers, the last observation was carried forward (LOCF).

Outcome measuresPrimary outcome-measures were the 17-item HDRS,18;19 the Maier subscale of the HDRS(containing items 1,2,7-10),28 the Bech subscale of the HDRS (items 1,2,7,8,10 and 13),37 the ClinicalGlobal Impression Severity (CGI-S) and Improvement (CGI-I) scale,24 and the Symptoms check list,90 items (SCL-90) depression subscale (SCL-90dep).51;52 Thus, three levels of information wereobtained: data from 1) an independent, trained, supervised and blinded research-assistant (HDRS-17 & Maier, Bech), 2) the treating clinician (CGI-S/I) and 3) the patient (SCL-90dep). The HDRS wasadministered using a semi-structured interview.53 Before participating in the study, the reliabilityof the HDRS-assessments was established. During the study, in order to avoid slippage,audiotaped assessments were discussed monthly.

In the analyses of treatment-efficacy, response was defined as a ≥50% HDRS-score reduction,partial response as ≥ 20-50% reduction in HDRS-score and remission as a HDRS score of 7 points orless.16;54

StatisticsCronbach-α coefficients and mean inter-item correlations were used to express internalconsistency. To check whether the increased number of items in the HDRS accounted for a higherCronbach-α coefficient than in the subscales (with only 6 items), we applied the Spearman-Brownformula.55 Next we calculated concurrent validity as Pearson correlation coefficients betweentotal HDRS, Maier and Bech subscale scores and SCL-90dep scores. Linear regression modelscalculated variance of HDRS-scores explained by the subscales.56 These analyses were performedin our diagnostic sample. Concurrent validity between CGI-S/I and HDRS subscale-ratings wasdetermined also, however, to avoid low correlations due to limited dispersion, this was done forthe last observation in the per protocol sample. The CGI improvement scale was compared withchanges expressed as percentages of the baseline score.

In order to compare differences in sensitivity to measure treatment effects (also referred toas responsiveness), in data from the per protocol sample effect sizes (E-S) for HDRS, Maier andBech subscales were calculated per subject as the within-subject changes in scale-scores dividedby the pooled standard deviation of the mean change in scale score .20 In this way,differences in effect sizes could be tested and 95% confidence intervals (95% CI) could becalculated. Differences in E-S between the scales were tested by paired T-tests. In order todetermine significant effect-modification, the above analyses were repeated while data werestratified. For stratification we used initial HDRS scores of at least 19 for severe depression,11

criteria for response as described above and treatment-condition. Differences in E-S betweenstrata were tested by Analysis of variance-models (ANOVAs).

The Partial Credit Item Response Theory (IRT) model57 was used to estimate the relationshipsbetween total scores on the HDRS and total scores on the Maier and Bech subscales of the HDRS.The scores were those obtained at exit (per protocol sample). The computer program OPLM58

was used to obtain a set of weights for each item in the HDRS using conditional maximumlikelihood methods. The same software and the item weights were used to obtain estimates ofthe latent trait associated with each score on the HDRS, the Maier subscale and the Bechsubscale. The total scores for the pairs of scales were equated by matching the total scores forwhich the latent trait scores were most similar.59 These methods are very similar to those used in aprevious publication about the Quick Inventory for Depressive Symptomatology IDS.60 The rangeof SCL scores associated with each HDRS score was obtained directly from the original data.

⎟⎠⎞

⎜⎝⎛ −

differencepooledSDTT LOCFend

_

)(0

Chap

ter

Chapter 3

3

48

Finally, Receiver operator characteristic (ROC) curves were constructed to summarize validityof cut-off points. Differences in areas under the curve (AUC) were tested with attention forinterrelation (because we studied these tests within the same subjects) as described by Hanley.61

For all data analysis except the IRT analysis, SPSS for Windows version 10.1 was used.62 For all teststwo-tailed significance levels were applied.

Table 3.1. Studied populations.*

Diagnostic sample (n= 485†)

Per protocol sample (n= 219)‡

Trial I Trial II CombinedI + II AD

(n= 57) AD+SPSP(16) (n= 72)

AD+SPSP(8)(n= 45)

AD+SPSP(16)(n= 45)

Sex (% female) 60.3 63.2 61.1 60.0 68.9 63.0

Age 35.3 ±9.9 34.9 ±8.2 34 ±9.4 38.1 ±10.5 36.2 ±10.5 35.5 ±9.7

Marital status (%): married divorced widowed never married other

97 (20.2)60 (12.5)3 (0.6)318 (66.1)3 (0.6)

12 (21.1)7 (12.3)-38 (66.7)-

10 (13.9)8 (11.1)-54 (75.0)-

19 (42.2) 5 (11.1) 1 (2.2) 20 (44.4) -

9 (20.5)9 (20.5)-26 (59.1)-

50 (22.9)29 (13.3)1 (0.5)138 (63.3)-

Educational level (%): low intermediate high

72 (15) 179 (37.3) 229 (47.7)

11 (19.3) 21 (36.8) 25 (43.9)

13 (18.3)22 (31.0)36 (50.7)

8 (17.8)20 (44.4) 17 (37.8)

8 (18.2)16 (36.4)20 (45.5)

40 (18.4)79 (36.4)98 (45.2)

Occupational (%): job on sickness social security disabled student other

166 (34.7)134 (28.0)84 (17.5)27 (5.6)41 (8.6)27 (5.6)

18 (31.6)14 (24.6)12 (21.1)3 (5.3)5 (8.8)5 (8.8)

22 (30.6)23 (31.9)10 (13.9)5 (6.9)10 (13.9)2 (2.8)

19 (42.2)13 (28.9)8 (17.8)1 (2.2)2 (4.4)2 (4.4)

14 (31.8)16 (36.4)3 (6.8)3 (6.8)3 (6.8)5 (11.4)

73 (33.5)66 (30.3)33 (15.1)12 (5.5)20 (9.2) 14 (6.4)

Duration of episode < 1 year 1-2 years > 2 years

314 (67.0) 70 (14.9) 85 (18.1)

39 (70.9)6 (10.9)10 (18.2)

49 (70.0) 9 (12.9) 12 (17.1)

27 (61.4) 7 (15.9)10 (22.7)

25 (56.8) 14 (31.8) 5 (11.4)

140 (65.7)36 (16.9)37 (17.4)

Psychiatric treatment during this episode (%)

95 (20.2) 13 (23.6) 9 (12.9) 5 (11.4) 10 (22.7) 37 (17.4)

Antidepressants ≤3 mths before (%) adequate (%)

77 (16.4)1 (0.2)

6 (10.9)-

10 (14.3)-

9 (20.5) -

9 (20.5) -

34 (16.0) -

Depressive episodes (prev. 5 years) none 1-2 3 or more

296 (63.1)138 (29.4) 35 (7.5)

26 (47.3)21 (38.2)8 (14.5)

4 (58.6) 22 (31.4)7 (10.0)

33 (75.0)10 (22.7) 1 (2.3)

28 (65.1)13 (30.2)2 (4.7)

128 (60.4)66 (31.1)18 (8.5)

Ethnicity (%) North-west Europe Mediterranean Caribbean Other

414 (86.3)18 (3.8)22 (4.6)26 (5.4)

51 (89.5)3 (5.3)1 (1.8)2 (3.5)

63 (87.5)2 (2.8)2 (2.8) 5 (6.9)

38 (84.4) 1 (2.2) 5 (11.1) 1 (2.2)

40 (90.9) 1 (2.3) 2 (4.5) 1 (2.3)

192 (88.1)7 (3.2) 10 (4.6) 9 (4.1)

Baseline scores of rating scales

HDRS-17 17.1 ±6.5 21.0 ±4.8 20 ±4.9 19.4 ±3.8 20.3 ±4.4 20.2 ±4.6§

Maier 9.2 ±3.6 11.0 ±2.9 10.9 ±2.8 10.7 ±2.3 10.9 ±2.9 10.9 ±2.8§

Bech 9.4 ±3.7 11.5 ±2.8 11.2 ±2.7 10.7 ±2.3 11.0 ±3.0 11.1 ±2.3§

CGI-S ¶ 4.7 ±0.7 4.8 ±0.6 4.7 ±0.7 4.5 ±0.7 4.6 ±0.6 4.7 ±0.7

SCL-90 depression subscale 45.9 ±11.8 48.7 ±11.7 47.8 ±9.8 49.3 ±8.7 52.0 ±10.1 49.2 ±10.2§

* Data represent means (±SD) unless indicated. Denominators of percentages vary due to missing values.† Total diagnostic sample.‡ No significant differences between treatment-groups (per protocol sample) (ANOVA or χ2).§ Significant differences (p< 0.05) between in- and excluded patients (indep. T-Test).¶ n= 241 References to studies: Trial I45 and Trial II46


3

49

Results

Patient characteristicsTable 3.1 shows demographics for the diagnostic and per protocol samples. There were nosignificant differences observed between the diagnostic and per protocol sample (tested asexcluded versus included), except from a lower mean HDRS-score (and Maier, Bech and SCL90depression scores) in the diagnostic sample. This difference was due to application of the entrancecriterion (HDRS ≥14) for randomization. No significant differences existed between the differenttreatment-groups. The studied population existed of mainly unmarried, mid-thirty, moderately tohighly educated, female, Caucasian adults, with moderate to severe depressive episodes of lessthan 1 year of duration. More than 75% of the subjects were not treated for the current depressiveepisode before, 16% percent received an inadequate trial of an antidepressant.

Internal and concurrent validityData for internal and concurrent validity are presented in Table 3.2. Cronbach-alphas were slightlylower for the Maier and Bech subscales. If a 17-item scale is reduced to 6 items, the expectedalpha is 0.49 (Spearman-Brown formula). Thus the observed values of 0.62 and 0.60 showincreased internal validity for the subscales. The mean inter-item correlation was markedly higherfor the Maier and Bech subscales. The correlation between Maier and Bech subscales was high.Both Maier and Bech subscales explained about 75% of the variance of the total HDRS-score. Theself-rated SCL-90dep was reasonably well correlated with the HDRS (r= 0.67) and the Maier andBech subscales (r= 0.64). Concurrent validity of the scales was overall slightly less in the moredepressed sub-group (HDRS ≥19; n= 194) compared to moderately depressed subjects, except forthe correlation between HDRS and Maier subscale.The CGI-S at study-endpoint was moderatelycorrelated with the HDRS (r= 0.57), as with the Maier and Bech subscales. The CGI-S showedhigher correlation with the Bech subscale, especially in those severely depressed. The CGI-I atstudy-endpoint was less well correlated with the percentage change in HDRS (r= 0.42) and thesubscales.

Table 3.2. Internal validity and concurrent validity of HDRS-17, Maier, Bech, SCL-90 depression subscale and CGI-S.

Internal consistency Concurrent validity: Pearson's r (% explained variance)

Cron-bach-α Mean Inter-item corr.

Overall Moderate depression* Severe depression*

Maier Bech Maier Bech Maier Bech

Diagnostic sampleMaier†

0.62 0.21 - - - - - -

Bech† 0.67 0.25 0.95(91%)

- 0.91(83%)

- 0.90(81%)

-

HDRS-17† 0.73 0.13 0.86(75%)

0.86(73%)

0.57(57%)

0.76(58%)

0.65(42%)

0.60(36%)

SCL-90 depression subscale†

0.88 0.33 0.64(41%)

0.64(40%)

0.45(21%)

0.45(20%)

0.40(16%)

0.42(18%)

Per protocol sampleCGI-S endpoint‡ NA 0.54

(29%) 0.58(34%)

0.55(31%)

0.55(30%)

0.59(34%)

0.65(42%)

CGI-I endpoint§ NA 0.42 (18%)

0.43(18%)

0.38(14%)

0.37(13%)

0.48(23%)

0.49(24%)

* Severe depression defined as initial HDRS-17 ≥19 (n= 221).† Maier, Bech, HDRS: n= 482; SCL-90dep: n= 473.‡ CGI-S: n= 229. § Compared with change expressed as percentage of baseline rating

Chap

ter

Chapter 3

3

50

Sensitivity to changeIn Table 3.3 and 3.4, overall and stratified E-S in the per protocol sample are presented. In thesetables the 95% CI of the E-S indicates whether the E-S significantly deviates from 0 (no effectmeasured). Comparisons between E-S may produce significant differences between E-S, evenwhen the 95% CIs between the two E-S overlap.

Of the 9 comparisons between the Maier and Bech subscales made in these tables, 5 were notsignificant. The Maier was significantly different from the HDRS in 4 out of 9 comparisons, whilethe Bech was significantly different from the HDRS in only 1 of the 9 comparisons. Differencesbetween E-S were small.

Table 3.3. Pre- and post-treatment Maier, Bech and HDRS-scores with corresponding effect sizes in per protocol sample. Stratification by depression-severity and final treatment response.

Mean ±SD baseline

Mean ±SD endpoint (LOCF)

Mean decrease (95% CI)

SD of decrease

Effect size [E-S](95% CI)

All subjects (n= 219)

Maier 10.9 ±2.75 6.2 ±4.46 4.7 (4.1 - 5.3) 4.54 1.03 (0.89 - 1.16)*†

Bech 11.1 ±2.69 6.2 ±4.50 4.9 (4.3 - 5.5) 4.54 1.08 (0.95 - 1.22)

HDRS-17 20.2 ±4.56 12.0 ±7.62 8.2 (7.2 - 9.2) 7.45 1.10 (0.96 - 1.23)

Moderate depression (initial HDRS-17 <19; n= 93)‡

Maier 8.9 ±2.13 5.3 ±3.86 3.7 (2.8 - 4.5) 4.19 0.81 (0.62 - 1.00)†

Bech 9.2 ±2.12 5.1 ±3.89 4.1 (3.2 - 5.0) 4.28 0.91 (0.71 - 1.10)

HDRS-17 16.2 ±1.42 9.8 ±6.15 6.4 (5.1 - 7.8) 6.47 0.86 (0.68 - 1.04)

Severe depression (initial HDRS-17 ≥19; n= 126)‡

Maier 12.3 ±2.25 6.9 ±4.75 5.4 (4.6 - 6.2) 4.66 1.19 (1.01 - 1.37)*

Bech 12.5 ±2.16 7.0 ±4.75 5.5 (4.7 - 6.3) 4.65 1.21 (1.03 - 1.39)

HDRS-17 23.1 ±3.79 13.7 ±8.18 9.5 (8.1 - 10.8) 7.88 1.27 (1.08 - 1.46)

Final Non-responders (n= 65)§

Maier 10.6 ±2.82 10.6 ±2.86 0.0 (-0.7 - 0.6) 2.65 -0.01 (-0.15 - 0.14)¶

Bech 11.1 ±2.76 10.7 ±2.89 0.4 (-0.2 - 1.1) 2.66 0.09 (-0.05 - 0.24)¶

HDRS-17 19.5 ±4.25 19.9 ±4.55 -0.4 (-1.2 - 0.3) 3.08 -0.06 (-0.16 - 0.05)¶

Final Partial-responders (n= 64)§

Maier 11.5 ±2.86 7.4 ±3.07 4.1 (3.4 - 2.7) 2.58 0.90 (0.76 - 1.04)

Bech 11.4 ±2.59 7.4 ±3.19 4.1 (3.4 - 4.7) 2.71 0.90 (0.75 - 1.04)

HDRS-17 21.4 ±5.06 14.2 ±4.41 7.2 (6.7 - 7.7) 1.88 0.96 (0.90 - 1.03)

Final Responders (n= 90)§

Maier 10.6 ±2.59 2.1 ±2.00 8.5 (7.8 - 9.1) 3.12 1.87 (1.72 - 2.01)**

Bech 10.9 ±2.71 2.1 ±1.89 8.8 (8.1 - 9.4) 3.14 1.93 (1.79 - 2.08)

HDRS-17 19.9 ±4.28 4.8 ±3.46 15.1 (14.1 - 16.1) 4.89 2.02 (1.89 - 2.16)**

Note that the overlap of two 95%CI of E-S does not rule out a statistical significant difference between these E-S (see text).* Significantly different from E-SHDRS (paired T-test; p <0.05) † Significantly different from E-SBech (paired T-test; p <0.05) ‡ Significant differences of E-SMaier, E-SBech and E-SHDRS between moderate and severe depression (ANOVA; p< 0.05)§ Response criteria: decrease in HDRS-scores: <20%= Non-response, 20-50%= Partial response and ≥50%= Response. Significant differences of E-SMaier, E-SBech and E-SHDRS between categories of response (ANOVA; p< 0.001)¶ Significant differences between E-SMaier-E-SBech, E-SHDRS-E-SMaier and E-SHDRS-E-SBech (paired T-test; p <0.05)** Significant difference between E-SHDRS-E-SMaier (paired T-test; p <0.05)


3

51

In the total per protocol sample the Maier subscale was significantly less powerful to observetreatment effects: the E-S assessed by the Maier was significantly lower than the E-S of the Bechand HDRS. When stratified for depression severity, the E-S of Maier, Bech subscales and HDRSwere larger in severe compared to moderate depression. A significant difference between thesestrata was observed for all E-S (ANOVA). Within the group of severely depressed subjects, theMaier was significantly less sensitive compared to the HDRS (paired T-test). Within themoderately depressed group, the Bech outperformed the Maier (paired T-test). Across differentstrata of final response significant differences in E-S were found (ANOVA). Within strata, the Bechsubscale performed less in final non-responders, while the Maier performed significantly less thanthe total HDRS in final responders (paired T-tests).

In Table 3.4 it is shown that no overall differences in effect sizes were found betweentreatment modalities (ANOVA). Within the group of patients treated with anti-depressants only,the Maier subscale was significantly less sensitive to detect treatment differences than the HDRS,however the Maier did not differ significantly from the Bech subscale (paired T-test).

Conversion of HDRS scores, criteria for remission and depression severityTable 3.5 shows the conversion between HDRS scores and Maier, Bech and SCL90dep scores.Maier and Bech cut-off scores to define remission (bold), mild, moderate and severe depression)11

can be identified (italic). Figure 3.1 shows the ROC-curves for Maier, Bech CGI-S, CGI-I and SCL-90dep cut-off scores, with HDRS ≤7 as the reference criterion. The difference in AUC for the Maierand Bech subscales was not significant (z= 1.25; p= 0.21). The difference in AUC between Maierand Bech subscales compared to SCL90dep and both CGIs was highly significant (z> 3.8; p< 0.001).In the table below Figure 3.1 sensitivity and specificity for cut-off scores ≤3 and ≤4 for the Maierand Bech subscales are given.

Table 3.4. Pre- and post-treatment Maier, Bech and HDRS-scores with corresponding effect sizes in per protocol sample. Stratification by treatment modality.

Mean ±SD baseline Mean ±SD endpoint (LOCF)

Mean decrease(95% CI)

SD of decrease

Effect size [E-S](95% CI)

All subjects (n= 219)

Maier 10.9 ±2.75 6.2 ±4.46 4.7 (4.1 - 5.3) 4.54 1.03 (0.89 - 1.16)*†

Bech 11.1 ±2.69 6.2 ±4.50 4.9 (4.3 - 5.5) 4.54 1.08 (0.95 - 1.22)

HDRS-17 20.2 ±4.56 12.0 ±7.62 8.2 (7.2 - 9.2) 7.45 1.10 (0.96 - 1.23)

AD (n= 57)‡

Maier 11.0 ±2.95 7.2 ±4.97 3.8 (2.4 - 5.1) 5.12 0.83 (0.53 - 1.13)§

Bech 11.5 ±2.78 7.1 ±5.01 4.4 (3.1 - 5.7) 5.01 0.97 (0.68 - 1.26)§

HDRS-17 21.0 ±4.77 13.9 ±8.36 7.1 (4.9 - 9.3) 8.38 0.95 (0.65 - 1.25)

AD + 8 SPSP (n= 45)‡

Maier 10.7 ±2.34 5.9 ±4.31 4.8 (3.6 - 5.9) 3.81 1.05 (0.80 - 1.31)

Bech 10.7 ±2.28 6.0 ±4.21 4.6 (3.5 - 5.8) 3.87 1.02 (0.76 - 1.28)

HDRS-17 19.4 ±3.80 11.1 ±6.80 8.3 (6.4 - 10.2) 6.37 1.12 (0.86 - 1.38)

AD + 16 SPSP (n= 117)‡

Maier 10.9 ±2.82 5.8 ±4.21 5.1 (4.2 - 5.9) 4.47 1.12 (0.93 - 1.30)

Bech 11.1 ±2.78 5.8 ±4.32 5.3 (4.4 - 6.1) 4.54 1.16 (0.98 - 1.35)

HDRS-17 20.1 ±4.69 11.5 ±7.44 8.6 (7.3 - 10.0) 7.37 1.16 (0.98 - 1.34)

AD= Antidepressants, SPSP = Short Psychodynamic Supportive Psychotherapy Note that the overlap of two 95%CI of E-S does not rule out a statistical significant difference between these E-S (see text).* Significantly different from E-SHDRS (paired T-test; p <0.05) † Significantly different from E-SBech (paired T-test; p <0.05) ‡ No significant differences of E-SMaier, E-SBech and E-SHDRS between treatment modalities (ANOVA)§ Significant differences between E-SMaier-E-SBech (paired T-test; p <0.05)

Chap

ter

Chapter 3

3

52

Discussion

Major findingsThis study examined the relative effectiveness of the HDRS subscales as developed by Maier etal.28 and Bech et al.37 in monitoring severity and treatment effects in depression. We found thatthe Maier and Bech subscales gave results comparable to the original 17-item HDRS, with highconcurrent validity and increased mean inter-item correlations and internal consistency. Maierand Bech subscales were highly comparable to each other in the measurement of treatmentchanges. Differences between E-S were rather small, and clinically irrelevant. For interpretation aconversion table linking HDRS-scores and Maier and Bech scores are provided. The Maier had aslightly (non-significant) higher sensitivity and specificity to predict the reference criterion forremission (HDRS ≤7). Both Maier and Bech subscales differentiated non-responders from partialand final responders.

A significant difference in sensitivity to change existed between the Bech and Maier withinthe group treated with antidepressants only. We were unable to find the reason for thisdifference compared to other treatment modalities, where the difference between Maier andBech was not found or was not significant. The question arises whether there is a difference insensitivity between the Maier and Bech subscales across different treatment-modalities or thatother (post-randomization) differences between the groups or mere chance explain thisobservation. Because this difference was not found in the other groups (treated with bothantidepressants and psychotherapy) we think it cannot be ascribed to a difference in detectingpharmacological (side-) effects. If a Bonferroni correction would be applied for the number ofcomparisons tested (p < 0.01), the observed difference would not remain its significance.

Figure 3.1. Receiver operating characteristic curves for Maier and Bech subscales, SCL90 depression and CGI-S/I at endpoint compared to HDRS-17 'Remission' in per protocol sample.

Reference: HDRS-17 score ≤7 (Remission). Sens.= sensitivity, Spec.= specificityAUC (SE): Maier 0.972 (.009), Bech 0.963 (.011), SCL-90dep 0.862 (.028), CGI-S 0.743 (.036), CGI-I 0.738 (.035)

��

��

��

��

��

!�!�


3

53

The relevance of the difference between the Maier and Bech subscalesThe only difference between the Maier and Bech subscales is the inclusion of agitation (e.g.running thoughts or restlessness, 0-4 points) in the Maier versus the inclusion of general somaticsymptoms (e.g. tiredness, 0-2 points) in the Bech. It could be argued that one scale is comparablewith the other scale without the different item, e.g. the Maier subscale would then becomparable to the Bech subscale minus the 'general somatic' item. In our (diagnostic) sample theitem agitation contributed 1.3 (SD= 0.9) points to the total Maier-score (9.2, SD= 3.6). The generalsomatic item contributed 1.5 (SD= 0.8) points to the total Bech-score (9.4, SD= 3.7). Thus, overall

Table 3.5. The conversion between the HDRS total scores and the Maier subscale, Bech subscale and the range of SCL scores using IRT analysis (per protocol sample).*

HDRS Maier Bech Range SCL90dep

0 0 0 -

1 – 2 1 1 -

3 – 4 2 2 -

5 3 3 -

6 3 4 -

7 (remission†) 4 4 -

8 – 9 5 5 -

10 – 11 6 6 -

12 7 7 -

13 (mild†) 7 8 -

14 8 8 31 – 61

15 8 9 22- 59

16 9 9 26 - 61

17 9 10 25 - 59

18 (moderate†) 10 10 30 - 62

19 10 10 28 - 60

20 11 11 38 - 67

21 11 11 30 - 72

22 12 12 39 - 61

23 12 12 36 - 61

24 (severe†) 13 13 38 - 67

25 (very severe†) 13 13 45 - 64

26 13 13 47 - 71

27 14 14 46 - 72

28 14 14 42 - 79

29 15 14 55 - 71

30 15 15 43 - 43

31 15 15 63 - 70

32 16 15 62 - 65

33 16 16 57 - 71

34 – 35 17 16 -

36 17 17 -

37 – 39 18 17 -

40 – 42 19 18 -

43 – 44 20 19 -

45 21 19 -

46 21 20 -

47 – 48 22 20 - * The only valid conversions that can be made from this table are between 1) HDRS and Maier, 2) HDRS and Bech and 3) HDRS and SCL90dep.† Cut-offs as provided by Yonkers.11

Chap

ter

Chapter 3

3

54

tiredness was more present than agitation in this sample, and agitation was not rated near itsmaximum like tiredness. Both items occurred intraindividually at the same time, but were notinterchangeable. This means that the Maier and Bech subscales show different perspectives ondepressive symptomatology. In this respect, it is noteworthy to mention that the agitation itemwas dropped beforehand when the Bech subscale was developed and validated, because thisitem showed limited variance (i.e. was not found to be scored) in the two studied samples.36;37

Furthermore, in the Maier subscale the items psychomotor agitation and psychomotorretardation are included, which -at first sight- seem to represent two opposed polarities.However, these items also co-occurred within the same individuals. This can be explained by thebroad definitions of agitation (both restlessness or running thoughts) and retardation (bothretardation in activities or in thinking) in our semi-structured interview.

The original HDRS is often criticized to measure these somatic symptoms.11;15;27;28 Although theBech subscale was designed as an unidimensional scale, the 'general somatic' item is still amongthe 6 items. However, in the Rasch-analysis this item was the least contributive and showed aceiling effect for moderate and severe depression.37 Though both a DSM-IV and ICD-10 criterionfor diagnosis,the aspecific 'tiredness' symptom may also be caused by physical illnesses. TheMaier subscale does not include this item. Thus, the Maier subscale might especially be useful inpatients suffering from somatic complaints or illnesses. Additional, methodological supportcomes from the exclusion of this somatic item by Santor et al.21 This hypothesis of betterperformance in patients with somatic complaints or illness needs further investigation, e.g. incomparison with the Hospital Anxiety and Depression Scale.63

Previous comparative studiesOur findings are in line with findings of previous studies,11;20;26;40-44 and extends the evidence tosupport the Maier and Bech subscale as a valid alternative for the HDRS. This is relevant for theplanning and conduction of clinical trials,40-42 but also for clinical practice.20;26 Hooper et al. 26 foundequal performance of the Bech subscale compared to the Montgomery Asberg Depression RatingScale (MADRS).30 Because the MADRS was not used in our trials, we were unable to examine theperformance of the Maier subscale compared to the MADRS. Hooper et al. questioned whether apossible ceiling effect in the Bech subscale would limit its usefulness in severely depressedpatients.26 In our study more than 57% of the per protocol patients had an initial HDRS greater than18 (indicative for severe depression).11 We did not find a ceiling effect in our diagnostic sample(data not shown), and found consequently higher effect sizes for the Maier and Bech subscales ininitially severely depressed patients, indicative for an adequate sensitivity to measure (larger)changes due to treatment. In addition to the observed ability to predict remission,41 we proposedcut-off scores for remission and various ranges for classification of depression severity.

In two publications Bech proposed the Bech subscale as an alternative measure to overcomethe confounding influence of drug related side-effects in the comparison with placebo or activedrugs.43;44 However this problem is not fully solved, as tiredness may be induced by histaminergiceffects from antidepressants (e.g. tricyclics and mirtazapine).43 On the other hand, agitation(included in the Maier subscale) is known as a (mostly transient) SSRI-induced side-effect .

An extra dimension of our study is that it extends the data for use of the Maier and Bechsubscales in populations treated with psychotherapy. Hooper et al. and O'Sullivan et al. alreadydemonstrated the usefulness of the Bech subscale in pharmacological treatment of melancholia,dysthymia and typical and atypical depression.20;26

An alarming point of our study is the moderate correlation of the Maier, Bech and HDRS withthe CGI-S and the CGI-I. Previous reports mentioned correlations between HDRS and CGI varyingbetween 0.65 and 0.90.11;28;40 Our results underscore the need of a HDRS or subscale rating insteadof the CGI. We consider the validity of the CGI to be questionable, as most CGI-raters (subjectively)evaluate their own treatment. Apparently, the clinician’s judgment does not coincide with scalescores.In this respect, the performance of the (self-rated) SCL-90dep is better. This was alsoillustrated in the ROC-curves regarding the criterion of remission. Further research is needed toinvestigate whether correlations with the HDRS of other self-rated scales (e.g. the Beck


3

55

Depression Inventory64) are higher than the SCL-90dep. In addition to this, a major limitation in ourstudy and in any study investigating depression-'severity', is that there is no definite goldstandard. We used HDRS data as the gold standard, which means that scales under investigationcan never be judged to be better than the HDRS, however this would be reversed if the CGI wasused as a gold standard.65

ConclusionWe conclude that both Maier and Bech subscales of the HDRS are equivalent to the HDRS, andcan easily be used to increase efficiency to measure treatment response in clinical practice. Ontheoretical grounds we have a slight preference for the Maier subscale. The use of subscaleswould improve the efficiency and objectivity of measuring response in clinical practice, whereoften no scale (instead of a Clinical Global Impression) is used at all. This would further bridge thegap between clinical practice and research based treatment-recommendations for non-responsein depression. Maier and Bech subscales should be compared in patients suffering from co-morbidsomatic illnesses, or patients treated with psychotherapy only. The impact of the difference of theone somatic item versus the agitation-item between the Maier and Bech subscales and theconsequences for their applicability in clinical sub-groups needs further research.

AcknowledgementThe original randomized controlled trials were supported by an unrestricted educational grantfrom Eli Lilly Netherlands. All studies were carried out by the Mentrum Depression ResearchGroup. We thank all psychotherapists, psychiatrists and residents for their excellent work.

Conflicts of interestExternal funding did not support these post-hoc analyses.

References

1. Murray CJ, Lopez AD. Evidence-based health policy--lessons from the Global Burden of Disease Study. Science. 1996; 274:740-743.

2. Greenberg PE, Stiglin LE, Finkelstein SN, Berndt ER. Depression: a neglected major illness. J Clin Psychiatry. 1993; 54: 419-424.




6. Centraal Begeleidingsorgaan voor de Intercollegiale Toetsing. Consensusbijeenkomst depressie bij volwassenen: vrijdag 16september 1994 Utrecht. Utrecht: Centraal Begeleidingsorgaan voor de Intercollegiale Toetsing; 1994.



9. Rush AJ, Crismon ML, Toprac MG, Trivedi MH, Rago WV, Shon S et al. Consensus guidelines in the treatment of majordepressive disorder. J Clin Psychiatry. 1998; 59: 73-84.

10. Trivedi MH, Rush AJ, Crismon ML, Kashner TM, Toprac MG, Carmody TJ et al. Clinical results for patients with majordepressive disorder in the Texas Medication Algorithm Project. Arch Gen Psychiatry. 2004; 61: 669-680.

11. Yonkers KA, Samson J. Mood disorders measures. In: Rush AJ, Pincus HA, First MB, et al., eds. Handbook of psychiatricmeasures. 1st ed. Washington DC: American Psychiatric Association; 2000:515-548.


Chap

ter

Chapter 3

3

56

13. Biggs MM, Shores-Wilson K, Rush AJ, Carmody TJ, Trivedi MH, Crismon ML et al. A comparison of alternative assessmentsof depressive symptom severity: a pilot study. [republished from Psychiatry Res 2000 Jul 24;95(1):55-65]. Psychiatry Res.2000; 96: 269-279.

14. Carroll BJ, Fielding JM, Blashki TG. Depression rating scales. A critical review. Arch Gen Psychiatry. 1973; 28: 361-366.15. Anonymous. Scales for assessment of diagnosis and severity of mental disorders. Acta Psychiatr Scand Suppl. 1993; 372: 1-

87.16. Frank E, Prien RF, Jarrett RB, Keller MB, Kupfer DJ, Lavori PW et al. Conceptualization and rationale for consensus

definitions of terms in major depressive disorder. Remission, recovery, relapse, and recurrence. Arch Gen Psychiatry. 1991;48: 851-855.


18. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 23: 56-61.19. Hamilton M. Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol. 1967; 6: 278-296.20. O'Sullivan RL, Fava M, Agustin C, Baer L, Rosenbaum JF. Sensitivity of the six-item Hamilton Depression Rating Scale. Acta

Psychiatr Scand. 1997; 95: 379-384.21. Santor DA, Coyne JC. Examining symptom expression as a function of symptom severity: Item performance on the

Hamilton Rating Scale for Depression. Psychol Assess. 2001; 13: 127-139.22. Bagby RM, Ryder AG, Schuller DR, Marshall MB. The Hamilton Depression Rating Scale: has the gold standard become a

lead weight? Am J Psychiatry. 2004; 161: 2163-2177.23. Williams JB. Standardizing the Hamilton Depression Rating Scale: past, present, and future. Eur Arch Psychiatry Clin

Neurosci. 2001; 251 Suppl 2: II6-12.24. Guy W, Clearly PA, Close JH, Conners CK, Covi L, et al. ECDEU Assessment manual for psychopharmacology. DHEW

publication (ADM) 76-338. 1976. Washington, DC, US Department of Health, Education, and Welfare. 25. American Psychiatric Association. Handbook of psychiatric measures. 1st ed. Washington DC: American Psychiatric

Association; 2000.26. Hooper CL, Bakish D. An examination of the sensitivity of the six-item Hamilton Rating Scale for Depression in a sample of

patients suffering from major depressive disorder. J Psychiatry Neurosci. 2000; 25: 178-184.27. Linden M, Borchelt M, Barnow S, Geiselmann B. The impact of somatic morbidity on the Hamilton Depression Rating

Scale in the very old. Acta Psychiatr Scand. 1995; 92: 150-154.28. Maier W, Philipp M. Improving the assessment of severity of depressive states: a reduction of the Hamilton Depression

Scale. Pharmacopsychiatry. 1985; 18: 114-115.29. Gibbons RD, Clark DC, Kupfer DJ. Exactly what does the Hamilton Depression Rating Scale measure? J Psychiatr Res. 1993;

27: 259-273.30. Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979; 134: 382-

389.31. Maier W, Philipp M. Comparative analysis of observer depression scales. Acta Psychiatr Scand. 1985; 72: 239-245.32. Moller HJ. Methodological aspects in the assessment of severity of depression by the Hamilton Depression Scale. Eur

Arch Psychiatry Clin Neurosci. 2001; 251: 13-20.33. Hughes JR, O'Hara MW, Rehm LP. Measurement of depression in clinical trials: an overview. J Clin Psychiatry. 1982; 43: 85-

88.34. Walczak DD, Apter JT, Halikas JA, Borison RL, Carman JS, Post GL et al. The oral dose-effect relationship for fluvoxamine:

a fixed-dose comparison against placebo in depressed outpatients. Ann Clin Psychiatry. 1996; 8: 139-151.35. Moller HJ, Glaser K, Leverkus F, Gobel C. Double-blind, multicenter comparative study of sertraline versus amitriptyline in

outpatients with major depression. Pharmacopsychiatry. 2000; 33: 206-212.36. Bech P, Gram LF, Dein E, Jacobsen O, Vitger J, Bolwig TG. Quantitative rating of depressive states. Acta Psychiatr Scand.

1975; 51: 161-170.37. Bech P, Allerup P, Gram LF, Reisby N, Rosenberg R, Jacobsen O et al. The Hamilton depression scale. Evaluation of

objectivity using logistic models. Acta Psychiatr Scand. 1981; 63: 290-299.38. Bech P, Allerup P, Reisby N, Gram LF. Assessment of symptom change from improvement curves on the Hamilton

depression scale in trials with antidepressants. Psychopharmacology (Berl). 1984; 84: 276-281.39. Bech P. The Bech-Rafaelsen Melancholia Scale (MES) in clinical trials of therapies in depressive disorders: a 20-year review

of its use as outcome measure. Acta Psychiatr Scand. 2002; 106: 252-264.40. Faries D, Herrera J, Rayamajhi J, DeBrota D, Demitrack M, Potter WZ. The responsiveness of the Hamilton Depression

Rating Scale. J Psychiatr Res. 2000; 34: 3-10.41. Entsuah R, Shaffer M, Zhang J. A critical examination of the sensitivity of unidimensional subscales derived from the

Hamilton Depression Rating Scale to antidepressant drug effects. J Psychiatric Res. 2002; 36: 437-448.42. Bech P, Tanghoj P, Andersen HF, Overo K. Citalopram dose-response revisited using an alternative psychometric

approach to evaluate clinical effects of four fixed citalopram doses compared to placebo in patients with majordepression. Psychopharmacol. 2002; 163: 20-25.

43. Bech P. Meta-analysis of placebo-controlled trials with mirtazapine using the core items of the Hamilton Depression Scaleas evidence of a pure antidepressive effect in the short-term treatment of major depression. Int J Neuropsychopharmacol.2001; 4: 337-345.

44. Bech P, Ciadella P, Haugh MC, Birkett MA, Hours A, Boissel JP et al. Meta-analysis of randomised controlled trials offluoxetine v. placebo and tricyclic antidepressants in the short-term treatment of major depression. Br J Psychiatry. 2000;176: 421-428.


3

57

45. De Jonghe F, Kool S, Van Aalst G, Dekker J, Peen J. Combining psychotherapy and antidepressants in the treatment ofdepression. J Affect Disord. 2001; 64: 217-229.

46. Dekker J, Molenaar PJ, Kool S, Van Aalst G, Peen J, De Jonghe F. Dose-effect relations in time-limited combined psycho-pharmacological treatment for depression. Psychological Medicine. 2005; 35: 47-58.

47. Werman DS. The practice of supportive psychotherapy. New York: Brunner/Mazel; 1984.48. Strupp HH, Binder JL. Psychotherapy in a new key. A guide to Time-limited Dynamic Psychotherapy. New York: Basic Books;

1984.49. Rockland LH. Supportive Therapy. A psychodynamic approach. New York: Basic Books; 1989.50. De Jonghe F, Rijnierse P, Janssen R. Psychoanalytic supportive psychotherapy. J Am Psychoanal Assoc. 1994; 42: 421-446.51. Derogatis LR, Lipman RS, Covi L. SCL-90: an outpatient psychiatric rating scale--preliminary report. Psychopharmacol Bull.

1973; 9: 13-28.52. Arindell WA, Ettema JM. Handleiding bij een multidimensionele psychopathologie-indicator.1986. Lisse, Swets &

Zeitlinger. 53. de Jonghe F. Leidraad voor het scoren van de Hamilton Depression Rating Scale: HDRS leidraad. Amsterdam: Benecke

Consultants; 1994.54. Fava M, Davidson KG. Definition and epidemiology of treatment-resistant depression. Psychiatr Clin North Am. 1996; 19:

179-200.55. Drenth PJ, Sijtsma K. Testtheorie. Inleiding in de theorie van de psychologische test en zijn toepassingen. Houten: Bohn

Stafleu Van Loghum; 1990.56. van Straten A, de Haan RJ, Limburg M, Schuling J, Bossuyt PM, van den Bos GA. A stroke-adapted 30-item version of the

Sickness Impact Profile to assess quality of life (SA-SIP30). Stroke. 1997; 28: 2155-2161.57. Masters GN. A Rasch model for partial scoring. Psychometrika. 1982; 47: 149-174.58. Verhelst ND, Glas CAW, Verstralen HHFM. OPLM: computer program and manual.1995. Arnhem, The Netherlands, CITO. 59. Orlando M, Sherbourne CD, Thissen D. Summed-score linking using item response theory: application to depression

measurement. Psychol Assess. 2000; 12: 354-359.60. Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, Klein DN et al. The 16-Item Quick Inventory of Depressive


61. Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from thesame cases. Radiology. 1983; 148: 839-843.

62. SPSS. SPSS for Windows. [Release 10.1]. 2000. Chicago, Ill., SPSS Inc. 63. Zigmond AS, Snaith RP. The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand. 1983; 67: 361-370.64. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry. 1961; 4:

561 571: 561.65. Mulder RT, Joyce PR, Frampton C. Relationships among measures of treatment outcome in depressed patients. J Affect

Disord. 2003; 76: 127-135.

Chap

ter

Chapter 3

3

58

CHA

PTER

DOSE-ESCALATION FOR INSUFFICIENT RESPONSE

TO A STANDARD DOSE OF A SELECTIVE SEROTONIN

REUPTAKE INHIBITOR IN MAJOR DEPRESSIVE

DISORDER: A SYSTEMATIC REVIEW.

British Journal of Psychiatry 2006; 189: 309-316

Henricus G. Ruhé1

Jochanan Huyser1

Jan A. Swinkels1

Aart H. Schene1

1 Program for Mood Disorders, Department of Psychiatry, Academic Medical Center, University ofAmsterdam, Amsterdam, The Netherlands

Chap

ter

Chapter 4

4

60

Abstract

BackgroundAlthough SSRIs are frequently used for major depressive disorder, only 50-60% of patientsrespond to a standard dose. For non-responders dose-escalation is often applied.

AimTo systematically review the evidence for dose-escalation of SSRIs.

MethodsA systematic literature search in MEDLINE, EMBASE, CINAHL and PsycInfo was performed.Randomised controlled trials and meta-analyses investigating dose-escalation of SSRIs wereidentified. Relevant articles were retrieved and critically appraised. Results were summarized inan evidence table. Pooling was not justified due to heterogeneity of the identified studies.

ResultsEight true dose-escalation studies and 3 meta-analyses were identified. The available dataprovided no unequivocal base for dose-escalation. Dose-escalation before 4 weeks of treatmentat a standard dose appeared to be ineffective.

ConclusionsDose-escalation of SSRIs is equivocally supported by evidence of RCTs, but methodologicaldifficulties in the studies may account for this lack of evidence. Clinical implications andmethodological considerations for future studies are discussed.

ChapterReview of SSRI dose-escalation in MDD

4

61

Introduction

Many countries developed national clinical guidelines for the treatment ofMajor DepressiveDisorder (MDD).1-8 In these guidelines pharmacotherapy is among the most important treatments,while in many of the countries, Selective Serotonin Reuptake Inhibitors (SSRIs) have become thefirst line antidepressants. It is less clear what should be done in those 40-50% of the patients whodo not respond to the first antidepressant given.9;10 Strategies in case of non-response have beenpublished in several narrative reviews11-24 and in one systematic review.25 Three major strategiesfor non-response are recommended: 1) increase of the dose of the antidepressant (dose-escalation), 2) augmenting the antidepressant by adding a second drug, 3) switching to anotherantidepressant of the same or a different class.

Available dose-finding studies do not provide evidence to initiate pharmacotherapy for MDDwith SSRIs in higher than standard doses.26-30 For non-responders, all guidelines recommend doseescalation as the appropriate strategy, instead of continuing an apparently insufficient regimen.1-7

Only the recent NICE guideline is less pronounced in this recommendation,8 and advises that if‘there are no significant side effects, a gradual increase in dose should be considered’. Moreover,surprisingly little systematic evidence is provided to underscore these recommendations.

Due to the above recommendations and its simplicity, dose-escalation is widely practiced andoften the first strategy applied.31-34 The aim of our study was to systematically review the evidencefor dose-escalation of SSRIs in MDD.

Methods

Design of studies to be includedIdeally the design of dose-escalation studies is randomisation of non-responders to higher dosesof an antidepressant or placebo after some weeks of a standard dose. In this review we considerthree other methodological requirements for those studies. First, dose-escalation should bedeferred, e.g. 3-6 weeks after the initiation of the treatment because antidepressants needseveral weeks to have a clinical effect.35 The practice of dose-escalation and the possibility todemonstrate a dose-response relationship is based upon a selection of ‘true’ non-responders.36 Asthis might take 6-10 weeks,37 dose-escalation studies with early randomisation unintentionallydiminish the possibility to prove the usefulness of dose-escalation. The inclusion of unidentifiedlate responders in both arms of the study reduces the contrast between the intervention andcontrol. Second, an outstanding study will have sufficient power to be able to demonstrate aclinically relevant difference (e.g. 20%) between treatment arms, and third will describe themethod of dose-escalation and describe the early drop-out rates due to dose-escalation.

Identification and selection of articlesFirst, systematic literature searches (updated until February 10th 2005) were performed in fourdatabases (MEDLINE, EMBASE, CINAHL, PsychInfo; all indexed years). As there are no specifickeywords for dose-escalation studies, ‘sensitive’ searches were performed with the followingterms: (((dose[textword(tw)] or dosage[tw]) and increase[tw]) or ((dose[tw] or dosage[tw]) andmaxim*[tw]) or (upward[tw] and titrat*[tw])) OR dose-response relationship, drug[MeSH] incombination with the Cochrane Collaboration search-filter for RCTs and systematic reviews, theCochrane Collaboration Depression Anxiety and Neurosis group (CCDAN) search-filter for MDDand MeSH-terms and text words for SSRIs. Primary selection (independently by the first andsecond author) was based on design and focus on dose-response relationships for SSRIs, byscreening title and abstract of the article. Agreement on predominantly exclusion of irrelevantarticles was 99.1%, with Cohen’s Kappa for interrater agreement of 0.62 (a 'substantial'agreement).38 Discrepancies between initial selection was resolved by discussion and concensus.

Chap

ter

Chapter 4

4

62

Second, all potentially relevant articles were judged according to specific inclusion andexclusion criteria (criteria available on request). In case of doubt an article was read fully andassigned afterwards. Additionally, relevant cross-references were retrieved. Double-publicationswere considered together to reveal the maximum of available information.

Critical Appraisal and summary Selected articles were critically appraised and abstracted by the first author, using standardizedforms derived from the Dutch Institute of Healthcare Improvement39 and the AHCPR.4 The itemsused for critical appraisal were the same as proposed by SIGN40 and Sackett et al.41 Each study wasassigned a ‘level of evidence’ (LoE; Table 4.1).39 Levels of evidence are based upon themethodological robustness of studies. For the results, the highest LoE of the supporting scientificevidence (A1-D) is indicated.

To assess judgement-bias by one person who performed the critical appraisal, interratervariation was determined in a slightly different set of 12 publications. Every other author criticallyappraised 4 publications, agreement for the appraisal-items was expressed by Cohen's Kappa.Kappa values were 0.49 (for 'validity of the study'), 0.86 (for 'concealment of allocation'), whilecomplete agreement existed for the appraisal-items 'randomization of the study', 'level ofevidence' and 'data extraction' (Kappa = 1.0). This is in line with other reports of interrateragreement in appraisal of psychiatric research.42

A qualitative summary with discussion of the results, restrictions, methodological flaws andexternal validity of the studies was described in an evidence table and a separate document, ofwhich a summary is provided in this paper. Because of the apparent heterogeneity in timing of thedose-escalation between the studies, results were not pooled in a meta-analysis.

Results

Search results and selection of studies are presented in Figure 4.1. The 11 studies selected for thisreview are summarized in Table 4.2. A table of excluded studies is available on request.

Table 4.1. Levels of Evidence: Therapeutic studies.

A1 Systematic review including at least some studies of A2-level. Consistent results (homogeneity) across the included trials.

A2 Randomized controlled (double-blind) trial of good methodological quality, adequate size and consistence of results.

B Randomized clinical trial of lower methodological quality or inadequate size. Other comparative research (e.g. non-randomized trial, comparative cohort-study, case-control study)

C Uncontrolled, open study

D Expert opinion, e.g. guideline-panel members

Source: Dutch Institute of Healthcare Improvement.39

Figure 4.1. Selection of reported studies.

Articles identified by systematicsearches (n= 1876)

Articles retrieved for moredetailed evaluation (n= 42)

Appropriate studies to beincluded in the review (n= 11)

Included dose-escalationstudies (n= 8)

Excluded, not meeting in- & ex-clusion criteria, based on title

and abstract (n= 1834)

Excluded articles:narrative review (n= 8);different objective (n= 5);no dose-response data (n= 3);retrospective data (n= 4);fixed-dose study (n= 11)

Included systematic reviews(n= 3)


4

63

Characteristics of the studies Our searches identified 8 dose-escalation studies that increased dosages after at least 3 weeks ofstandard dosages.43-50 We further found 3 systematic reviews about dose response relationships,which included respectively three,51 three,52 and four53 of the eight identified dose-escalationstudies.

Across the studies different outcome definitions for end-points were used. In 7 articlesresponse was defined as a reduction of ≥50% in the Hamilton depression rating scale(HDRS).43;44;47;50;53 A Clinical global impression (CGI) improvement or severity score ≤2 was used forresponse in one study.48 Partial response was used in 3 studies and defined as 25-50% decrease inHDRS.45;46;49 In 7 studies remission-rates were reported. These were defined as a HDRS ≤7,46;49;50 orHDRS ≤8.48

Different criteria were used to decide whether a patient should be randomized: non-response(by CGI,47 <50% decrease in HDRS43;44;46;49), or no remission (HDRS ≤848). In the present studies nogenetic information of the CYP P450 system, nor drug blood levels were reported.

The three previous reviews all had some methodological problems: Bollini et al. pooled studieswith completely different designs and drug-classes, and applied a dose-equivalence strategy thatlumped differential doses of SSRIs together.51 Baker et al. also pooled heterogeneous studies withdifferent moments of dose-escalation, and used an unusual low reference dose of fluoxetine(5mg).53 Corruble et al. did not use an adequate search strategy and only described the dose-response relationships found in their identified studies as ‘flat’ ‘curvilinear’ or ‘linear’.52

Outline of dose-escalation studiesWe will briefly outline the dose-escalation studies. Dorseif et al. first investigated week 3 non-responders (n= 371 outpatients) to fluoxetine who were randomized to continuation with 20 mgor increase to 60 mg/day for 5 weeks. Response rates were 40.5% and 44.7% respectively andremission rates 33.3% and 36.2%. Drop-out rates due to side-effects were significantly differentwith 5.3% and 11.6% respectively.43 Schweizer et al. investigated 77 non-responsive outpatientsafter 3 weeks of fluoxetine 20 mg/day, with a randomization to placebo-increase or dose-escalation up to 60 mg/day for 5 weeks. Response rates were 51.2% and 50% respectively, withnonsignificant drop-out rates of 4.9% versus 16.7%.44 In a similar study Schweizer et al. studieddose-escalation of sertraline in outpatient nonremitters after 3 weeks of sertraline 50 mg/day (n=75). Doses were randomly either kept at 50 mg/day or increased to 150 mg/day. Remission ratesafter 5 weeks were 32% and 47% respectively (non-significant). Specified drop out rates due to sideeffects were not reported.48

Fava et al. first openly treated 15 outpatients (who were week 8 non-responders to fluoxetineat 20 mg/day) with increased doses of fluoxetine titrated up to 80 mg/day for 4 weeks. Noresponse rates were given, but the mean HDRS17-score decreased 6.2 points in week 8 non-responders and 10.1 points in partial responders.45 In a second study Fava et al. randomized week8 non-responders to fluoxetine 20 mg/day (n= 41) to either fluoxetine 40-60 mg, desipramineaddition or lithium-addition for 4 weeks. No placebo-increase was practised. Remission rates were53%, 25% and 29% respectively, but these differences were nonsignificant. Initial partial respondersappeared to benefit most from fluoxetine dose-increases (data nonsignificant). Drop-out ratesfor side effects were 0%, 17%, and 7% respectively.46 In a third study Fava et al. repeated the 3-armrandomized design from their 1994 study with a stratification for partial or non-response at week8 (n= 101). After 4 weeks, the high-dose fluoxetine group showed increased but nonsignificantremission rates (42.4%) compared to desipramine addition (29.4%) and lithium-addition (23.5%).Again initial partial responders appeared to benefit more from fluoxetine dose-increasescompared to initial non-responders (differences nonsignificant). No specific data on dropout dueto side effects were given.49

64

Chap

ter

Chapter 4

4

Tabl

e 4.

2. E

ffec

tiven

ess

of in

crea

sing

the

dose

: Sel

ecte

d st

udie

s.

Stud

y (y

ear)

refe

renc

eLo

E N

D

esig

n (f

ollo

w-u

p)

Inte

rven

tion*

Com

paris

on*

Out

com

e†Re

mar

ks

4.2a

. Dos

e-es

cala

tion

stu

dies

Benk

ert e

t al.

(199

7)47

B 54

4

MD

D, M

inD

Out

P

RCT

of w

eek

3 no

nres

pond

ers‡

(n=

86)

(3 w

eeks

)

PAR

40m

gPA

R 2

0mg

Res

pons

e (≥

50%

in H

DR

S 17)

all:

NN

T PA

R40m

g=

100

(5.1-∞

) For

MD

D o

nly:

NN

T PA

R40

mg

= 8

(2.4

- ∞

) For

bas

elin

e H

DR

S ≥2

4: N

NT P

AR4

0mg

= 6

(1.7

- ∞

)

Stud

y in

clud

es M

inD

(42%

of r

ando

miz

ed

patie

nts)

. Dos

e-tit

ratio

n re

sulte

d in

18-2

0% n

ew

SE. N

o sp

ecif

ied

rate

s of

Dro

p-ou

t. R

espo

nse

rate

in ‘p

lace

bo-g

roup

s’ 7

5%. S

tudy

als

o in

vest

igat

es m

apro

tilin

e do

se-e

scal

atio

n (d

ata

N/A

)

Dor

nsei

f et a

l. (1

989)

43B

572

MD

D

Out

P

RCT

of w

eek

3 no

nres

pond

ers§

(n=

371)

(5

wee

ks)

FLX

60m

g FL

X 20

mg

Res

pons

e (≥

50%

in H

DR

S 21)

NN

T =

25 (6

.5-∞

) R

emis

sion

(HD

RS21≤7

) NN

T =

36 (7

.3-∞

) R

espo

nse

(CG

I-I ≤

2) N

NT

= 20

(6.5

-∞) D

rop-

out S

E N

NH

16 (8

.3-14

4)

Blin

ding

is u

ncle

ar. M

ore

SE in

FLX

60m

g (n

s).

Resp

onse

rate

in ‘p

lace

bo-g

roup

’ 40.

5%.

Fava

et a

l. (1

992)

45C

15

MD

D

Out

P

Ope

n tr

ial o

f no

nres

pond

ers§ to

8-12

w

eeks

of F

LX 2

0mg

(4

wee

ks)

FLX

40-8

0mg

(if

tole

rate

d)

- D

ecre

ase

in H

DR

S 17-

scor

es in

NR

(-6.

2) a

nd P

R

(-10

.1) (

p<0.

05) D

ecre

ase

in C

GI-S

NR

(-0.

9; n

.s.)

an

d PR

(-2.

0 p<

0.05

)

Hig

hly

sele

cted

pop

ulat

ion

(ter

tiary

car

e). N

o pl

aceb

o co

ntro

l. Li

mite

d po

wer

.

Fava

et a

l. (1

994)

46B

41

MD

D

Sett

ing?

RCT

of n

onre

spon

ders

§ to

8 w

eeks

of F

LX

20m

g (4

wee

ks)

FLX4

0-60

mg

(if

tole

rate

d)

FLX

20m

g +

DES

25-

50m

g FL

X 20

mg

+ Li

30

0-60

0mg

Rem

issi

on (H

DRS

17≤7

) NN

T all =

4 (1

.6-∞

) N

NH

NR

= 6

(1.5

-∞) (

vs. L

i) N

NT P

R =

2 (0

.9-2

.8)

(vs.

Li)

Dro

p-ou

t SEN

NH

= 6

(2.6

-∞) (

vs. L

i)

Lim

ited

pres

enta

tion

of s

tudy

-pop

ulat

ion.

No

plac

ebo

cont

rol.

Lim

ited

pow

er, e

spec

ially

in

subg

roup

ana

lyse

s.

Fava

et a

l. (2

002)

49B

101

MD

D

Out

P

RCT

of n

onre

spon

ders

§ to

8 w

eeks

of F

LX

20m

g (4

wee

ks)

FLX4

0-60

mg

(if

tole

rate

d)

FLX

20m

g +

DES

25-

50m

g FL

X 20

mg

+ Li

30

0-60

0mg

Rem

issi

on (H

DRS

17≤7

) NN

T all =

6 (2

.4-∞

) (vs

. Li)

NN

T NR

= 5

(2.0

-∞) (

vs. L

i) N

NT P

R =

6 (2

.0-∞

) (vs

. Li

) Dro

p-ou

t SE

NA

No

plac

ebo

cont

rol.

Lim

ited

pow

er, e

spec

ially

in

sub

grou

p an

alys

es.

Lich

t et a

l. (2

002)

50A

2 16

29

MD

D

Out

P

RCT

of w

eek

6 no

nres

pond

ers§

(n=

295)

(5

wee

ks)

SER

200

mg

SER

100

mg

+ PL

AC

Res

pons

e (≥

50%

in H

DRS

17) N

NH

=7

(3.6

-74.

4)

Res

pons

e (C

GI-I

≤2)

NN

H =

6 (3

.4-16

.4)

Rem

issi

on (H

DRS

17≤7

) NN

H =

12 (4

.5-∞

) Dro

p-ou

t SE

not s

ign.

diff

eren

t

Dos

age

SER

was

incr

ease

d fr

om 5

0mg

to

100m

g 2

wee

ks p

rior t

o ra

ndom

izat

ion.

R

espo

nse

rate

in p

lace

bo-g

roup

70.

4%.

Schw

eize

r et a

l. (1

990)

44B

108

MD

D

Out

P

RCT

of w

eek

3 no

nres

pond

ers§

(n=

77)

(5 w

eeks

)

FLX

60m

g FL

X 20

mg

Res

pons

e (≥

50%

in H

DR

S 17)

NN

T =

82 (4

.2-

∞)D

rop-

out S

E N

NH

= 9

(3.9

-∞)S

E in

crea

sed

in

FLX

60m

g (t

rend

)

Gen

eral

izat

ion

to o

ther

SSR

Is m

ight

be

diff

icul

t be

caus

e of

long

hal

f-life

of F

LX a

nd it

s m

etab

olite

. Res

pons

e ra

te in

pla

cebo

-gro

up

51.2

%.

Schw

eize

r et a

l. (2

001)

48B

91

MD

D

Out

P

RCT

of w

eek

3 no

n-re

mitt

ers¶

(n

= 75

) (5

wee

ks)

SER

150m

g SE

R 50

mg

Rem

issi

on (H

DRS

17≤8

) NN

T 7

(2.7

-∞) R

espo

nse

(CG

I-I ≤

2) N

NT

5 (2

.3-6

1.5)

Dro

pout

NN

H =

34

(8.5

-∞) S

E bo

th in

crea

sed

or d

ecre

ased

(t

rend

s) in

SER

150m

g

Stud

y is

mar

gina

lly d

escr

ibed

. Blin

ding

unc

lear

. Re

mis

sion

-rat

e in

pla

cebo

-gro

up 3

2%.


4

65

4.2b

. Sys

tem

atic

revi

ews

and

met

a-an

alys

es

Bake

r et a

l. (2

002)

53A

2 11

02 +

57

3 M

DD

Se

ttin

g?

Met

a-an

alys

1s o

f 1) 4

fix

ed d

ose

RCT

s (3

-7

wee

ks) a

nd 2

) 4 d

ose-

esca

latio

n RC

Ts (3

-5

wee

ks) o

f NR

in w

eek

3-8.

(SSR

Is o

nly)

.

1) M

ediu

m /

Hig

h do

se 2

) D

ose-

esca

latio

n

1) L

ow d

ose

2)

addi

ng P

LAC

(or L

i/DES

)

Incr

ease

in R

espo

nse

rate

(≥50

% in

HD

RS)

acro

ss d

ose-

rang

e: 1)

ITT:

-9.5

% (n

.s.)

DT:

7.8

% (p

<0.0

1) 2

) ITT

: 6%

(n.s

.). D

T: 9

.3%

(n.s

.) C

hang

e of

HD

RS-d

ecre

ase

acro

ss d

ose-

rang

e: 1)

ITT:

-2.

0 (p

<0.0

001)

. DT:

una

vaila

ble.

2) I

TT: 1

.93

(p<0

.01)

. DT:

una

vaila

ble.

Dat

e of

sys

tem

atic

sea

rch

not g

iven

. No

appr

aisa

l of s

tudi

es. H

eter

ogen

eity

of d

ose-

esca

latio

n st

udie

s ig

nore

d in

met

a-an

alys

is.

Dos

e-re

spon

se re

latio

nshi

p ca

lcul

ated

as

regr

essi

on s

lope

usi

ng S

SRI m

g eq

uiva

lent

s (S

ME)

. Low

sta

ndar

d-do

se (5

mg)

for F

LX u

sed

as re

fere

nce

for S

ME.

ITT

popu

latio

n us

ed to

es

timat

e ex

pres

sed

dose

-resp

onse

re

latio

nshi

p. P

oten

tial d

ose-

resp

onse

re

latio

nshi

p es

timat

ed in

a d

ose-

tole

rant

(DT)

sa

mpl

e, o

mitt

ing

drop

outs

due

to s

ide

effe

cts

36.

Bolli

ni e

t al.

(199

9)51

A2

5844

M

DD

In

P &

Out

P

Met

a-an

alys

is o

f 33

RCTs

(197

5-19

97) w

ith

vario

us

antid

epre

ssan

ts. (

3-15

6 w

eeks

)

Hig

her d

oses

IM

I-equ

ival

ent

201-

250

(a) a

nd

>250

mg

(b)

Ave

rage

dai

ly

dose

(IM

I-eq

uiva

lent

100-

200m

g)

Effi

cacy

in IT

T an

alys

is: 5

3.3%

(ref

), 4

6.3%

(a),

48.3

% (b

) Com

plet

ers

anal

ysis

: 69%

(ref

), 6

7.3%

(a

), 7

6% (b

) Dro

pout

rate

s: 2

2% (r

ef),

28%

(a),

35

% (b

). S

E: 3

0% (r

ef),

36%

(a),

48%

(b)

Met

a-an

alys

is u

sing

regr

essi

on m

odel

s. H

ighl

y he

tero

geno

us s

tudi

es (i

.e. d

esig

ns) p

oole

d, n

o se

para

tion

of v

ario

us a

ntid

epre

ssan

t cla

sses

, no

nsys

tem

atic

bia

s es

peci

ally

for S

SRIs

by

conv

ersi

on to

IMI-e

quiv

alen

ts.

Corr

uble

et a

l. (2

000)

52C

? MD

D

Sett

ing

?

Revi

ew o

f lite

ratu

re o

f RC

Ts in

vest

igat

ing

dose

-res

pons

e re

latio

n fo

r SSR

Is

Hig

h do

se S

SRI

Stan

dard

dos

e Q

ualit

ativ

e de

scrip

tion

of d

ose-

resp

onse

re

latio

nshi

p pe

r dru

g. O

nly

for C

IT a

nd F

LX

poss

ibly

cur

vilin

ear,

oth

er S

SRIs

flat

dos

e-re

spon

se re

latio

nshi

p.

Sear

ch o

f MED

LIN

E on

ly. N

o ap

prai

sal o

f st

udie

s. N

o po

olin

g or

exp

lora

tion

of

diff

eren

ces

betw

een

iden

tifie

d st

udie

s an

d do

se-re

spon

se re

latio

nshi

p-da

ta p

er d

rug.

*All

dosa

ges

in m

g/da

y. †

Inte

ntio

n to

trea

t (IT

T)-r

esul

ts u

nles

s sp

ecif

ied.

‡ CG

I-eff

icac

y in

dex:

min

imal

or n

o ch

ange

in d

epre

ssio

n w

ith n

o or

non

-inte

rfer

ing

side

eff

ects

. § <

50%

redu

ctio

n in

HD

RS-

scor

e.

¶ H

DR

S 17

>8

Abb

revi

atio

ns: C

GI-I

/S =

Clin

ical

glo

bal i

mpr

essi

on im

prov

emen

t/se

verit

y, D

T =

Dos

e to

lera

nt s

ampl

e, H

DR

S xx

= H

amilt

on d

epre

ssio

n ra

ting

scal

e (x

x de

note

num

ber o

f ite

ms

used

), In

P =

Inpa

tient

s, IT

T =

Inte

ntio

n to

trea

t, M

DD

= M

ajor

dep

ress

ive

diso

rder

, Min

D=

Min

or d

epre

ssio

n, N

/A =

not

app

licab

le, N

NH

/T =

Num

ber n

eede

d to

har

m/t

reat

, NR

= N

onre

spon

ders

, Out

P =

Out

patie

nts,

PR

= Pa

rtia

l re

spon

ders

, RCT

= R

ando

miz

ed c

ontr

olle

d tr

ial,

SE =

sid

e ef

fect

s.D

rugs

: CIT

= c

italo

pram

, FLX

= fl

uoxe

tine,

IMI =

imip

ram

ine,

PA

R =

par

oxet

ine,

PLA

C =

plac

ebo,

SER

= s

ertr

alin

e, T

CA =

Tric

yclic

ant

idep

ress

ants

.

Tabl

e 4.

2. E

ffec

tiven

ess

of in

crea

sing

the

dose

: Sel

ecte

d st

udie

s. (C

ontin

ued)

Stud

y (y

ear)

refe

renc

eLo

E N

D

esig

n (f

ollo

w-u

p)

Inte

rven

tion*

Com

paris

on*

Out

com

e†Re

mar

ks

Chap

ter

Chapter 4

4

66

Benkert et al. investigated dose-escalation of paroxetine 20 mg/day in outpatients that weredepressed or had minor depression. Those who did not respond after 3 weeks of treatment (n=86) were randomized to receive 40 mg paroxetine for 3 additional weeks or placebo-increase.Response rates were 75% in the placebo-increased group and 74% in the 40 mg group.47 Licht et al.investigated randomized dose-escalation of sertraline (up to 200 mg/day) versus sertraline 100mg/day (placebo-increase) or mianserin addition in 295 outpatiens non-responsive to sertraline 50mg for 4 weeks and additionally increased to 100 mg for 2 more weeks. Response rates 5 weeksafter randomization were significantly lower in the dose-increase group (56%) than in thesertraline 100 mg group (70%) and the mianserin-addition group (67%). Drop-outs due to sideeffects were not specified.50

Strengths, flaws and other details of all selected studies are provided in table 4.2. In summarywe mention several methodological problems we encountered: absence of placebocontrols,45;46;49 inclusion of minor depression,47 insufficient data-presentation,44;46 insufficientpower,44-49 uncertainty about blinding,43;48 earlier dose-escalation before the randomisation,50

inadequate pooling of heterogeneous data51;53 and problems with conversion to dose-equivalents.51;53 None of the studies provided information about the method of dose-escalationnor described the early drop-out rates due to dose-escalation.

Evidence for dose-escalation?From 4 of the 8 dose-escalation studies it appeared that dose-increments before 4 weeks werenot effective (LoE: A2).43;44;47;48;51-53 However, in the meta-analysis of some of these studies byBaker et al., a potential dose-response relationship was found for dose-escalation if dropouts dueto side effects were excluded from the analysis (a so called dose-tolerant sample).53 Baker et al.proposed that differential dropout due to side effects in the dose-escalation group (compared toplacebo-increase) gave a substantial (negative) bias of the potential dose-response relationship.They argued that by applying a last-observation carried forward approach (often used in theoriginal studies), more early drop-outs (due to side-effects) in the high-dose groups wouldunequally increase average severity scores and decrease response rates compared to the lowerdose (or placebo) groups. This methodological problem could be overcome by analysing onlydose-tolerant subjects (those not dropping out due to side effects).36

In the well performed study with sertraline by Licht et al. (not included in the three reviews),dose-escalation after 6 weeks was found to be less effective than continuation of the standarddose, or augmentation with mianserin (LoE: A2).50 After 8 weeks of treatment, increased dosagesof fluoxetine were more effective than augmentation with lithium or desipramine,46;49 although inthe latter study this was not significant (LoE: B). In these studies no placebo dose-escalation wasperformed. Both studies showed a non-significant trend of increased efficacy of dose-escalationcompared to augmentation (lithium or desipramine) especially for partial responders(LoE: B).46;49

Across all studies, higher doses were related to increased dropout rates, which wereassociated with more side effects in some studies (LoE: A2).51 It appeared that the occurrence ofside effects did not increase equally when dosages were gradually escalated in initial non-responders compared to fixed-dose trials. However, this could not be comparedstraightforwardly between the studies, and was not investigated specifically.

Additional concerns for cliniciansWe identified no evidence to recommend on how dose-increase should be practiced. Also themaximum dosage to be achieved was not investigated well.


4

67

Discussion

Our systematic review provided 8 studies about dose-escalation in SSRIs. Only one of theseapproached our rather stringent criteria.50 We found no evidence for increased efficacy of dose-escalation within the first 4 weeks. Dose-escalation after 6 weeks appeared less effective thancontinuing the same dose. We found some, but limited evidence for efficacy of dose-escalationafter 8 weeks, particularly in partial responders. This effect was seen within 4 weeks after dose-escalation. Irrespective of efficacy, dose-escalation unequivocally increased side-effects, buteffects on drop-out rates due to side effects were less straightforward. Thus, in the absence ofmethodologically well designed studies we can neither unequivocally state that dose-escalation isuseful, nor discard it as useless.

These findings may challenge the current beliefs and recommendations about dose-escalation, underscored by the fact that dose-escalation is generally practiced.31-34 Contrarily tothis challenge, many patients that only partially responded are too often treated with long-termobviously insufficient treatments (e.g. standard doses of SSRIs). For these patients, one couldargue that it is better to try dose-escalation than to continue inadequate treatment. Presumably,in the absence of clear guidance by trial-data, clinicians do not have many alternatives for non-responders or partial responders, and clinicians all have their case-histories of improvement afterdose-escalation. A more sophisticated question must therefore also be asked: which subgroup ofpatients will benefit from dose-escalation?

So far, only the NICE guideline displayed some reserve in the general recommendation aboutdose-escalation.8 The British Medicines and Healthcare products Regulatory Agency’s Committeeon Safety of Medicines examined the available evidence for dose-escalation as provided bypharmaceutical companies, and recommended the lowest efficacious dose (MHRA, 2004Internet: www.mhra.gov.uk). From this report it was unclear which studies were taken asevidence. Three previous reviews concerning higher doses of antidepressants were published,51-53

of which the methodological shortcomings were already mentioned. The findings in thesereviews previously challenged the belief of a dose-response relationship, but Baker et al.proposed a potential dose-response relationship, based on their dose-tolerant analysis. All reportssummarized studies performed until 1997, thereafter the study by Licht et al. further challengedthe efficacy of dose-escalation.50

Limitations of the identified studiesFour major issues of concern in the 8 identified studies should be mentioned. First, themethodological quality of these studies varied between poor and good according to ourclassification. We summarized these methodological problems in the results section and table 4.2.

Second, and more in general, all dose-escalation studies (except the studies of Fava andcollegues, that lacked a placebo-control)45;46;49 suffered a methodological problem of the timingof dose-escalation.36 Even the most robust study by Licht et al. hampered its own design by anonrandomized dose-increase 2 weeks prior to randomization.50 This problem might explain thehigh placebo response rates in some of the dose-escalation studies (up to 75%).

Third, in most studies no data were provided on the selective drop-out, nor the schedule ofdose-increments.36 Because patients randomised to true dose-escalation might drop out morefrequently and earlier after randomisation (with associated high severity scores) compared tothose receiving placebo, this might have biased the intention to treat analyses in which last-observations are usually carried forward to study endpoints. This especially happens when dose-escalation is performed rapidly. The analysis of a dose-tolerant sample in such studies wouldindeed provide additional information, but these data were not provided.

Fourth, in the selected trials, mostly response was used as primary outcome, while currentlyremission of depression is the clinical aim of treatment.54 If dose-escalation would be effective,the question remains whether dose-escalation will also further improve initial responders thatwere nonremitters. So far only Schweizer et al. addressed this topic with equivocal effects ofdose-escalation.48

Chap

ter

Chapter 4

4

68

Possible explanations for a dose-response relationshipA possible explanation of the clinical observation that response might occur after dose-escalation,can be initial lower blood levels. This may be related to increased metabolism due to geneticpolymorphisms of the cytochrome P450 (CYP P450) enzyme system.55-58 The incidence ofincreased metabolism by (multi-)duplicated genes of the CYP P450 2D6 genes varies between 1-2%in Swedish Caucasians, 3.6% in Germany and 7-10% in Spain and Sicily, and varies between ethnicgroups (e.g. 29% in Black Ethiopians).59 A few studies showed equivocal evidence for theinvolvement of CYP P450 polymorphisms (responsible for rapid metabolism) as an explanation ofnon-response to a standard dose of SSRIs.56;58-61 However, a clear relationship between bloodlevels of SSRIs and response was never found.26;62-66 Perhaps, genetic variability of the centraltarget of these drugs, the serotonin reuptake transporter, may be responsible more directly forthe effects of SSRIs.67;68

From in vitro and ex-vivo studies it appears that at higher doses selective antidepressants likeSSRIs may become dual action agents that will also affect other monoamine systems likenorepinephrine.69-71 From the current data of dose-escalation in SSRIs this theoretical hypothesiscannot be falsified nor proven. In addition, we are unaware of an acceptable method to testwhether specific sites of action are responsible for the observed treatment-effects.

Limitations of the review No meta-analysis was performed because the differences in timing of dose-escalation betweenthe identified studies introduced substantial heterogeneity. An extension of the meta-regressionapproach as performed by Baker et al.53 was considered impossible to acknowledge this problemas the number of studies to explore this heterogeneity will lack power, moreover as age, sex,outcome definition and type of SSRI ideally should also be included in such a model.

The grading system for studies is not representing the appraised methodological dimensionsof evidence. This improved the applicability of the results for busy clinicians, but reduced theirstrength.

Finally, patients studied in trials are generally selected populations, reducing external validityfor the 'real world' clinical practice. All identified studies excluded psychotic depression, bipolardepression, depression in children or adolescents, and depressive disorder with severe psychiatricand somatic comorbidity.

Future dose-escalation studies For future dose-escalation trials methodological issues should be considered. First, for optimalcontrast in the study, an appropriate group of non-responders should be selected by postponingrandomization and refraining from (additional) interventions before dose-escalation is applied.Six weeks is the minimum which can be reconciled with recommendations in current guidelinesand is acceptable for clinical practice. Second, studies should have enough power to detectsignificant differences. This implies a large sample to start with, as approximately 50% of patientswill show a response in the first 6 weeks. Third, the method of dose-escalation should bedescribed and applied in a way that few patients drop-out. Fourth, adequate results should bepresented: response and remission rates in intention to treat analyses and for the group thatcould be described as dose tolerant. Fifth, efficacy should be tested in predefined subgroups (e.g.partial responders at week 6). Sixth, genetic sampling (e.g. CYP P450 and SERT polymorphisms)and plasma SSRI level sampling would be interesting to further examine potential explanationsfor the clinically observed efficacy of dose-escalation and to identify potential prognosticvariables.

Clinical implications and conclusionFor clinicians this review may challenge the belief in dose-escalation of SSRIs for patients whoshow insufficient response to a standard dose. This review points out that dose-escalation ofSSRIs in a standard dose before the fourth week of treatment does not seem to be more


4

69

beneficial for patients. Thereafter dose-escalation is one of the easiest thing to do, but – to ouropinion – should not be performed obligatory, as the evidence for this strategy is not compelling.Contrary dose-escalation appears more adequate to try than to continue an inadequatetreatment, maybe partial responders will benefit most. Side-effects will most likely increase andmight induce drop-out of treatment. For more clinical guidance well-performed new dose-escalation studies are needed. In order to fill this gap, for the introduction and registration of(new) antidepressants, clinical studies with an adequate dose-escalation design should berequired by the U.S. Food and Drug Administration (FDA) and the European Agency for theEvaluation of Medicinal Products (EMEA).

AcknowledgementsThis systematic review was realised by a grant for the development of a local evidence-basedclinicalpractice guideline (No SFA.07.012) from the Academic Medical Center, Amsterdam, theNetherlands. This guideline considering strategies for non-response to a standard dose of a firstSSRI is available on request.

Conflicts of interestNone

References









9. Thase ME, Rush AJ. Treatment-Resistant Depression. In: Bloom FE, Kupfer DJ, eds. Psychopharmacology: The FourthGeneration of Progress. New York: Raven Press Ltd.; 1995:1081-1097.


11. Anderson IM. Drug treatment of depression: Reflections on the evidence. Adv Psychiatr Treat. 2003; 9: 11-20.12. Kennedy N, McDonough M. Pharmacological management of treatment resistant depression: A clinical review. Irish J

Psychol Med. 2003; 20: 18-23.13. Nelson JC. Managing treatment-resistant major depression. J Clin Psychiatry. 2003; 64 Suppl 1: 5-12.14. McIntyre RS, Muller A, Mancini DA, Silver ES. What to do if an initial antidepressant fails? Can Fam Physician. 2003; 49: 449-

457.15. Hirschfeld RM, Montgomery SA, Aguglia E, Amore M, Delgado PL, Gastpar M et al. Partial response and nonresponse to

antidepressant therapy: current approaches and treatment options. J Clin Psychiatry. 2002; 63: 826-837.16. Kennedy S, McIntyre R, Fallu A, Lam R. Pharmacotherapy to sustain the fully remitted state. J Psychiatry Neurosci. 2002;

27: 269-280.17. Marangell LB. Switching antidepressants for treatment-resistant major depression. J Clin Psychiatry. 2001; 62 Suppl 18: 12-

17.18. Trivedi MH, Kleiber BA. Algorithm for the treatment of chronic depression. J Clin Psychiatry. 2001; 62 Suppl 6: 22-29.19. Fava M. New approaches to the treatment of refractory depression. J Clin Psychiatry. 2000; 61: 26-32.20. Fava M. Management of nonresponse and intolerance: Switching strategies. J Clin Psychiatry. 2000; 61: 10-12.

Chap

ter

Chapter 4

4

70

21. O'Reardon JP, Brunswick DJ, Amsterdam JD. Treatment-resistant depression in the age of serotonin: Evolving strategies.Curr Opin Psychiatry. 2000; 13: 93-98.

22. Crismon ML, Trivedi M, Pigott TA, Rush AJ, Hirschfeld RMA, Kahn DA et al. The Texas medication algorithm project:Report of the Texas consensus conference panel on medication treatment of major depressive disorder. J Clin Psychiatry.1999; 60: 142-156.

23. Nelson JC. Treatment of antidepressant nonresponders: augmentation or switch? J Clin Psychiatry. 1998; 59 Suppl 15: 35-41.

24. Thase ME, Rush AJ. When at first you don't succeed: sequential strategies for antidepressant nonresponders. J ClinPsychiatry. 1997; 58: 23-29.

25. Stimpson N, Agrawal N, Lewis G. Randomised controlled trials investigating pharmacological and psychologicalinterventions for treatment-refractory depression. Systematic review. Br J Psychiatry. 2002; 181: 284-294.

26. Beasley CM, Jr., Bosomworth JC, Wernicke JF. Fluoxetine: relationships among dose, response, adverse events, andplasma concentrations in the treatment of depression. Psychopharmacol Bull. 1990; 26: 18-24.

27. Altamura AC, Montgomery SA, Wernicke JF. The evidence for 20mg a day of fluoxetine as the optimal dose in thetreatment of depression. Br J Psychiatry Suppl. 1988; 109-112.

28. Tignol J, Stoker MJ, Dunbar GC. Paroxetine in the treatment of melancholia and severe depression. Int ClinPsychopharmacol. 1992; 7: 91-94.

29. Dunner DL, Dunbar GC. Optimal dose regimen for paroxetine. J Clin Psychiatry. 1992; 53 Suppl: 21-26.30. Montgomery SA, Pedersen V, Tanghoj P, Rasmussen C, Rioux P. The optimal dosing regimen for citalopram--a meta-

analysis of nine placebo-controlled studies. Int Clin Psychopharmacol. 1994; 9 Suppl 1: 35-40.31. Byrne S, Rothschild AJ. Psychiatrists' responses to failure of maintenance therapy with antidepressants. Psychiatr Serv.

1997; 48: 835-837.32. Fredman SJ, Fava M, Kienke AS, White CN, Nierenberg AA, Rosenbaum JF. Partial response, nonresponse, and relapse

with selective serotonin reuptake inhibitors in major depression: a survey of current "next-step" practices. J ClinPsychiatry. 2000; 61: 403-408.

33. Mischoulon D, Nierenberg AA, Kizilbash L, Rosenbaum JF, Fava M. Strategies for managing depression refractory toselective serotonin reuptake inhibitor treatment: a survey of clinicians. Can J Psychiatry. 2000; 45: 476-481.

34. Shergill SS, Katona CL. Pharmacological choices after one antidepressant fails: a survey of UK psychiatrists. J AffectDisord. 1997; 43: 19-25.

35. Mischoulon D. Why do antidepressants take so long to work? Am Soc Clin Psychopharmacology Progress Notes. 1997; 8: 9-11.

36. Baker CB, Woods SW. Is there a SSRI dose response in treating major depression? The case for re-analysis of current dataand for enhancing future study design. Depress Anxiety. 2003; 17: 10-18.

37. Quitkin FM, Petkova E, McGrath PJ, Taylor B, Beasley C, Stewart J et al. When should a trial of fluoxetine for majordepression be declared failed? Am J Psychiatry. 2003; 160: 734-740.

38. Munoz SR, Bangdiwala SI. Interpretation of Kappa and B statistics measures of agreement. J Applied Stat. 1997; 24: 105-111.

39. Richtlijnontwikkeling binnen het Kwaliteitsinstituut voor de Gezondheidszorg CBO. Handleiding voor werkgroepleden.Kwaliteitsinstituut voor de Gezondheidszorg CBO.2000. Utrecht, Kwaliteitsinstituut voor de Gezondheidszorg CBO.

40. SIGN 50: A guideline developers' handbook. 50. 2001. Edinburgh, Scottish Intercollegiate Guideline Network. 41. Sackett DL, Straus SE, Richardson WS, Rosenberg W, Haynes RB. Evidence-Based Medicine. How to practice and teach EBM.

2 ed. Edinburgh: Churchill Livingstone; 2000.42. Moncrieff J, Churchill R, Drummond DC, McGuire H. Development of a quality assessment instrument for trials of

treatments for depression and neurosis. Int J Methods Psychiatr Res. 2001; 10: 126-133.43. Dornseif BE, Dunlop SR, Potvin JH, Wernicke JF. Effect of dose escalation after low-dose fluoxetine therapy.

Psychopharmacol Bull. 1989; 25: 71-79.44. Schweizer E, Rickels K, Amsterdam JD, Fox I, Puzzuoli G, Weise C. What constitutes an adequate antidepressant trial for

fluoxetine? J Clin Psychiatry. 1990; 51: 8-11.45. Fava M, Cohen L, Rosenbaum JF, Reiter S, et a. High-dose fluoxetine in the treatment of depressed patients not

responsive to a standard dose of fluoxetine. J Affect Disord. 1992; 25: 229-234.46. Fava M, Rosenbaum JF, McGrath PJ, Stewart JW, Amsterdam JD, Quitkin et al. Lithium and tricyclic augmentation of

fluoxetine treatment for resistant major depression: a double-blind, controlled study. Am J Psychiatry. 1994; 151: 1372-1374.

47. Benkert O, Szegedi A, Wetzel H, Staab HJ, Meister W, Philipp M. Dose escalation vs. continued doses of paroxetine andmaprotiline: a prospective study in depressed out-patients with inadequate treatment response. Acta Psychiatr Scand.1997; 95: 288-296.

48. Schweizer E, Rynn M, Mandos LA, DeMartinis N, Garcia-Espana F, Rickels K. The antidepressant effect of sertraline is notenhanced by dose titration: results from an outpatient clinical trial. Int Clin Psychopharmacol. 2001; 16: 137-143.

49. Fava M, Alpert J, Nierenberg A, Lagomasino I, Sonawalla S, Tedlow J et al. Double-blind study of high-dose fluoxetineversus lithium or desipramine augmentation of fluoxetine in partial responders and nonresponders to fluoxetine. J ClinPsychopharmacol. 2002; 22: 379-387.

50. Licht RW, Qvitzau S. Treatment strategies in patients with major depression not responding to first-line sertralinetreatment: A randomised study of extended duration of treatment, dose increase or mianserin augmentation.Psychopharmacology (Berl). 2002; 161: 143-151.

51. Bollini P, Pampallona S, Tibaldi G, Kupelnick B, Munizza C. Effectiveness of antidepressants. Meta-analysis of dose-effectrelationships in randomised clinical trials. Br J Psychiatry. 1999; 174: 297-303.


4

71

52. Corruble E, Guelfi JD. Does increasing dose improve efficacy in patients with poor antidepressant response: a review.Acta Psychiatr Scand. 2000; 101: 343-348.

53. Baker CB, Tweedie R, Duval S, Woods SW. Evidence that the SSRI dose response in treating major depression should bereassessed: A meta-analysis. Depress Anxiety. 2003; 17: 1-9.

54. Nierenberg AA, DeCecco LM. Definitions of antidepressant treatment response, remission, nonresponse, partialresponse, and other relevant outcomes: a focus on treatment-resistant depression. J Clin Psychiatry. 2001; 62 Suppl 16: 5-9.

55. Bertilsson L, Aberg-Wistedt A, Gustafsson LL, Nordin C. Extremely rapid hydroxylation of debrisoquine: a case report withimplication for treatment with nortriptyline and other tricyclic antidepressants. Ther Drug Monit. 1985; 7: 478-480.

56. Brosen K. Some aspects of genetic polymorphism in the biotransformation of antidepressants. Therapie. 2004; 59: 5-12.57. Charlier C, Broly F, Lhermitte M, Pinto E, Ansseau M, Plomteux G. Polymorphisms in the CYP 2D6 gene: association with

plasma concentrations of fluoxetine and paroxetine. Ther Drug Monit. 2003; 25: 738-742.58. Steimer W, Muller B, Leucht S, Kissling W. Pharmacogenetics: a new diagnostic tool in the management of antidepressive

drug therapy. Clin Chim Acta. 2001; 308: 33-41.59. Bertilsson L, Dahl ML, Dalen P, Al Shurbaji A. Molecular genetics of CYP2D6: clinical relevance with focus on psychotropic

drugs. Br J Clin Pharmacol. 2002; 53: 111-122.60. Kawanishi C, Lundgren S, Agren H, Bertilsson L. Increased incidence of CYP2D6 gene duplication in patients with

persistent mood disorders: ultrarapid metabolism of antidepressants as a cause of nonresponse. A pilot study. Eur J ClinPharmacol. 2004; 59: 803-807.

61. Bertilsson L, Dahl ML, Tybring G. Pharmacogenetics of antidepressants: clinical aspects. Acta Psychiatr Scand Suppl. 1997;391: 14-21.

62. Amsterdam JD, Fawcett J, Quitkin FM, Reimherr FW, Rosenbaum JF, Michelson D et al. Fluoxetine and norfluoxetineplasma concentrations in major depression: a multicenter study. Am J Psychiatry. 1997; 154: 963-969.

63. Baumann P. Pharmacokinetic-pharmacodynamic relationship of the selective serotonin reuptake inhibitors. ClinPharmacokinet. 1996; 31: 444-469.

64. Bourdeaux R, Pannetier P, Younos C, Desor D, Lehr PR, Capolaghi B. Fluoxetine : Relationships among plasmaconcentrations and therapeutic effects in the treatment of 32 patients with major depressive disorder at 20 mg/day.Encephale. 1998; 24: 57-61.

65. DeVane CL. Translational pharmacokinetics: Current issues with newer antidepressants. Depress Anxiety. 1998; 8: 64-70.66. Norman TR, Gupta RK, Burrows GD, Parker G, Judd FK. Relationship between antidepressant response and plasma

concentrations of fluoxetine and norfluoxetine. Int Clin Psychopharmacol. 1993; 8: 25-29.67. Hahn MK, Blakely RD. Monoamine transporter gene structure and polymorphisms in relation to psychiatric and other

complex disorders. Pharmacogenomics J. 2002; 2: 217-235.68. Smits KM, Smits LJ, Schouten JS, Stelma FF, Nelemans P, Prins MH. Influence of SERTPR and STin2 in the serotonin

transporter gene on the effect of selective serotonin reuptake inhibitors in depression: a systematic review. MolPsychiatry. 2004; 9: 433-441.

69. Gorman JM, Sullivan G. Noradrenergic approaches to antidepressant therapy. J Clin Psychiatry. 2000; 61 Suppl 1: 13-16.70. Gilmor ML, Owens MJ, Nemeroff CB. Inhibition of norepinephrine uptake in patients with major depression treated with

paroxetine. Am J Psychiatry. 2002; 159: 1702-1710.71. Owens MJ, Morgan WN, Plott SJ, Nemeroff CB. Neurotransmitter receptor and transporter binding profile of

antidepressants and their metabolites. J Pharmacol Exp Ther. 1997; 283: 1305-1322.

Chap

ter

Chapter 4

4

72

CHA

PTER

SWITCHING ANTIDEPRESSANTS AFTER A FIRST

SELECTIVE SEROTONIN REUPTAKE INHIBITOR IN

MAJOR DEPRESSIVE DISORDER: A SYSTEMATIC

REVIEW

Journal of Clinical Psychiatry 2006; 67: 1836-1855

Henricus G. Ruhé1

Jochanan Huyser1

Jan A. Swinkels1

Aart H. Schene1

1 Program for Mood Disorders, Department of Psychiatry, Academic Medical Center, University ofAmsterdam, Amsterdam, The Netherlands

Chap

ter

Chapter 5

5

74

Abstract

BackgroundSelective Serotonin Reuptake Inhibitors (SSRIs) are frequently used as a first antidepressant formajor depressive disorder, but have response-rates of 50-60% in daily practice. For patients withinsufficient response to SSRIs, switching is often applied.

AimTo systematically review the evidence for switching pharmacotherapy after a first SSRI.

MethodsA systematic literature search (updated until Feb. 10, 2005) in MEDLINE, EMBASE, CINAHL andPsycInfo (all indexed years) identified randomised controlled trials (RCTs) and open studiesinvestigating switching strategies. In the absence of specific keywords for switching, weperformed “sensitive” searches using free text words with w0ildcards ($): “switch$” or(“alternat$” adj5 “treat$”) or (“alternat$” adj5 “therap$”) in combination with the CochraneCollaboration search filter for RCTs, the Cochrane Collaboration Depression Anxiety and NeurosisGroup search filter for major depressive disorder, and MeSH terms for antidepressants (incombination with additional text words for all antidepressive agents). Additionally, we included 4recent STAR*D publications. We limited searches to adults and humans but did not applylanguage restrictions. Relevant articles were retrieved and critically appraised. The methodologyof the studies, the results on efficacy and dropout s due to side effects, and remarks weresummarized in an evidence table. Three studies comparing a switch tovenlafaxine or SSRIs werepooled.

ResultsEight RCTs and 23 open studies were identified, studying populations with different levels oftreatment resistance. Definitions of response and remission rates varied between studies.Observed response rates after switching to any of the classes of antidepressants varied between12% and 86%. Remission rates varied between 7% and 82%. The number of previous treatments withantidepressants was negatively correlated with treatment outcome. Rates of dropout due to side-effects varied considerably across agents (5-39%). Switching to venlafaxine showed a modest andclinically equivocal benefit over SSRIs (Number Needed to Treat = 13 (9.1-25.0))

ConclusionsAfter a first SSRI any switch within or between classes of antidepressants appear legitimate(second SSRI, novel dual acting antidepressants, selective noradrenergic or noradrenergic/dopaminergic agents, or TCA or mianserin). No unequivocal evidence is available to prove anadvantage of a between class switch. More guidance by randomized empirical studies is needed.Clinical implications and methodological considerations for future studies are discussed.

ChapterReview of switching in MDD

5

75

Introduction

Major depressive disorder (MDD) is one of the most prevalent and disabling illnesses inpsychiatry.1 For the treatment ofMDD, several national clinical guidelines were developed.2-9 Inthese guidelines, pharmacotherapy is among the most important treatments; mostly SelectiveSerotonin Reuptake Inhibitors (SSRIs) are the antidepressants of first choice. However, only 50 to60% of patients respond to the first antidepressant given.10;11 In a case of non-response, alltreatment guidelines recommend three major strategies: 1) increasing the dose of theantidepressant (dose-escalation), 2) switching to another antidepressant of the same or differentclass, and 3) augmenting the antidepressant by adding a second drug that by itself is not anantidepressant. By various authors, a fourth strategy of combination of antidepressants isproposed.12-15

Surprisingly very little systematic evidence exists to date to underscore the recommendationsfor non-responders. One Cochrane review summarizes randomized, controlled trials (RCTs) ofstrategies in patients non-responsive to at least 4 weeks of an antidepressant at therecommended dose.16 With a thorough methodology, 16 RCTs were selected. Unintentionally, thestudies included in this review represented more heterogeneous, difficult-to-treat populations,referred to as treatment-resistant depression (TRD). Although little information on previoustreatments was found, the included studies especially considered tricyclic antidepressant (TCA)non-responders. The switch-options that were investigated in the included studies did not reflectclinical practice of switching to another antidepressant (one of the above recommendations), butused a variety of other drugs (oestrogen, benzodiazepines, ketoconazole, olanzapine). For theaugmentation studies, meta-analyses were performed with 2 trials of lithium-augmentation and 3pindolol-trials. A clinically significant benefit was found only for lithium augmentation. Thus, thisreview does not provide helpful information for clinicians in the case of non-response to a (first)SSRI.

Strategies for non-response have been summarized in several narrative reviews, focusing onall strategies together,17-28 switching,12;13;29-32 augmentation,12;13;31;32 or combination.12-15;33 Doseescalation was summarized in 2 meta-analyses,34;35 1 narrative 36 and 1 recent systematic review.37

The evidence for lithium augmentation was also summarized in meta-analyses by Bauer et al.38;39

After dose-escalation, switching antidepressants is widely practiced.40-42 Switching to adifferent pharmacological class seems to be preferred by clinicians.43 The above narrative reviewsof switching strategies altogether provided a substantial overview. However, each reviewindividually was limited in its presentation, predominantly by a lack of a well defined searchstrategy, and none of the reviews presented data on critical appraisal of the identified studies asproposed by the Cochrane collaboration.44 The general conclusion today is that there is limitedevidence available for switching antidepressants, and that there is no clear proven advantage ofone switch option over the others. Additionally, recently the results of a large study designed toelucidate sequential treatment strategies after non-response became available (SequencedTreatment Alternatives to Relieve Depression, STAR*D).45 This study provided prospective dataon response and remission rates after randomized treatment allocations in patients who were notin remission after one to four sequential steps of treatment (further referred to as STAR*D level I-IV).

Therefore, our primary objective was to systematically review and appraise the availableresearch focusing on switching strategies for SSRI non-responders in MDD, including the recentSTAR*D results. A secondary aim was to acknowledge and investigate the expected differentlevels of TRD as a source of variation between studies. Our principal question was whether theavailable evidence justifies distinct recommendations for next-step strategies after non-responseto a first SSRI. We performed a systematic review following the Cochrane methodology andperformed a meta-analysis of two switch options after a first SSRI: a second SSRI versus aserotonin-norepinephrin reuptake inhibitor (SNRI).

Chap

ter

Chapter 5

5

76

Methods

Studies included in the reviewWe expected very few randomised, controlled, switch-studies a-priori, despite the widespreadavailability of SSRIs during the last decade. As best-available evidence, we included open andrandomized studies in which at least 50% of participants used an SSRI previously in the currentdepressive episode. Thus, we excluded studies describing switching from TCAs to SSRIs. Studiesperformed in populations with TRD were also included if previous use of an SSRI (in ≥50% ofsubjects) was unambiguously documented.

Identification and selection of articlesWe performed systematic literature searches (updated until February 10th 2005) in four databases(MEDLINE, EMBASE, CINAHL, and PsychInfo; all indexed years). In the absence of specifickeywords for switching, we performed ‘sensitive’ searches using free text words with wildcards($): “switch$” or (“alternat$” adj5 “treat$”) or (“alternat$” adj5 “therap$”) in combination withthe Cochrane Collaboration search-filter for RCTs, the Cochrane Collaboration Depression Anxietyand Neurosis group search-filter for MDD and MeSH-terms for antidepressants (in combinationwith additional text words for all antidepressive agents). We limited searches to adults andhumans, but did not apply language restrictions. Full queries are available on request. In addition,we included four identified studies released after these searches, including three studies from theSTAR*D trial.46-49

The first and second authors (H.G.R., J.H.) independently screened titles and abstracts andselected articles on the basis of design and focus on switching antidepressants after SSRI-treatment. Agreement on exclusion of irrelevant articles was 99.1%, with a Cohen’s κ for interrateragreement of 0.62 (κ values between 0.45-0.75 indicate 'substantial' agreement; values above0.75 indicate 'almost perfect' agreement).50 We resolved discrepancies between initial selectionby discussion and consensus.

The first author (H.G.R.) judged all potentially relevant articles according to specific inclusionand exclusion criteria (full criteria available on request). In case of doubt the article was fully readand assigned thereafter. We retrieved additional cross-references, and checked reference lists ofidentified narrative reviews. We considered double-publications together to reveal the maximumof available information.

Critical Appraisal and summary The first author (H.G.R.), a certified epidemiologist, critically appraised and abstracted thearticles, using standardized forms derived from the Dutch Institute of Healthcare Improvement51

and the Agency of Healthcare Policy and Research (AHCPR).5 We used the same items for criticalappraisal as proposed by the Scottish Intercollegiate Guidelines Network52 and Sackett et al.53 Weassigned a ‘level of evidence’ (LoE; see table 4.1) to each study.51 Levels of evidence are basedupon the methodological robustness of studies. In the results section, the LoE of the supportingscientific evidence (A1-D) is indicated. We extracted data on efficacy and tolerability from eachstudy. As primary efficacy outcome we took the percentage response or remission on anintention to treat (ITT) basis. If several scales were used, we a priori preferred data for theHamilton depression rating scale (HDRS;54 17-item version or other versions); otherwise we useddata from the Montgomery Åsberg depression rating scale (MADRS),55 clinical global impressionsscale (CGI),56 or other applied scales (e.g. the 16-item quick inventory for depressive symptoms-self-rated (QIDS-SR16).57 For tolerability, we took the dropout rate due to side effects as primarymeasure, followed by the overall dropout rate.

To assess judgement-bias by one person who performed the critical appraisal, we measuredinterrater variation in a slightly different set of 12 publications. Every other author (J.H., J.A.S.,A.H.S.) critically appraised 4 publications. Cohen's κ values for the appraisal-items were 0.49 (for'validity of the study') and 0.86 (for 'concealment of allocation'), while complete agreement


5

77

existed for the appraisal-items 'randomization of the study', 'level of evidence' and 'dataextraction' (κ= 1.0). These results are in line with other reports of interrater agreement inappraisal of psychiatric research.58

We first described a qualitative summary with discussion of the results, restrictions,methodological flaws and external validity of the studies in an evidence table and a separatedocument, of which a summary is provided in this article. For each study we indicated the level oftreatment resistance as proposed by Thase et al.10;24 If possible, we calculated risk-differences andcorresponding Numbers Needed to Treat (NNT) and Harm (NNH), with 95% confidence intervals(95% CIs). Because of the lack of homogenous randomised studies, we refrained from pooling in ameta-analysis, except for the three studies comparing the venlafaxine (SNRI) vs. second SSRIswitch. We grouped antidepressants into six classes following the classification of the AHCPR.5

Results

We selected thirty-one studies for this review. Figure 5.1 shows the search results and selection ofstudies. Table 5.1 summarizes the included studies. A table of 8 excluded studies59-66 is available onrequest.

Figure 5.1. Selection of reported studies.

Abbreviations: BUP = bupropion, MAO-I= irreversible inhibitor of monoamine-oxidase, MIAN = mianserin, MIR = mirtazapine,NEF = nefazodone, REB = reboxetine, RIMA = reversible inhibitor of MAO, SSRI = selective serotonin reuptake inhibitors, TCA= tricyclic antidepressant, VLX = venlafaxine

�� "#$%&"'"$#�(L�)L)&$*+&",�)$+-,.$)��

�� M�� M��

�� M�N��

��

�O�� !�� " #�O��$�� $��$��%��

�O�� &� �� &�� '�()��'�*�+��'�,-.��%�'�/�0 "��%�� &��M��'-��1�N�� %��2�'��$�!��

��/�� 345 �*�+��%��

��()� �/��*�� P�

� (N��$ P ��$��$��O�%�$��LQ ��2�O��$�N��$$��$�� $L$��$��$�

��, �,-.�� P�

��/�0 "��

��/��

�O��345 �/��Q

�O��,-.��%� Q �O��/�0"��%�Q

�O��()��%� Q�O�� Q

�O��3456/��%�Q

�O��()��%� Q�O�� Q

�O��3456/��%�Q

�O��()��%� Q�O�� Q

�O��/�0"��%�Q

�O��3456/��%�Q

�O��()��%� Q�O�� Q

�O��,-.��%� Q �O��/�0"��%�Q

�O��3456/��%�Q

�O��()��%� Q�O�� Q

�O��345 �/��Q

�O��,-.��%� Q �O��/�0"��%�Q

�O��3456/��%�Q

�O��()��%� Q�O�� Q

�O��345 �/��Q

�O�� Q �O��()��%� Q

�O��345 �/��Q

�O�� Q

�O��3456/��%�Q

�O��()��%� Q

�O��345 �/��Q

�O�� Q

�O��/�0"��%�Q

�O��3456/��%�Q

�O��()��%� Q

�O��345 �/��Q

�O�� Q

�O��,-.��%� Q �O��/�0"��%�Q

�O��()��%� Q

�O��345 �/��Q

�O�� Q�O�� Q �O��()��%� Q�O�� Q

�O��3456/��%�Q

�O��()��%� Q�O�� Q

�O��/�0"��%�Q

�O��3456/��%�Q

�O��()��%� Q�O�� Q

�O��,-.��%� Q �O��/�0"��%�Q

�O��3456/��%�Q

�O��()��%� Q�O�� Q

78

Chapter 5

Chap

ter

5

Tabl

e 5.

1. Ef

fect

iven

ess

of s

witc

hing

aft

er ≥

1 SSR

I: Se

lect

ed s

tudi

es.

Stud

y (y

ear)

refe

renc

eLo

E N

D

esig

n (f

ollo

w-u

p)

Inte

rven

tion*

Co

mpa

rison

* O

utco

me†

R

emar

ks

5.1.A

. to

2nd

SSRI

Bald

omer

o et

al.

(200

5)46

See

unde

r 5.2

.C.

Brow

n et

al.

(199

5)71

C 11

2 M

DD

Out

P TR

D-I

Ope

n m

ultic

ente

r tria

l of

FLX

into

lera

nt

subj

ects

(8

wee

ks)

SER

50-2

00m

g -

Res

pons

e (C

GI-I

≤2)

71.

8%

Dro

pout

ove

rall =

21.4

% D

ropo

ut S

E =

9.8%

No

plac

ebo

cont

rol.

FLX-

into

lera

nce

befo

re

switc

h on

ly.

Cala

bres

e et

al.

(200

3)73

C 55

M

DD

Out

PTR

D-I

Ope

n m

ultic

ente

r tria

l of

FLX

into

lera

nt

subj

ects

(6

wee

ks)

CIT

20-4

0mg

- R

espo

nse

(CG

I-I ≤

2) 6

5.5%

Dro

pout

ove

rall =

5%

Dro

pout

SE

= 0%

No

plac

ebo

cont

rol.

FLX-

into

lera

nce

befo

re

switc

h on

ly. N

o m

inim

um le

vel o

f HD

RS

requ

ired

for s

tudy

-ent

ranc

e. S

tart

of C

IT

afte

r pla

cebo

was

h-ou

t of 2

-4 w

eeks

.

Joff

e et

al.

(199

6)67

C 55

M

DD

O

utP

TRD

-I

Ret

rosp

ectiv

e st

udy

of

patie

nts

who

wer

e tr

eate

d w

ith a

sec

ond

SSRI

in 2

yea

rs. (≥5

w

eeks

)

FLX

(n=

12)

FLV

(n=

9)

PAR

(n=

11)

SER

(n=

23)

- R

espo

nse

(CG

I-I ≤

2) 5

1% N

o si

gnif

ican

t di

ffer

ence

s be

twee

n va

rious

com

bina

tions

of

sw

itchi

ng. N

o SE

repo

rted

.

No

plac

ebo

cont

rol.

Met

hodo

logi

cally

poo

r:

retr

ospe

ctiv

e st

udy,

unc

lear

def

initi

on o

f in

itial

non

-rep

onse

, sm

all n

umbe

rs, n

o ch

arac

teris

tics

of p

opul

atio

n. U

ncle

ar a

fter

ho

w m

any

wee

ks re

spon

se w

as d

eter

min

ed.

Poiri

er e

t al.

(199

9)74

;87

See

unde

r 5.1.

C.

Rus

h et

al.

(200

6)47

A2

727

MD

D

Out

P TR

D-I

3 ar

m M

ulti

cent

er

unbl

inde

d R

CT

(14

wee

ks)

1. VL

X-XR

75-

375m

g 2.

BU

P 15

0-40

0mg

SER

50-

200m

g R

emis

sion

(HD

RS17

≤7)

VLX

= 2

4.8%

, BU

P =

21.3

%, S

ER =

17.6

% N

NT V

LX-S

ER =

14 (7

.0-∞

),

NN

T BU

P-SE

R =

28 (9

.3-∞

) NN

T VLX

-BU

P =

29 (9

.2-

∞) R

espo

nse

(≥50

% in

QID

S 16)

VLX

= 2

8.2%

, BU

P =

26.1%

, SER

= 2

6.7%

NN

T VLX

-SER

= 6

6 (1

0.6-∞

), N

NT B

UP-

SER

= 18

9 (1

3.4-∞

) NN

T VLX

-BU

P =

49 (1

0.1-∞

) Dro

pout

SE V

LX =

21.2

%, B

UP

= 27

.2%,

SER

= 2

1.0%

(Non

sign

ific

ant

diff

eren

ces)

Unb

linde

d st

udy,

blin

ded

asse

ssor

s, n

o pl

aceb

o-gr

oup.

Met

hodo

logi

cally

wel

l pe

rfor

med

eff

ectiv

enes

s tr

ial.

Part

icip

ants

all

rece

ived

CIT

20-

60m

g as

prio

r tre

atm

ent

(leve

l II S

TAR

*D).

Due

to h

igh

dose

s of

CIT

in

‘leve

l 1’ 4

07/7

27 (5

6%) o

f sub

ject

s in

this

stu

dy

wer

e cl

assi

fied

as C

IT-in

tole

rant

. No

was

hout

ap

plie

d.

Thas

e et

al.

(199

7)68

C 10

6 M

DD

O

utP

TRD

-I

Ope

n m

ultic

ente

r tria

l of

SER

non

resp

onde

rs

or in

tole

rant

sub

ject

s (6

wee

ks)

FLX

20-6

0mg

- R

espo

nse

(≥50

% in

HD

RS 1

7) O

vera

ll 62

%;

initi

alin

tole

rant

sub

ject

s 71

%, in

itial

no

nres

pond

ers

58%.

No

plac

ebo

cont

rol.

Non

resp

onse

to S

ER w

as

dete

rmin

ed re

tros

pect

ivel

y (in

mos

t pa

tient

s).

Thas

e et

al.

(200

1)69

C 57

M

DD

O

utP

TRD

-I

Ope

n st

udy

of

pros

pect

ivel

y de

term

ined

FLX

no

nres

pond

ers

(12

wee

ks)

CIT

20m

g (d

osag

es

coul

d be

incr

ease

d to

60m

g)

- R

espo

nse

(CG

I-I ≤

2) 6

3% (≥

50%

in H

DRS

24)

46%

Dro

pout

ove

rall =

18%;

Dro

pout

SE

= 5%

W

ell p

erfo

rmed

ope

n st

udy.

Unk

now

n pl

aceb

o re

spon

se ra

te. T

oler

ance

for C

IT

afte

r FLX

is g

ood,

des

pite

dire

ct s

witc

h. N

o in

crea

sed

rate

of S

E in

firs

t wee

ks o

f stu

dy.

Review of switching in MDD

79

Chapter5

Thas

e et

al.

(200

2)72

C 61

M

DD

O

utP

TRD

-I

Ope

n st

udy

in P

AR

into

lera

nt p

atie

nts

(6 w

eeks

)

CIT

20m

g (d

osag

es

coul

d be

incr

ease

d to

40m

g)

- R

espo

nse

(CG

I-I ≤

2) 5

6% D

ropo

ut o

vera

ll =

13%;

D

ropo

ut S

E =

10%

Ope

n st

udy,

no

min

imal

HD

RS-

scor

e on

en

tran

ce. U

nkno

wn

plac

ebo

resp

onse

rate

. To

lera

nce

for C

IT a

fter

PA

R is

goo

d, 1

wee

k w

asho

ut. R

ecur

renc

e of

the

sam

e SE

as

durin

g PA

R tr

eatm

ent i

s 5-

30%.

Zara

te e

t al.

(199

6)70

C 39

M

DD

,Bi

pD,

SchA

, O

CD

InP

TRD

->I

Ret

rosp

ectiv

e st

udy

of

patie

nt c

hart

s in

pa

tient

s pr

evio

usly

us

ing

FLX

(var

ious

du

ratio

ns o

f fol

low

-up)

SER

mea

n do

se 9

3 ±6

2 m

g -

Res

pons

e (C

GI-I

≤2)

41.

9% o

n di

scha

rge,

26%

at

follo

w-u

p.M

etho

dolo

gica

lly v

ery

poor

stu

dy:

retr

ospe

ctiv

e, n

onsy

stem

atic

follo

w-u

p, S

ER

was

not

pre

scrib

ed fo

llow

ing

FLX,

he

tero

gene

ous

stud

ypop

ulat

ion

with

he

tero

geno

us h

isto

ry o

f pre

viou

s tre

atm

ents

(in

clud

ing

MA

O-I,

ECT

). C

onfo

undi

ng b

y no

n-co

mpl

ianc

e an

d ad

ditio

nal p

harm

acot

hera

py

(in 6

9% o

f pat

ient

s). P

ossi

ble

reca

ll-bi

as.

5.1.B

. to

a TC

A o

r mia

nser

in

Fava

et a

l (20

06)48

A2

234

MD

D

Out

P TR

D-II

Mul

ti ce

nter

unb

linde

d RC

T (1

4 w

eeks

)

NO

R 50

-200

mg

MIR

15-6

0mg

Rem

issi

on (H

DRS

17 ≤

7) N

OR

= 19

.8%,

MIR

=

12.4

% N

NT N

OR-

MIR

= 14

(6.0

-∞) N

ot d

iffer

ent

in le

vel I

I int

oler

ant g

roup

. Res

pons

e (≥

50%

in

QID

S 16)

NO

R =

16.5

%, M

IR =

13.5

% N

NT N

OR-

MIR

= 3

2 (8

.1-∞

) Dro

pout

SE N

OR

= 34

.7%,

MIR

=

33.3

%

Unb

linde

d st

udy,

blin

ded

asse

ssor

s, n

o pl

aceb

o-gr

oup.

Met

hodo

logi

cally

wel

l pe

rfor

med

eff

ectiv

enes

s tr

ial.

Part

icip

ants

re

ceiv

ed C

IT 2

0-60

mg

and

VLX

or B

UP

or S

ER

or a

ugm

enta

tion

of C

IT w

ith B

UP

or B

US

or

CBT

(leve

l III

STA

R*D

). 5

2.1%

was

con

side

red

leve

l II i

ntol

eran

t. B

lood

-leve

ls o

f NO

R al

low

ed b

ut n

ot o

blig

ator

y fo

r dos

ing

(33.

9%

mea

sure

d). N

o w

asho

ut a

pplie

d.

Ferr

eri e

t al.

(200

1)75

B 10

3 M

DD

In

P +

Out

P TR

D-I

3 ar

m m

ultic

ente

r RCT

of

FLX

20m

g-N

R (6

wee

ks)

1. M

IAN

60m

g 2.

FLX

20m

g +

MIA

N

60m

g

3. F

LX 2

0mg

Res

pons

e (≥

50%

in H

DR

S 17)

Mia

nser

ine

= 48

.5%,

Flu

oxet

ine

= 37

% N

NT M

IAN

-FLX

=9 (2

.9-

∞);

NN

T MIA

N+F

LX-F

LX=4

(2.1-

34.1)

Rem

issi

on

(HD

RS 1

7≤8)

Mia

nser

ine

= 36

%, F

luox

etin

e =

18%

NN

T MIA

N-F

LX=6

(2.6

-∞);

NN

T MIA

N+F

LX-

FLX=

4 (2

.2-2

3.9)

Dro

pout

SE

NN

HM

IAN

-FLX

=5

(2.6

-10.4

); N

NH

MIA

N+F

LX-F

LX=1

6 (6

.8-∞

) D

ropo

ut o

vera

ll NN

HM

IAN

-FLX

=6 (2

.6-∞

);

NN

HM

IAN

+FLX

-FLX

=64

(4.9

-∞)

Met

hodo

logi

cally

sou

nd, s

mal

l sam

ple

size

fo

r thr

ee a

rms,

lim

ited

pow

er; i

n M

IAN

-gro

up

due

to lo

ng h

alf-l

ife o

f FLX

, firs

t wee

ks a

lso

sort

of ‘

com

bina

tion’

-the

rapy

. Low

dos

e of

co

ntin

ued

FLX

in re

fere

nce

grou

p. N

o w

asho

ut a

pplie

d, d

irect

sw

itch

asso

ciat

ed

with

incr

ease

d in

tole

ranc

e an

d dr

opou

t .

Nie

renb

erg

et a

l. (2

003)

76C

92

MD

D

Out

P TR

D->

II

Ope

n ph

ase

of N

OR

tr

eatm

ent p

rece

ding

a

2nd R

CT14

0

(6 w

eeks

)

NO

R 10

0mg

(adj

uste

d to

ac

hiev

e 10

0 ng

/ml)

- R

espo

nse

(≥50

% in

HD

RS 1

7) =

42.

4%

Rem

issi

on (H

DRS

17≤7

) = 11

.9%

Dro

pout

ove

rall

= 34

.7%

TRD

def

ined

as

1-5 fa

iled

adeq

uate

tria

ls

durin

g cu

rren

t epi

sode

(mea

n ±S

D 2

.3

±1.5

). 9

5.7%

of p

atie

nts

was

trea

ted

with

≥1

SSR

I. U

nkno

wn

plac

ebo

resp

onse

rate

.

Nol

en e

t al.

(198

8)77

C 31

M

DD

InP

TRD

-III

Blin

ded

cons

ecut

ive

ther

apy

afte

r 4 w

eeks

of

FLV

(4

wee

ks)

OXA

100-

300m

g -

Res

pons

e (≥

50%

in H

DRS

17) =

38.

7% R

elap

se

= 19

.4%

with

in 6

mon

ths

Mod

ifie

d IT

T an

alys

is, e

xclu

ding

pat

ient

s w

ho d

ropp

ed-o

ut in

firs

t 2 w

eeks

. TR

D:

patie

nts

used

≥1 T

CA b

efor

e tr

eatm

ent w

ith

FLV.

No

data

on

drop

outs

.

Tabl

e 5.

1. Ef

fect

iven

ess

of s

witc

hing

aft

er ≥

1 SSR

I: Se

lect

ed s

tudi

es. (

Cont

inue

d)

Stud

y (y

ear)

refe

renc

eLo

E N

D

esig

n (f

ollo

w-u

p)

Inte

rven

tion*

Co

mpa

rison

* O

utco

me†

R

emar

ks

80

Chapter 5

Chap

ter

5

Pese

low

et a

l. (1

989)

78C

15 +

10

MD

D

Out

P TR

D-I

Blin

ded

cros

s-ov

er

desi

gn w

ith o

rigin

al

rand

omiz

atio

n (6

wee

ks ?)

IMI 6

5-27

5mg

- R

espo

nse

(≥50

% in

HD

RS u

nkno

wn

or C

GI-I

≤2

) PA

R

IMI =

73%

M

etho

dolo

gica

lly p

oor:

unc

lear

des

crip

tion

of s

tudi

ed p

opul

atio

n, li

mite

d pr

esen

tatio

n of

dat

a. D

ata

of in

itial

non

resp

onde

rs to

IMI

switc

hed

to P

AR

als

o pr

ovid

ed.

Thas

e et

al.

(200

2)79

;80

C 11

7 Ch

ron

MD

D,

MD

D +

D

ys

Out

P TR

D-I

Blin

ded,

mul

ticen

ter

cros

s-ov

er d

esig

n w

ith

orig

inal

rand

omiz

atio

n(1

2 w

eeks

)

IMI 5

0-30

0mg

- R

espo

nse

(CG

I-I ≤

2, ≥

50%

in H

DRS

24,

HD

RS 2

4≤15

and

CG

I-S ≤

3) S

ER

IMI =

44.

4%

Rem

issi

on (H

DRS

24≤7

and

CG

I-I ≤

2) S

ER

IM

I = 2

3% D

ropo

ut o

vera

ll = 2

5%; D

ropo

ut S

E =

9%

Wel

l-per

form

ed s

tudy

.Unk

now

n pl

aceb

o re

spon

se ra

te. D

ata

of in

itial

non

resp

onde

rs

to 12

wee

ks IM

I sw

itche

d to

SER

als

o pr

ovid

ed. D

ue to

abs

ence

of s

econ

d ra

ndom

izat

ion

only

tent

ativ

e co

mpa

rison

s w

ith s

witc

h IM

I S

ER a

vaila

ble.

5.1.C

. to

mir

taza

pine

, nef

azod

one,

ven

lafa

xine

(dua

l act

ion

agen

ts)

Bald

omer

o et

al.

(200

5)46

B 30

97

MD

D +

M

inD

+

Dys

O

utP

TRD

-I

Mul

ticen

ter,

ope

n de

sign

; RCT

of V

LX v

s CA

in S

SRI-N

R or

in

tole

rant

pat

ient

s (2

4 w

eeks

)

VLX

75-2

25m

g CA

: FLX

, PA

R,

CIT

20-6

0mg

SER

50-

200m

g M

IR 15

-45m

g

Res

pons

e (≥

50%

in H

DR

S 17)

: VLX

= 7

7.3%

, SS

RIs

= 7

1.1%

NN

T VLX

-SSR

I = 17

(10.

5-35

.0)

Rem

issi

on (H

DRS

17 ≤

7): V

LX =

59.

3%, S

SRIs

=5

2.1%

NN

T VLX

-SSR

I = 14

(9.1-

29.3

) HD

RS17

-sc

ores

diff

er s

igni

fica

ntly

but

clin

ical

ly

irrel

evan

t at w

eek

12 a

nd 2

4Dro

pout

ove

rall:

VL

X =

19.6

%, C

A =

23.

3% N

NH

VLX-

CA =

27

(15.

1-12

0) D

ropo

ut S

E: V

LX =

2.3

%, C

A =

1.7%

N

NH

VLX-

CA 16

1 (62

.1-∞

)

Larg

e ra

ndom

ized

but

unb

linde

d st

udy.

So

me

met

hodo

logi

cal p

robl

ems.

63.

3% o

f in

clud

ed p

atie

nts

prev

ious

ly u

sed

a SS

RI.

Incl

usio

n of

8.7

% M

inD

. No

diff

eren

tia

tion

of

firs

t-SS

RI i

ntol

eran

t and

unr

espo

nsiv

e pa

tient

s. M

odif

ied

ITT

anal

ysis

of ≥

wee

k 4

com

plet

ers;

in C

A-t

reat

ed s

witc

h gr

oup

22.7

% re

ceiv

ed a

non

-SSR

I. Ba

selin

e H

DR

S-sc

ores

si

gn. h

ighe

r in

VLX-

grou

p; d

iffer

entia

l los

s to

fo

llow

up

26.2

% VL

X vs

36.

2% C

A O

nly

3 tim

e-po

ints

ove

r 24

wee

ks. N

o se

para

te

dich

otom

ous

data

for w

eek

12 re

spon

se/

rem

issi

on

Fava

et a

l. (2

001)

81C

94

MD

D

Out

P TR

D-I

Mul

ticen

ter,

ope

n de

sign

(RCT

of d

irect

sw

itch

vs. w

asho

ut)

(8 w

eeks

)

MIR

15-4

5mg

- R

espo

nse

(≥50

% in

HD

RS 1

7) S

SRI N

R =

48%,

SS

RI I

ntol

.= 5

3% D

ropo

ut o

vera

ll = 4

3%; D

ropo

ut

SE =

26%

Met

hodo

logi

cally

sou

nd. U

nkno

wn

plac

ebo

resp

onse

rate

. Was

hout

-pha

se o

ffer

s no

ad

vant

ages

.

Fava

et a

l (20

06)48

See

unde

r 5.1.

B

Kapl

an (2

002)

82C

73

MD

D

Out

P TR

D-I

Ret

rosp

ectiv

enat

ural

istic

stu

dy o

f SSR

I-NR

or

non-

sust

aini

ng S

SRI

resp

onde

rs

(6-8

wee

ks +

follo

w-

up)

VLX

50-4

00m

g -

Res

pons

e (H

DR

S 25≤

10 a

nd P

GI-2

1 ≥5)

= 8

6%

Rem

issi

on (H

DRS

25≤8

) = 8

2% D

ropo

ut S

E =

5.5%

Met

hodo

logi

cally

ver

y po

or: o

pen,

unb

linde

d de

sign

, 1 re

sear

cher

, ret

rosp

ectiv

e da

ta

obta

ined

. Mild

dep

ress

ion

incl

uded

als

o.

Unc

lear

, but

pro

babl

e se

lect

ion-

bias

. R

ecru

itmen

t of S

SRI r

espo

nder

s w

ho d

id n

ot

sust

ain

thei

r res

pons

e (5

2%) m

ight

incr

ease

re

spon

se ra

te. I

TT-r

esul

ts n

ot m

entio

ned

in

stud

y.

Tabl

e 5.

1. Ef

fect

iven

ess

of s

witc

hing

aft

er ≥

1 SSR

I: Se

lect

ed s

tudi

es. (

Cont

inue

d)

Stud

y (y

ear)

refe

renc

eLo

E N

D

esig

n (f

ollo

w-u

p)

Inte

rven

tion*

Co

mpa

rison

* O

utco

me†

R

emar

ks


81

Chapter5

Mis

chou

lon

et a

l. (2

004)

83C

13

MD

D

Out

P TR

D->

I

Ope

n de

sign

of S

SRI-

NR

or i

n to

l era

nt

patie

nts

(12

wee

ks)

NEF

300

-600

mg

- R

espo

nse

(>50

% H

DRS

6 an

d/or

CG

I-S ≤

2) 3

1%

Dro

pout

SE

= 39

%Sm

all s

ampl

e (p

ilot s

tudy

). U

nkno

wn

plac

ebo

resp

onse

rate

; 61.5

% at

triti

on;

espe

cial

ly in

pre

viou

s FL

X-us

ers.

Was

hout

of

4-7

days

app

lied.

Het

ero

ge n

eous

gro

up o

f TR

D a

pp. 4

6% ≥

stag

e II.

No

sign

. diff

eren

ces

in re

spon

se ra

tes

and

side

eff

ects

com

pare

d to

13 p

atie

nts

trea

ted

with

NEF

as

first

ap

plie

d an

tidep

ress

ant (

but l

ow p

ower

).

Mitc

hell

et a

l. (2

000)

84C

312

MD

D

Sett

ing?

TR

D-≥

I

Mul

ticen

ter,

ope

n un

blin

ded

desi

gn

(8 w

eeks

)

VLX

75-3

00m

g-

Res

pons

e (>

50%

in M

AD

RS) =

52.

6%

Rem

issi

on (M

AD

RS

<12)

= 4

0.7%

Dro

pout

SE

= 11

%

Met

hodo

logi

cally

wel

l. U

ncle

ar s

ettin

g.

Unk

now

n pl

aceb

o re

spon

se ra

te. U

ncle

ar

whi

ch p

ropo

rtio

n us

ed ≥

1 SSR

I (41

-68%

). Pr

obab

ly c

hron

ical

ly d

epre

ssed

sub

ject

s.

De

Mon

tigny

et a

l. (1

999)

85C

152

MD

D In

P &

O

utP

TRD

-I

Mul

ticen

ter,

ope

n de

sign

(8

wee

ks)

VLX

75-3

75m

g -

Res

pons

e (≥

50%

in H

DR

S 21)

= 5

8%; (≥5

0%

in

MA

DR

S) =

62%

; (CG

I-S ≤

3) =

66%

Rem

issi

on

(≥75

% in

HD

RS21

) = 2

1% D

ropo

ut S

E =

7.9%

TRD

: at l

east

1 pr

evio

us T

CA o

r SSR

I or M

OC

or T

RA

Z or

NEF

, maj

ority

of p

atie

nts

used

a

SSR

I. M

ean

dura

tion

of e

piso

de 2

yea

rs

(ran

ge 2

mon

ths

- 12.

5 ye

ars)

. Unk

now

n pl

aceb

o re

spon

se ra

te.

Nie

renb

erg

et a

l. (1

994)

86C

70

MD

DIn

P &

O

utP

TRD

-III

2 ce

nter

, ope

n de

sign

(1

2 w

eeks

) VL

X 50

-450

mg

- R

espo

nse

(≥50

% in

HD

RS21

) = 3

2.9%

; (≥5

0%

in

MA

DR

S) =

30%

; (CG

I-I ≤

2) =

40%

Rem

issi

on

(HD

RS 2

1≤8)

= 15

.7%;

(MA

DRS

≤12

) = 18

.6%;

(C

GI-I

= 1)

= 2

2.9%

Dro

pout

SE

= 9.

6%

TRD

: at l

east

≥3

drug

s of

≥2

diff

eren

t cla

sses

, ≥1

TCA

and

≥1 a

ugm

enta

tion.

Unc

lear

wha

t pr

opor

tion

of p

atie

nts

used

≥1 S

SRI.

Chro

nica

lly d

epre

ssed

gro

up (m

edia

n du

ratio

n 2.

5 ye

ars)

. Unk

now

n pl

aceb

o re

spon

se ra

te.

Poiri

er e

t al.

(199

9)74

;87

A2

123

MD

DIn

P &

O

utP

TRD

-II

Mul

ticen

ter R

CT

(4 w

eeks

) VL

X 75

-300

mg

PAR

20-4

0mg

Res

pons

e (≥

50%

in H

DR

S 17)

: VLX

= 4

5.0%

, PA

R =

29.

0% N

NT V

LX-P

AR

= 7

(3.0

-∞)

Rem

issi

on (H

DRS

17≤1

0) :

VLX

= 36

.7%,

PA

R =

17

.7%

NN

T VLX

-PA

R =

6 (2

.9-2

8.9)

Dro

pout

SE:

VL

X =

8.2%

, PA

R =

4.8

%NN

HVL

X-PA

R =

30 (8

.3-

∞)

Met

hodo

logi

cally

wel

l. Sh

ort f

ollo

w-u

p du

ring

stud

y. D

osin

g-sc

hedu

les

are

diff

eren

t be

twee

n VL

X an

d PA

R.

Reyn

aert

-Dup

uis

et a

l. (2

002)

88C

688

MD

DIn

P &

O

utP

TRD

-I

Mul

ticen

ter,

na

tura

listic

, ope

n de

sign

(6

wee

ks)

VLX

75-3

75m

g -

Res

pons

e (≥

50%

in H

DR

S 21)

= ±

61%

(inpr

evio

us S

SRI-t

reat

ed p

atie

nts)

86

.3%

of p

atie

nts

wer

e sw

itche

d to

VLX

due

to

inef

fica

cy, o

f 41.7

% w

ho w

ere

switc

hed

from

a S

SRI s

epar

ate

resp

onse

rate

s ar

e gi

ven.

Imm

edia

te s

witc

hing

app

lied

(exc

ept

from

MA

O-Is

). U

ncle

ar p

rese

ntat

ion

of d

ata.

D

ropo

ut ra

te n

ot m

entio

ned.

Typ

e of

pr

evio

us S

SRI n

ot s

igni

fica

ntly

aff

ecte

d VL

X-ef

ficac

y.

Rus

h et

al.

(200

6)47

See

unde

r 5.1.

A

Tabl

e 5.

1. Ef

fect

iven

ess

of s

witc

hing

aft

er ≥

1 SSR

I: Se

lect

ed s

tudi

es. (

Cont

inue

d)

Stud

y (y

ear)

refe

renc

eLo

E N

D

esig

n (f

ollo

w-u

p)

Inte

rven

tion*

Co

mpa

rison

* O

utco

me†

R

emar

ks

82

Chapter 5

Chap

ter

5

Saiz

-Rui

z et

al.

(199

8)89

C 69

M

DD

O

utP

TRD

-I

Mul

ticen

ter,

na

tura

listic

, ope

n de

sign

(2

4 w

eeks

)

VLX

75-3

75m

g -

Res

pons

e (≥

50%

in H

DR

S 17)

= 6

9.6%

; (CG

I-I

≤2) =

63.

8% D

ropo

ut o

vera

ll = 3

0.4%

Dro

pout

SE

= 8.

7% S

E oc

cur i

n 54

%

Sele

ctio

n of

NR

to a

pre

viou

s SS

RI in

at l

east

a

stan

dard

dos

e fo

r 4 w

eeks

. Lim

ited

pres

enta

tion

of d

ata.

In th

e pa

per a

mod

ifie

d IT

T an

alys

is is

app

lied

for w

eek

4 co

mpl

eter

s.

Endp

oint

of s

tudy

is o

nly

repo

rted

for w

eek

24.

Wan

et a

l. (2

003)

90C

24

MD

D

Sett

ing?

TR

D-≥

II

Retr

ospe

ctiv

e ch

art

revi

ew o

f con

secu

tive

sub

ject

s w

ho fa

iled

resp

onse

to ≥

1 TCA

and

≥1

SSR

I (2

wee

ks -

3 ye

ars)

MIR

15-4

5mg

- R

espo

nse

(CG

I-I ≤

2) =

16.7

% Pa

rtia

l res

pons

e (C

GI-I

=3)

= 2

0.8%

Dro

pout

SE =

20.

8%

Met

hodo

logi

cally

ver

y po

or. I

n 17

.2%

of

elig

ible

pat

ient

s da

ta w

ere

insu

ffic

ient

for

incl

usio

n. H

ighl

y tr

eatm

ent r

esis

tant

po

pula

tion

(ave

rage

7 p

revi

ous

drug

-tria

ls

[ran

ge 2

-13])

. Unc

lear

wha

t res

pons

e in

dica

ted

switc

h to

MIR

. CG

I-dat

a de

term

ined

by

char

t rev

iew

. Chr

onic

de

pres

sion

in 4

5.8%

. Hig

h le

vel o

f co

mor

bidi

ty w

ith a

nxie

ty d

isor

ders

. Co

med

icat

ion

with

ant

idep

ress

ants

and

an

tipsy

chot

ics

in 4

1.7%.

5.1.D

. to

rebo

xeti

ne o

r bup

ropr

ion

(sel

ecti

ve N

RI o

r no

radr

ener

gic

& d

opam

iner

gic

agen

ts)

Fava

et a

l. (2

003)

94C

128

MD

D

Out

P TR

D-I

Mul

ticen

ter,

ope

n de

sign

of F

LX-N

R (8

wee

ks)

REB

8-10

mg

- R

espo

nse

(≥50

% in

HD

RS 2

5 and

CG

I-I ≤

2) =

44

.5%

Resp

onse

(≥50

% in

HD

RS 1

7) =

45.

3%

Dro

pout

SE =

13.3

%

Met

hodo

logi

cally

wel

l per

form

ed o

pen

stud

y; u

nkno

wn

plac

ebo

resp

onse

rate

. FLX

-N

R de

term

ined

at e

nd o

f FLX

-tre

atm

ent;

di

rect

sw

itch

to R

EB w

ell t

oler

ated

. Due

to

long

hal

f-life

of F

LX, f

irst 4

wee

ks

‘com

bina

tion’

-the

rapy

.

Fava

et a

l. (2

003)

92C

29

MD

D

Out

P TR

D-I

Two-

cent

er o

pen

desi

gn o

f pr

ospe

ctiv

ely

dete

rmin

ed F

LX-N

R

(8 w

eeks

)

BUP-

SR 15

0-40

0mg

- M

odif

ied

ITT

(n=

26) R

espo

nse

(≥50

% in

H

DR

S 17)

= 3

4.6%

Par

tial r

espo

nse

(25-

49%

in

HD

RS 1

7) =

30.

8% R

emis

sion

(HD

RS 1

7 ≤7)

=

23.1%

Met

hodo

logi

cally

wel

l per

form

ed, b

ut s

mal

l op

en st

udy;

unk

now

n pl

aceb

o re

spon

se ra

te.

No

spec

ifie

d dr

opou

trat

es. F

LX-N

R

pros

pect

ivel

y de

term

ined

; dire

ct s

witc

h to

BU

P; n

o do

cum

enta

tion

of e

ffec

ts o

f thi

s sw

itch.

Due

to lo

ng h

alf-l

ife o

f FLX

, firs

t 4

wee

ks ‘c

ombi

natio

n’-t

hera

py.

Rus

h et

al.

(200

6)47

See

unde

r 5.1.

A

Wal

ker e

t al.

(199

3) 9

3 C

39

MD

D

Out

P TR

D-I

Ope

n de

sign

of

patie

nts

with

sex

ual

side

-eff

ects

on

FLX

(par

tially

FLX

-NR

; n=

16)

(8 w

eeks

)

BUP

150-

450m

g -

All

(n=

36):

H

DR

S 28:

16.6

(±7.

8)

8.4

(±

8.3)

Dro

pout

SE =

10.3

% D

ropo

utin

effi

cacy

=

10.3

% Bs

l HD

RS28≥1

8 (n

= 16

):

HD

RS28

: 25.

4 (±

5.8)

10

.9 (±

10.8

)

Initi

al d

esig

n of

sw

itchi

ng d

ue to

sex

ual s

ide-

effe

ct; l

imite

d da

ta p

rovi

ded

for d

epre

ssed

su

bjec

ts; e

.g. n

o re

spon

se-r

ates

; unk

now

n pl

aceb

o re

spon

se ra

te. I

mpr

ovem

ent o

f or

gasm

func

tion

(84%

), sa

tisfa

ctio

n (7

8%) a

nd

libid

o (7

8%) a

fter

sw

itch.

2 w

eeks

was

hout

ap

plie

d; d

isap

pear

ance

of s

exua

l dys

func

tion

linea

r with

FLX

-was

hout

Tabl

e 5.

1. Ef

fect

iven

ess

of s

witc

hing

aft

er ≥

1 SSR

I: Se

lect

ed s

tudi

es. (

Cont

inue

d)

Stud

y (y

ear)

refe

renc

eLo

E N

D

esig

n (f

ollo

w-u

p)

Inte

rven

tion*

Co

mpa

rison

* O

utco

me†

R

emar

ks


83

Chapter5

5.1.E

. to

a M

AO

-I

McG

rath

et a

l. (2

006)

49B

109

MD

D

Out

P TR

D-≥

II

Mul

ticen

ter u

nblin

ded

RCT

(12

wee

ks)

TCP

10-6

0mg

VLX-

XR 7

5-30

0mg

+ M

IR 15

-45

mg

Rem

issi

on (H

DRS

17≤7

) TCP

: 6.9

%, V

LX-M

IR:

13.7

% N

NH

= 15

(5.5

-∞)

Res

pons

e (≥

50%

in

QID

S-SR

) TCP

: 12.

1%, V

LX-M

IR: 2

3.5%

NN

H =

9

(3.9

-∞) D

ropo

utSE

TCP

: 41.4

%, V

LX-M

IR: 2

1.6%

NN

H =

6 (2

.7-3

5.2)

Unb

linde

d st

udy,

blin

ded

asse

ssor

s.

Part

icip

ants

rece

ived

at l

east

1 SS

RI,

a 2nd

SS

RI o

r VLX

or B

UP

or C

IT-a

ugm

enta

tion

(BU

P or

BU

S) a

nd s

ome

CBT

and

a 3rd

tr

eatm

ent w

ith N

OR

or M

IR (l

evel

IV

STA

R*D

).

Nol

en e

t al.

(198

5)96

B 26

M

DD

InP

TRD

-≥II

Rand

omiz

ed

unbl

inde

d cr

osso

ver

desi

gn

(4 w

eeks

)

TCP

20-10

0mg

L5H

TP 2

0-20

0mg

Res

pons

e (≥

50%

in H

DR

S 17)

TCP

: 42.

9%,

L5H

TP: 0

% N

NT

= 3

(1.5

-5.9

) tra

nylc

ypra

min

e si

de e

ffec

ts =

62%

car

diov

ascu

lar;

15%

inso

mni

a

Smal

l stu

dy, a

lloca

tion

of in

itial

trea

tmen

t no

t cle

arly

des

crib

ed. S

tudy

-gro

ups

diff

er

sign

ifica

ntly

in b

asel

ine

HD

RS-s

core

. Sta

ge II

-III

TR

D-p

atie

nts.

In th

e pa

per d

ata

of s

econ

d 4

wee

ks (c

ross

-ove

r pha

se) a

re a

lso

give

n.

Lim

ited

pres

enta

tion

of d

ata.

Nol

en e

t al.

(198

8)95

B 21

MD

D

InP

TRD

-≥I

I

RCT

with

sec

onda

ry

cros

s-ov

er

(4 w

eeks

)

TCP

40-1

00m

g N

OM

150-

250m

g R

espo

nse

(≥50

% in

HD

RS 1

7) T

CP: 4

5.5%

N

OM

: 10.

0% N

NT

= 3

(1.4

-154

.8) T

CPSE

= 58

% ca

rdio

vasc

ular

, 33%

inso

mni

a

Stag

e II-

III T

RD

-pat

ient

s. U

ncle

ar m

etho

d of

ra

ndom

izat

ion.

In th

e pa

per d

ata

of s

econ

d 4

wee

ks (c

ross

-ove

r pha

se) a

re a

lso

give

n.

Lim

ited

pres

enta

tion

of d

ata.

* A

ll do

sage

s in

mg/

day.

† In

tent

ion

to tr

eat (

ITT)

-res

ults

unl

ess

spec

ified

. A

bbre

viat

ions

: AD

= a

ntid

epre

ssan

t, C

A =

‘Con

vent

iona

l ant

idep

ress

ants

’ (PA

R, F

LX, S

ER, C

IT, M

IR a

nd o

ther

ant

idep

ress

ants

not

spe

cifi

ed),

CBT

= Co

gniti

ve B

ehav

ior t

hera

py, C

GI-I

= C

linic

al g

loba

l im

pres

sion

–im

prov

emen

t sca

le, D

ys =

dys

thym

ia, H

DRS

? = H

amilt

on d

epre

ssio

n ra

ting

scal

e (?

indi

cate

s nu

mbe

r of i

tem

s), I

nP =

In-p

atie

nts,

MA

DR

S= M

ontg

omer

y Å

sber

g D

epre

ssio

n R

atin

g Sc

ale,

MD

D

= m

ajor

dep

ress

ive

diso

rder

,Min

D =

Min

or D

epre

ssio

n, N

NH

/T =

num

ber n

eede

d to

har

m/t

reat

, NR

= n

onre

spon

ders

, OCD

= O

bses

sive

com

puls

ive

diso

rder

, Out

P =

Out

-pat

ient

s, P

GI =

Pat

ient

glo

bal

impr

ovem

ent s

cale

, QID

S-SR

= Q

uick

inve

ntor

y of

dep

ress

ive

sym

ptom

atol

ogy

self-

rate

d, S

chA

= S

chiz

oaff

ectiv

e di

sord

er, S

E =

Side

-eff

ects

, TRD

= tr

eatm

ent r

esis

tant

dep

ress

ion,

sta

ges

I: fa

ilure

of 1

ad

equa

te tr

ial o

f 1 m

ajor

AD

cla

ss; I

I: St

age

I + fa

ilure

of a

n ad

equa

te tr

ial o

f 1 d

istin

ctly

diff

eren

t AD

cla

ss ;

III: s

tage

II +

failu

re o

f an

adeq

uate

tria

l of T

CA; I

V: S

tage

III +

failu

re o

f an

adeq

uate

tria

l of a

n M

AO

I; V:

Sta

ge IV

+fa

ilure

of a

cou

rse

of b

ilate

ral E

CT.10

;24 D

rugs

: BU

P =

bupr

opio

n, B

US

= bu

spiro

ne, C

IT =

cita

lopr

am, F

LV=

fluvo

xam

ine,

FLX

= fl

uoxe

tine,

IMI =

imip

ram

ine,

L5H

TP =

L-5

-hyd

roxy

-tr

ypto

phan

, MA

O-I

= irr

ever

sibl

e in

hibi

tor o

f mon

oam

ine-

oxid

ase,

MIR

= m

irtaz

apin

e, M

OC

= m

oclo

bem

ide,

NEF

= n

efaz

odon

e, N

OM

= n

omife

nsin

e, N

OR

= n

ortr

ipty

line,

NR

I = n

orad

rene

nalin

e re

upta

ke

inhi

bito

r, O

XA =

oxa

prot

iline

, PA

R =

par

oxet

ine,

PLA

C =

plac

ebo,

REB

= re

boxe

tine,

SER

= s

ertr

alin

e, S

SRI =

sel

ectiv

e se

roto

nine

reup

take

inhi

bito

r, T

CA =

tric

yclic

ant

idep

ress

ants

, TCP

= tr

anyl

cypr

omin

e,

TRA

Z =

traz

odon

e, V

LX =

ven

lafa

xine

.

Tabl

e 5.

1. Ef

fect

iven

ess

of s

witc

hing

aft

er ≥

1 SSR

I: Se

lect

ed s

tudi

es. (

Cont

inue

d)

Stud

y (y

ear)

refe

renc

eLo

E N

D

esig

n (f

ollo

w-u

p)

Inte

rven

tion*

Co

mpa

rison

* O

utco

me†

R

emar

ks

Chap

ter

Chapter 5

5

84

Second SSRIWe identified 7 open studies investigating a switch to a second SSRI.67;68;68-73 In one of thesestudies, the non-response to the initial SSRI was determined prospectively, and switching wasapplied immediately.69 In 4 studies, intolerance was determined retrospectively, with an (unclear)interval between the end of the previous SSRI and the next.67;68;70;72 In the remaining 2 (SSRI-intolerance) studies, patients either started a second SSRI soon after the first SSRI or had an SSRI-free interval.71;73

Response rates of switching in SSRI non-responders varied between 46% and 58% in threeuncontrolled studies of variable methodological quality.67-69 The response rate was lower (42%) ina fourth study with a heterogeneous group of inpatients.70 However, response rates to a secondSSRI varied between 56% and 72% when patients were intolerant to the first SSRI (4 studies).68;71-73

Dropout rates due to side effects were between 5% and 21%, in studies with initial non-responders69and between 0% and 10% in SSRI-intolerant samples71-73 (LoE: C).

In the SSRI-arms of three RCTs, response rates varied between 26.7% and 71.1%, while remissionrates were between 17.6-52.1%.46;47;74 Dropout rates due to side effects varied between 4.8-21.0%.For results on the comparisons with other arms see below (LoE: A2-B).

In summary, the data from the open studies and one of the RCTs46 suggest that, after 1 SSRI,non-responders and, notably, also SSRI-intolerant patients can benefit from a switch to a secondSSRI with response rates of approximately 50% and 70%, respectively. However, the results in twoRCTs47;74 indicated much less advantageous response and remission rates for a second SSRI (26.7-29.0% and ~17.6% respectively).

Tricyclic Antidepressants and mianserinWe identified 2 RCTs with a switch to a TCA,48;75 with one having limited power due to arandomization into 3 arms.75 Four open studies investigated a switch from an SSRI to a TCA.76-80

The methodology of the open studies varied: one large cross-over study was methodologicallysound,79;80 one small study was unequivocally poor,78 and two studies were of reasonable quality(investigating populations with TRD).76;77

In the RCT of Ferreri et al. switching to mianserin (a noradrenergic tetracyclic) versuscontinuation of fluoxetine was investigated, with a third arm for their combination.75 Nosignificant difference was found between switching to mianserin and continuation of fluoxetine(response 48.5% and 36.8% respectively; NNT= 9 (95% CI 2.9-∞)) in an ITT analysis. The combinationof fluoxetine and mianserin performed better than continuation of fluoxetine (response 62.5% inthe combination group; NNT= 4 (95% CI 2.1-34.1)). Dropouts due to side effects were highest in theswitch-group (24%; NNH vs continuation= 5 (95% CI 2.6-10.4)) (LoE: B).

The STAR*D level III study compared a switch to nortriptyline versus mirtazapine in arandomized unblinded design.48 All participants received citalopram plus either a switch tosertraline, venlafaxine or bupropion or citalopram augmentation with buspirone or bupropion.Response rates (≥50% decrease in QIDS-SR16 score) were 16.5% for nortriptyline and 13.5% formirtazapine (NNT= 32 (95% CI 8.1-∞)). Remission rates (HDRS17 ≤7) were 19.8% vs12.4% fornortriptyline and mirtazapine, respectively (NNT= 14 (95% CI 6.0-∞)). There were no differences inremission rates for those intolerant to the level II treatments versus those who tolerated theirsecond trial of antidepressants. Dropoutrates due to side effects were high both for nortriptyline(34.7%) and mirtazapine (33.3%) (LoE: A2).

Thase et al. investigated a switch to imipramine in non-responders to sertraline in chronicdepressive out-patients. They found a 44% ITT response rate, with a dropout rate due tointolerable side effects of 9%.79;80 The methodologically poor study by Peselow et al. (includingalso SSRI-intolerant patients) found a 73% response rate after a switch to imipramine inoutpatients.78 In the studies that recruited TRD populations, response rates after switching tonortriptyline76 and oxaprotiline77 decreased to 39% in in-patients77 and 42% in outpatients,76 with a35% overall dropout rate in the latter study (LoE: C).


5

85

In summary, for the switch to a TCA response rates of approximately 16.5 to 48.5% werefound. Lower response rates were observed in studies that included more treatment resistantpatients.

Mirtazapine, nefazodon or venlafaxine (novel dual acting agents)We identified 13 switch studies to novel dual acting agents.46-48;74;81-90 The methodological qualityvaried. Four studies were randomized controlled trials: Poirier and Boyer compared a switch toparoxetine versus venlafaxine,74 Baldomero et al. compared a switch to venlafaxine extendedrelease versus switching to any other antidepressant (77% of these switches used paroxetine,citalopram, sertraline, or fluoxetine) in an unblinded design,46 and the level II and III STAR*Dswitch studies, which were also unblinded studies, compared a switch after citalopram tovenlafaxine extended release, bupropion, or sertraline47 and the switch thereafter to nortriptylineor mirtazapine.48 Other studies described open studies with mirtazapine,81;90 nefazodone,83 andvenlafaxine.82;84-86;88;89 In seven of the studies, all patients received an SSRI before switching.47;48;81-

83;89;90 Five studies included patients with variable but higher levels of treatmentresistance;48;74;83;86;90 in 2 studies, this was unclear.81;84 In contrast, one study included patients(52%) who initially responded to an SSRI but did not sustain their response.82

In the RCT performed by Poirier and Boyer, switching to venlafaxine was more efficaciousthan paroxetine when remission (HDRS17≤10) was considered (remission rates: 36.7% and 17.7%respectively), with a NNT of 6 (95% CI 2.9-28.9).74 For a response criterion (≥50% reduction inHDRS17), the difference was insignificant (response rates: venlafaxine= 45% and paroxetine= 29%;NNT= 7 (95% CI 3.0-∞)). Dropout rates due to side effects were comparable (8.2% for venlafaxineand 4.8% for paroxetine; NNH= 30 (95% CI 8.3-∞)) (LoE: A2).

In the randomized, unblinded study by Baldomero et al., venlafaxine showed a significantlyincreased remission (HDRS17≤7) rate (59.3%) compared with conventional antidepressants (51.5%)after 24 weeks of treatment, with a NNT of 13 (95% CI 8.9-23.7).46 In the conventionalantidepressants group, 77.3% of the patients used a second SSRI; for SSRIs, the remission rate was52.1% (NNT= 14 (95% CI 9.1-29.3)). Response (≥50% reduction in HDRS17) rates also showed amodest but significant advantage: 77.3% for venlafaxine versus 71.1% for SSRIs (NNT= 17 (95% CI10.5-35.0)). Overall dropout was slightly lower in the venlafaxine group when compared with allconventional antidepressants (28.3% vs 32.8%; NNH= 27 (95% CI 15.1-120). Dropout rates due to sideeffects were not significantly different between venlafaxine and conventional antidepressants(12% vs 7.3% respectively; NNH= 161 (95% CI 62.1-∞)) (LoE: B).

The level II STAR*D trial did not find significant differences between the switches tovenlafaxine, bupropion, and sertraline.47 Before the switch, all participants received citalopram(20-60 mg for a maximum of 14 weeks). Patients were randomized over different randomizationpossibilities for which they were at equipoise.91 The assessors of the primary outcome (HDRS17)were blind to the treatment. After 14 weeks of treatment, response rates (≥50% decrease in QIDS-SR16) were 28.2% for venlafaxine, 26.1% for bupropion and 26.7% for sertraline (not significant).Remission rates (HDRS17 ≤7) were not significantly different for venlafaxine, bupropion, andsertraline (24.8%, 21.3% and 17.6% respectively). For corresponding NNTs see table 5.1. The dropoutrate due to side effects was not statistically different for venlafaxine (21.2%), bupropion (27.2%),and sertraline (21.0%) (LoE: A2).

The level III stwitch study was described earlier.48 Mirtazapine response, remission and side-effects related dropout rates were 13.5%, 12.4% and 33.3%, respectively (LoE: A2).

In open studies, mirtazapine, nefazodone, and venlafaxine showed response rates between17% and 86%, with decreased response rates at increased levels of treatment resistance (LoE:C).81-86;88-90 Dropout rates due to adverse effects varied between 5.5% and 11% for venlafaxine,between 20.8% and 26% for mirtazapine and was 39% in one study with nefazodone (LoE: A2, C).

Chap

ter

Chapter 5

5

86

We performed a meta-analysis of the three RCTs that compared switching to venlafaxineversus SSRIs,46;47;74 although the differences in duration of follow-up introduced someheterogeneity (ranging from 4 weeks by Poirier and Boyer74 to 24 weeks by Baldomero et al.46). Asshown in figure 5.2, the weighted difference in remission-rates (fixed effects model) was 8% (4 –11%) in favour of venlafaxine (NNT= 13 (95% CI 9.1 – 25.0), and for response 6% (1 – 10%), (NNT= 17(95% CI 10.0 – 100.0)). Omission of the methodologically poorer study of Baldomero et al.increased the difference in remission rates (10% (95% CI 3-16%) fixed effects model; NNT= 10 (95% CI6.3-33.3)), but decreased the difference in response rates (4% (-3-12%) fixed effects model; NNT=25 (95% CI 8.3-∞)). The dropoutrate due to side effects was only reported in two studies47;74 ; theweighted difference was 1% (95% CI -5-7%) (fixed effects model) with more dropouts forvenlafaxine.

In summary, heterogeneous studies considering switching to mirtazapine, nefazodone andvenlafaxine showed response rates of approximately 28-50% in subjects without obvious TRD,while in subjects with increased levels of TRD response percentages dropped (investigated forvenlafaxine and mirtazapine). Pooling of results showed a modest and clinically equivocallyadvantageous increased remission rate for venlafaxine over SSRIs (NNT= 13 (95% CI 9.1-25.0).

Bupropion and Reboxetine (agents specifically affecting dopaminergic and/ornoradrenergic neurotransmission)We identified one RCT and two small open studies of switching to bupropion.47;92;93 The STAR*Dlevel II switch study including bupropion was described earlier.47 There were no significantdifferences in remision or response rates for bupropion compared to venlafaxine or sertraline. Inthis study, bupropion had the (statistically unsignificant) highest dropout rate (27.2%) due to sideeffects (LoE : A2).

Figure 5.2. Meta-analysis of switch-studies comparing a switch to venlafaxine versus a second SSRI.

A: Remission. B: Response. VLX = venlafaxine. References to studies: Poirier et al.,74 Baldomero et al.,46 Rush et al.47

�

�


5

87

In two open studies with bupropion, Fava et al prospectively determined fluoxetinenoneresponse in a small but well performed study,92 and Walker et al. recruited patients that wereprimarily suffering sexual side-effects of fluoxetine, and only reported a decrease in 28-itemHDRS-scores.93 One larger, well-performed, open study investigated the switch to reboxetine influoxetine non-responders.94

Thus, switching from fluoxetine was investigated, with reported response rates of 34.6% forbupropion92 and 45.3% for reboxetine.94 For bupropion, specified dropout rates were not reportedin one study.92 The side effect-related dropoutrate was 10.3% in subjects with sexual dysfunctionwhile taking fluoxetine.93 For reboxetine, the dropout rate due to side effects was 13.3% (LoE: C).

In summary, switching to bupropion or reboxetine was scarcely studied but was a possibleoption with response-rates of 26.1%-34.6% and 45.3% respectively. The remission rate of switchingto bupropion was not different compared to venlafaxine or sertraline.

Reversible Inhibitor of Monoamine-oxidase A We identified no studies that investigated switching from a SSRI to a reversible inhibitor ofmonoamine-oxidase A.

Monoamine-oxidase A inhibitorWe identified one RCT from STAR*D49 (level IV) and two small, interrelated randomized studiesafter 4 weeks of treatment with at least one SSRI (fluvoxamine) and oxaprotiline.95;96 Weidentified no studies of SSRI non-responders in atypical depression. Two studies were RCTs49;95

and one an unblended, randomized, cross-over study.96 The STAR*D study investigatedoutpatients; the studies by Nolen et al.95;96 were performed in treatment resistant in-patients.

Nolen et al. found tranylcypromine to be more efficacious than nomifensine,95 in both studiesthe response rate for tranylcypromine was 42.9 and 45.5%.95;96 All patients previously received atleast fluvoxamine and oxaprotiline. Fifty-eight to 62% had side effects affecting their bloodpressure levels (LoE: B).

The STAR*D level IV study included patients that had not been in remission after citalopram(level I), either venlafaxine, bupropion, sertraline or citalopram augmentation with buspirone orbupropion (level II), and additionally received nortriptyline or mirtazapine (level III).49 Thesepatients were randomized between tranylcypromine and a combination of venlafaxine withmirtazapine. Of the included patients 32.1% were intolerant for the level III medication. Remissionrates (HDRS17 ≤7) were low for tranylcypromine (6.9%) and the combination treatment (13.7%;NNH= 15 (95% CI 5.5-∞)). Response rates (≥50% decrease in QIDS-SR16) were also not significantlydifferent: 12.1% vs 23.5% for tranylcypromine and venlafaxine with mirtazapine, respectively (NNH=9 (95% CI 3.9-∞)). Dropout rates due to side effects were higher for tranylcypromine: 41.4% versus21.6% for venlafaxine with mirtazapine (NNH= 6 (95% CI 2.7-35.2)).

Additional concerns for clinicians regarding switchingLittle evidence is available about the optimal way to switch.20;29;69;88;97 Abrupt reduction ordiscontinuation of SSRIs may produce somatic and psychological withdrawal symptoms, of whichoccurrence is inversely related to the plasma half-life of the initial SSRI.98;99 Overlap ofantidepressants during switching is generally avoided.20;29

Direct switching (without a washout phase) from an initial SSRI (fluoxetine at the standarddose or citalopram at high dosages) to another SSRI (paroxetine, citalopram, sertralin),47;69;97

nortriptyline,48 mirtazapine,48;81 bupropion,47 reboxetine,94 or venlafaxine47;88) was well tolerated.Also, direct switching reduced the emergence of side effects compared with placebo in a 1-weekwashout phase (which might have been discontinuation symptoms).97

In case of higher than standard doses of SSRIs, some data for tolerance of direct switchingwere generated by STAR*D.47-49 However, the results published so far do not specify dropoutrates in the first 2 weeks after switching. Also, because tapering of high doses of previousantidepressants was not applied in STAR*D, this trial was not designed to examine the optimal

Chap

ter

Chapter 5

5

88

switch strategy if higher than standard doses were used before the switch. Thus, if necessary,direct switching of high-dose antidepressant therapy appears possible after citalopram as a firstSSRI.47

In a case of switching from an SSRI to a TCA, other reviewers did not recommend a washoutperiod.20;29 In one included study a direct switch to mianserin was less well tolerated.75 Forswitching to nefazodone, in one study a 4-day to 7-day washout was applied but notinvestigated.83 A 1-week washout period is suggested for switching to a reversible inhibitor ofmonoamine-oxidase A, and a 1-week to 2-week washout period is recommended for switching to aMAO-I.20;29 For fluoxetine these washout-periods should be prolonged to 5 weeks because of thelong half-life of fluoxetine and norfluoxetine.20 The inhibition of cytochrome P450 subenzymes bySSRIs may increase the levels of some TCAs during the first to fifth (fluoxetine) week.20

Discussion

This report systematically reviewed and appraised the available research focusing on switchingstrategies for SSRI non-responders in MDD, including the recent STAR*D results. We found thatthe available evidence does not justify distinct recommendations for next-step strategies afternon-response to a first SSRI. Thepooled difference in remission rates of switching to venlafaxine(an SNRI) versus a second SSRI showed a modest and clinically equivocal advantage ofvenlafaxine (NNT= 13 (95% CI 9.1-25.0)), this difference increased when the largest andmethodologically poorest study was omitted (NNT= 10 (95% CI 7-34)).

In summary, after a first SSRI, switching to any of the current classes of antidepressants hasapproximately a 50% chance of response. Still, a direct comparison of the rates across thepredominantly open studies is methodologically not justified. In STAR*D response and remissionrates were lower (respectively 26.8% and 21.3% at level II,47 15% and 16.2% at level III,48 and 17.4% and10.1 at level IV49). Rush et al. attributed these lower remission rates to the inclusion of patientswho were more chronically depressed, had lower socio-economic status and suffered from moreco-morbid somatic and psychiatric diseases.47 In general, the level of TRD10 of included studies wasinversely correlated with treatment outcome. Although this finding carries the risk of anecological fallacy, it is worrisome, which further appears from the STAR*D results. After thesecond antidepressant the chances of response or remission by switching again are becomingrather low, challenging us to find new approaches.100-102

Dropout rates due to side-effects varied between 5-21% for a second SSRI and venlafaxine; 10-35% for TCAs, bupropion, and reboxetine; 20-33% for mirtazapine; 39% for nefazodone and 41.4%for tranylcypromine. It should be noted that these percentages cannotsimply be compared witheach other because of heterogeneous populations and open-study designs. In randomizedcomparisons, no significant differences in side-effect related dropoutwere found, except fortranylcypromine versus a combination of venlafaxine with mirtazapine.49

With eight RCTs,46-49;74;75;95;96 switching-options after a first SSRI were generally investigatedwith open studies. In these open studies, switching to a second SSRI (7 studies) and venlafaxine (7studies) were studied most frequently. Furthermore, the studies were of variable methodologicalquality. In our opinion, the available evidence for switching strategies allows generalrecommendations only. Switching is open to all studied antidepressant classes (second SSRI,novel dual-acting antidepressants, selective noradrenergic and noradrenergic/dopaminergicagents, or TCA or mianserin) without clear recommendations other than those that apply for theselection of initial treatment. In the choice of an initial antidepressant, some reports promotedTCAs for treatment of inpatients;103-106 however, it is unclear what special feature is associatedwith inpatients (e.g. severity), and studies investigating switching strategies after an SSRI ininpatients were not identified. From the available studies it must be emphasized that side-effectsto a first SSRI did not reduce the chance of response or increase the chance of intolerance for a


5

89

second SSRI. Because of side effects, we think that MAO-I should not be prescribed as a secondantidepressant after a first SSRI. A possible exception – but not investigated after a first SSRI – isfor atypical depression.

Switching from a failed TCA treatment was reviewed earlier.7;10;32;80;107 The response rates forwithin-class switching with SSRIs appear more favourable than a TCA-TCA switch: in two smalltrials,108;109 response-rates of a within-class TCA switch were 9108 and 30%.109 The SSRI resultschallenge the belief that any within class switch should be considered illogical. The betweenclasses switching strategies from a TCA to an SSRI (investigated in 10 trials;77-79;109-115 response ratesvarying between 4% (inpatients) and 75% (out-patients)), to a heterocyclic antidepressant (e.g.bupropion, trazodone, nomifensine, oxaprotiline; 6 studies;77;95;116-119 response rates between 10%-56%) and to a MAO-I (6 trials;95;96;120-123 response rates between 29-83%) showed similar broadranges of response rates. These ranges reflect differences in heterogeneous study-populations aswell. Again, it is inappropriate to simply compare these rates determined in different studies.

On theoretical grounds, it is logical (and often recommended) to switch to an antidepressantwith different or combined sites of action (e.g. norepinephrine uptake inhibition afterunsuccessful serotonergic uptake inhibition).124-126 Others pointed out the complex interaction ofmonoamine systems alone, proposed other possible etiologic mechanisms, and considered themonoamine hypothesis only partially explanative for depression and the response toantidepressants.127-131 Six RCTs so far compared different pharmacologic approaches in non-responders (venlafaxine versus paroxetine,74 venlafaxine versus an SSRI46, venlafaxine versussertraline or bupropion,47 nortriptyline versus mirtazapine,48 fluoxetine versus mianserin or amianserin-fluoxetine combination75, and tranylcypromine versus a venlafaxin-mirtazapinecombination49). These RCTs found equivocal superiority of dual-action pharmacotherapy.However, in STAR*D the empirical proof of this theoretical strategy was not found.

Apart from switching, augmentation or combination, and addition of (or switching to)psychotherapy are possible options. Only Ferreri et al.75 and McGrath et al.49 compared switchingversus combination (the latter at a higher level of TRD). In STAR*D, a switch to or augmentationwith cognitive behaviour therapy was possible after citalopram (unpublished yet), andaugmentation of citalopram with buspirone or bupropion was also studied.132 A direct comparisonbetween switching and augmentation after citalopram was not feasible.47 Therefore, clearrecommendations about choosing one of these strategies relative to each other are not possible.In most countries, SSRIs are generally prescribed as first line treatment, often provided in primarycare. We think that switching-strategies after a first SSRI will be preferred, especially in primarycare, in which augmentation and combination strategies may be unfamiliar to physicians. Thishypothesis is supported by audits, even among psychiatrists.40-42

Limitations of the identified studiesWell-designed switch-studies are difficult to carry out, and therefore, it does not surprise that theevidence to date is limited in several ways. We found predominantly open, uncontrolled studies,with a risk of more positive results than in blinded studies and without a possibility to activelycompare strategies. There were few studies that clearly described the inclusion ofprospectivelydetermined SSRI non-responders.47-49;69;75;79;92 This finding is of importance, as in retrospectivelydetermined non-responders, current depression may cause recall-bias. Furthermore, in somestudies non-responders were not treated directly after cessation of the unsuccessful drug, whichmight have biased results; for example depression worsened after cessation, or –the other wayaround– depression may have improved because of the natural course of depression.133;134

Several other problems were encountered: unclear criteria for initial non-response,67;72;73;82;90

inclusion of mild or minor depression,46;82 possible selection-bias,82 limited presentation ofresults,46;78;88;89;93;95;96 absence of ITT-data,82 small sample sizes (n< 40),70;77;78;83;90;92;93;95;96 and lowstatistical power.75;83 In general, less robust studies found more positive results for the drug ofinterest. Table 5.2 presents a summary of these problems.

Chap

ter

Chapter 5

5

90

The STAR*D trials47-49 were randomized but unblinded effectiveness trials. The primaryoutcome (remission by HDRS17) was determined by blind assessors; the secondary outcomes bythe QIDS-SR16 were self-rated by the unblinded patients. The a priori definition of nonremission formissing data will have decreased remission rates because of attrition, but this a priori definitionwas considered noninfluential after sensitivity analyses. The aggressive dose increases in STAR*Dtrials prevented undertreatment, but might have increased attrition, and definitely increased thepercentages of treatment-intolerant patients at all levels. Especially in the level IV trials thetreating physicians might have been unfamiliar with the prescribed medication (tranylcypromine,venlafaxine-mirtazapine combination), reducing the vigour of the applied pharmacologicintervention.49

Future switching studies After the STAR*D trials, the question arises as to whether many randomized direct comparisonsbetween switches among drug classes are fruitful to develop fully evidence-basedrecommendations for switching. Results of some studies might still be published.135-137 Also, thepredictors of poor response and nonremission need to be further clarified. In order to structuredirections of research, the recommended approaches in guidelines should be evaluated for eachtreatment step. The Texas Medication Algorithm project proved that algorithms are beneficial forpatient care;138 however, our next challenge is to investigate which steps within these algorithmsare better compared with each other.

Ideally, three or more armed studies should be designed. Switching within the same class or todifferent classes of drugs should be compared with an augmentation or new approach, while alsoan arm for continuation of the initial therapy should be included. The latter arm would thenrepresent a form of placebo control.Naturally, these studies are hard to carry out, may have toovercome resistance and doubts concerning the ethics of the continuation arm, or may sufferfrom selective patient withdrawal from this continuation arm. The STAR*D-project has been amajor step in this direction, especially by proving the feasibility of such large multicenter trials andthe methodology of (equipoise) randomization. At the same time the effectiveness approachwith many centers, high levels of co-morbidity, chronicity and many arms of treatment might havereduced the ability to find differences.

We found that the response rates in switch-studies decreased with increased levels of TRD.Therefore, future studies must consider the level of TRD as an important effect-modifyingvariable. Ideally, in future research, clear populations of prospectively determined treatmentresistance should be selected, or analyzed in a priori defined subgroups to increase ourknowledge about confounding or effect-modifying variables. Finally, to improve the acceptanceof switching in daily clinical practice, more studies of patients’ perspectives of switching ofantidepressants are needed.

Limitations of the reviewSeveral limitations of this review should be mentioned. First, a review like this cannot overcomethe paucity of high-quality evidence to date. The Cochrane Collaboration primarily rejects openstudies as high-quality evidence. If this criterion had been applied, only 8 studies would havequalified for the review, obviously limiting its applicability. The majority of included studies hadmethodological flaws, two studies were excluded for clear invalidity.61;62 We decided a priori toinclude open studies, and–even more–to include studies in which 50-100% of patients initially usedan SSRI, introducing different levels of TRD. Of course, the latter decision is debateable from amethodological point ofview.

Second, in the selected trials, mostly response was used as the primary outcome, whilecurrently remission of depression is the clinical aim of treatment.139 Only 13 of 31 studies (42%)included remission as an outcome criterion.46-49;74-76;79;82;84-86;92 Only STAR*D primarily investigatedthe practice of switching in order to achieve remission.47-49


5

91

Third, patients studied in the included trials represented selected populations, reducing thegeneralizability of the findings to the 'real world' clinical practice; as an effectiveness trial, theSTAR*D results overcame this problem. Fourth, critical appraisal was performed by one reviewer(H.G.R.), while ideally this should have been performed by two raters. However, we found ourinterobserver agreement to be moderate to good and no worse than in previous interraterattempts in psychiatry.58 Fifth, the grading system for studies does not represent the appraisedmethodological dimensions of evidence. This improved the applicability of the results for busyclinicians, but reduced their strength.

Strengths of the reviewThis is the first review that applied the thorough methodology to search for, identify, andappraise articles as used in Cochrane reviews. The applied methodology and transparentpresentation of data allow clinicians to make their own judgements and, if necessary, to retrievethe source of data. Apart from the relevant up-to-date information for clinicians, this review couldwell serve national guideline committees as a building stone for the development oftreatmentguidelines for MDD.

ConclusionThis systematic review about switching identified 8 RCTs and mostly open switch studies ofvariable methodological quality in heterogeneous populations. The STAR*D results largelyincreased the amount and quality of the available evidence, but did not show differential classeffects to guide switching. After a first SSRI switching is open to all studied antidepressant classes(except irreversible MAO-inhibitors), without clear recommendations other than those that applyfor the selection of initial treatment. For recommendations about when to choose betweenswitching, augmentation, combination, or psychotherapeutic strategies as a next step, hardly anyevidence of comparisons of these strategies relative to each other exists. Future algorithm-basedswitch studies and studies of patient perspectives regarding switching will have to improve ourknowledge to guide treatment for SSRI non-responders.

Acknowledgements

This systematic review was realised by a grant for the development of a local evidence-basedclinicalpractice guideline (No SFA.07.012) from the Academic Medical Center, Amsterdam, theNetherlands. This guideline considering strategies for non-response to a standard dose of a firstSSRI is available on request from the first author.

Conflicts of interest

None

Chap

ter

Chapter 5

5

92

References










10. Thase ME, Rush AJ. Treatment-Resistant Depression. In: Bloom FE, Kupfer DJ, eds. Psychopharmacology: The FourthGeneration of Progress. New York: Raven Press Ltd.; 1995:1081-1097.


12. Nelson JC. Managing treatment-resistant major depression. J Clin Psychiatry. 2003; 64 Suppl 1: 5-12.13. Hirschfeld RM, Montgomery SA, Aguglia E, Amore M, Delgado PL, Gastpar M et al. Partial response and nonresponse to

antidepressant therapy: current approaches and treatment options. J Clin Psychiatry. 2002; 63: 826-837.14. Lam RW, Wan DD, Cohen NL, Kennedy SH. Combining antidepressants for treatment-resistant depression: a review. J Clin

Psychiatry. 2002; 63: 685-693.15. Shelton RC. The use of antidepressants in novel combination therapies. J Clin Psychiatry. 2003; 64 Suppl 2: 14-18.16. Stimpson N, Agrawal N, Lewis G. Randomised controlled trials investigating pharmacological and psychological

interventions for treatment-refractory depression. Systematic review. Br J Psychiatry. 2002; 181: 284-294.17. Anderson IM. Drug treatment of depression: Reflections on the evidence. Adv Psychiatric Treat. 2003; 9: 11-20.18. Kennedy N, McDonough M. Pharmacological management of treatment resistant depression: A clinical review. Irish J

Psychol Med. 2003; 20: 18-23.19. McIntyre RS, Muller A, Mancini DA, Silver ES. What to do if an initial antidepressant fails? Can Fam Physician. 2003; 49: 449-

457.20. Kennedy S, McIntyre R, Fallu A, Lam R. Pharmacotherapy to sustain the fully remitted state. J Psychiatry Neurosci. 2002;

27: 269-280.21. Trivedi MH, Kleiber BA. Algorithm for the treatment of chronic depression. J Clin Psychiatry. 2001; 62 Suppl 6: 22-29.22. Fava M. New approaches to the treatment of refractory depression. J Clin Psychiatry. 2000; 61: 26-32.23. Crismon ML, Trivedi M, Pigott TA, Rush AJ, Hirschfeld RMA, Kahn DA et al. The Texas medication algorithm project:

Report of the Texas consensus conference panel on medication treatment of major depressive disorder. J Clin Psychiatry.1999; 60: 142-156.

24. Thase ME, Rush AJ. When at first you don't succeed: sequential strategies for antidepressant nonresponders. J ClinPsychiatry. 1997; 58: 23-29.

25. Thase ME. Therapeutic alternatives for difficult-to-treat depression: A narrative review of the state of the evidence. CNSSpectr. 2004; 9: 808-821.

26. Parikh RM, Lebowitz BD. Current perspectives in the management of treatment-resistant depression. Dial Clin Neurosci.2004; 6: 53-60.

27. Pridmore S, Turnier-Shea Y. Medication options in the treatment of treatment-resistant depression. Aust N Z J Psychiatry.2004; 38: 219-225.

28. Mann JJ. The medical management of depression. N Engl J Med. 2005; 353: 1819-1834.29. Marangell LB. Switching antidepressants for treatment-resistant major depression. J Clin Psychiatry. 2001; 62 Suppl 18: 12-

17.30. Fava M. Management of nonresponse and intolerance: Switching strategies. J Clin Psychiatry. 2000; 61: 10-12.31. O'Reardon JP, Brunswick DJ, Amsterdam JD. Treatment-resistant depression in the age of serotonin: Evolving strategies.

Curr Opin Psychiatry. 2000; 13: 93-98.32. Nelson JC. Treatment of antidepressant nonresponders: augmentation or switch? J Clin Psychiatry. 1998; 59 Suppl 15: 35-

41.33. Rojo JE, Ros S, Aguera L, de la GJ, de Pedro JM. Combined antidepressants: clinical experience. Acta Psychiatr Scand

Suppl. 2005; 25-31, 36.


5

93




37. Ruhe HG, Huyser J, Swinkels JA, Schene AH. Dose escalation for insufficient response to standard-dose selectiveserotonin reuptake inhibitors in major depressive disorder: Systematic review. Br J Psychiatry. 2006; 189: 309-316.

38. Bauer M, Dopfmer S. Lithium augmentation in treatment-resistant depression: meta-analysis of placebo-controlledstudies. J Clin Psychopharmacol. 1999; 19: 427-434.

39. Bauer M, Forsthoff A, Baethge C, Adli M, Berghofer A, Dopfmer S et al. Lithium augmentation therapy in refractorydepression-update 2002. Eur Arch Psychiatry Clin Neurosci. 2003; 253: 132-139.

40. Byrne S, Rothschild AJ. Psychiatrists' responses to failure of maintenance therapy with antidepressants. Psychiatr Serv.1997; 48: 835-837.



43. Fredman SJ, Fava M, Kienke AS, White CN, Nierenberg AA, Rosenbaum JF. Partial response, nonresponse, and relapsewith selective serotonin reuptake inhibitors in major depression: a survey of current "next-step" practices. J ClinPsychiatry. 2000; 61: 403-408.

44. Bero L, Rennie D. The Cochrane Collaboration. Preparing, maintaining, and disseminating systematic reviews of theeffects of health care. JAMA. 1995; 274: 1935-1938.

45. Rush AJ, Fava M, Wisniewski SR, Lavori PW, Trivedi MH, Sackeim HA et al. Sequenced treatment alternatives to relievedepression (STAR*D): rationale and design. Control Clin Trials. 2004; 25: 119-142.

46. Baldomero EB, Ubago JG, Cercos CL, Ruiloba JV, Calvo CG, Lopez RP. Venlafaxine extended release versus conventionalantidepressants in the remission of depressive disorders after previous antidepressant failure: ARGOS study. DepressAnxiety. 2005; 22: 68-76.

47. Rush AJ, Trivedi MH, Wisniewski SR, Stewart JW, Nierenberg AA, Thase ME et al. Bupropion-SR, sertraline, or venlafaxine-XR after failure of SSRIs for depression. N Engl J Med. 2006; 354: 1231-1242.

48. Fava M, Rush AJ, Wisniewski SR, Nierenberg AA, Alpert JE, McGrath PJ et al. A comparison of mirtazapine andnortriptyline following two consecutive failed medication treatments for depressed outpatients: a STAR*D report. Am JPsychiatry. 2006; 163: 1161-1172.

49. McGrath PJ, Stewart JW, Fava M, Trivedi MH, Wisniewski SR, Nierenberg AA et al. Tranylcypromine versus venlafaxineplus mirtazapine following three failed antidepressant medication trials for depression: a STAR*D report. Am J Psychiatry.2006; 163: 1531-1541.

50. Munoz SR, Bangdiwala SI. Interpretation of Kappa and B statistics measures of agreement. J Applied Stat. 1997; 24: 105-111.

51. Richtlijnontwikkeling binnen het Kwaliteitsinstituut voor de Gezondheidszorg CBO. Handleiding voor werkgroepleden.Kwaliteitsinstituut voor de Gezondheidszorg CBO.2000. Utrecht, Kwaliteitsinstituut voor de Gezondheidszorg CBO.

52. SIGN 50: A guideline developers' handbook. 50. 2001. Edinburgh, Scottish Intercollegiate Guideline Network. 53. Sackett DL, Straus SE, Richardson WS, Rosenberg W, Haynes RB. Evidence-Based Medicine. How to practice and teach EBM.

2 ed. Edinburgh: Churchill Livingstone; 2000.54. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 23: 56-61.55. Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979; 134: 382-

389.56. Guy W, Clearly PA, Close JH, Conners CK, Covi L, et al. ECDEU Assessment manual for psychopharmacology. DHEW

publication (ADM) 76-338. 1976. Washington, DC, US Department of Health, Education, and Welfare. 57. Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, Klein DN et al. The 16-Item Quick Inventory of Depressive


58. Moncrieff J, Churchill R, Drummond DC, McGuire H. Development of a quality assessment instrument for trials oftreatments for depression and neurosis. Int J Methods Psychiatr Res. 2001; 10: 126-133.

59. Posternak MA, Zimmerman M. Switching versus augmentation: a prospective, naturalistic comparison in depressed,treatment-resistant patients. J Clin Psychiatry. 2001; 62: 135-142.

60. Smeraldi E. Amisulpride versus fluoxetine in patients with dysthymia or major depression in partial remission: a double-blind, comparative study. J Affect Disord. 1998; 48: 47-56.

61. Emrich HM, Berger M, Riemann D, von Zerssen D. Serotonin reuptake inhibition vs. norepinephrine reuptake inhibition: adouble-blind differential-therapeutic study with fluvoxamine and oxaprotiline in endogenous and neurotic depressives.Pharmacopsychiatry. 1987; 20: 60-63.

62. Weintraub D. Nortriptyline in geriatric depression resistant to serotonin reuptake inhibitors: case series. J GeriatrPsychiatry Neurol. 2001; 14: 28-32.

63. Keller MB, McCullough JP, Klein DN, Arnow B, Dunner DL, Gelenberg AJ et al. A comparison of nefazodone, the cognitivebehavioral-analysis system of psychotherapy, and their combination for the treatment of chronic depression. N Engl JMed. 2000; 342: 1462-1470.

64. Sajatovic M, DiGiovanni S, Fuller M, Belton J, DeVega E, Marqua S et al. Nefazodone therapy in patients with treatment-resistant or treatment-intolerant depression and high psychiatric comorbidity. Clin Ther. 1999; 21: 733-740.

Chap

ter

Chapter 5

5

94

65. Lam RW, Hossie H, Solomons K, Yatham LN, Lam RW, Hossie H et al. Citalopram and bupropion-SR: combining versusswitching in patients with treatment-resistant depression. J Clin Psychiatry. 2004; 65: 337-340.

66. Birkenhager TK, van den Broek WW, Mulder PG, Bruijn JA, Moleman P. Efficacy and tolerability of tranylcypromine versusphenelzine: a double-blind study in antidepressant-refractory depressed inpatients. J Clin Psychiatry. 2004; 65: 1505-1510.

67. Joffe RT, Levitt AJ, Sokolov STH, Young LT. Response to an open trial of a second SSRI in major depression. J ClinPsychiatry. 1996; 57: 114-115.

68. Thase ME, Blomgren SL, Birkett MA, Apter JT, Tepner RG. Fluoxetine treatment of patients with major depressivedisorder who failed initial treatment with sertraline. J Clin Psychiatry. 1997; 58: 16-21.

69. Thase ME, Feighner JP, Lydiard RB. Citalopram treatment of fluoxetine nonresponders. J Clin Psychiatry. 2001; 62: 683-687.

70. Zarate CA, Kando JC, Tohen M, Weiss MK, Cole JO. Does intolerance or lack of response with fluoxetine predict the samewill happen with sertraline? J Clin Psychiatry. 1996; 57: 67-71.

71. Brown WA, Harrison W. Are patients who are intolerant to one serotonin selective reuptake inhibitor intolerant toanother? J Clin Psychiatry. 1995; 56: 30-34.

72. Thase ME, Ferguson JM, Lydiard RB, Wilcox CS. Citalopram treatment of paroxetine-intolerant depressed patients.Depress Anxiety. 2002; 16: 128-133.

73. Calabrese JR, Londborg PD, Shelton MD, Thase ME. Citalopram treatment of fluoxetine-intolerant depressed patients. JClin Psychiatry. 2003; 64: 562-567.

74. Poirier MF, Boyer P. Venlafaxine and paroxetine in treatment-resistant depression. Double-blind, randomisedcomparison. Br J Psychiatry. 1999; 175: 12-16.

75. Ferreri M, Lavergne F, Berlin I, Payan C, Puech AJ. Benefits from mianserin augmentation of fluoxetine in patients withmajor depression non-responders to fluoxetine alone. Acta Psychiatr Scand. 2001; 103: 66-72.

76. Nierenberg AA, Papakostas GI, Petersen T, Kelly KE, Iacoviello BM, Worthington JJ et al. Nortriptyline for treatment-resistant depression. J Clin Psychiatry. 2003; 64: 35-39.

77. Nolen WA, van de Putte JJ, Dijken WA, Kamp JS, Blansjaar BA, Kramer HJ et al. Treatment strategy in depression. I. Non-tricyclic and selective reuptake inhibitors in resistant depression: a double-blind partial crossover study on the effects ofoxaprotiline and fluvoxamine. Acta Psychiatr Scand. 1988; 78: 668-675.

78. Peselow ED, Filippi AM, Goodnick P, Barouche F, Fieve RR. The short- and long-term efficacy of paroxetine HCL: B. Datafrom a double-blind crossover studyand from a year-long term trial vs. imipramine and placebo. Psychopharmacol Bull.1989; 25: 272-276.

79. Thase ME, Rush AJ, Howland RH, Kornstein SG, Kocsis JH, Gelenberg AJ et al. Double-blind switch study of imipramine orsertraline treatment of antidepressant-resistant chronic depression. Arch Gen Psychiatry. 2002; 59: 233-239.

80. Thase ME, Keller MB, Gelenberg AJ, Hirschfeld RM, Schatzberg AF. Double-blind crossover antidepressant study:sertraline versus imipramine [Abstract]. Psychopharmacol Bull. 1995; 31: 535.

81. Fava M, Dunner DL, Greist JH, Preskorn SH, Trivedi MH, Zajecka J et al. Efficacy and safety of mirtazapine in majordepressive disorder patients after SSRI treatment failure: an open-label trial. J Clin Psychiatry. 2001; 62: 413-420.

82. Kaplan EM. Efficacy of venlafaxine in patients with major depressive disorder who have unsustained or no response toselective serotonin reuptake inhibitors: An open-label, uncontrolled study. Clin Ther. 2002; 24: 1194-1200.

83. Mischoulon D, Opitz G, Kelly K, Fava M, Rosenbaum JF, Mischoulon D et al. A preliminary open study of the tolerabilityand effectiveness of nefazodone in major depressive disorder: comparing patients who recently discontinued an SSRIwith those on no recent antidepressant treatment. Depress Anxiety. 2004; 19: 43-50.

84. Mitchell PB, Schweitzer I, Burrows G, Johnson G, Polonowita A. Efficacy of venlafaxine and predictors of response in aprospective open-label study of patients with treatment-resistant major depression. J Clin Psychopharmacol. 2000; 20:483-487.

85. de Montigny C, Silverstone PH, Debonnel G, Blier P, Bakish D. Venlafaxine in treatment-resistant major depression: aCanadian multicenter, open-label trial. J Clin Psychopharmacol. 1999; 19: 401-406.

86. Nierenberg AA, Feighner JP, Rudolph R, Cole JO, Sullivan J. Venlafaxine for treatment-resistant unipolar depression. J ClinPsychopharmacol. 1994; 14: 419-423.

87. Poirier MF. The concept of resistant depression and therapeutic strategies, particularly with venlafaxine. Encephale. 1999;25: 55-57.

88. Reynaert-Dupuis C, Zdanowicz N, Leyman S, Mignon A, Seghers S. Efficacy and tolerance of venlafaxine in depressedpatients switched from prior antidepressant treatment. Prim Care Psychiatry. 2002; 8: 63-68.

89. Saiz-Ruiz J, Ibanez A, Diaz-Marsa M, Arias F, Padin J, Martin-Carrasco M et al. Efficacy of venlafaxine in major depressionresistant to selective serotonin reuptake inhibitors. Prog Neuropsychopharmacol Biol Psychiatry. 2002; 26: 1129-1134.

90. Wan DD, Kundhur D, Solomons K, Yatham LN, Lam RW. Mirtazapine for treatment-resistant depression: a preliminaryreport. J Psychiatry Neurosci. 2003; 28: 55-59.

91. Lavori PW, Rush AJ, Wisniewski SR, Alpert J, Fava M, Kupfer DJ et al. Strengthening clinical effectiveness trials: Equipoise-stratified randomization. Biol Psychiatry. 2001; 15 NOV0150: 792-801.

92. Fava M, Papakostas GI, Petersen T, Mahal Y, Quitkin F, Stewart J et al. Switching to bupropion in fluoxetine-resistantmajor depressive disorder. Ann Clin Psychiatry. 2003; 15: 17-22.

93. Walker PW, Cole JO, Gardner EA, Hughes AR, Johnston JA, Batey SR et al. Improvement in fluoxetine-associated sexualdysfunction in patients switched to bupropion. J Clin Psychiatry. 1993; 54: 459-465.

94. Fava M, McGrath PJ, Sheu WP. Switching to reboxetine: an efficacy and safety study in patients with major depressivedisorder unresponsive to fluoxetine. J Clin Psychopharmacol. 2003; 23: 365-369.

95. Nolen WA, van de Putte JJ, Dijken WA, Kamp JS, et a. Treatment strategy in depression: II. MAO inhibitors in depressionresistant to cyclic antidepressants: Two controlled crossover studies with tranylcypromine versus L-5-hydroxytryptophanand nomifensine. Acta Psychiatr Scand. 1988; 78: 676-683.


5

95

96. Nolen WA, van de Putte JJ, Dijken WA, Kamp JS. L-5HTP in depression resistant to re-uptake inhibitors. An opencomparative study with tranylcypromine. Br J Psychiatry. 1985; 147: 16-22.

97. Kreider MS, Bushnell WD, Oakes R, Wheadon DE. A double-blind, randomized study to provide safety information onswitching fluoxetine-treated patients to paroxetine without an intervening washout period. J Clin Psychiatry. 1995; 56:142-145.

98. Rosenbaum JF, Fava M, Hoog SL, Ascroft RC, Krebs WB. Selective serotonin reuptake inhibitor discontinuation syndrome:a randomized clinical trial. Biol Psychiatry. 1998; 44: 77-87.

99. Judge R, Parry MG, Quail D, Jacobson JG. Discontinuation symptoms: Comparison of brief interruption in fluoxetine andparoxetine treatment. Int Clin Psychopharmacol. 2002; 17: 217-225.

100. Fava M, Rush AJ. Current status of augmentation and combination treatments for major depressive disorder:a literaturereview and a proposal for a novel approach to improve practice. Psychother Psychosom. 2006; 75: 139-153.

101. George MS, Rush AJ, Marangell LB, Sackeim HA, Brannan SK, Davis SM et al. A one-year comparison of vagus nervestimulation with treatment as usual for treatment-resistant depression. Biol Psychiatry. 2005; 58: 364-373.

102. Berton O, Nestler EJ. New approaches to antidepressant drug discovery: beyond monoamines. Nat Rev Neurosci. 2006; 7:137-151.

103. Anderson IM. SSRIS versus tricyclic antidepressants in depressed inpatients: A meta- analysis of efficacy and tolerability.Depress Anxiety. 1998; 7: 11-17.

104. Anderson IM. Meta-analytical studies on new antidepressants. Br Med Bull. 2001; 57: 161-178.105. Bruijn JA, Moleman P, Mulder PG, van den Broek WW, van Hulst AM, van der Mast RC et al. A double-blind, fixed blood-

level study comparing mirtazapine with imipramine in depressed in-patients. Psychopharmacology (Berl). 1996; 127: 231-237.

106. van den Broek WW, Birkenhager TK, Mulder PG, Bruijn JA, Moleman P. A double-blind randomized study comparingimipramine with fluvoxamine in depressed inpatients. Psychopharmacology (Berl). 2004; 175: 481-486.

107. Howland RH, Thase ME. Switching strategies for the treatment of unipolar major depression. In: Rush AJ, ed. Mooddisorders: systematic medication management. Basel, Switzerland: Karger; 1997:56-65.

108. Charney DS, Price LH, Heninger GR. Desipramine-yohimbine combination treatment of refractory depression.Implications for the beta-adrenergic receptor hypothesis of antidepressant action. Arch Gen Psychiatry. 1986; 43: 1155-1161.

109. Reimherr FW, Wood DR, Byerley B, Brainard J, Grosser BI. Characteristics of responders to fluoxetine. PsychopharmacolBull. 1984; 20: 70-72.

110. Amsterdam JD, Maislin G, Potter L. Fluoxetine efficacy in treatment resistant depression. Prog NeuropsychopharmacolBiol Psychiatry. 1994; 18: 243-261.

111. Beasley CM, Jr., Sayler ME, Cunningham GE, Weiss AM, Masica DN. Fluoxetine in tricyclic refractory major depressivedisorder. J Affect Disord. 1990; 20: 193-200.

112. Delgado PL, Price LH, Charney DS, Heninger GR. Efficacy of fluvoxamine in treatment-refractory depression. J AffectDisord. 1988; 15: 55-60.

113. White K, Wykoff W, Tynes LL, Schneider L, Zemansky M. Fluvoxamine in the treatment of tricyclic-resistant depression.Psychiatr J Univ Ott. 1990; 15: 156-158.

114. Gagiano CA, Mueller PG, Fourie J, LeRoux JF. The therapeutic efficacy of paroxetine: (a) An open study in patients withmajor depression not responding to antidepressants; (b) a double-blind comparison with amitriptyline in depressedoutpatients. Acta Neurol Scand. 1989; 80: 130-131.

115. Ghaziuddin N, Naylor MW, King CA. Fluoxetine in tricyclic refractory depression in adolescents. Depression. 1995; 2: 287-291.

116. Ferguson J, Cunningham L, Merideth C, Apter J, Feighner J, Ionescu-Pioggia M et al. Bupropion in tricyclic antidepressantnonresponders with unipolar major depressive disorder. Ann Clin Psychiatry. 1994; 6: 153-160.

117. Stern WC, Harto-Truax N, Bauer N. Efficacy of bupropion in tricyclic-resistant or intolerant patients. J Clin Psychiatry. 1983;44: 148-152.

118. Cole JO, Schatzberg AF, Sniffin C, Zolner J, Cole JP. Trazodone in treatment-resistant depression: an open study. J ClinPsychopharmacol. 1981; 1 (Suppl.): 49-54.

119. Schmauss M, Laakmann G, Dieterle D. Nomifensine: a double-blind comparison of intravenous versus oral administrationin therapy-resistant depressed patients. Pharmacopsychiatry. 1985; 18: 88-90.

120. Nolen WA, Haffmans PM, Bouvy PF, Duivenvoorden HJ. Monoamine oxidase inhibitors in resistant major depression. Adouble-blind comparison of brofaromine and tranylcypromine in patients resistant to tricyclic antidepressants. J AffectDisord. 1993; 28: 189-197.

121. McGrath PJ, Stewart JW, Harrison W, Quitkin FM. Treatment of tricyclic refractory depression with a monoamine oxidaseinhibitor antidepressant. Psychopharmacol Bull. 1987; 23: 169-172.

122. McGrath PJ, Stewart JW, Nunes EV, Ocepek-Welikson K, Rabkin JG, Quitkin FM et al. A double-blind crossover trial ofimipramine and phenelzine for outpatients with treatment-refractory depression. Am J Psychiatry. 1993; 150: 118-123.

123. Thase ME, Frank E, Mallinger AG, Hamer T, Kupfer DJ. Treatment of imipramine-resistant recurrent depression, III:Efficacy of monoamine oxidase inhibitors. J Clin Psychiatry. 1992; 53: 5-11.

124. Leonard BE. Neuropharmacology of antidepressants that modify central noradrenergic and serotonergic function: Ashort review. Hum Psychopharmacol. 1999; 14: 75-81.

125. Shelton RC. The dual-action hypothesis: does pharmacology matter? J Clin Psychiatry. 2004; 65 Suppl 17: 5-10.126. Stahl SM. Symptoms and circuits, part 1: major depressive disorder. J Clin Psychiatry. 2003; 64: 1282-1283.127. Hindmarch I. Beyond the monoamine hypothesis: mechanisms, molecules and methods. Eur Psychiatry. 2002; 17 Suppl 3:

294-299.

Chap

ter

Chapter 5

5

96

128. Owens MJ. Selectivity of antidepressants: from the monoamine hypothesis of depression to the SSRI revolution andbeyond. J Clin Psychiatry. 2004; 65 Suppl 4: 5-10.

129. Burke WJ. Selective versus multi-transmitter antidepressants: are two mechanisms better than one? J Clin Psychiatry.2004; 65 Suppl 4: 37-45.

130. Fava M, Kendler KS. Major depressive disorder. Neuron. 2000; 28: 335-341.131. Delgado PL. How antidepressants help depression: mechanisms of action and clinical response. J Clin Psychiatry. 2004; 65

Suppl 4: 25-30.132. Trivedi MH, Fava M, Wisniewski SR, Thase ME, Quitkin F, Warden D et al. Medication augmentation after the failure of

SSRIs for depression. N Engl J Med. 2006; 354: 1243-1252.133. Keller MB, Lavori PW, Mueller TI, Endicott J, Coryell W, Hirschfeld RM et al. Time to recovery, chronicity, and levels of

psychopathology in major depression. A 5-year prospective follow-up of 431 subjects. Arch Gen Psychiatry. 1992; 49: 809-816.

134. Mueller TI, Leon AC. Recovery, chronicity, and levels of psychopathology in major depression. Psychiatr Clin North Am.1996; 19: 85-102.

135. Lenox-Smith AJ, Schaeffer P, Reynolds A, Williard L. A double blind trial of venlafaxine XR versus citalopram in patientswith severe depression. [Abstract]. J Psychopharmacol. 2001; 15 Suppl 3: A11.

136. Thase ME, Kremer C, Rodrigues HE, et al. Mirtazapine versus sertraline after SSRI non-response. [Abstract] 39th annualmeeting of the American College of Neuropsychopharmacology. Dec 10-14, 2000.San Juan, Puerto Rico. 2000.

137. Thase ME, Kremer C, Rodrigues H. Mirtazapine versus sertraline after SSRI nonresponse [Abstract]. American PsychiatricAssociation.2001 annual meeting: New Orleans LA . 2001.

138. Trivedi MH, Rush AJ, Crismon ML, Kashner TM, Toprac MG, Carmody TJ et al. Clinical results for patients with majordepressive disorder in the Texas Medication Algorithm Project. Arch Gen Psychiatry. 2004; 61: 669-680.

139. Nierenberg AA, DeCecco LM. Definitions of antidepressant treatment response, remission, nonresponse, partialresponse, and other relevant outcomes: a focus on treatment-resistant depression. J Clin Psychiatry. 2001; 62 Suppl 16: 5-9.

140. Nierenberg AA, Papakostas GI, Petersen T, Montoya HD, Worthington JJ, Tedlow J et al. Lithium augmentation ofnortriptyline for subjects resistant to multiple antidepressants. J Clin Psychopharmacol. 2003; 23: 92-95.


5

97

SEROTONIN IN THE ETIOLOGY OF MAJOR DEPRESSIVE DISORDER

PART

Serotonin in the etiology of Major Depressive Disorder

CHA

PTER

MOOD IS INDIRECTLY RELATED TO SEROTONIN, NOREPINEPHRINE AND DOPAMINE LEVELS IN HUMANS: A META-ANALYSIS OF MONOAMINE DEPLETION STUDIES

Molecular Psychiatry 2007; 12: 331-359

Henricus G. Ruhé1

Nada S. Mason1

Aart H. Schene1

1Program for Mood Disorders, Department of Psychiatry, Academic Medical Center, University ofAmsterdam, Amsterdam, The Netherlands

Chap

ter

Chapter 6

6

102

Abstract

BackgroundDysfunction in the monoamine systems of serotonin (5-HT), norepinephrine (NE) and dopamine(DA) may causally be related to Major Depressive Disorder (MDD). Monoamine depletion studiesinvestigate the direct effects of monoamines on mood. Acute tryptophan depletion (ATD) orpara-chlorophenylalanine (PCPA) deplete 5-HT, acute phenylalanine/tyrosine depletion (APTD) oralpha-methyl-para-tyrosine (AMPT) deplete NE/DA. Available depletion studies found conflictingresults in heterogeneous populations: healthy controls, patients with previous MDD in remissionand patients suffering from MDD.

AimTo review and quantify the decrease in mood after 5-HT and NE/DA depletion in humans.

MethodsSystematic search of MEDLINE and EMBASE (1966-October 2006) and cross-references.Randomized studies applying ATD, PCPA, APTD or AMPT versus control depletion were included.Pooling of results by meta-analyses was stratified for studied population and design of the study(within or between subjects).

ResultsSeventy-three ATD, 2 PCPA, 10 APTD and 8 AMPT studies were identified of which 45 ATD and 8APTD studies could be meta-analyzed. 5-HT or NE/DA depletion did not decrease mood in healthycontrols. 5-HT or NE/DA depletion slightly lowered mood in healthy controls with a family historyof MDD. In drug-free patients with MDD in remission, a moderate mood decrease was found forATD, without an effect of APTD. ATD induced relapse in patients with MDD in remission who usedserotonergic antidepressants.

ConclusionsMonoamine depletion studies demonstrate decreased mood in subjects with a family history ofMDD and in drug-free patients with MDD in remission, but do not decrease mood in healthyhumans. Although depletion studies usefully investigate the etiological link of 5-HT and NE withMDD, they fail to demonstrate a causal relation. They presumably clarify a vulnerability trait tobecome depressed. Directions for further investigation of this vulnerability trait are proposed.

ChapterMonoamine depletion in humans

6

103

Introduction

Major depressive disorder (MDD) is characterized by a lowered mood. MDD is a disabling diseasewhich affects 20% of the world’s population.1 MDD is often treated with antidepressants (AD),mostly Selective Serotonin Reuptake Inhibitors (SSRIs), Serotonin Norepinephrin Reuptakeinhibitors (SNRIs), Norepineprine Reuptake Inhbitors (NERIs) or Tricyclic Antidepressants (TCAs).2-5

The working mechanism of AD is believed to be either by (1) increased neurotransmission byincreased synaptic levels of serotonin, norepinephrine and dopamine (monoamines) or (2)specific agonistic effects on serotonin or norepinephrine (sub-)receptors. The increased levels ofmonoamines were discovered in the late fifties, when the TCAs and Monoamine-oxidase Ainhibitors (MAO-I) appeared to effectively treat MDD. This discovery led to the monoaminehypothesis: MDD might etiologically be explained by a deficiency in monoamineneurotransmitters: serotonin (5-HT), norepinephrine (NE) or (to a lesser degree) dopamine (DA).The monoamine systems in the brain have complex interactions. Therefore, the current, lesspertinent view is that the monoamine hypothesis only partially explains MDD and the response toAD.6-10

Depletion of the available 5-HT, NE and/or DA is used as a model to test the involvement ofmonoaminergic systems in MDD. Two reviews recently described and reviewed the techniques ofmonoamine precursor depletion and enzyme-blocking methods.11;12 In brief, 5-HT depletion can beachieved by rapidly lowering the levels of the essential amino-acid tryptophan which cannot besynthesized by the body and must be ingested to enable formation of 5-HT. To achieve depletion,a tryptophan free amino-acid mixture is administered (acute tryptophan depletion (ATD)).13

Depletion of NE and DA uses the same concept (acute depletion of the essential amino-acidsphenylalanine/tyrosine (APTD)).14 As an alternative to induce a state of depletion, enzyme-blocking agents decrease the production of the monoamines. Para-chlorophenylalanine (PCPA)blocks 5-HT synthesis,15;16 and alpha-methyl-para-tyrosine (AMPT) NE and DA synthesis.17 Since1975 an increasing number of depletion studies have been conducted. Monoamine depletionshowed differential effects in different study populations: healthy controls, healthy controls witha family-history of MDD, patients with MDD in remission using AD or after cessation of AD, andpatients who had a current episode of MDD.

Previous reviews summarized the methodology of depletion tests.18 Others reviewed thefindings of specific depletion tests across different psychiatric illnesses,11;18-20 or the prediction ofresponse to depletion.21;22 Booij et al. presented a study that pooled results across studiesindividual subject data of six ATD studies ('mega-analysis') in order to investigate the mediatingrole of clinical, demographic and biochemical characteristics in the mood response to ATD inremitted subjects who previously had MDD.21 However, a clear summary of the mood-effects ofmonoamine depletion across different populations is lacking. Except from the study of Booij et al.,we are unaware of any attempt of pooling studies. Pooling is important because the small-sizeddepletion studies might not have detected small differences by a lack of power. Finally–as in drugresearch–pooling will quantify the weight of positive versus negative studies.

Therefore, we aimed to review the evidence for mood lowering properties of monoaminedepletion studies as a model for MDD. Our main question was: does the depletion of monoamine(5-HT and NE/DA) systems lower mood in humans? A secondary question was: is the lowering ofmood different across different populations? This paper reports our systematic review with astratified meta-analysis of the mood effects of monoamine depletion studies.

Chap

ter

Chapter 6

6

104

Methods

Design of the study We included all available randomized prospective monoamine depletion studies (tryptophan andphenylalanine/tyrosine) and enzyme blocking studies (PCPA or AMPT) in humans. Included studiesmeasured a change in mood after depletion. Two study designs were included. First, randomizedwithin-subjects studies, where each subject was exposed to a true depletion versus a shamintervention at a second occasion. This way, each subject served as his/her own control. Second,randomized between-subjects studies, where two groups of subjects were compared, with theintervention applied to one group and a control intervention to the other group. We excludedanimal studies and studies that selected patients with apparent co-morbidity (e.g. substanceabuse or dependence, studies with only smokers, psychotic disorders, anxiety disorders).Additionally, we excluded studies in depression subtypes with a supposed different etiology (e.g.seasonal affective disorder, bipolar disorder). We also excluded studies that combined depletionwith other interventions (e.g. sleep deprivation) or studies that did not report mood effects.Finally, we excluded studies recruiting a patient group with both unipolar and bipolar depressionwhen >25% of subjects had bipolar depression and when no separate data for the unipolar groupwere provided.

Literature searches and selectionWe searched PubMed from 1966 to October 1st 2006 and EMBASE from 1980 to October 1st 2006using a comprehensive search strategy (Table 6.1). We retrieved additional references identified inprevious reviews11;12;18-20;23-25 and cross-references from identified studies. Two reviewers (H.G.R.and N.S.M.) independently selected articles based on the a priori inclusion and exclusion criteriaas stated above. In case of doubt an article was retrieved and the full content was considered. In85.4% the reviewers agreed on selection instantly. The initial kappa for agreement was 0.61 (95%confidence interval (95% CI): 0.52 - 0.71). Agreement was mainly influenced by a differentialinclusion of research letters, studies in depression subtypes with a supposed different etiologyand studies without apparent mood-scores in the abstract. Discrepancies in included studies werediscussed between both reviewers until full consensus was reached.

Data extraction and Assessment of the studies Two authors (H.G.R. and N.S.M.) abstracted the retrieved studies as follows: population studied,study design (within-subjects or between subjects), applied mood scale(s), number of subjects(and–if applicable–number of subjects for each sequence of intervention and control depletion),male/female ratio, the intervention and its control condition (including dosage and duration).Furthermore, we extracted plasma levels of tryptophan, tyrosine or relevant levels ofmonoamine-metabolites before and after intervention and control condition and other relevant

Table 6.1. Search terms.

# Search terms used

1 Depressive disorder[MeSH] OR depression[MeSH] OR (depressive[TW] AND symptoms[TW]) OR mood[TW]

2 Indoleamine$ OR serotonin[MeSH] OR 5-hydroxytryptamine[TW] OR 5-HT[TW]

3 ((Tryptophan[MeSH] OR Tryptophan[TW]) AND depletion[All Fields]) OR (fenclonine[MeSH] OR PCPA[TW] OR parachlorophenylalanine[TW])

4 #1 AND #2 AND #3

5 Catecholamines[MeSH] OR norepinephrine[MeSH] OR epinephrine[MeSH] OR Dopamine[MeSH]

6 ((Tyrosine[MeSH] OR Tyrosine[TW]) AND depletion[All Fields]) OR (alpha-methyltyrosine[MeSH] OR AMPT[TW] OR alpha-methyl-para-tyrosine[TW])

7 #1 AND #5 AND #6

8 (#4 OR #7) NOT (Animal[MeSH] NOT Human[MeSH])

MeSH = Medical Subject Heading, TW = textword Note: PubMed search terms are provided; these terms were slightly modified for EMBASE-searches.


6

105

data that could influence the outcome of the study (e.g. different consistency of control drink,unclearly reported data, combined interventions). Primary outcome variables were changes inmood-scale scores before and after intervention and control, and the number of relapses in theintervention and control group. If more than one mood-scale was used, we primarily used theProfile of Mood States (POMS).26 Otherwise, we used the Multiple Affect Adjective Checklist(MAACL),27 Visual Analogue [Mood] Scales (VA[M]S),28 Hamilton-depression rating-scale (HDRS),29

and Montgomery-Åsberg Depression Rating Scale (MADRS).30 However, in studies in patientswith previous MDD in remission we primarily chose the HDRS or MADRS followed by the POMS,because most of these studies used depression rating scales to measure effects of depletion. Ifsubscales were used, we took the subscale representing depressed mood.

We validated the abstracted studies (criteria available on request). We based criteria onprevious reviews and the Cochrane handbook.11;12;18;31 We assessed studies as poor, moderate, orgood in the context of this meta-analysis. We particularly assessed quality as good when studiesapplied a randomized double-blind design, when the achieved depletion was judged to besufficient, and when mood-scale ratings were provided.11;18 If more than one of these items wasrated inadequate, we assessed the study as moderate. If studies had one of these items missingand at least one other aspect of the study was not well reported, we also assessed the study asmoderate. If crucial data (e.g. scores and SDs) were missing the study was assessed as poor.

Data synthesisWe qualitatively summarized all included studies in Tables 6.2-6.5 (ATD, PCPA, APTD and AMPTrespectively), irrespective whether the studies were pooled in the meta-analysis. Weacknowledged three clinically heterogeneous study populationsapriori: A. healthy controls (A1negative/ A2 positive family history of MDD), B. patients with a previous MDD currently inremission (B1 currently not using AD / B2 currently using AD) and C. patients with a currentepisode of MDD (with or without AD). These populations were considered and analyzedseparately.

Statistical pooling Mostly, continuous scores for changes in mood rating scales were reported. Some studies inpatients with MDD in remission presented relapse rates. In order to pool continuous effectestimates from different scales, we applied a standardization (providing Hedges’ g). The effect-estimates (1 per study) with the corresponding standard error (SE) were entered in the inverse-variance statistics for pooling in Review Manager 4.2.32;33 The supplementary appendix gives theformulas used to determine Hedge’s g, the difference in relapse rates and corresponding SEs fordifferent study designs.

Different study designs were not combined in pooling. Especially the within subjects designneeds attention in meta-analysis. In this design, differences between the experimental andcontrol condition are statistically paired. As paired data improve power, calculations with datafrom a within subjects (or cross-over) design in a weighted mean differences model would not bejustified.33

Heterogeneity, effect modification, sensitivity-analysis and assessment ofpublication biasWe first performed the meta-analyses with fixed effects models. We assumed morehomogeneous results after our a priori attempt to reduce clinical heterogeneity by stratification.However, if effect-estimates and 95% CI for the individual studies showed graphical poor overlapor consistency, we interpreted this an indication of statistical heterogeneity. We used the chi-squared test and I2 in addition. I2 represents a chi-squared statistic relative to its degree offreedom. An I2 value >50% is indicative of heterogeneity.34 We applied a conservative randomeffects model35 when we suspected statistical heterogeneity.

Chap

ter

Chapter 6

6

106

We investigated effect modification by gender, and the influence of a positive family historyof MDD in healthy controls. Furthermore, we investigated whether different mood scales causeddifferences in outcomes. Therefore, we stratified analyses for these variables, and presentedstratified Hedges’ g, with 95% CI. Differences between strata were tested by subtracting the chi-squared heterogeneity statistic per stratum from the total chi-squared heterogeneity statistic.This residual (Qres) has a chi-squared distribution, with the total number of strata-1 degrees offreedom.

We imputed R= 0.5 to calculate missing pooled SDs within each intervention/control and forthe changes in mood-scores between interventions (supplementary appendix). We checked theimpact of this imputation on the calculated effect estimates. Therefore, we increased R to 0.8(less conservative) and decreased R to 0.2 (more conservative) to compare the new effectestimates with the original findings for R = 0.5.

We assessed publication bias graphically with a funnel-plot, plotting Hedges’ g versus theprecision of the study (the inverse of the standard error of Hedges’ g). Additionally, we testedpublication bias with Galbraith’s radial plot, which regresses the standard normal deviate(Hedges’ g divided by its standard error) against the precision. For a set of studies not distortedby selection bias the intercept of the regression model will be close to zero.36

Results

Identified studiesOur systematic searches identified 392articles. In total we selected 90 studies. Three studiesapplied both ATD and APTD and a control.37-40 Therefore, 73 studies with ATD, 2 with PCPA, 10 withAPTD and 8 with AMPT as monoamine depletion method were identified (summarized in Tables6.2-6.5 respectively). Several studies investigated contrasts in different populations.41-51 A list ofexcluded studies is available on request. Most studies had a within-subjects design (n= 74). Themajority of studies investigated healthy controls (n=64).

Qualitative summary The majority of the 90 studies also investigated other effects of monoamine depletion: 24 studiesmeasured effects on cognitive functions,13;38;43;45;52-67;68-72 11 measured effects on other behavioralmeasures (pain, impulsivity, panic attacks, appetite, noise-stress, aggression),73-83 8 measuredeffects on neuroendocrine parameters,37;60;84-89 11 measured electric encephalogram (EEG)alterations and/or sleep effects,72;90-99 14 measured changes in brain function with PositronEmission Tomography (PET),47;49;51;100-104 Single Photon Emission Computed Tomography(SPECT),105 or Magnetic Resonance Imaging (MRI).66;68-70;106-108 Two studies stratified moodresponse to ATD by genetic polymorphism of the 5-HT transporter promoter region.46;51

We judged the overall methodological quality of the identified studies as good, withappropriate application of the ATD, APTD or AMPT depletion. Two studies with PCPA reportedcase series only and were rated ‘poor’.15;109 In 8 studies no data on the adequacy of the achieveddepletion was reported.15;73;81;88;97;103;109;110 In 2 studies insufficient depletion was attained.63;111 In 10studies no clear SDs or SEs were given for the observed mood-effects.13;76;85;88;92;103;110;112-115 Sixstudies included patients (<25%of total) with Bipolar I or II Disorder.15;65;86;94;114;115 In total, this were14 patients. The reason that several apparently highquality studies were rated as moderate in thisreview was due to the lack of presentation of the mood effects, which in those studies was notthe primary outcome.

In 34 of 90 studies the POMS was used, in 2studies the MAACL, in 34 studies a visual analoguescale, in 29 studies a version of the HDRS, and in 5 of the 90 studies the MADRS. In 5 studies onlyother or undefined mood-scales were used.79;80;97;98;113 Especially for the POMS and the VAS thedirection of a decrease of mood in subjects was not always clearly reported.53;57;77;113 In our meta-analysis no differential effect of the applied mood-scale was observed (Qres= 3.89, df= 2, p>0.05;data not shown).

Chapter6

107

Monoamine depletion in humans

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

A1.

Hea

lthy

con

trol

s

Abb

ott (

1992

)73R

and.

, DB,

Be

tw. S

S2*

30

MN

ot c

lear

ly

repo

rted

POM

S-de

pres

sion

: ATD

: BL:

2.5

0 (±

0.58

SE)

, pos

t-te

st: 1

.73

(±0.

71).

CO

NT:

BL:

2.5

0 (±

0.58

), p

ost-

test

: 1.9

8 (±

0.58

).St

udy

prim

arily

inve

stig

ates

eff

ect o

f ATD

on

anal

gesi

a.M

oder

ate

Alle

n (2

006)

66R

and.

, DB,

W

ithin

SS,

FH

-10

M

+FG

ood

VAS:

No

exac

t sco

res

repo

rted

, no

sign

. diff

eren

ce in

m

ood

foun

dSt

udy

inve

stig

ates

eff

ects

of A

TD o

n co

gniti

on (v

erba

l flu

ency

and

wor

king

m

emor

y) a

nd p

refr

onta

l act

ivity

by

fMRI

Mod

erat

e

Am

in (2

006)

67R

and.

, DB,

W

ithin

SS

19

FG

ood**

POM

S an

d H

DR

S: N

o ex

act s

core

s re

port

ed, n

o si

gn.

diff

eren

ce in

moo

d fo

und

on to

tal P

OM

S or

HD

RS

Stud

y in

vest

igat

es e

ffec

t of A

TD o

n co

gniti

on (m

emor

y, v

isuo

spat

ial l

earn

ing)

an

d th

e pr

otec

tive

effe

cts

of e

stro

gen

ther

apy

(ET)

in p

eri-/

post

- men

opau

sal

wom

en. O

nly

pre

ET-d

ata

used

.

Mod

erat

e

Barr

(199

7)14

5R

and.

, DB,

W

ithin

SS,

FH

-6 M

+FG

ood

VAS-

happ

y: A

TD: B

L: 5

1.2 (±

9.1)

, pos

t-te

st: 4

2.7

(±13

.8).

CO

NT:

51.7

(±25

.9),

pos

t-te

st: 6

2.7

(±12

.2);

PO

MS

not

repo

rted

Stud

y in

vest

igat

es e

ffec

t of f

luox

etin

e tr

eatm

ent o

n AT

D in

hea

lthy

cont

rols

, onl

y pr

e-flu

oxet

ine

data

use

d.

Goo

d

Benk

elfa

t (19

94)41

Ran

d., D

B,

With

in S

S, F

H-

19 MG

ood

POM

S-de

pres

sed:

ATD

: BL:

57.

2 (±

2.0

SE),

pos

t-te

st: 5

8.9

(±2.

0), c

hang

e: 1.

7 (±

1.4).

CON

T: B

L: 5

6.3

(±2.

6), p

ost-

test

: 60

.0 (±

2.1)

, cha

nge:

4.3

(±2.

3). n

o on

e re

ache

d 10

poi

nt

decl

ine

in P

OM

S-D

.

POM

S ch

ange

d in

FH

+ sa

mpl

e, n

ot in

FH

- sa

mpl

e in

sam

e st

udy

(dat

a pr

ovid

ed

sepa

rate

ly).

Goo

d

Bhat

ti (1

998)

90R

and.

, DB,

W

ithin

SS,

FH

-10

M

Goo

dPO

MS-

depr

esse

d: A

TD: B

L:0.

9 (±

1.3),

post

-tes

t: 0

.7 (±

3.3)

, CO

NT:

BL:

0.9

(±1.3

), p

ost-

test

: 0.4

(±1.0

).

Cont

rol-d

rink

is 2

5% in

terv

entio

n dr

ink,

ATD

an

d CO

NT

diff

er >

48hr

s. S

tudy

inve

stig

ates

ef

fect

s of

ATD

on

slee

p EE

G.

Mod

erat

e

Crae

n (2

002)

74R

and.

?, D

B?,

With

in S

S,85

% FH

-

20

M

Goo

d PO

MS-

depr

esse

d: N

o ex

act s

core

s re

port

ed, n

o si

gn.

diff

eren

ce in

moo

d fo

und

Stud

y in

vest

igat

es d

iffer

ence

in im

puls

ivity

af

ter A

TD b

etw

een

cont

rols

with

FH

+ an

d FH

- for

alc

ohol

ism

Poor

Dan

jou

(199

0)84

Ran

d., D

B,

Betw

.SS,

2*

10

MG

ood**

VAS

and

adap

tatio

n of

MA

ACL

, exa

ct s

core

s no

t rep

orte

d,

no s

ign.

diff

eren

ce in

moo

d fo

und

Stud

y fo

cuse

s on

psy

chom

otor

and

ne

uroe

ndoc

riene

eff

ects

of A

TDM

oder

ate

Deb

ener

(200

2)97

Ran

d., D

B,

With

in S

S, F

H-

18

F N

ot te

sted

M

ood

‘que

stio

nnai

res’

did

not

reve

al d

iffer

ence

s in

aff

ect

for A

TD

Stud

y in

vest

igat

es th

e ef

fect

of A

TD o

n au

dito

ry e

voke

d po

tent

ials

with

EEG

Po

or

Elle

nbog

en (1

996)

134

Ran

d., D

B,

With

in S

S, F

H-

21

FG

ood

POM

S-de

pres

sed**

: ATD

: cha

nge

-4 (±

1.6 S

E), C

ON

T:

chan

ge: 0

.8 (±

1.55)

; VA

MS:

Stud

y al

so e

xam

ines

sta

bilit

y of

ATD

re

spon

se w

hich

app

ears

to b

e po

orG

ood

Ever

s (2

005)

107;

108

Ran

d., D

B,

With

in S

S, F

H-

12

M

Goo

d VA

S sa

dnes

s: A

TD: B

L: 1.

8 (±

1.6),

pos

t-te

st: 2

.0 (±

1.6).

CO

NT:

BL:

1.0

(±0.

7), p

ost-

test

: 1.8

(±1.6

).

Stud

y in

vest

igat

es e

ffec

t of A

TD o

n am

ygda

le re

activ

ity to

fear

ful f

aces

with

fM

RI

Mod

erat

e

108

Chap

ter

Chapter 6

6

Ever

s (2

006a

)68;6

9R

and.

, DB,

W

ithin

SS

13

M

Goo

d PO

MS:

No

exac

t sco

res

repo

rted

, no

sign

. diff

eren

ce in

m

ood

foun

d St

udy

inve

stig

ates

eff

ects

of A

TD o

n co

gniti

on (p

erfo

rman

ce m

onito

ring

and

resp

onse

inhi

bitio

n an

d m

emor

y) a

nd

pref

ront

al /

hipp

ocam

pal a

ctiv

ity b

y fM

RI

Mod

erat

e

Ever

s (2

006b

)70R

and.

, DB,

W

ithin

SS,

FH

- 15

F

Goo

d PO

MS:

No

exac

t sco

res

repo

rted

, no

sign

. diff

eren

ce in

m

ood

foun

d St

udy

inve

stig

ates

eff

ects

of A

TD o

n a

com

bine

d co

gniti

ve a

nd e

mot

iona

ltask

by

fMRI

Mod

erat

e

Gal

lagh

er (2

003)

52R

and.

, DB,

W

ithin

SS

15 MG

ood

VAM

S-co

nten

tedn

ess-

scal

e: s

core

s no

t com

plet

ely

repo

rted

. ATD

: pos

t-te

st: 8

3.4

(±7.

5),C

ON

T: p

ost-

test

: 86.

5 (±

7.8)

No

sign

. cha

nges

in m

ood.

Stud

y fo

cuse

s on

exe

cutiv

e ne

uroc

ogni

tive

func

tions

Mod

erat

e

Har

riso

n (2

002)

37R

and.

, DB,

W

ithin

SS,

FH

-13 F

Goo

d PO

MS-

depr

esse

d:AT

D: B

L: 2

.07

(±0.

63 S

E), p

ost-

test

: 0.7

6 (±

0.63

) CO

NT:

BL:

1.61

(±0.

52),

post

-tes

t: 1.

15 (±

0.56

).

Stud

y al

so re

port

s ab

senc

e of

eff

ects

of A

TD

and

APT

D o

n in

terle

ukin

-6 a

ctiv

atio

nG

ood

Har

rison

(200

4)38

Ran

d., D

B,

With

in S

S, F

H-

13 FG

ood

VAM

S: e

xact

sco

res

not r

epor

ted,

no

sign

. diff

eren

ces

in

moo

d fo

und

Stud

y al

so in

vest

igat

es m

emor

y an

d co

gniti

ve e

ffec

ts o

f ATD

and

APT

DM

oder

ate

Hay

war

d (2

005)

43R

and.

, DB,

Be

tw. S

S, 2

0 FH

-2*

12

M+F

G

ood

HD

RS: A

TD: B

L: 0

.17

(±0.

11 S

E), p

ost-

test

: 0.6

7 (±

0.19

),

CON

T: B

L: 0

.33

(±0.

19),

pos

t-te

st: 0

.41 (

±0.19

). B

DI:

ATD

: BL

: 1.5

(±0.

56 S

E), p

ost-

test

: 0.5

8 (±

0.33

), C

ON

T: B

L: 2

.13

(±0.

57),

pos

t-te

st: 2

.25

(±0.

80).

PO

MS

and

VAS:

exa

ct

scor

es n

ot re

port

ed.

No

sign

ific

ant e

ffec

t on

any

scal

e fo

und.

Stud

y al

so in

vest

igat

es c

ogni

tive

proc

essi

ng

in fo

rmer

ly d

epre

ssed

ver

sus

heal

thy

cont

rols

dur

ing

ATD

Mod

erat

e

Hug

hes

(200

0)98

Ran

d., D

B,

With

in S

S, F

H-

20

M

Goo

d U

ndef

ined

sca

le: n

o si

gnif

ican

t cha

nges

in m

ood

foun

d St

udy

inve

stig

ates

eff

ects

of A

TD o

n au

dito

ry e

voke

d po

tent

ials

on

EEG

Po

or

Hug

hes

(200

3)53

Ran

d., D

B,

With

in S

S, F

H-

20 MG

ood

VAS-

sadn

ess:

ATD

: BL:

90.

7 (±

3.5

SE),

pos

t-te

st: +

6.7

(±4.

7),

CON

T: B

L: 9

4.8

(±2.

6), p

ost-

test

: -1.6

(±1.9

). N

o si

gnifi

cant

m

ain

effe

cts

on a

ny V

AM

S sc

ale.

Stud

y in

vest

igat

es e

ffec

ts o

f ATD

on

verb

al

and

visu

ospa

tial l

earn

ing

and

mem

ory,

at

tent

ion

and

exec

utiv

e fu

nctio

n

Goo

d

Kähk

önen

(200

5)91

Ran

d., D

B,

With

in S

S14

M

+FG

ood

VAS-

depr

essi

on: A

TD: c

hang

e 41

.5 (±

20.7

), C

ON

T: c

hang

e 21

.5 (±

17.6

)St

udy

inve

stig

ates

eff

ects

of A

TD o

n au

dito

ry re

spon

se, M

EG a

nd E

EG s

igna

lsM

oder

ate

Klaa

ssen

(199

9a)14

6R

and.

, DB,

W

ithin

SS

13 M+F

Goo

d PO

MS-

depr

esse

d:AT

D: B

L: 0

.6 (±

1.3),

pos

t-te

st: 1

.5 (±

3.3)

, CO

NT:

BL:

0.5

(±1.0

), p

ost-

test

: 0.2

(±0.

6). V

AS-

depr

esse

d:AT

D: B

L: 2

.7 (±

6.6)

, pos

t-te

st: 8

.2 (±

18.2

), CO

NT:

BL

: 3.2

(±6.

3), p

ost-

test

: 4.5

(±13

.0).

Stud

y in

vest

igat

es s

peci

fici

ty o

f ATD

vs

LYS-

depl

etio

n on

moo

d-ef

fect

s an

d m

emor

y.

Goo

d

Klaa

ssen

(199

9b),

44

Klaa

ssen

(200

2)45

Ran

d., D

B,

With

in S

S, F

H+

& F

H-

11 F

H-

M +

FG

ood

POM

S-de

pres

sed:

sco

res

not c

ompl

etel

y re

port

ed. A

TD-

CON

T di

ffer

ence

s: F

H- 0

.0 (±

0.45

)N

o ba

selin

e PO

MS-

scor

es p

rovi

ded

(‘no

n-si

gnifi

cant

diff

eren

t ces

’). S

tudy

com

pare

d FH

+ gr

oup

with

FH

-. Th

e FH

+ ve

s re

spon

ded

to tr

ypto

phan

dep

letio

n w

ith a

sig

nifi

cant

lo

wer

ing

of P

OM

S sc

ores

, but

FH

- ves

did

no

t. A

lso

mem

ory

effe

cts

mea

sure

d.

Mod

erat

e

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

Chapter6

109


Kosz

ycki

(199

6)75

Ran

d., D

B,

Betw

. SS,

FH

-2*

20M

Goo

d VA

MS-

depr

esse

d:AT

D: B

L: 1.

5 (±

1.0 S

E), p

ost-

test

: 0.8

(±

0.6)

. CO

NT:

BL:

3.4

(±1.8

), p

ost-

test

: 4.5

(±1.4

).

Stud

y in

vest

igat

es e

ffec

t of a

dmin

istr

atio

n of

cho

lecy

stin

e-te

trap

eptid

e af

ter A

TD o

n oc

curr

ence

of p

anic

-att

acks

Goo

d

Leyt

on (1

999)

,39

Leyt

on (2

000)

40R

and.

, DB,

Be

tw. S

S, F

H-

15 A

TD

14 C

ON

T F

Goo

d PO

MS**

: ATD

: cha

nge:

-3.2

8 (±

1.4 S

E). C

ON

T: c

hang

e -1.

04

(±1.2

SE)

. VA

MS-

depr

esse

d: A

TD: B

L: 1.

0 (±

0.9)

, pos

t-te

st:

0.8

(±0.

8), c

hang

e: -0

.3 (±

0.8)

. CO

NT:

BL:

0.4

(±0.

5), p

ost-

test

: 0.4

(±0.

4), c

hang

e: 0

.1 (±

0.4)

.

POM

S; n

o ex

act s

core

s re

port

ed. T

hird

arm

w

ith A

PTD

incl

uded

Mod

erat

e

McA

llist

er-W

illia

ms

(200

2)72

Ran

d., D

B,

With

in S

S, F

H-

14

M

Goo

d PO

MS:

No

exac

t sco

res

repo

rted

, no

sign

. diff

eren

ce in

m

ood

foun

d St

udy

inve

stig

ates

eff

ects

of A

TD o

n ev

ent

rela

ted

brai

n po

tent

ials

on

EEG

dur

ing

an

epis

odic

mem

ory

task

Mod

erat

e

Mill

er (2

000)

76R

and.

, DB,

W

ithin

SS,

FH

-19

M

+FG

ood

POM

S**: A

TD: B

L: 1.

8, p

ost-

test

: 0.3

. CO

NT:

BL:

2.9

, pos

t-te

st: 0

.7.V

AS-

sad**

: ATD

: BL:

8.0

, pos

t-te

st: 1

0.1.

CON

T: B

L:

12.2

, pos

t-te

st: 1

0.9.

No

SDs

repo

rted

, no

sign

. diff

eren

ces

in m

ood

foun

d

No

clea

r pre

sent

atio

n of

moo

d-sc

ores

; stu

dy

com

pare

s AT

D+

CO2-

chal

leng

e in

hea

lthy

cont

rols

vs.

pat

ient

s w

ith p

anic

dis

orde

r

Poor

Mor

eno

(199

9),11

6 M

oren

o (2

000)

130

Ran

d., D

B,

With

in S

S, F

H-

12 M+F

Goo

d H

DRS

-25,

PO

MS

and

IDS

scor

es n

ot c

lear

ly re

port

ed. A

TD

did

not c

ause

sig

nifi

cant

incr

ease

s in

dep

ress

ive

sym

ptom

s in

hea

lthy

cont

rol s

ubje

cts.

Stud

y co

mpa

red

ATD

eff

ects

in h

ealth

y co

ntro

ls v

s. p

atie

nts

with

MD

D in

rem

issi

on.

Inst

ead

of p

lace

bo a

25%

str

engt

h de

plet

ion

drin

k w

as u

sed

as a

con

trol

.

Poor

Mur

phy

(200

2)54

Ran

d., D

B,

With

in S

S11

F

Goo

d Sc

ores

repo

rted

sep

arat

ely

for g

roup

s w

ith A

TD f

irst a

nd

CON

T fi

rst.

VA

MS-

sadn

ess:

ATD

firs

t:AT

D: c

hang

e: -1

.4

(±5.

5), C

ON

T: c

hang

e: -0

.2 (±

0.5)

. CO

NT

firs

t: A

TD: c

hang

e:

0.8

(±26

.6),

CO

NT:

cha

nge:

-12.

5 (±

26.1)

Wei

ghte

d av

erag

ed s

core

(with

hig

hest

SD

s) u

sed

for m

eta-

anal

ysis

Stud

y in

vest

igat

es c

ogni

tive

and

emot

iona

l pr

oces

sing

aft

er A

TDG

ood

Neu

mei

ster

(200

2)46

Ran

d., D

B,

With

in S

S, F

H-

24

FG

ood

HD

RS-2

1: s/

s sub

type

(n=

4): A

TD: c

hang

e 8.

5 (±

2.9)

. CO

NT:

0.

3 (±

0.5)

s/l

subt

ype

(n=

10):

ATD

: cha

nge

4.6

(±2.

9).

CON

T: 0

.5 (±

0.9)

l/l s

ubty

pe**

(n=

10):

ATD

: cha

nge

-0.1

(±1.

3). C

ON

T: 0

.0 (±

1.1)

Stud

y in

vest

igat

es e

ffec

t of b

i-alle

lic

sero

toni

n tr

ansp

orte

r pro

mot

er (5

-HTT

PR)

geno

type

and

eff

ect o

f ATD

Goo

d

Neu

mei

ster

(200

4)47

Ran

d., D

B,

With

in S

S, F

H-

19

M+F

Goo

dH

DRS

-24**

: ATD

: BL:

0.5

(±0.

1), p

ost-

test

: 3.4

(±0.

6). C

ON

T:

BL: 0

.5 (±

0.1)

, pos

t-te

st: 2

.6 (±

0.5)

.St

udy

mai

nly

inve

stig

ates

regi

onal

cer

ebra

l bl

ood-

flow

cha

nges

dur

ing

ATD

with

PET

-sc

ans

Goo

d

Neu

mei

ster

(200

6)51

Ran

d., D

B,W

ithin

SS,

FH

- 26

M

+F

Goo

d H

DRS

**:

s/s

subt

ype

(n=

7): A

TD: B

L: 0

.4 (±

0.4

SE),

post

-te

st: 3

.6 (±

1.4).

CON

T: B

L: 0

.8 (±

0.4)

, pos

t-te

st: 2

.4 (±

0.8)

. s/

l sub

type

(n=

12):

ATD

: BL:

0.4

(±0.

4), p

ost-

test

: 5.2

(±1.1

).

CON

T: B

L: 0

.4 (±

0.4)

, pos

t-te

st: 2

.5 (±

0.6)

. l/l

subt

ype

(n=

7): A

TD: B

L: 0

.4 (±

0.4)

, pos

t-te

st: 2

.4 (±

1.3).

CO

NT:

BL:

0.2

(±

0.4)

, pos

t-te

st: 1

.2 (±

0.8)

.

Stud

y in

vest

igat

es e

ffec

t of t

ri-al

lelic

se

roto

nin

tran

spor

ter p

rom

oter

(5-H

TTPR

) ge

noty

pe a

nd e

ffec

t of A

TD o

n m

ood

and

regi

onal

cer

ebra

l met

abol

ism

with

PET

Goo

d

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

110

Chap

ter

Chapter 6

6

Old

man

(199

4)77

Ran

d., D

B,

With

in S

S12 F

Goo

d PO

MS-

depr

esse

d: A

TD: B

L: 1.

1 (±0

.52

SE),

post

-tes

t: 0

.9

(±0.

58).

CO

NT:

BL:

1.9

(±0.

92),

pos

t-te

st: 2

.3 (±

0.89

).VA

S-sa

d:AT

D: B

L: 15

.0 (±

4.99

SE)

, pos

t-te

st: 9

.2 (±

4.5)

. CO

NT:

BL

: 11.7

(±4.

73),

pos

t-te

st: 8

.3 (±

3.87

).

In th

is s

tudy

a s

econ

d pl

aceb

o-ar

m (p

lain

w

ater

) is

used

; com

paris

ons

here

are

for

bala

nced

drin

k as

CO

NT.

Eff

ects

of A

TD o

n ap

petit

e in

vest

igat

ed a

lso.

Goo

d

Park

(199

4)55

Ran

d., D

B,

With

in S

S12 M

Goo

d VA

MS-

sad:

ATD

: BL:

6.7

(±5.

7 SE

), po

st-t

est:

6.6

(±8.

1).

CON

T: B

L: 3

.2 (±

3.0)

, pos

t-te

st: 5

.6 (±

6.8)

. St

udy

inve

stig

ates

cog

nitiv

e pe

rfor

man

ce

afte

r ATD

Goo

d

Pras

chak

-Rie

der

(200

5)10

0R

and.

, DB,

W

ithin

SS,

FH

-14

M

+FG

ood

No

exac

t VA

S-ra

tings

pro

vide

d in

stu

dy; “

none

of t

he 14

su

bjec

ts e

xper

ienc

ed a

tran

sien

t det

erio

ratio

n in

moo

d [..

] le

vels

.

Stud

y pr

imar

ily in

vest

igat

es e

ffec

ts o

f ATD

on

ser

oton

in tr

ansp

orte

r den

sity

/aff

inity

du

ring

ATD

Mod

erat

e

Ravi

ndra

n (1

999)

,85

Knot

t (19

99)92

Ran

d., D

B,

Betw

. SS,

FH

-2*

13

MG

ood

POM

S-ch

ange

:ATD

-1.2

9 (±

0.94

) vs

CON

T +1

.41(

±0.5

5).

Sign

ific

ant c

hang

es in

moo

d, b

ut th

is o

ccur

red

in 5

of t

he

13 s

ubje

cts

only

, whi

le in

oth

ers

in th

e AT

D-g

roup

no

decr

ease

of m

ood

was

obs

erve

d.

No

clea

r pre

sent

atio

n of

dat

a on

moo

d-sc

ores

; pro

babl

y SD

s re

port

ed. S

tudy

als

o in

vest

igat

es e

ffec

t of f

enflu

ram

ine

adm

inis

trat

ion

afte

r ATD

and

eff

ect o

f ATD

on

imm

une

mea

sure

s an

d EE

G

Mod

erat

e

Ric

hell

(200

5)78

Ran

d., D

B,

Betw

. SS

15 +

13

M+F

Goo

dN

o ba

selin

e da

ta o

n PO

MS

or V

AS

repo

rted

, onl

y m

easu

res

durin

g th

e no

ise-

stre

ss p

arad

igm

repo

rted

aft

er

ATD

, no

sign

. diff

eren

ces

in m

ood

foun

d.

Stud

y es

peci

ally

inve

stig

ates

diff

eren

ce in

re

spon

se to

noi

ce-s

tres

s by

ATD

vs

CON

T.Po

or

Rubi

a (2

005)

106

Ran

d., D

B,

With

in S

S,FH

-9 M

+FG

ood

No

VAS-

scor

es re

port

ed, n

o si

gn. c

hang

es o

f moo

d ob

serv

ed b

y AT

DSt

udy

inve

stig

ates

func

tiona

l bra

in a

ctiv

ity

by fM

RI w

ith a

resp

onse

inhi

bitio

n-ta

skPo

or

Rub

insz

tein

(200

1)56

Ran

d., D

B,

Betw

. SS,

80%

FH

-2*

15

M+F

Goo

d PO

MS-

depr

esse

d: n

o ex

act s

core

s re

port

ed. S

core

s w

ere

sim

ilar b

etw

een

test

and

pla

cebo

gro

ups,

als

o on

VA

MS,

no

sig

n. d

iffer

ence

s in

moo

d fo

und.

Stud

y in

vest

igat

es e

ffec

ts o

f ATD

on

cogn

itive

test

con

cern

ing

atte

ntio

nPo

or

Schm

eck

(200

2)79

Ran

d., D

B,

Betw

. SS,

FH

-12 M

+FG

ood

Eige

nsch

afts

wör

terli

ste:

no

exac

t sco

res

repo

rted

. Hig

h ag

gres

sive

wom

en w

ere

sign

. mor

e de

pres

sed

afte

r ATD

th

an o

ther

gro

ups.

Hig

h ag

gres

sive

men

did

not

sho

w a

n ef

fect

of A

TD o

n m

ood.

The

moo

d of

non

-agg

ress

ive

men

an

d w

omen

did

not

cha

nge

afte

r ATD

.

Stud

y lo

oked

at d

iffer

ent r

eact

ion

to A

TD in

hi

gh a

nd lo

w-a

ggre

ssiv

e m

ales

and

fem

ales

.Po

or

Schm

itt (

2000

)57R

and.

,DB,

W

ithin

SS,

FH

-17

M

+FG

ood

POM

S-de

pres

sion

: ATD

: BL:

0.9

2 (±

0.11

SE)

,pos

t-te

st: 0

.93

(±0.

12).

CO

NT:

BL:

0.9

5 (±

0.11

), p

ost-

test

: 0.9

6 (±

0.13

) St

udy

inve

stig

ates

cog

nitiv

e pe

rfor

man

ce

afte

r ATD

Goo

d

Shan

sis

(200

0)58

Ran

d.,D

B,

With

in S

S, 7

/12

FH-

12 MG

ood

POM

S-de

pres

sed

& V

AS:

no

exac

t sco

res

repo

rted

. Moo

d ch

ange

s on

con

trol

day

not

sig

n. d

iffer

ent t

han

on A

TD

day

for a

ny o

f the

six

dim

ensi

ons

in th

e PO

MS

and

VAS.

Prev

ious

moo

d-di

sord

er in

one

sub

ject

(not

ex

clud

ed).

Stud

y in

ves

tigat

es e

ffec

ts o

f ATD

on

att

entio

n an

d m

emor

y

Poor

Smith

(198

7)11

2R

and.

,DB,

W

ithin

SS

80

MG

ood

MA

ACL

**:A

TD: B

L: 10

.9, p

ost-

test

: 18.

2. C

ON

T: B

L: 9

.5,

post

-tes

t: 8

.2. N

o SD

s re

port

ed. S

igni

fica

nt lo

wer

ing

of

moo

d sc

ores

by

ATD

.

Stud

y in

vest

igat

es A

TD in

neg

ativ

e an

d po

sitiv

e en

viro

nmen

t. B

oth

cond

ition

s sh

ow

a si

gnif

ican

t cha

nge

on m

ood

scor

es.

Mod

erat

e

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

Chapter6

111


Smit

h (1

997b

)147

Ran

d.,D

B,

With

in S

S, F

H-

11

M+F

Goo

d VA

S-sa

dnes

s:no

exa

ct s

core

s re

port

ed. N

o si

gnif

ican

t m

ain

effe

cts

wer

e se

en. V

AS-

happ

y fo

r fem

ales

(n=6

) re

port

ed s

epar

atel

y**: A

TD: B

L: 6

9.7

(±5.

6 SE

), po

st-t

est:

61

.6 (±

6.4)

. CO

NT:

BL:

71.4

(±5.

3), p

ost-

test

: 66.

1 (±6

.1).

For m

ales

(n=5

)**: A

TD: B

L: 7

2.2,

pos

t-te

st: 6

7.8.

CO

NT:

BL:

77

.2, p

ost-

test

: 70.

6. n

o SE

s re

port

ed

Stud

y al

so in

vest

igat

es e

ffec

t of A

TD w

hen

sad-

moo

d is

indu

ced.

Dat

a fo

r fem

ales

VA

S-ha

ppy

used

in m

eta-

anal

ysis

Mod

erat

e

Stew

art (

2002

)50R

and.

,DB,

W

ithin

SS,

FH

-27

M

+FG

ood

POM

S-de

pres

sion

:ATD

: BL:

1.7

(±2.

6), p

ost-

test

: 1.1

(±2.

3),

chan

ge: -

0.6

(±2.

1). C

ON

T: B

L: 3

.4 (±

6.0)

, pos

t-te

st: 2

.4

(±4.

8), c

hang

e: -1

.0 (±

4.6)

.

Stud

y al

so in

vest

igat

es e

ffec

t mod

ifica

tion

by h

igh

vs lo

w n

euro

ticis

m s

core

s du

ring

ATD

(on

moo

d an

d ps

ycho

met

ric

perf

orm

ance

).

Goo

d

Talb

ot (2

006a

)71R

and.

, DB,

Be

tw. S

S, F

H-

17 +

15

M+F

G

ood

VAM

S: N

o ex

act s

core

s re

port

ed, n

o si

gn. d

iffer

ence

in

moo

d fo

und

Stud

y in

vest

igat

es e

ffec

ts o

f ATD

on

deci

sion

m

akin

g an

d le

arni

ng

Mod

erat

e

Talb

ot (2

006b

)105

Ran

d., D

B,

With

in S

S, F

H-

16

M

Goo

d VA

MS:

ATD

: cha

nge

-0.5

(±6.

1). C

ON

T: c

hang

e -5

.7 (±

8.1)

BD

I: AT

D: c

hang

e -0

.0 (±

1.4)

. CO

NT:

cha

nge

0.4

(±1.

5)

Stud

y in

vest

igat

es e

ffec

t of A

TD o

n R

egio

nal

Cere

bral

Blo

od F

low

by

SPEC

T G

ood

Vode

rhol

zer

(199

8)99

Ran

d., D

B,

With

in S

S, F

H-

12

M+F

G

ood

HD

RS-6

: ATD

: BL:

0.8

3 (±

1.8),

pos

t-te

st: 0

.42

(±1.2

). CO

NT:

BL

: 0.17

(±0.

4), p

ost-

test

: 0.4

2 (±

0.7)

. St

udy

inve

stig

ates

eff

ects

of A

TD o

n sl

eep

EEG

G

ood

Wel

tzin

(199

4)11

3††

Ran

d.,D

B,

With

in S

S9 F

Goo

dO

nly

peak

-cha

nges

for A

TD a

nd C

ON

T ar

e pr

ovid

ed fo

r an

uncl

early

def

ined

moo

d-sc

ale:

ATD

: 0.7

8. C

ON

T 0.

0. S

Ds

not r

epor

ted.

SD

of d

iffer

ence

est

imat

ed c

onse

rvat

ivel

y fr

om p

< 0.

05: ±

1.01.

Stud

y in

vest

igat

es e

ffec

t of A

TD o

n m

ood

in

bulim

ic p

atie

nts

(n=1

3) a

nd h

ealth

y co

ntro

ls

(n=

9).

Mod

erat

e

Wel

tzin

(199

5)80

††R

and.

,DB,

W

ithin

SS

10

FG

ood

Onl

y pe

ak-c

hang

es b

etw

een

ATD

and

CO

NT

are

prov

ided

fo

r an

uncl

early

def

ined

moo

d-sc

ale:

0.0

(±1.3

), th

is v

alue

us

ed in

met

a-an

alys

is

Stud

y in

vest

igat

es e

ffec

t of A

TD o

n m

ood

and

shor

t-te

rm e

atin

g be

ha v

iour

in b

ulim

ic

patie

nts

(n=1

0) a

nd h

ealth

y co

ntro

ls (n

= 10

).

Mod

erat

e

Yath

am (2

001)

104

Ran

d., D

B,

With

in S

S, F

H-

10

F G

ood

HD

RS-2

0, P

OM

S: N

o ex

act s

core

s re

port

ed, n

o si

gn.

diff

eren

ce in

moo

d fo

und

Stud

y in

vest

igat

es e

ffec

ts o

f ATD

on

5-H

T 2

rece

ptor

s w

ith P

ETM

oder

ate

Youn

g (1

985)

13R

and.

, DB,

Be

tw. S

S3*

12

MG

ood

No

clea

r MA

ACL

-sco

res

or S

Ds

repo

rted

per

gro

up fo

r ATD

vs

CO

NT.

Sig

n. d

ecre

ase

of m

ood

afte

r ATD

. St

udy

inve

stig

ates

eff

ects

of A

TD o

n pr

oofr

eadi

ng ta

sk w

ith o

r with

out

dist

ract

ion

Poor

A2.

Hea

lthy

con

trol

s w

ith

+ve

fam

ily h

isto

ry fo

r MD

D

Benk

elfa

t (19

94)41

Ran

d., D

B,

With

in S

S, F

H+

20 MG

ood

POM

S-de

pres

sed:

ATD

: BL:

55.

2 (±

1.7 S

E), p

ost-

test

: 51.2

(±

1.6),

cha

nge:

-4.0

(±1.

6). C

ON

T: B

L: 5

4.0

(±1.6

), p

ost-

test

: 57

.3 (±

1.1),

cha

nge:

3.3

(±1.6

). S

ix (3

0%) o

f peo

ple

show

ed a

de

clin

e of

10 p

oint

s or

gre

ater

on

POM

S-D

.

POM

S ch

ange

d in

FH

+ sa

mpl

e, n

ot in

FH

- sa

mpl

e in

sam

e st

udy.

G

ood

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

112

Chap

ter

Chapter 6

6

Elle

nbog

en (1

999)

135

Ran

d., D

B,

With

in S

S, F

H+

13

FG

ood

POM

S-de

pres

sed**

: 1st

ATD

: cha

nge:

-0.3

(±1.

7 SE

). CO

NT:

ch

ange

: -1.1

(±1.5

). V

AM

S:no

exa

ct s

core

s re

port

ed. N

o si

gnif

ican

t cha

nge

in m

ood

on a

ny o

f the

PO

MS

or V

AM

S ite

ms.

Poss

ible

sel

ectio

n-bi

as b

y ex

clus

ion

of m

any

form

er p

atie

nts

men

tione

d: e

xtre

mel

y FH

+ sa

mpl

e, b

ut n

ever

dep

ress

ed. S

tudy

foun

d po

or te

mpo

ral s

tabi

lity

of A

TD in

repe

ated

te

stin

g.

Goo

d

Klaa

ssen

(199

9b),

44

Klaa

ssen

(200

2)45

Ran

d., D

B,

With

in S

S, F

H+

& F

H-

16 F

H+

M+F

Goo

d PO

MS-

depr

esse

d: s

core

s no

t com

plet

ely

repo

rted

. ATD

-CO

NT

diff

eren

ces:

FH

+ 1.9

4 (±

3.17

)Se

e ab

ove

in s

ectio

n “h

ealth

y co

ntro

ls”

Mod

erat

e

Neu

mei

ster

(200

2)46

Ran

d., D

B,

With

in S

S, F

H+

21

FG

ood

HD

RS-2

1: s

/s s

ubty

pe (n

= 5)

: ATD

: cha

nge

9.2

(±2.

1). C

ON

T:

0.8

(±0.

5) s

/l su

btyp

e (n

= 7)

: ATD

: cha

nge

10.0

(±4.

5).

CON

T: 0

.1 (±

0.4)

l/l s

ubty

pe**

(n=

9): A

TD: c

hang

e 0.

4 (±

1.2).

CON

T: -0

.2 (±

1.3)

Stud

y in

vest

igat

es e

ffec

t of b

i-alle

lic

sero

toni

n tr

ansp

orte

r pro

mot

er (5

-HTT

PR)

geno

type

and

eff

ect o

f ATD

Goo

d

Stew

art (

2002

)50R

and.

,DB,

W

ithin

SS,

FH

-5 M

+FG

ood

POM

S-de

pres

sion

:ATD

: BL:

5.2

(±10

.0),

pos

t-te

st: 5

.6

(±10

.4),

cha

nge:

0.4

(±0.

5). C

ON

T: B

L: 6

.4 (±

12.7

), p

ost-

test

: 4.

2 (±

7.4)

, cha

nge:

-2.2

(±5.

5).

See

abov

e in

sec

tion

“hea

lthy

cont

rols

”G

ood

B1. P

atie

nts

wit

h M

DD

pre

viou

sly,

but

cur

rent

ly in

rem

issi

on (w

itho

ut m

edic

atio

n)

Cass

idy

(199

7)11

7R

and.

, DB,

W

ithin

SS

1-4 d

ays

afte

r EC

T

5

M+F

R

x-fr

ee

<3 m

ths

Goo

dM

AD

RS:

ATD

: BL:

4.2

(±3.

5), p

ost-

test

: 3.6

(±3.

2). C

ON

T: B

L:

4.8

(±2.

9), p

ost-

test

: 4.0

(±3.

5).

Stud

y in

vest

igat

es re

curr

ence

of M

DD

aft

er

ATD

in p

atie

nts

succ

essf

ully

trea

ted

with

ECT

Goo

d

Del

gado

(199

1)42

Ran

d., D

B,

With

in S

S ≥2

wee

ks a

fter

‘s

tabi

lity’

69

M+F

R

x-fr

ee

?? m

ths

Goo

d H

DRS

-25:

exa

ct s

core

s no

t rep

orte

d; n

o si

gn. d

iffer

ence

by

ATD

. 30%

exp

erie

nced

an

impr

ovem

ent o

f moo

d th

e da

y af

ter A

TD. D

epre

ssiv

e re

laps

e: N

o ex

act d

ata

prov

ided

.

In o

rigin

al s

tudi

es B

ipol

ar p

atie

nts

incl

uded

; un

clea

r whe

ther

thes

e w

ere

excl

uded

in th

is

pool

ed a

naly

sis

Poor

Hay

nes

(200

4)93

Ran

d., D

B,

With

in S

S 1-2

wee

ks

resp

onse

13

M+F

Po

st- C

BT

Goo

d PO

MS-

depr

essi

on: A

TD: B

L: 6

.4 (±

7.4)

, pos

t-te

st (8

hr):

4.9

(±

6.4)

. CO

NT:

BL:

5.5

(±7.

7), p

ost-

test

: 5.0

(±7.

8).

HD

RS-

19:

ATD

: BL:

4.7

(±3.

7), p

ost-

test

: 4.9

(±3.

9). C

ON

T: B

L: 3

.3

(±2.

5), p

ost-

test

: 3.5

(±4.

3).

Stud

y in

vest

igat

es re

curr

ence

of M

DD

aft

er

ATD

in p

atie

nts

succ

essf

ully

trea

ted

with

CB

T, a

nd e

ffec

ts o

n RE

M-s

leep

Mod

erat

e

Hay

war

d (2

005)

43R

and.

, DB,

Be

tw. S

S, 11

FH

- ≥6

mon

ths

rem

issi

on

2*12

M

+F

Rx-

free

≥3

mth

s

Goo

dH

DRS

: ATD

: BL:

1.0

(±0.

41 S

E), p

ost-

test

: 1.3

8 (±

0.36

), CO

NT:

BL:

0.4

2 (±

0.23

), p

ost-

test

: 0.5

0 (±

0.19

). B

DI:

ATD

: BL

: 2.2

5 (±

0.76

SE)

, pos

t-te

st: 1

.75

(±0.

68),

CO

NT:

BL:

2.17

(±

0.63

), p

ost-

test

: 1.6

3 (±

0.53

). P

OM

S an

d VA

S: e

xact

sc

ores

not

repo

rted

. N

o si

gnif

ican

t eff

ect o

n an

y sc

ale

foun

d.

Stud

y al

so in

vest

igat

es c

ogni

tive

proc

essi

ng

in fo

rmer

ly d

epre

ssed

ver

sus

heal

thy

cont

rols

dur

ing

ATD

Mod

erat

e

Mor

eno

(199

9),11

6 M

oren

o (2

000)

130

Ran

d., D

B,

With

in S

S ≥3

mon

ths r

e m

is

sion

(ave

rage

29

mon

ths)

12

M+F

R

x-fr

ee

≥3 m

ths

Goo

dH

DRS

-25,

PO

MS

and

IDS

scor

es n

ot c

lear

ly re

port

ed. A

TD

caus

ed n

on-s

igni

fican

t inc

reas

es in

dep

ress

ive

sym

ptom

s in

his

tory

-pos

itive

sub

ject

sPos

t-te

st: m

ean

peak

HD

RS

scor

es A

TD: 1

4 (±

6), ‘

CON

T’: 1

0 (±

7).

Stud

y co

mpa

red

heal

thy

cont

rols

with

pa

tient

s w

ith M

DD

in re

mis

sion

. Ins

tead

of

plac

ebo

a 25

% st

reng

th d

eple

tion

drin

k w

as

used

as

a co

ntro

l.

Poor

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

Chapter6

113


Neu

mei

ster

(200

4)47

Ran

d., D

B,

With

in S

S ≥3

mon

ths r

e m

is

sion

27

M+F

R

x-fr

ee

≥3 m

ths

Goo

dH

DRS

-24**

: ATD

: BL:

1.2

(±0.

2), p

ost-

test

: 11.5

(±10

.3).

CON

T: B

L: 0

.7 (±

0.2)

, pos

t-te

st: 2

.1 (±

0.3)

. 16/

27 (5

9%)

patie

nts

expe

rienc

ed tr

ansi

ent r

etur

n of

MD

D d

urin

g AT

D

vs 0

/27

durin

g CO

NT

See

abov

e in

sec

tion

“hea

lthy

cont

rols

”G

ood

Neu

mei

ster

(200

6)51

Ran

d., D

B,

With

in S

S ≥3

mon

ths r

e m

is

sion

27

M+F

R

x-fr

ee

≥3 m

ths

Goo

d H

DRS

**:

s/s

subt

ype

(n=

7): A

TD: B

L: 0

.4 (±

0.4

SE),

post

-te

st: 7

.0 (±

1.4)

. CO

NT:

BL:

0.6

(±0.

4), p

ost-

test

: 2.6

(±0.

8).

s/l s

ubty

pe (n

= 12

): AT

D: B

L: 0

.8 (±

0.4)

, pos

t-te

st: 1

1.4

(±1.0

). C

ON

T: B

L: 0

.8 (±

0.4)

, pos

t-te

st: 2

.4 (±

0.6)

. l/l

subt

ype

(n=

8): A

TD: B

L: 2

.0 (±

0.3)

, pos

t-te

st: 1

4.6

(±1.2

).

CON

T: B

L: 0

.9 (±

0.4)

, pos

t-te

st: 1

.7 (±

0.6)

.

Stud

y in

vest

igat

es e

ffec

t of t

ri-al

lelic

se

roto

nin

tran

spor

ter p

rom

oter

(5-H

TTPR

) ge

noty

pe a

nd e

ffec

t of A

TD o

n m

ood

and

regi

onal

cer

ebra

l met

abol

ism

with

PET

Goo

d

O’R

eard

on (2

004)

48R

and.

, DB,

W

ithin

SS

5 m

onth

s re

mis

si

on

10 M+F

Po

st-C

BT

Goo

d H

DRS

-24**

: ATD

: cha

nge:

4.9

7 (±

1.0

SE).

CO

NT:

cha

nge:

3.

22 (±

1.0)

0/1

0 pa

tient

s re

laps

ed in

eith

er c

ondi

tion

Curr

ently

not

usi

ng a

ntid

epre

ssan

t m

edic

atio

n, re

spon

ders

to C

BT tr

eatm

ent.

G

ood

Smit

h (1

997a

)118

Ran

d., D

B,

With

in S

S ≥6

mon

ths

rem

issi

on

15 F Rx-

free

≥6

mth

s

Goo

d

HD

RS-12

: ATD

: cha

nge

(7hr

): 7

.3. C

ON

T: c

hang

e: 0

.1. S

D o

f di

ffer

ence

ATD

-CO

NT

calc

ulat

ed fr

om 9

5% C

I: 7.

2 (±

4.9)

5/

15 p

atie

nts

expe

rienc

ed re

laps

e du

ring

ATD

, 0/1

5 du

ring

CON

T

Curr

ently

not

usi

ng a

ntid

epre

ssan

t m

edic

atio

n (r

ecov

ered

MD

D)

Mod

erat

e

B2. P

atie

nts

wit

h M

DD

pre

viou

sly,

but

cur

rent

ly in

rem

issi

on (c

urre

ntly

usi

ng m

edic

atio

n)

Abe

rg-W

iste

dt

(199

8)11

1R

and.

, DB,

Be

tw. S

S A

fter

resp

onse

12+8

M

+F

CIT

Insu

ffic

ient

M

AD

RS:

exa

ct s

core

s no

t rep

orte

d. F

ive

of th

e 12

pat

ient

s w

ith A

TD s

how

ed a

Wor

seni

ng o

f dep

ress

ive

sym

ptom

s in

AT

D: 5

/12

(MA

DR

S 12

-71%

). C

ON

T: 0

/8.

Stud

y al

so in

vest

igat

es re

latio

n w

ith re

laps

e du

e to

ATD

and

oth

er v

aria

bles

(pla

sma

cort

isol

, pla

tele

t-se

roto

nin)

Mod

erat

e

Booi

j (20

05a)

59R

and.

, DB,

W

ithin

SS

Aft

er ≥

par

tial

rem

issi

on

20

M+F

SS

RI

SNR

I

Goo

dM

AD

RS:

ATD

: BL:

4.6

(±0.

9 SE

), po

st-t

est:

7.9

(±1.8

). C

ON

T:

BL: 3

.7 (±

0.9)

, pos

t-te

st: 3

.7 (±

0.9)

. Rel

apse

of M

DD

: ATD

: 7/

20; C

ON

T: 0

/20

assu

med

.

Incl

usio

n of

sub

ject

s w

ith H

DRS

<15

(par

tial

rem

issi

on).

A 2

5% s

tren

gth

tryp

toph

an

defi

cien

t mix

ture

was

use

d as

CO

NT.

Goo

d

Booi

j (20

06)83

Ran

d., D

B,

With

in S

S A

fter

≥ p

artia

l re

mis

sion

19

M+F

SS

RI S

NRI

Goo

d H

DRS

-17 s

core

s si

gn. l

ower

aft

er A

TD:

SI+

(n=

8): A

TD: B

L:

2.5

(±0.

8 SE

), p

ost-

test

: 7.6

(±2.

0). C

ON

T: N

ot g

iven

SI-

(n=

11):

ATD

: BL:

2.9

(±0.

7), p

ost-

test

: 3.4

(±0.

6). C

ON

T: N

ot

give

n M

AD

RS s

core

s si

gn. l

ower

aft

er A

TD:

SI+:

ATD

: BL:

5.

0 (±

1.2 S

E), p

ost-

test

: 12.

00 (±

2.4)

. CO

NT:

Not

giv

en S

I-:

ATD

: BL:

5.4

(±1.3

SE)

, pos

t-te

st: 9

.3 (±

1.7).

CO

NT:

Not

gi

ven

Incl

usio

n of

sub

ject

s w

ith H

DRS

<15

(par

tial

rem

issi

on).

A 2

5% s

tren

gth

tryp

toph

an

defi

cien

t mix

ture

was

use

d as

CO

NT.

No

data

fo

r ATD

moo

d-ra

tings

aft

er C

ON

T pr

ovid

ed.

Stud

y al

so in

vest

igat

es e

ffec

t of A

TD o

n H

eart

Rat

e Va

riabi

lity

(HRV

) and

impu

lsiv

ity

and

the

inte

ract

ion

by +

ve/-v

e hi

stor

y of

su

icid

al id

eatio

n (S

I+/S

I-)

Mod

erat

e

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

114

Chap

ter

Chapter 6

6

Brem

ner

(199

7)10

1R

and.

, DB,

W

ithin

SS

Ave

rage

rem

is

sion

45

wee

ks

21 M+F

FL

X PA

R

Goo

dR

elap

se o

f MD

D: A

TD: 7

/21;

CO

NT:

1/21

. HD

RS-

scor

es a

re

split

bet

wee

n re

laps

e an

d no

n-re

laps

e gr

oup.

With

rela

pse

(n =

7):

ATD

: BL:

6.8

6 (±

4.98

), po

st-t

est:

15.5

7 (±

6.92

). CO

NT:

BL:

7.0

0 (±

5.45

), p

ost-

test

: 6.7

1 (±6

.07)

. With

out-

rela

pse

(n =

14):

ATD

: BL:

4.7

9 (±

3.45

), p

ost-

test

: 6.3

6 (±

4.25

), C

ON

T: B

L: 6

.21 (

±4.6

9), p

ost-

test

: 9.2

1 (±5

.99)

.

19/2

1 of p

atie

nts u

sed

FLX.

Stu

dy in

vest

igat

es

brai

n m

etab

olis

m w

ith P

ET d

urin

g AT

D a

nd

asso

ciat

ions

with

MD

D re

laps

e

Goo

d

Del

gado

(199

1)42

Ran

d., D

B,

With

in S

S ≥2

wee

ks a

fter

‘s

tabi

lity’

46 M+F

va

rious

Goo

d H

DRS

-25:

exa

ct s

core

s no

t rep

orte

d. D

epre

ssiv

e re

laps

e:

ATD

: 24/

46, C

ON

T: 0

/46.

H

eter

ogen

eous

pop

ulat

ion

curr

ently

usi

ng

vario

us A

D. B

UP

or D

ES-t

reat

men

t low

er

chan

ce o

f rel

apse

In

orig

inal

stu

dies

Bip

olar

pa

tient

s in

clud

ed; u

ncle

ar w

heth

er th

ese

wer

e ex

clud

ed in

this

poo

led

anal

ysis

Mod

erat

e

Del

gado

(199

9)11

4R

and.

, DB,

W

ithin

SS ≥2

wee

ks

resp

onse

30

M+F

FL

X D

ES

Goo

dH

DRS

-25:

FLX

-tre

ated

pat

ient

s: A

TD: B

L: 7

.7, p

ost-

test

17.1.

CO

NT:

not

repo

rted

Rel

apse

: ATD

8/1

5, C

ON

T 0/

15. D

ES-

trea

ted

patie

nts:

ATD

: BL:

8.6

, pos

t-te

st 10

.7. C

ON

T: n

ot

repo

rted

. Rel

apse

: ATD

1/15

, CO

NT

0/15

.

Two

bipo

lar p

atie

nts

incl

uded

. Stu

dy

repl

icat

es p

revi

ous

findi

ng th

at D

ES-t

reat

ed

rem

itter

s do

not

suf

fer a

rela

pse

if im

prov

ed

on D

ES, b

ut d

id a

fter

FLX

-indu

ced

rem

issi

on.

Mod

erat

e

Evan

s (2

002)

94R

and.

, DB,

W

ithin

SS

Aft

er ≥

par

tial

rem

issi

on

8 M+F

BU

P

Goo

d**PO

MS-

depr

esse

d: A

TD: B

L: 5

.25

(±1.6

2 SE

), p

ost-

test

: 3.12

(±

1.37)

. CO

NT:

BL:

5.2

5 (±

1.62)

, pos

t-te

st: 3

.50

(±1.

80)

HD

RS-19

: ATD

: BL:

4.3

8 (±

1.21

SE)

, pos

t-te

st: 5

.00

(±1.0

2).

CON

T: B

L: 4

.38

(±1.2

1), p

ost-

test

: 4.12

(±0.

81)

Stud

y al

so in

vest

igat

es e

ffec

ts o

f ATD

on

slee

p-EE

G. O

ne B

ipol

ar II

pat

ient

incl

uded

. Fi

ve p

atie

nts

in p

artia

l rem

issi

on (H

DR

S 7-

9)

Goo

d

Land

olt (

2003

)95R

and.

, DB,

W

ithin

SS,

PHZ

Aft

er

rem

issi

on

5 M+F

PH

Z

Goo

dPO

MS-

depr

essi

on: n

o BL

mea

sure

s w

ere

give

n se

para

tely

fo

r ATD

and

CO

NT:

BL:

0.4

(± 0

.4 S

E). A

TD: p

ost-

test

: 2.0

(±

1.5).

CO

NT:

pos

t-te

st: 1

.0 (±

0.4)

. Pos

t-te

st m

easu

res

take

n ne

xt m

orni

ng a

fter

ATD

or C

ON

T at

16:0

0h.

Stud

y pr

imar

ily in

vest

igat

es e

ffec

t of A

TD o

n re

curr

ence

of R

EM-s

leep

whi

le u

sing

PH

ZM

oder

ate

Moo

re (1

998)

96R

and.

, DB,

W

ithin

SS ≥2

mon

ths

rem

issi

on

10 M

FLX

SER

PAR

Goo

d PO

MS-

depr

esse

d:AT

D: B

L: 3

.0 (±

1.3 S

E), p

ost-

test

: 2.7

(±

1.4),

CO

NT:

BL:

3.0

(±1.3

), p

ost-

test

: 3.3

(±1.6

). H

DR

S-25

ra

tings

not

exa

ctly

repo

rted

but

did

not

cha

nge

sign

ific

antly

Stud

y al

so in

vest

igat

es e

ffec

ts o

f ATD

on

REM

-sle

ep. A

25%

str

engt

h tr

ypto

phan

de

fici

ent m

ixtu

re w

as u

sed

as C

ON

T.

Subj

ects

in re

mis

sion

>2

mon

ths

Goo

d

Mor

ris

(199

9)14

8R

and.

, DB,

W

ithin

SS

Full

rem

issi

on

8 M

SSR

I A

MI

TRA

Z PH

Z TC

P Li

Goo

dM

odifi

ed H

DRS

-12, B

DI.

No

exac

t sco

res

give

n. H

DRS

in

crea

se >

5 po

ints

(rel

apse

): A

TD: 7

/8, C

ON

T: 0

/8

Stud

y in

vest

igat

es b

rain

met

abol

ism

with

PE

T du

ring

ATD

and

ass

ocia

tions

with

MD

D

rela

pse.

Six

of 8

par

ticip

ants

rece

ived

≥6

mon

ths

antid

epre

ssan

ts. N

o in

dica

tion

whi

ch d

rugs

wer

e cu

rren

tly u

sed.

Mod

erat

e

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

Chapter6

115


O’R

eard

on (2

004)

48R

and.

, DB,

W

ithin

SS

5 m

onth

s re

mis

sion

9 M+F

FL

X PA

R

Goo

d H

DRS

-24**

: ATD

: cha

nge:

8.2

2 (±

1.2),

CON

T: c

hang

e: 2

.56

(±1.2

). 3/

9 pa

tient

s re

laps

ed d

urin

g AT

D v

s 0/

9 du

ring

cont

rol

Curr

ently

usi

ng a

ntid

epre

ssan

t med

icat

ion,

re

spon

ders

to S

SRI t

reat

men

t.

Goo

d

Pras

chak

-Rie

der

(200

4)49

Ran

d., D

B,

With

in S

S 2

mon

ths

rem

issi

on

8 M

CIT

Goo

dH

DRS

-21:

ATD

: BL:

3.8

8 (±

2.6)

, pos

t-te

st: 1

2.3

(±5.

4).

CON

T:no

sco

res

repo

rted

. Rel

apse

: ATD

: 6/8

pat

ient

s,

CON

T: n

o nu

mbe

rs re

port

ed 0

/8 a

ssum

ed fo

r ana

lysi

s

Stud

y pr

imar

ily in

vest

igat

es e

ffec

ts o

f ATD

on

regi

onal

5-H

T 1A

bin

ding

pot

entia

l with

PE

T.

Poor

Smit

h (1

999)

102

Ran

d., D

B,

With

in S

S ≥6

mon

ths

rem

issi

on

8 M

vario

us

Goo

dH

DRS

: no

exac

t sco

res

repo

rted

Rel

apse

: ATD

: 6/8

pa

tient

s, C

ON

T: 1/

8St

udy

prim

arily

inve

stig

ates

eff

ects

of A

TD

on re

gion

al b

rain

act

ivity

with

PET

. Sub

ject

s in

rem

issi

on >

6 m

onth

s. 2

pat

ient

s di

d no

t us

e an

tidep

res-

sant

s, 6

use

d va

rious

type

s (S

SRI,

TCA

, MA

O-I)

Mod

erat

e

Spill

man

n (2

001)

119

Ran

d., D

B,

With

in S

S ≥3

mon

ths

rem

issi

on

10

M+F

FL

X SE

R VL

X PA

R

Goo

d H

DRS

-6:A

TD: B

L: 1.

1 (±1

.2) p

ost-

test

(7h)

:4.7

(±3.

4), c

hang

e:

3.6

(±3.

1); C

ON

T: B

L: 1.

4 (±

1.7)

pos

t-te

st: 1

.0 (±

1.4),

chan

ge:

-0.4

(±1.2

). In

crea

se in

sco

res

on th

e H

DRS

-6 w

as

sign

ific

antly

hig

her f

or A

TD th

an fo

r CO

NT.

In re

mis

sion

aft

er S

SRI,

curr

ently

usi

ng

antid

epre

ssan

t med

icat

ion.

5 in

itial

pa

rtic

ipan

ts d

ropp

ed o

ut.

Goo

d

C. P

atie

nts

curr

entl

y w

ith

MD

D (w

ith

or w

itho

ut a

ntid

epre

ssan

ts)

Booi

j (20

05b)

120

Ran

d., D

B,

With

in S

S14

M

+F

VLX

MIR

TC

A

Li

Goo

dH

DRS

: ATD

: BL:

12.9

(±1.5

SE)

, pos

t-te

st: 1

3.1 (

±1.9

). CO

NT:

BL

: 11.4

(±1.4

), p

ost-

test

: 13.

6 (±

1.3).

MA

DRS

: ATD

: BL:

22.

7 (±

3.7)

, pos

t-te

st: 2

3.1 (

±4.7

). CO

NT:

BL:

19.6

(±3.

3), p

ost-

test

: 22.

9 (±

2.0)

.

A 2

5% s

tren

gth

tryp

toph

an d

efic

ient

mix

ture

w

as u

sed

as C

ON

T. Im

prov

emen

t of m

ood-

ratin

gs 2

4hr a

fter

full

ATD

in V

LX/M

IR tr

eate

d pa

tient

s, b

ut n

ot in

pat

ient

s tr

eate

d ot

herw

ise.

Goo

d

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

116

Chap

ter

Chapter 6

6

Del

gado

(199

4)12

1R

and.

, DB,

W

ithin

SS

43

M+F

≥2

wks

R

x-fr

ee

Goo

dH

DRS

-25:

No

exac

t moo

d-sc

ores

pro

vide

d O

ne d

ay a

fter

AT

D 16

/43

patie

nts

expe

rienc

ed im

prov

emen

t of m

ood,

10/

46 d

ecre

ase

in m

ood

rela

tive

to C

ON

T

Impr

ovem

ent o

f moo

d-ra

tings

24h

r aft

er fu

ll AT

D a

ssoc

iate

d w

ith tr

eatm

ent r

espo

nse

afte

r ant

idep

ress

ant t

reat

men

t

Poor

Pric

e (1

997)

86R

and.

, DB,

W

ithin

SS

22

M+F

R

x-fr

ee

Goo

dVA

S, H

DR

S-25

: No

exac

t sco

res

prov

ided

; sco

res

unaf

fect

ed b

y AT

D.

Stud

y in

vest

igat

es th

e (b

ioch

emic

al) e

ffec

ts

of m

-Chl

oro

phen

yl p

iper

azin

e (m

CCP;

se

roto

nin-

agon

ist)

on

HD

RS,

cor

tisol

, pr

olac

tin a

nd g

row

th-h

orm

one

durin

g AT

D

Poor

Pric

e (1

998)

87R

and.

, DB,

W

ithin

SS

36

M+F

≥3

wks

Rx-

free

Goo

dH

DRS

-25:

ATD

: BL:

28

(±7)

, pos

t-te

st 2

8 (±

7). C

ON

T: 2

7 (±

9),

post

-tes

t 26

(±9)

Stud

y in

clud

ed 4

pat

ient

s w

ith B

D. S

tudy

in

vest

igat

es th

e (b

ioch

emic

al) e

ffec

ts o

f ra

pid

tryp

toph

an in

fusi

on o

n H

DRS

, cor

tisol

, pr

olac

tin a

nd g

row

th-h

orm

one

afte

r ATD

Goo

d

AD

= a

ntid

epre

ssan

ts; A

MI =

Am

itrip

tylin

e; A

MPT

= A

lpha

-met

hyl-p

ara-

tyro

sine

;ATD

= A

cute

Try

ptop

han

Dep

letio

n; B

D =

Bip

olar

Dis

orde

r; B

DI =

Bec

k D

epre

ssio

n In

vent

ory;

Bet

w.S

S =

Betw

een-

subj

ects

pa

ralle

l gro

ups

desi

gn; B

L =

Base

line;

BU

P =

bupr

oprio

n; C

BT =

Cog

nitiv

e Be

havi

our t

hera

py; C

IT =

cita

lopr

am; C

ON

T =

Cont

rol;

DB

= do

uble

-blin

d; D

ES =

des

ipra

min

e; E

CT =

Ele

ctro

-con

vuls

ive

ther

apy;

EEG

=

elec

tro-

ence

phal

ogra

phy;

FLX

= fl

uoxe

tine;

HA

MA

= H

amilt

on A

nxie

ty S

cale

; HD

RS

= H

amilt

on D

epre

ssio

n Sc

ale;

IMI =

imip

ram

ine;

LYS

= ly

sine

; Li =

lith

ium

; MA

ACL

= M

ultip

le A

ffec

t Adj

ectiv

e Ch

eckL

ist;

M

AD

RS =

Mon

tgom

ery

Åsb

erg

Dep

ress

ion

Rat

ing

Scal

e; M

AO

-I =

Mon

oam

ine

oxid

ase

inhi

bito

r; M

DD

= M

ajor

Dep

ress

ive

Dis

orde

r; M

EG =

mag

neto

-enc

epha

logr

aphy

; MIR

= m

irtaz

apin

e; N

OR

=

nort

ripty

line;

PA

R =

paro

xetin

e; P

HZ

= ph

enel

zine

; PO

MS

= Pr

ofile

of M

ood

Stat

es; R

and.

= ra

ndom

ized

; SD

= S

tand

ard

devi

atio

n; S

E =

Stan

dard

Err

or; T

CA =

tric

yclic

ant

idep

ress

ants

; TCP

=

Tran

ylcy

prom

ine;

TRA

Z =

Traz

odon

e; V

A(M

)S =

Vis

ual A

nalo

gue

(Moo

d) S

cale

; VLX

= v

enla

faxi

ne;

With

in S

S =

With

in-s

ubje

cts

cros

s ov

er d

esig

n*

stud

ies

incl

uded

in th

e m

eta-

anal

ysis

in b

old

† Pla

sma

tryp

toph

an d

eclin

ed b

y at

leas

t 50%

‡ S

D g

iven

unl

ess

indi

cate

d ot

herw

ise

(e.g

. SE)

§

Judg

emen

t in

the

cont

ext o

f app

licab

ility

for t

his

met

a-an

alys

is

** E

stim

ated

from

figu

re

†† U

ncle

ar w

heth

er p

artia

lly th

e sa

me

popu

latio

n is

repo

rted

(Wel

tzin

(199

4 an

d 19

95))

Tabl

e 6.

2. In

clud

ed s

tudi

es A

cute

Try

ptop

han

Dep

letio

n (A

TD).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

q. o

f de

plet

ion†

Res

ults

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

emen

t§

Chapter6

117


Tabl

e 6.

3. In

clud

ed s

tudi

es w

ith P

ara-

chlo

roph

enyl

alan

ine

(PCP

A).

Aut

hor

(yea

r)

Des

ign

N

Ade

quac

y of

de

plet

ion

Resu

lts

moo

d sc

ores

(sca

le, s

core

s (±

SD))

Re

mar

ks

Judg

emen

t§

Shop

sin

(197

5)15

With

in S

S, M

DD

in

rem

issi

on b

y IM

I1

Not

mea

sura

ble;

poo

r PC

PA g

iven

in s

mal

l dos

es fo

r rel

ativ

ely

shor

t per

iods

cau

sed

a re

vers

al o

f the

ant

idep

ress

ant e

ffec

t of I

MI.

HD

RS, n

o ex

act

moo

d sc

ores

repo

rted

.

Not

e: n

= 1,

1 pa

tient

with

BD

and

3

patie

nts

trea

ted

with

AM

PT n

ot

incl

uded

her

e.

Poor

Shop

sin

(197

6)10

9W

ithin

SS,

MD

D in

re

mis

sion

by

TCP

3 N

ot re

port

ed; p

oor

All

patie

nts

expe

rienc

ed a

rela

pse.

HD

RS,

no

exac

t moo

d sc

ores

re

port

ed.

2 pa

tient

s w

ith B

D n

ot in

clud

ed

here

. Po

or

PCPA

= P

ara-

chlo

roph

enyl

alan

ine.

For

oth

er s

ymbo

ls a

nd a

bbre

viat

ions

see

foot

note

in T

able

6.2

.

118

Chap

ter

Chapter 6

6

Tabl

e 6.

4. In

clud

ed s

tudi

es A

cute

Phe

nyla

lani

ne/T

yros

ine

Dep

letio

n (A

PTD

).

Aut

hor

(yea

r)*

Des

ign

N

Ade

quac

y of

de

plet

ion†

Resu

lts

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

e m

ent§

A1.

Hea

lthy

con

trol

s

Coup

land

(200

1)81

Rand

., D

B,

With

in S

S5

MN

ot re

port

edPO

MS-

depr

esse

d:A

PTD

: cha

nge:

-0.2

(±2.

7) C

ON

T:

chan

ge: 0

.2 (±

2.6)

.St

udy

inve

stig

ates

eff

ect o

f pen

taga

strin

e to

indu

ce

anxi

ety

afte

r APT

DM

oder

ate

Har

mer

(200

1)60

Rand

., D

B,

With

in S

S12

M+F

Goo

dVA

S-de

pres

sed:

APT

D: c

hang

e: 2

.03

(±4.

98).

CO

NT:

ch

ange

: -4.

20 (±

15.2

3).

Stud

y in

vest

igat

es e

ffec

t of A

PTD

on

prol

actin

, ne

urop

sych

olog

ical

and

sub

ject

ive

mea

sure

sG

ood

Har

riso

n (2

002)

37Ra

nd.,

DB,

W

ithin

SS,

FH

-13

FG

ood

POM

S-de

pres

sed:

APT

D: B

L: 0

.92

(±0.

56 S

E), p

ost-

test

: 1.1

5 (±

0.55

) CO

NT:

BL:

1.61

(±0.

52),

pos

t-te

st: 1

.15 (±

0.56

).St

udy

also

repo

rts

abse

nce

of e

ffec

ts o

f ATD

and

A

PTD

on

inte

rleuk

in-6

act

ivat

ion

Goo

d

Har

rison

(200

4)38

Rand

., D

B,

With

in S

S, F

H-

13 F

Goo

dVA

MS,

no

sign

ific

ant c

hang

es o

bser

ved;

exa

ct sc

ores

not

re

port

edSt

udy

also

inve

stig

ates

mem

ory

and

cogn

itive

eff

ects

of

ATD

and

APT

D

Mod

erat

e

Leyt

on (1

999)

, Le

yton

(2

000)

39;4

0Ra

nd.,

DB,

Bet

w.

SS, F

H-

12 A

PTD

14

CO

NT

FG

ood

POM

S**: A

PTD

: cha

nge:

-2.3

9 (±

2.1 S

E). C

ON

T: c

hang

e -

1.04

(±1.

2). V

AM

S-de

pres

sed:

APT

D: B

L: 1.

3 (±

1.6 S

D),

po

st-t

est:

1.7

(±2.

7), c

hang

e: 0

.4 (±

1.6).

CO

NT:

BL:

0.4

(±

0.5)

, pos

t-te

st: 0

.4 (±

0.4)

, cha

nge:

0.1

(±0.

4).

POM

S; n

o ex

act s

core

s re

port

ed. T

hird

arm

with

ATD

in

clud

edM

oder

ate

Lyth

e (2

005)

61Ra

nd.,

DB,

W

ithin

SS,

12 M

+F

Goo

dVA

MS,

no

sign

ific

ant c

hang

es o

bser

ved;

exa

ct sc

ores

not

re

port

edSt

udy

inve

stig

ates

eff

ect o

f APT

D o

n ne

uro

psyc

ho

logi

cal a

nd s

ubje

ctiv

e m

easu

res

Mod

erat

e

McL

ean

(200

4)62

Rand

., D

B, B

etw

. SS

19+2

0M+F

Goo

d**VA

MS-

cont

ente

dnes

s: A

PTD

: BL:

10.2

(±0.

7 SE

), p

ost-

test

: 11

.5 (±

0.5)

. CO

NT:

BL:

11.8

(±0.

5), p

ost-

test

: 11.8

(±0.

5).

Stud

y in

vest

igat

es e

ffec

t of A

PTD

on

neur

o ps

ycho

lo

gica

l and

sub

ject

ive

mea

sure

s.G

ood

A2.

Hea

lthy

con

trol

s w

ith

+ve

fam

ily h

isto

ry fo

r MD

D

Gre

vet (

2002

)63Ra

nd.,

DB,

W

ithin

SS,

6/1

0 FH

+

10 M

+FIn

suff

icie

ntPO

MS-

depr

esse

d: A

PTD

: BL:

23.

40 (±

7.21

), p

ost-

test

: 23

.50

(±6.

48).

CON

T: B

L: 2

1.30

(±8.

05),

pos

t-te

st: 2

5.20

(±

4.73

).VA

MS-

cont

ente

dnes

s: A

PTD

: BL:

2.7

8 (±

1.43)

, po

st-t

est:

3.2

4 (±

1.49)

. CO

NT:

BL:

3.12

(±1.2

6), p

ost-

test

: 2.

79 (±

1.29

).

Stud

y in

vest

igat

es e

ffec

t of A

PTD

on

neur

o ps

ycho

lo

gica

l and

sub

ject

ive

mea

sure

s.M

oder

ate

Chapter6

119


B1. P

atie

nts

wit

h M

DD

pre

viou

sly,

but

cur

rent

ly in

rem

issi

on (w

itho

ut m

edic

atio

n)

McT

avis

h (2

005)

64Ra

nd.,

DB,

W

ithin

SS ≥6

m

onth

s re

mis

si

on, a

ver a

ge 4

1 m

onth

s

15 F

Rx-

free

≥6

mth

s

Goo

d O

nly

(TYR

+ PH

E)/N

AA

ratio

pr

ovid

ed

HD

RS-

6**: A

PTD

: BL:

1.7

(±0.

7 SE

), p

ost-

test

: 2.4

(±0.

9).

CON

T: B

L: 1.

0 (±

0.4)

, pos

t-te

st: 2

.7 (±

1.1).V

AS-

depr

esse

d no

t sig

nifi

cant

ly c

hang

ed b

y A

PTD

vs

CON

T; n

o ex

act

scor

es p

rovi

ded.

Stud

y in

vest

igat

es e

ffec

ts o

f ATP

D o

n su

bjec

tive

and

a sp

atia

l rec

ogni

tion

mem

ory-

task

Goo

d

Rois

er (2

005)

65Ra

nd.,

DB,

W

ithin

SS ≥6

m

onth

s re

mis

si

on

17 M

+F

Rx-fr

ee

≥6 m

ths

Insu

ffic

ient

TYR

de

plet

ion;

(TYR

+ PH

E)/N

AA

ratio

de

crea

ses

80%)

HD

RS-

19: A

PTD

: BL:

2.1

(±2.

0), p

ost-

test

: 2.2

(±2.

0). C

ON

T:

BL: 1

.4 (±

1.3)

, pos

t-te

st: 1

.1 (±

1.1).V

AS-

cont

ente

dnes

s:

APT

D: B

L: 12

.4 (±

1.5)

, pos

t-te

st: 1

2.8

(±1.

5). C

ON

T: B

L: 12

.3

(±1.8

), p

ost-

test

: 12.

8 (±

1.8).

BD-II

pat

ient

s in

clud

ed (n

=?).

Stu

dy in

vest

igat

es

effe

cts

of A

TPD

on

subj

ectiv

e an

d ne

uroc

og n

itive

m

easu

res

Goo

d

B2. P

atie

nts

wit

h M

DD

pre

viou

sly,

but

cur

rent

ly in

rem

issi

on (c

urre

ntly

usi

ng m

edic

atio

n)

- C. P

atie

nts

curr

entl

y w

ith

MD

D (w

ith

or w

itho

ut a

ntid

epre

ssan

ts)

- APT

D =

Acu

te P

heny

lala

nine

/Tyr

osin

e D

eple

tion;

NA

A =

neu

tral

am

ino

acid

s; P

HE

= ph

enyl

alan

ine;

TYR

= ty

rosi

ne† t

yros

ine

leve

ls d

eclin

ed b

y at

leas

t 60%

For o

ther

sym

bols

and

abb

revi

atio

ns s

ee fo

otno

te

in T

able

6.2

.

Tabl

e 6.

4. In

clud

ed s

tudi

es A

cute

Phe

nyla

lani

ne/T

yros

ine

Dep

letio

n (A

PTD

). (C

ontin

ued)

Aut

hor

(yea

r)*

Des

ign

N

Ade

quac

y of

de

plet

ion†

Resu

lts

moo

d sc

ores

(sca

le, s

core

s (±

SD‡ ))

Re

mar

ks

Judg

e m

ent§

120

Chap

ter

Chapter 6

6

Tabl

e 6.

5. In

clud

ed s

tudi

es A

lpha

-met

hyl-p

ara-

tyro

sine

(AM

PT).

Aut

hor

(yea

r)*

Des

ign

N

Ade

quac

y of

dep

leti

on†

Resu

lts

moo

d sc

ores

(s

cale

, sco

res

(±SD

‡ ))

Rem

arks

Judg

emen

t

A1.

Hea

lthy

con

trol

s

Krah

n (1

999)

88Ra

nd.,

DB,

With

in

SS, F

H-

10 M

+FN

ot re

port

edH

DRS

sco

res

incr

ease

d si

gnifi

cant

(cha

nge:

3 S

D

not r

epor

ted)

in A

MPT

com

pare

d to

CO

NT

(no

mor

e da

ta p

rovi

ded)

PO

MS

no s

igni

fica

nt

chan

ges

(no

mor

e da

ta p

rovi

ded)

Stud

y in

vest

igat

es e

ffec

ts o

f AM

PT o

n m

elat

onin

sec

retio

n. P

rom

etha

zine

25m

g as

CO

NT

Poor

McC

ann

(199

3)11

0Ra

nd.,

DB,

Bet

w. S

S40

MN

ot d

eter

min

ed; u

nspe

cifie

d si

gn. p

rola

ctin

rise

in A

MPT

gr

oup

VAS-

sadn

ess**

; AM

PT: B

L: 3

.2, p

ost-

test

: 13.

1. CO

NT:

BL:

11.2

, pos

t-te

st: 1

6.9.

(no

SDs

prov

ided

) (o

nly

rest

ing

cond

ition

pro

vide

d he

re) n

o si

gn.

Diff

eren

ce A

MPT

and

pla

cebo

in re

sted

con

ditio

n

Subj

ects

wer

e ra

ndom

ized

to 1.

AM

PT +

rest

, 2.

AM

PT+S

leep

Dep

r., 3

. CO

NT

+ Sl

eepD

epr.,

4.

CO

NT

+ re

st

Mod

erat

e

McC

ann

(199

5)82

Rand

., D

B, B

etw

. SS

41 M

Not

det

erm

ined

; pro

lact

in ri

se o

f 18

8% in

AM

PT g

roup

POM

S- s

adne

ss, V

AS-

sadn

ess;

no

exac

t sco

res

prov

ided

; fig

ure

for c

alm

ness

sho

wn.

AM

PT

caus

ed m

axim

al ra

tings

of t

ensi

on, w

hich

was

re

vers

ed b

y ad

ditio

n of

+ L

-dop

a/ c

arbi

dop

a.

Non

-sig

nific

ant i

ncre

ase

in s

adne

ss re

port

ed in

A

MPT

-gro

up (r

egar

dles

s L-

dopa

/ car

bi d

opa)

.

Subj

ects

wer

e ra

ndom

ized

to 1.

AM

PT +

PLA

C,

2. A

MPT

+L-d

opa/

carb

i dop

a, 3

. CO

NT

+ L-

dopa

/ car

bi d

opa,

4. C

ON

T +

PLA

C, in

ord

er to

in

vest

igat

e th

e sp

ecif

icity

of A

MPT

-cha

nges

by

eff

ects

on

cate

chol

amin

es

Mod

erat

e

Zim

mer

man

n (1

996)

89Ra

nd.,

DB,

With

in

SS, F

H-

10 M

+FSu

ffic

ient

MH

PG d

ecre

ase

in M

+FH

DRS

, PO

MS:

No

sign

ifica

nt d

iffer

ence

s. N

o ex

act s

core

s re

port

edSt

udy

inve

stig

ates

eff

ects

of A

MPT

on

prol

actin

and

mel

aton

in s

ecre

tion,

incl

udin

g ge

nder

-eff

ects

. Pro

met

hazi

ne 5

0 m

g as

CO

NT

Mod

erat

e

A2.

Hea

lthy

con

trol

s w

ith

+ve

fam

ily h

isto

ry fo

r MD

D

- B1. P

atie

nts

wit

h M

DD

pre

viou

sly,

but

cur

rent

ly in

rem

issi

on (w

itho

ut m

edic

atio

n)

Berm

an (1

999)

122

Rand

., D

B, W

ithin

SS

≥4

mon

ths

rem

issi

on, A

ver a

ge

18 m

onth

s

15M

+F R

x fr

ee ≥

3 m

onth

s

Goo

d H

DRS

-23;

AM

PT: B

L: 3

.1 (±

3.2)

, Pea

k: 2

3.9

(±12

.0).

CO

NT:

BL:

2.8

(±3.

4), p

eak:

9.0

(±9.

7).ID

S; A

MPT

: BL

: 8.2

(±7.

2). P

eak:

23.

6 (±

14.8

). CO

NT:

BL:

6.1

(±6.

6), p

eak:

10.0

(±8.

1).R

elap

se in

10/1

4 A

MPT

-pa

tient

s an

d in

1/13

CO

NT.

No

sepa

rate

dat

a fo

r n=1

2 w

ho u

nder

wen

t bo

th te

sts.

Giv

en s

core

s ar

e pe

ak s

core

s.

Post

-tes

t sco

resn

ot re

port

ed.

Mod

erat

e

Chapter6

121


B2. P

atie

nts

wit

h M

DD

pre

viou

sly,

but

cur

rent

ly in

rem

issi

on (c

urre

ntly

usi

ng m

edic

atio

n)

Brem

ner (

2003

)103

Rand

., D

B, W

ithin

SS

Aft

er re

mis

sion

18 M

+F D

ES

NO

RN

ot re

port

edH

DRS

**; n

o ex

act s

core

s an

d SD

s re

port

ed

AM

PT: W

ith A

MPT

rela

pse:

BL:

7.4

,pos

t-te

st:

24.8

. With

out r

elap

se: B

L: 11

.5,p

ost-

test

: 12.

5 CO

NT:

With

AM

PT re

laps

e: B

L: 8

.5,p

ost-

test

: 8.5

. W

ithou

t rel

apse

: BL:

6.4

,pos

t-te

st: 1

3.2

Rela

pses

: A

MPT

: 11/

18, C

ON

T: n

ot p

rovi

ded,

0/1

8 as

sum

ed.

Stud

y pr

imar

ilyin

vest

igat

es b

rain

met

abol

ic

corr

elat

es o

f AM

PT in

duce

d de

pres

siv

e re

laps

es w

ith P

ET

Mod

erat

e

Del

gado

(199

3)17

Rand

., D

B, W

ithin

SS

‘Var

ious

’ len

gth

of re

mis

sion

14 M

+F D

ES

MA

ZFLX

SE

R

Goo

d**H

DRS

-25;

No

exac

t sco

res

repo

rted

. Sig

nifi

cant

ef

fect

s of

AM

PT v

s CO

NT

in D

ES a

nd M

AZ

trea

ted

patie

nts,

not

in F

LX o

r SER

trea

ted

patie

nts.

Patie

nts

wer

e cu

rren

tly u

sing

nor

adre

nerg

ic

or s

erot

oner

gic

antid

epre

ssan

t med

icat

ion.

M

oder

ate

C. P

atie

nts

curr

entl

y w

ith

MD

D (w

ith

or w

itho

ut a

ntid

epre

ssan

ts)

Mill

er (1

996)

115

Rand

., D

B, W

ithin

SS

17 M

+FG

ood

HD

RS-2

5**; A

MPT

: BL:

29,

pos

t-te

st: 2

8.5.

CO

NT:

BL

: 30.

5, p

ost-

test

: 29.

5. S

Ds

not r

epor

ted.

VAS:

ex

act d

ata

not p

rovi

ded;

sig

n de

crea

se in

su

bsca

les

‘tire

d’ a

nd ‘e

nerg

etic

’ for

AM

PT v

s CO

NT

Patie

nts

wer

e no

t usi

ng a

ntid

epre

ssan

t m

edic

atio

n. 2

/17

patie

nts

had

bipo

lar

depr

essi

on, 6

/17

had

co-m

orbi

d an

xiet

y,

dyst

hym

ic o

r eat

ing

diso

rder

s

Mod

erat

e

AM

PT =

Alp

ha-m

ethy

l-par

a-ty

rosi

ne; I

DS

= In

vent

ory

of D

epre

ssiv

e Sy

mpt

oms;

MA

Z =

maz

indo

l; PL

AC

= ad

ditio

nal p

lace

bo, S

leep

Dep

r. =

Slee

p de

priv

atio

n† D

eclin

e in

cat

echo

lam

ine

met

abol

ites

hom

ovan

illic

aci

d (H

VA) b

y at

leas

t 60%

or 3

-met

hoxy

-4-h

ydro

xyph

enyl

glyc

ol (M

HPG

) by

at le

ast 4

0% F

or o

ther

sym

bols

and

abb

revi

atio

ns s

ee fo

otno

te in

Tab

le 6

.2.

Tabl

e 6.

5. In

clud

ed s

tudi

es A

lpha

-met

hyl-p

ara-

tyro

sine

(AM

PT).

(Con

tinue

d)

Aut

hor

(yea

r)*

Des

ign

N

Ade

quac

y of

dep

leti

on†

Resu

lts

moo

d sc

ores

(s

cale

, sco

res

(±SD

‡ ))

Rem

arks

Judg

emen

t

Chap

ter

Chapter 6

6

122

Quantitative summary (pooling) Fifty-eight percent (52/90) of the studies supplied enough data to be included in the meta-analysis. Meta-analysis was possible for ATD and APTD studies only. Table 6.6 gives an overview ofthe number of identified studies for each pre-defined population and eligibility for meta-analysis.Due to remaining heterogeneity we used random effects models for all meta-analyses.

ATDIn figure 6.1 the pooled results of ATD in healthy controls in studies with a within subjects designare presented. Overall Hedges' g (95% CI) was -0.27(-0.45 to -0.09). We stratified results by familyhistory for MDD (negative, positive or not reported in the studies). Pooled Hedges’ g for healthycontrols with a negative family history (-0.19 (-0.43 to 0.05)) was significantly different (Qres=6.59, df= 1, p= 0.01) compared to controls with a positive family history (-0.56 (-1.00 to -0.13)). Thepooled result in studies that did not report the family history status resembled the studies with anegative family history (-0.28 (-0.57 to 0.00) Qres= 1.36, df= 1, p= 0.24). In between subjects studiessimilar results were found. Pooled Hedges’ g was -0.63 (-1.95 to 0.70) for controls with a negativefamily history and -0.06 (-0.57 to 0.45) in one study that did not report family history status (Qres=0.14, df= 1, p= 0.71). The large Hedges’ g in controls with a negative family history was largelydetermined by one study.85 Leaving this study out reduced Hedges’ g to 0.16 (-0.43 to 0.76) (datanot shown).

Table 6.6. Identified studies for different types of depletion and design and eligibility for meta-analysis.

Identified (in meta-analysis)

Within Subjects Between Subjects

Acute Tryptophan Depletion

A. Healthy controls

FH- 41 (22) 10 (5)

FH+ 5 (5) 0

B. Patients with MDD in remission

Without AD 8 (6) 1

With AD 13 (12) 1

C. Patients with Current MDD

Without AD 3 (1) 0

With AD 1 (1) 0

Para-chlorophenylalanine


With AD 2 0

Acute Phenylalanine/Tyrosine Depletion

A. Healthy controls

FH- 5 (3) 2 (2)

FH+ 1 (1) 0

B. Patients with MMD in remission

Without AD 2 (2) 0

Alpha-methyl-para-tyrosine

A. Healthy controls

FH- 2 2


Without AD 1 0

With AD 2 0

C. Patients with Current MDD

Without AD 1 0

AD = antidepressant; FH = family history; MDD = Major Depressive Disorder


6

123

Figures 6.2A and 6.2B show the modification of Hedges’ g by gender for healthy controls(within subjects design) with a negative or positive family history. The difference in Hedges’ gbetween males and females was most prominent in healthy controls with a negative familyhistory (0.23 (-0.10 to 0.57) vs. -0.44 (-0.81 to -0.06) respectively; Qres= 11.92, df= 1, p<0.001). Incontrols with a positive family history, males experienced a larger decrease in mood after ATD inonly one study (Hedges’ g -0.98 (-1.53 to -0.42) Qres= 11.92, df= 1, p<0.001). In contrast, femaleswith a positive family history only had a slightly larger Hedges’ g (-0.56 (-1.43 to 0.31)) thanfemales with a negative family history (Qres= 0.62, df= 1, p= 0.43).

Figures 6.3A and 6.3B present the pooled Hedges’ g for patients with MDD in remissionwithout or with current AD (within subjects design). In the remitted patients without current ADwe stratified results by length of time without AD. Only 2 studies with 3-6 months without AD47;51

largely differed in Hedges’ g (-4.35 (-7.39 to -1.31) compared to 1 study directly after successfulelectroconvulsive therapy (Hedges’ g 0.04 (-0.85 to 0.94),117 and 3 studies with at least 6 monthswithout AD (pooled Hedges’ g -0.60 (-1.38 to 0.18) (Qres= 29.34, df= 2, p< 0.0001).48;93;118 Leavingone possible outlier47 out diminished the overall pooled Hedges’ g for remitted patients withoutAD from -1.90 (-3.02 to -0.78) to -1.06 (-1.83 to -0.29), but the observed effect modification bylength of time without AD remained highly significant (p< 0.001; data not shown).

Figure 6.1.

Acute Tryptophan Depletion (ATD) in healthy controls studied in a within subjects design, stratified by status of family historyfor depression (negative or positive for Major Depressive Disorder). References to studies as indicated, different subgroupsper study handled as separate studies with appropriate pooling weights. FH- = family history negative, FH+ = family historypositive, N/A = not reported.

�� T �� !� � � "��#�"�� "�� "��#�"�� $��"��T ! ! � � "��#�"��%� �&'(��) �( �&'(��)

��* ��"��R��* $�+��&&,�� +��-��.��&&/�* $�� +��&&0�� %��"��&&1�� %R��'�� "��2�� 3��&&&�� !��4�� !��4�� !��4�� !��/�� !��1��4�� !��1��4�� !��1��4�� &&,�� S��* $�� 1�� U��&&0��

��&'(��)� �� .��"��T��V�5�''6�17� .�5��&��8�9��6��7�)V�5�1'61(��.��R��..��:�5��6'��8�5��6�2�

��* ��R��* ;�+��-��.��&&/�* ;�� %��"��&&&�� 3��&&&�� !��4�� !��4�� !��4�� S��* ;��

��&'(��)� �� .��"��T��V�5��6�07� .�5�1��8�5��6�1�7�)V�5�'�60(��.��R��..��:�5��6'��8�5��6��

�2�* �� !4��3��&&&�� 3W�-X��'�� T�� &&/�� 8��-��&&/�� U��&&/�� U��&&'��

��&'(��)� �� .��"��T��V�5�06007� .�5�1��8�5��6�0�7�)V�5�2�6'(��.��R��..��:�5��6&/��8�5��6�'�

��&'(��)� �� .��"��T��V�5�026�&7� .�5�22��8�9��6��7�)V�5�1�6/(��.��R��..��:�5��6&1��8�5��6��2�

�$/ �$� �� /

�� T�� T��!�

Chap

ter

Chapter 6

6

124

Figure 6.2A/B.

Acute Tryptophan Depletion (ATD) in healthy controls studied in a within subjects design, stratified by status of family historyfor depression and gender included in the studies. Family history negative for MDD (A) and family history positive for MDD(B). References to studies as indicated, different subgroups per study handled as separate studies with appropriate poolingweights. FH- = family history negative, FH+ = family history positive, MDD = major depressive disorder.

<�R��S� �� T�� T�� T��* $��

�� T � � "��#�"�� "�� "��#�"�� $��"��T �&'(��) �( �&'(��)

��+��-��.��&&/�* $�� +��&&0�� %R��'�� "��2�� 1��

��&'(��)� �� .��"��T��V�5�,60&7� .�5�/��8�5��6��7�)V�5�/&62(��.��R��..��:�5��620��8�5��6�,�

��*��%��"��&&1�� !��4�� !��4�� !��4�� &&,��

��&'(��)� �� .��"��T��V�5�,6/27� .�5�'��8�5��6�&�7�)V�5�2�6,(��.��R��..��:�5��6�&��8�5��6��

�2��;*��+��&&,�� 3��&&&�� !��/�� !��1��4�� !��1��4�� !��1��4�� S��* $�� U��&&0��

��&'(��)� �� .��"��T��V�5��16&27� .�5�0��8�5��6��,�7�)V�5�,�62(��.��R��..��:�5��61'��8�5��6��

��&'(��)� �� .��"��T��V�5�''6�17� .�5��&��8�9��6��7�)V�5�1'61(��.��R��..��:�5��6'��8�5��6�2�

�$/ �$� �� /

�� T�� T��!�

<�R��S� �� T�� T�� 2�� T��* ;��

�� T � � "��#�"�� "�� "��#�"�� $��"��T �&'(��) �( �&'(��)

��+��-��.��&&/�* ;��

��&'(��)� �� .��"��T��.��R��..��:�5�26/1��8�5��6��'�

��*��%��"��&&&�� !��4�� !��4�� !��4��

��&'(��)� �� .��"��T��V�5�,6'17� .�5�2��8�5��6�1�7�)V�5�1�62(��.��R��..��:�5��6�,��8�5��6��

�2��;�*��3��&&&�� S��* ;��

��&'(��)� �� .��"��T��V�5��6�/7� .�5��8�5��6,��7�)V�5��(��.��R��..��:�5��6�,��8�5��6�2�

��&'(��)� �� .��"��T��V�5��6�07� .�5�1��8�5��6�1�7�)V�5�'�60(��.��R��..��:�5��6'��8�5��6��

�$/ �$� �� /

�� T�� T��!�

�

�


6

125

Figure 6.3A/B.

Acute Tryptophan Depletion (ATD) in former depressed patients in remission, studied in a within subjects design without orwith current medication. Figure 6.3A is stratified by length of time without medication; 6.3B is stratified by type ofmedication used by patients. References to studies as indicated, different subgroups per study handled as separate studieswith appropriate pooling weights. BUP = bupropion; SNRI = Serotonin Norepinephrin Reuptake Inhibitor; SSRI = SelectiveSerotonin Reuptake Inhibitor; PHZ = phenelzine.

2�R��S� �� T��S��R�� T�� R�� T��S�� !T��"��##��

�� T �� "��$�"�� "�� "��$�"�� !%��"��T �&'(��) �( �&'(��)

�*�+�,�� T�*&&-��

� !��&'(��)� � �� #��"��T��!��#��R��##��.�/��0*��/��0&��

��,%1��2� ��3�� 2� ��1�4� �� 2� ��1��4� � �� 2� ��1��4��

� !��&'(��)� �� #��"��T��5�/��60,1�� #�/�,��+��0��*��)5�/�6&03(��#��R��##��.�/��06*��/��0��'�

�,�7/1��T��3�� $8�� 3�� *&&-��

� !��&'(��)� �� #��"��T��5�/�60,'�� #�/��/��0��)5�/�-10�(��#��R��##��.�/�*0'��/��0*,�

��&'(��)� �� #��"��T��5�/�110�'�� #�/�-��+��0��*��)5�/�6&03(��#��R��##��.�/�,0,,��/��0��&�

�%*� �%' �� ' �*�

�� !T�� !T��2�

8�R��S� �� T��S��R�� T�� ,��R�� T��!T��

�� T �� "��$�"�� "�� "��$�"�� !%��"��T �&'(��) �( �&'(��)

�*�� 8)�� 9��*&&-�2��8�0 � �� 9��*&&-�8�� *&&6�� $8�� 3�� *��

� !��&'(��)� �� #��"��T��5�/�*30�&�� #�/�3��/��0��-��)5�/�-*01(��#��R��##��.�/�*0-��/��0�6�

�� 8)4�28)�� 9��:��'��

� !��&'(��)� � �� #��"��T��!��#��R��##��.�/��0*1��/��0�,�

�,�� 9;�� <R��

� !��&'(��)� � �� #��"��T��!��#��R��##��.�/��016��/��0'��

�3�� .�� =�� ,��

� !��&'(��)� �� #��"��T��!��#��R��##��.�/�*0*��/��0�-�

��&'(��)� �� #��"��T��5�/�*'0�*�� #�/�-��/��0�3��)5�/�',0,(��#��R��##��.�/��03,��/��0�*�

�%3 �%� �� 3

�� !T�� !T��2�

�

�

Chap

ter

Chapter 6

6

126

In remitted patients previously depressed and currently using AD, ATD caused a decrease inmood (pooled Hedges’ g -0.49 (-0.89 to -0.10). Hedges’ g varied slightly depending on type of AD(Qres= 0.92, df= 3, p= 0.82). Surprisingly, the pooled Hedges’ g for SSRIs showed a moderate pointestimate, which did not reach significance (-0.60 (-1.28 to 0.08). Especially for bupropiontreatment Hedges’ g was small and not significant (-0.25 (-0.96 to 0.47). No ATD studies withother ADs without a 5-HT mechanism of action were available for this comparison.

We stratified the results of ATD studies by length of remission, which revealed significanteffect modification. In remitted patients using AD decreased mood after ATD was especially seenin the first 5 months after the achievement of remission (pooled Hedges’ g -0.55 (-0.90 to -0.21)Qres= 47.18, df= 2, p< 0.0001).48;59;94-96;119 In contrast, remitted patients without AD showed moredecrease in mood after ≥2 months of remission(Hedges’ g -1.65 (-2.60 to -0.69) Qres= 13.81, df= 2,p= 0.001)48;51;118 (figure available on request).

Relapse rates in remitted patients with AD were increased after ATD compared to controldepletion (pooled difference in relapse rate 47% (28% to 66%); Figure 6.4). This increase in relapserate was especially seen in patients using SSRIs (47% (27% to 67%)) or a SNRI (35% (14% to 56%)). TheNE acting drug desipramine showed no significant difference in relapse rate (7% (-6% to 19%)) afterATD. This effect modification by drug was statistically significant (Qres= 18.02, df= 2, p< 0.001).

In patients who were depressed at the time of ATD we found two studies for meta-analysis.These studies included patients who used,120 or did not use87 AD (Figure 6.5). The effects of ATDwere opposed: Hedges’ g was 0.32 (-0.22 to 0.86) for patients using different types of ADs, and -0.12 (-0.45 to 0.21) for patients without AD. Two studies in depressed patients without AD werenot suitable for meta-analysis. ATD did not decrease mood during depletion in these studies.86;121

Contra-intuitively, in one study mood improved the day after depletion in 16/43 patients,121 a resultwhich was also found by one study in the meta-analysis.120

Figure 6.4.

Acute Tryptophan Depletion (ATD) in former depressed patients in remission, differences of relapse rates versus controldepletion studied in a within subjects design in remitted patients stratified by current medication use. References to studiesas indicated. SNRI = Serotonin Norepinephrine Reuptake Inhibitor; SSRI = Selective Serotonin Reuptake Inhibitor.

�� !��"#� �� "�� $%�&' �% �$%�&'

(��')��($$*�� "��($$$�� +��,-�� #�. ��,- � ��

��$%�&'� �� /��#��"��&#�0�1�2�2$��1�3��1��4��'0�1��,%/�� 5�1�-�3��6��(�

,��'7�8�')��9�,��

��$%�&'� � �� /��#��"��/�� 5�1�3�,4��1��(�

3��:�'��;�'8:��"��($$$��

��$%�&'� � �� /��#��"��/�� 5�1�(�-��1��3�

-�<��=��"��($$(�� ($$$�� #�($$$��

��$%�&'� �� /��#��"��&#�0�1�2�-��1�,��1��-��'0�1�2$�-%/�� 5�1��(��6��(�

/��$%�&'� �� /��#��"��&#�0�1�-�3(��1�4��6��(��'0�1�4�3%/�� 5�1�-�4��6��(�

� ( � � � �� (

��&�8/ ��;/�


6

127

APTDFigure 6.6A and 6.6B show results from APTD studies in healthy controls (within and betweensubjects). APTD did not decrease mood: pooled Hedges’ g were 0.10 (-0.23 to 0.43) in withinsubjects, and 0.12 (-0.43 to 0.68) in between subjects studies. However, in one study with healthycontrols with a positive family history for MDD, a moderate but non-significant effect on moodwas found (Hedges’ g -0.49 (-1.17 to 0.19)).63 In patients with MDD in remission without AD (Figure6.7), no effect of APTD was observed in 2 studies (pooled Hedges’ g -0.02 (-0.50 to 0.47).64;65 Nostudies inpatients with current MDD were identified.

Sensitivity analysis and publication biasWe examined our assumptions for intercorrelation ofbefore-after mood ratings per condition inATD studies in healthy controls without a family history of MDD. At the same examination, we alsoexamined the assumed intercorrelation between two test conditions (ATD vs. CONT). Asexpected, increasing the assumed R to 0.8 (less conservative) increased the pooled Hedges’ gfrom -0.19 (-0.43 to 0.05) to -0.31 (-0.64 to 0.02). Reducing R to 0.2 (more conservative) decreasedthe pooled Hedges’ g to -0.14 (-0.34 to 0.06). The calculated R’s were higher than 0.5 in 80% of thestudies that reported all relevant data. Therefore, we judged the imputed value of 0.5 for R asacceptable.

Inspection of the funnel-plot of the within subjects ATD studies in healthy controls without afamily history of MDD revealed asymmetry (figure available on request). More studies with adecrease in mood after ATD vs. control depletion were published. In a Galbraith-plot of studies,the intercept in the regression equation was -2.08 (SE 0.90; p= 0.030). Therefore, we concludedthat publication bias probably distorted our findings. We did not inspect funnel-plots in otherpopulations of our review due to the limited number of studies.

Discussion

In this systematic review we performed the first meta-analysis of the mood-effects in ATD andAPTD studies. The depletion of monoamine systems (both 5-HT and NE/DA) does not decreasemood in healthy controls. However, in healthy controls with a family history of MDD the resultssuggest that mood is slightly decreased, both by ATD and APTD. Additionally, healthy females aremore affected by ATD than healthy males, especially in controls without a family history of MDD.In patients who were previously depressed but in remission without AD, ATD moderatelydecreases mood, while APTD does not significantly decrease mood. ATD has comparable moodlowering effects in patients with MDD in remission who are still using ADs. The site of action of

Figure 6.5.

Acute Tryptophan Depletion (ATD) in depressed patients stratified by use of concurrent medication (within subjects design).References to studies as indicated.

�� "� �� #� ��"$��%�$��"�� $�� "$��%�$��"�� &'��$�� # # ��"$��%�$��(� �+)*��+ �* �+)*��+

�!��"��,��-��)&�� &��+)*��+� �� .��$��&��.��M��..��/�0�!1!)��0��1�)�

�� "��!++2�� &��+)*��+� �� .��$��&��.��M��..��/�0��13!��0��142�

��+)*��+� �� .��$��5�0�!12�6�".�0�!��0��1!2�6�+5�0�4)1�*��.��M��..��/�0��1��0��124�

�'4 �'� �� 4

��"�&�� "�&�� #�

Chap

ter

Chapter 6

6

128

these ADs (5-HT or NE/DA) predicts the occurrence of a lowering of mood or a short relapse inMDD after depletion of the corresponding monoamine. Our findings are in line with thesummaries in previous reviews.11;18-20;24

The most consistent finding from this review is the decrease of mood and relapse into adepressed state after ATD and APTD in remitted MDD patients who use AD. In remitted patientswithout AD relapses are less prominent. Because after remission medication is often continued,the difference in mood responses after depletion might be related to the duration of the achievedremission. Previous reviews discussed the relationship between the duration of the remission andthe effect of monoamine depletion, with opposite conclusions. Bell et al. concluded that effectsof ATD were more pronounced early in recovery.123 In contrast, Booij et al. concluded thatduration of remission was not associated with mood response to ATD.21

Booij et al. investigated predictors of relapse in a pooled analysis of the individual patient dataof some of the studies included in this review.21 This is often referred to as a ‘mega-analysis’. Theyfound that recurrent depressive episodes (≥2), female gender, and a history of suicidal thoughts/attempts predicted relapse. Duration of remission did not contribute to this prediction whenconfounding was considered. We found a modest relationship between relapse and the durationof the remission after ATD. We defined the duration of remission as the reported averageduration, or – if not stated – the minimum duration of remission used as inclusion criterion for thestudies. We found that especially in the first 5 months after remission, ATD caused lowering of

Figure 6.6A/B.

Acute phenylalanine/tyrosine depletion (APTD) in healthy controls (6.6A. within subjects design; 6.6B. between subjectsdesign), stratified by status of family history for depression. References to studies as indicated. FH- = family history negative,FH+ = family history positive, N/A = not reported.

<�7��8� �,�� T��T�� T�� T�� !��

��!T �� "� ��!#� $�#��%�&�!� ��#� ��!#� $�#��%�&�!�� '��#��T " " ��!#� $�#��(� �)*+��, �+ �)*+��,

��-��#��.��-�'��

�� )*+��,� �� %��#��T�� %��.��%%��/�0��1�2��0��1�2�

��-�� .��-�3�4��.��

�� )*+��,� �� %��#��T�� %��.��%%��/�0��156��0��1�*�

�6�-��!��"��!��

�� )*+��,� �� %��#��T��7�0��1)68�!%�0��0��165�8�,7�0��+�� %��.��%%��/�0��1��0��16��

��)*+��,� �� %��#��T��7�0�519)8�!%�0�6��0��1�)�8�,7�0�6915+�� %��.��%%��/�0��1:��0��1**�

�'5 �'� �� 5

�;��!�� !�� ;��!�� !��"�

<�.��=� ;�� 5��T��T�� T�� >�=1�� T�� !��

��!T �� "� ��!#� $�#��!�� #� ��!#� $�#��!�� '��#��T " " ��!#� $�#��(� �)*+��, �+ �)*+��,

��-�'?�T��)))��

�� )*+��,� �� %��#��T�� %��.��%%��/�0��1*6��0��1:��

��-��"��;�?��5��

�� )*+��,� �� %��#��T�� %��.��%%��/�0��1�6��0��1�:�

��)*+��,� �� %��#��T��7�0��1�98�!%�0��0��1�:�8�,7�0��1�+�� %��.��%%��/�0��156��0��1:9�

�'5 �'� �� 5

�;��!�� !�� ;��!�� !��"�

�

�


6

129

mood in remitted patients still using AD. However, the problem with our and Bell’s comparison isthe intraindividual spread in duration within the studies. This spread is not considered at the samelevel of detail as in a ‘mega-analysis’. In addition, confounding can only be considered in ‘mega-analysis’. Therefore – although their study did not include all available studies – we think thatBooij et al. provide the best available indication of riskfactors for mood lowering effects by ATD.

Do depletion studies elucidate the pathogenesis of MDD?The absence of robust mood effects in healthy controls indicates that mood is not a directcorrelate of 5-HT or NE levels in the brain. The only healthy controls who are modestly affected bymonoamine depletion studies are healthy controls with a positive family history for MDD. Thismight be indicative of a biological vulnerability which is revealed by depletion studies. Of interestare the findings of recent studies that combined ATD with neuroimaging or geneticsampling,47;49;51;66;68-70;100-10846;51 reviewed by Fusar-Poli et al.25 The intelligent approach of combiningdepletion with imaging or genotyping appears very promising. A summary of the complex resultsof these studies goes beyond the scope of our review. However, these neuroimaging andgenotyping studies also suggest that monoamine depletion discloses rather a ‘trait’ vulnerabilitythan a pure ‘state’ dependent change due to depletion.

Additionally, a depressive relapse after monoamine depletion in remitted patients who useAD, occurs only if the target of the depletion (5-HT, NE) coincides with the working mechanism ofthe antidepressant used. This emphasizes that AD indeed specifically affect their supposed targetsystems. However, we may only conclude that an undepleted 5-HT system is required forserotonergic AD. The same holds for the NE system and norepinephrinergic AD. Delgadoproposed an alternative explanation for the decrease in mood after monoamine depletion inpatients: depletion of e.g. 5-HT may give the same effect as abrupt discontinuation of SSRIs.124

Rapid discontinuation is also associated with mood-effects, which are considered to be differentfrom a depressive relapse.125

What certainly cannot be concluded is that MDD is caused by low levels of 5-HT and/or NE/DA.This simplification, which is often used to promote the use of AD specifically affecting 5-HT or NEor both systems, represents a Catch-22 argument, and ignores the notion that serotonergic andnorepinephrinergic AD presumably act by a final common pathway. In this pathway post-synapticchanges at the cellular level are supposed to be responsible for the remission of MDD.126 Cellularchanges include up or down regulated receptors, increased neuronal interconnections andsprouting and changes in levels of neuropeptides (e.g. corticotrophin releasing hormone). Theclinical question of how to distinguish a patient who will respond to an AD with e.g. NE effects hasnot been solved. Descriptive variables at the symptom level have not yet sufficiently predictedthe response to any selective agent. Therefore, a pragmatic approach to affect this final commonpathway might be to prescribe ADs which target both 5-HT and NE. However, the effectiveness ofthis approach is still equivocal.127-129

In patients who are depressed at the time of monoamine depletion, no further decrease inmood is observed. A ceiling effect could be responsible for this finding. But, a morestraightforward conclusion is that there is no simple relation between 5-HT or NE deficiency andmood or MDD. Nevertheless, this finding is complicated by the finding of some authors120;121 thatmood is lowered or elevated the day after ATD. A delayed decrease of mood was indicative fortreatment refractoriness,121 a finding that was not yet replicated by others.86;87;120 The number of

Figure 6.7.

Acute phenylalanine/tyrosine depletion (APTD) in former depressed patients in remission without medication, studied in awithin subjects design. References to studies as indicated.

#��%� � �� %&��'�&��%�� "��&�� %&��'�&��%��()��&�� * * ��%&��'�&��#+� �,-.��/ �. �,-.��/

�� -�� 0��-��

�� ,-.��/� �� 1��&��2�3��456!�%1�3��3��4�5�!�/2�3�7-46. ��1�� 11��8�3��4�9��3��4,7�

�)7 �)� �� 7

��%��%�� %��%��$*

Chap

ter

Chapter 6

6

130

comparable studies to date is limited. Therefore, we think no clear conclusions or explanationscan be made, except that even in MDD patients mood is not a correlate of 5-HT or NE levels in thebrain.

We agree with the conclusion of Booij et al. and Bell et al. that a relapse of depressivesymptoms in remitted patients after depletion probably reflects a biological vulnerability of the 5-HT system in remitted patients.11;21;123 This vulnerability increases their risk to become depressed.This increased risk was further demonstrated in two prospective studies that used the responseto ATD in remitted patients to predict later relapse/recurrence.130;131 However, also these resultsneed replication. Moreover, the cause of an increased vulnerability remains uncertain. Probablythe cause is a combination of genetic, environmental and otherdeterminants (e.g. ‘scarring’ thebrain after multiple depressive episodes).123

In conclusion, monoamine depletion by ATD and APTD does not elucidate a causal factor inthe pathogenesis of MDD. However, ATD and APTD remain useful models to safely and directlymanipulate 5-HT, NE and DA function in living humans, and to study the behavioral consequencesof this manipulation, especially in subgroups of humans with an apparent vulnerability.11;12;24

Still, several methodological issues need to be addressed. First, because of the competition ofaminoacids to pass the blood-brain barrier, ATD might unwillingly result in an intracerebralincrease of tyrosine/phenylalanine.24;132 Vice-versa, APTD might increase intracerebral tryptophanavailability. Levels of other amino-acids than those depleted are not provided in the studies.Second, ATD provides a specific net lowering of 5-HT. Contrary, depletion of tyrosine andphenylalanine lowers both NE and DA, which are synthesized in the same cascade (with DAthereafter transformed into NE by dopamine beta-hydroxylase). Also AMPT interferes early in thiscascade. Although evidence from animal studies points to more DA depletion by APTD and moreNE depletion by AMPT,133 it is impossible to truly distinguish between net NE and DA depletion.Third, test re-test reliability for monoamine depletion paradigms was only tested for ATD and wasrather limited, which limits the robustness of the method.134;135 Fourth, also in healthy controlssubtle cognitive effects of monoamine depletion in the brain occur: deficits in learning andmemory consolidation and improvement in focused attention and executive function.136 Theseeffects show similarity with symptoms of MDD. The question remains whether these effectsmight represent mild first symptoms of MDD or the starting symptoms of a cascade of alteredbrain functions leading to MDD. Fifth, in line with the fourth issue, MDD does not develop withinone day. Therefore, the changes by experimental monoamine depletion by ATD and APTD may betoo short to really induce a complex biological and psychological deregulation which is recognizedas MDD. Patients suffering from gastrointestinal carcinoid tumors – 5-HT producing tumors withexpected prolonged states of secondary tryptophan depletion – are generally not depressed butdo show improved focused attention.137 However, carcinoid findings were not yet related totryptophan levels over time. Sixth, the single depletion of one monoamine system by ATD orAPTD/AMPT may be too simplistic, especially because of the complex interaction of monoaminesystems. Five dual depletion studies were not included in this review.38;138-141 In healthy controlscontrasting results were found. Hughes et al. found some decrease of mood on 3 VAMS-subscales,139 which was also found in another open study.140 However, no effects were found in 2other studies.38;141 In unmedicated patients with MDD no significant increase of MDD was foundafter dual depletion.138 It would be interesting to investigate the effects of simultaneous depletionof 5-HT and NE/DA in other populations.

Limitations of the studies and the reviewSome limitations should be mentioned. First, female gender is a risk-factor for MDD, and was alsofound to be a predictor of relapse after ATD in remitted patients. Based on several studies, genderdifferences in 5-HT metabolismare probable, and hormonal factors may play a role in 5-HTfunction.21 In our results we examined the effect of gender in healthy controls. The differencebetween males and females was most prominent in healthy controls without a family history ofMDD. In the included studies hormonal status (pre or postmenopausal state) was notdistinguished in the results nor used as inclusion criterion. Second, many small differences


6

131

between the studies existed: different composition of depletion and control drinks, differentpresentation of tryptophan/tyrosine/phenylalanine or their ratios to other neutral amino acids,different presentation of free vs. total tryptophan/tyrosine/phenylalanine values, differentmeasurements, different scales. Differences between studies undoubtedly introducedheterogeneity between the studies, which may bias the results of this review. Therefore, werecommend an international consensus protocol. Third, limited presentation of the data forced usto make assumptions, that may have influenced our findings. However, the assumptionsappeared to have little influence on the results in a sensitivity analysis. Future studies need toaddress clear presentation of their data, preferably including a description of the directions of theeffects, SDs, in cross-over designs the numbers of subjects having the control or sham first, andpreferably include a SD of the pooled difference between the depletion and the control condition.For example, only 10 studies adequately reported the number of patients treated in eachsequence in the within subjects design.54;57;64-66;68;70;95;122;135 Fourth, 6 studies included BipolarPatients. Three of these studies were also included in the meta-analysis.65;94;114 This involved 6patients of the total of 265 patients (2.2%) in the concerning meta analytical comparisons.Therefore, we consider the possibility of bias by the inclusion of this etiologically differentdisorder unlikely. Fifth, the rate of agreement in the selection of studies still may rise questionsabout the clarity of our selection criteria. However, discrepancies in agreement could easily besolved between the two reviewers. Therefore, we do not think our sensitive searches missedrelevant studies. Finally, an indication of publication bias was found. If publication bias truly exists,the studies which found no effect or an increase in mood after ATD would not have beenpublished. Indeed, the mood effects of studies that could not be included in the meta-analysiswere mostly very small and non-significant. Furthermore, the exact information in studiesrequired for meta-analysis forced us to exclude many studies. It seems natural that a non-significant result will not be given this level of detailed attention (change-scores with SDs),especially when mood-effects are not the primary outcome in these studies. Without thispublication bias the pooled effect of ATD in healthy controls would probably have been lower.Because our conclusion already is that there is no apparent mood effect of ATD in healthycontrols, we conclude that the observed publication bias will not severely distort the generalconclusion. Therefore, despite these limitations, we consider the results of our review after ourstrict methods as valid.

Conclusion and future studiesWe conclude that although ATD and APTD are important in the investigation of the monoaminesystems, monoamine depletion does not directly decrease mood. Although previously themonoamine systems were considered to be responsible for the development of MDD, theavailable evidence to date does not support a direct causal relationship with MDD. There is nosimple direct correlation of 5-HT or NE levels in the brain and mood. The depletion of 5-HT by ATDand NE/DA by APTD most clearly decreases mood in vulnerable patients who are in remission fromtheir MDD, while still using AD. Furthermore, depletion affects mood in unmedicated patients inremission or healthy controls with a family history positive for MDD. Therefore, the monoaminesystems are probably important systems in the vulnerability to become depressed. The changes inbrain metabolism in remitted patients who relapse after ATD or AMPT suggest that 5-HT and NEsystems give input to a final common pathway which needs further research to be clarified.

We suggest some lines for future research. First: three or four-armed depletion studiescomparing ATD, APTD (and – if possible – their combination) versus sham depletion to investigatethe differential effects of the 5-HT and NE systems on mood in remitted patients or controls withpositive family history for MDD. Second: a further exploration of the relation between knowngenetic polymorphisms of the 5-HT, NE and DA systems (e.g. 5-HTTPR) and biological cerebralresponses to depletion paradigms, as measured by PET/fMRI (e.g. like Neumeister et al.46;51).Third: the relations between monoamine depletion indicating biological vulnerability andpsychological vulnerability for MDD (see Booij et al.)11 Fourth: Replication, further validation andstandardization of the properties of ATD and other depletion paradigms as a diagnostic test for

Chap

ter

Chapter 6

6

132

recurrences in remitted patients (see Moreno et al.130 and Neumeister et al.131). New studies willincrease our knowledge of the 5-HT and NE systems, which are important targets in the currenttreatments of MDD. This knowledge will finally improve the treatment for MDD.

Acknowledgement

The authors wish to thank Mrs Natasha Wiebe, MMath PStat, Research Associate at theUniversity of Alberta, Canada and especially Dr. Rob J.P.M. Scholten, M.D. PhD, epidemiologistand director of the Dutch Cochrane Centre at the Academic Medical Centre, Amsterdam for theirindispensable help with the statistic methods used in this review. This study was partially fundedby the program Opleiding Onderzoekers GGZ (OOG) (project number 100-002-002) by ZonMw,the Hague, the Netherlands.

Conflict of interests

None

References






6. Hindmarch I. Beyond the monoamine hypothesis: mechanisms, molecules and methods. Eur Psychiatry. 2002; 17 Suppl 3:294-299.


8. Burke WJ. Selective versus multi-transmitter antidepressants: are two mechanisms better than one? J Clin Psychiatry.2004; 65 Suppl 4: 37-45.

9. Fava M, Kendler KS. Major depressive disorder. Neuron. 2000; 28: 335-341.10. Delgado PL. How antidepressants help depression: mechanisms of action and clinical response. J Clin Psychiatry. 2004; 65

Suppl 4: 25-30.11. Booij L, Van der Does AJ, Riedel WJ. Monoamine depletion in psychiatric and healthy populations: review. Mol Psychiatry.

2003; 8: 951-973.12. Hood SD, Bell CJ, Nutt DJ. Acute tryptophan depletion. Part I: Rationale and methodology. Aust N Z J Psychiatry. 2005; 39:

558-564.13. Young SN, Smith SE, Pihl RO, Ervin FR. Tryptophan depletion causes a rapid lowering of mood in normal males.

Psychopharmacol (Berl). 1985; 87: 173-177.14. Moja EA, Lucini V, Benedetti F, Lucca A. Decrease in plasma phenylalanine and tyrosine after phenylalanine-tyrosine free

amino acid solutions in man. Life Sci. 1996; 58: 2389-2395.15. Shopsin B, Gershon S, Goldstein M, Friedman E, Wilk S. Use of synthesis inhibitors in defining a role for biogenic amines

during imipramine treatment in depressed patients. Psychopharmacol Commun. 1975; 1: 239-249.16. Gratton A. Time course analysis of para-chlorophenylalanine induced suppression of self-stimulation behavior. Pharmacol

Biochem Behav. 1982; 17: 597-602.17. Delgado PL, Miller HL, Salomon RM, Licinio J, Heninger GR, Gelenberg AJ et al. Monoamines and the mechanism of

antidepressant action: effects of catecholamine depletion on mood of patients treated with antidepressants.Psychopharmacol Bull. 1993; 29: 389-396.


6

133

18. Van der Does AJ. The effects of tryptophan depletion on mood and psychiatric symptoms. J Affect Disord. 2001; 64: 107-119.

19. Bell C, Abrams J, Nutt D. Tryptophan depletion and its implications for psychiatry. Br J Psychiatry. 2001; 178: 399-405.20. Moore P, Landolt HP, Seifritz E, Clark C, Bhatti T, Kelsoe J et al. Clinical and physiological consequences of rapid



Gen Psychiatry. 2001; 58: 200-202.23. Riedel WJ, Klaassen T, Schmitt JA. Tryptophan, mood, and cognitive function. Brain Behav Immun. 2002; 16: 581-589.24. Reilly JG, McTavish SF, Young AH. Rapid depletion of plasma tryptophan: a review of studies and experimental

methodology. J Psychopharmacol. 1997; 11: 381-392.25. Fusar-Poli P, Allen P, McGuire P, Placentino A, Cortesi M, Perez J. Neuroimaging and electrophysiological studies of the

effects of acute tryptophan depletion: a systematic review of the literature. Psychopharmacol (Berl). 2006; 188:131-143.26. McNair DM, Lorr M, Droppelman LF. Manual for the Profile of Mood States. San Diego: Educational and Industrial Testing

Service; 1988.27. Zuckerman M, Lubin B. Manual for the Multiple Affect Adjective Checklist. San Diego CA: Educational and testing service;

1965.28. Bond A, Lader M. The use of analogue scales in rating subjective feelings. Br J Med Psychol. 1974; 47: 211-218.29. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 23: 56-61.30. Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979; 134: 382-

389.31. Kawanishi C, Lundgren S, Agren H, Bertilsson L. Increased incidence of CYP2D6 gene duplication in patients with

persistent mood disorders: ultrarapid metabolism of antidepressants as a cause of nonresponse. A pilot study. Eur J ClinPharmacol. 2004; 59: 803-807.

32. Review Manager. [Version 4.2 for Windows]. 2002. Oxford, UK, The Cochrane Collaboration. 33. The Cochrane Collaboration. RevMan 4.2 User Guide, version 4.2 for Windows. 4.2 ed. Oxford, UK: The Cochrane

Collaboration; 2002.34. Cochrane Reviewers' Handbook. 4.2.2 [Updated March 2004] ed. Chichester, UK: John Wiley & Sons, Ltd; 2004.35. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986; 7: 177-188.36. Egger M, Davey SG, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997; 315: 629-

634.37. Harrison BJ, Olver JS, Norman TR, Nathan PJ. Effects of serotonin and catecholamine depletion on interleukin-6

activation and mood in human volunteers. Hum Psychopharmacol. 2002; 17: 293-297.38. Harrison BJ, Olver JS, Norman TR, Burrows GD, Wesnes KA, Nathan PJ. Selective effects of acute serotonin and

catecholamine depletion on memory in healthy women. J Psychopharmacol. 2004; 18: 32-40.39. Leyton M, Young SN, Pihl RO, Etezadi S, Lauze C, Blier P et al. A comparison of the effects of acute tryptophan depletion

and acute phenylalanine/tyrosine depletion in healthy women. Adv Exp Med Biol. 1999; 467: 67-71.40. Leyton M, Young SN, Pihl RO, Etezadi S, Lauze C, Blier P et al. Effects on mood of acute phenylalanine/tyrosine depletion

in healthy women. Neuropsychopharmacology. 2000; 22: 52-63.41. Benkelfat C, Ellenbogen MA, Dean P, Palmour RM, Young SN. Mood-lowering effect of tryptophan depletion. Enhanced

susceptibility in young men at genetic risk for major affective disorders. Arch Gen Psychiatry. 1994; 51: 687-697.42. Delgado PL, Price LH, Miller HL, Salomon RM, Licinio J, Krystal JH et al. Rapid serotonin depletion as a provocative

challenge test for patients with major depression: relevance to antidepressant action and the neurobiology ofdepression. Psychopharmacol Bull. 1991; 27: 321-330.

43. Hayward G, Goodwin GM, Cowen PJ, Harmer CJ. Low-dose tryptophan depletion in recovered depressed patients induceschanges in cognitive processing without depressive symptoms. Biol Psychiatry. 2005; 57: 517-524.

44. Klaassen T, Riedel WJ, van Someren A, Deutz NE, Honig A, Van Praag HM. Mood effects of 24-hour tryptophan depletionin healthy first-degree relatives of patients with affective disorders. Biol Psychiatry. 1999; 46: 489-497.

45. Klaassen T, Riedel WJ, Deutz NE, Van Praag HM. Mood congruent memory bias induced by tryptophan depletion. PsycholMed. 2002; 32: 167-172.

46. Neumeister A, Konstantinidis A, Stastny J, Schwarz MJ, Vitouch O, Willeit M et al. Association between serotonintransporter gene promoter polymorphism (5HTTLPR) and behavioral responses to tryptophan depletion in healthywomen with and without family history of depression. Arch Gen Psychiatry. 2002; 59: 613-620.

47. Neumeister A, Nugent AC, Waldeck T, Geraci M, Schwarz M, Bonne O et al. Neural and behavioral responses totryptophan depletion in unmedicated patients with remitted major depressive disorder and controls. Arch Gen Psychiatry.2004; 61: 765-773.

48. O'Reardon JP, Chopra MP, Bergan A, Gallop R, DeRubeis RJ, Crits-Christoph P. Response to tryptophan depletion in majordepression treated with either cognitive therapy or selective serotonin reuptake inhibitor antidepressants. BiolPsychiatry. 2004; 55: 957-959.

49. Praschak-Rieder N, Hussey D, Wilson AA, Carella A, Lee M, Dunn E et al. Tryptophan depletion and serotonin loss inselective serotonin reuptake inhibitor-treated depression: an [18F] MPPF positron emission tomography study. BiolPsychiatry. 2004; 56: 587-591.

50. Stewart ME, Deary IJ, Ebmeier KP. Neuroticism as a predictor of mood change: the effects of tryptophan depletion. Br JPsychiatry. 2002; 181: 242-247.

Chap

ter

Chapter 6

6

134

51. Neumeister A, Hu XZ, Luckenbaugh DA, Schwarz M, Nugent AC, Bonne O et al. Differential effects of 5-HTTLPR genotypeson the behavioral and neural responses to tryptophan depletion in patients with major depression and controls. Arch GenPsychiatry. 2006; 63: 978-986.

52. Gallagher P, Massey AE, Young AH, McAllister-Williams RH. Effects of acute tryptophan depletion on executive functionin healthy male volunteers. BMC Psychiatry. 2003; 3: 10.

53. Hughes JH, Gallagher P, Stewart ME, Matthews D, Kelly TP, Young AH. The effects of acute tryptophan depletion onneuropsychological function. J Psychopharmacol. 2003; 17: 300-309.

54. Murphy FC, Smith KA, Cowen PJ, Robbins TW, Sahakian BJ. The effects of tryptophan depletion on cognitive andaffective processing in healthy volunteers. Psychopharmacol (Berl). 2002; 163: 42-53.

55. Park SB, Coull JT, McShane RH, Young AH, Sahakian BJ, Robbins TW et al. Tryptophan depletion in normal volunteersproduces selective impairments in learning and memory. Neuropharmacology. 1994; 33: 575-588.

56. Rubinsztein JS, Rogers RD, Riedel WJ, Mehta MA, Robbins TW, Sahakian BJ. Acute dietary tryptophan depletion impairsmaintenance of "affective set" and delayed visual recognition in healthy volunteers. Psychopharmacol (Berl). 2001; 154:319-326.

57. Schmitt JA, Jorissen BL, Sobczak S, van Boxtel MP, Hogervorst E, Deutz NE et al. Tryptophan depletion impairs memoryconsolidation but improves focussed attention in healthy young volunteers. J Psychopharmacol. 2000; 14: 21-29.

58. Shansis FM, Busnello JV, Quevedo J, Forster L, Young S, Izquierdo I et al. Behavioural effects of acute tryptophandepletion in healthy male volunteers. J Psychopharmacol. 2000; 14: 157-163.

59. Booij L, Van der Does AJ, Haffmans PM, Riedel WJ, Fekkes D, Blom MJ. The effects of high-dose and low-dose tryptophandepletion on mood and cognitive functions of remitted depressed patients. J Psychopharmacol. 2005; 19: 267-275.

60. Harmer CJ, McTavish SF, Clark L, Goodwin GM, Cowen PJ. Tyrosine depletion attenuates dopamine function in healthyvolunteers. Psychopharmacol (Berl). 2001; 154: 105-111.

61. Lythe KE, Anderson IM, Deakin JFW, Elliott R, Strickland PL. Lack of behavioural effects after acute tyrosine depletion inhealthy volunteers. J Psychopharmacol. 2005; 19: 5-11.

62. McLean A, Rubinsztein JS, Robbins TW, Sahakian BJ. The effects of tyrosine depletion in normal healthy volunteers:implications for unipolar depression. Psychopharmacol (Berl). 2004; 171: 286-297.

63. Grevet EH, Tietzmann MR, Shansis FM, Hastenpflugl C, Santana LC, Forster L et al. Behavioural effects of acutephenylalanine and tyrosine depletion in healthy male volunteers. J Psychopharmacol. 2002; 16: 51-55.

64. McTavish SF, Mannie ZN, Harmer CJ, Cowen PJ. Lack of effect of tyrosine depletion on mood in recovered depressedwomen. Neuropsychopharmacology. 2005; 30: 786-791.

65. Roiser JP, McLean A, Ogilvie AD, Blackwell AD, Bamber DJ, Goodyer I et al. The subjective and cognitive effects of acutephenylalanine and tyrosine depletion in patients recovered from depression. Neuropsychopharmacology. 2005; 30: 775-785.

66. Allen PP, Cleare AJ, Lee F, Fusar-Poli P, Tunstall N, Fu CHY et al. Effect of acute tryptophan depletion on pre-frontalengagement. Psychopharmacology. 2006; 187:486-497.

67. Amin Z, Gueorguieva R, Cappiello A, Czarkowski KA, Stiklus S, Anderson GM et al. Estradiol and Tryptophan DepletionInteract to Modulate Cognition in Menopausal Women. Neuropsychopharmacology. 2006;31:2489-2497.

68. Evers EA, van der Veen FM, van Deursen JA, Schmitt JA, Deutz NE, Jolles J. The effect of acute tryptophan depletion onthe BOLD response during performance monitoring and response inhibition in healthy male volunteers. Psychopharmacol(Berl). 2006; 187: 200-208.

69. van der Veen FM, Evers EA, van Deursen JA, Deutz NE, Backes WH, Schmitt JA. Acute tryptophan depletion reducesactivation in the right hippocampus during encoding in an episodic memory task. Neuroimage. 2006; 31: 1188-1196.

70. Evers EA, van der Veen FM, Jolles J, Deutz NE, Schmitt JA. Acute tryptophan depletion improves performance andmodulates the BOLD response during a Stroop task in healthy females. Neuroimage. 2006; 32: 248-255.

71. Talbot PS, Watson DR, Barrett SL, Cooper SJ. Rapid tryptophan depletion improves decision-making cognition in healthyhumans without affecting reversal learning or set shifting. Neuropsychopharmacology. 2006; 31: 19.

72. McAllister-Williams RH, Massey AE, Rugg MD. Effects of tryptophan depletion on brain potential correlates of episodicmemory retrieval. Psychopharmacol (Berl). 2002; 160: 434-442.

73. Abbott FV, Etienne P, Franklin KB, Morgan MJ, Sewitch MJ, Young SN. Acute tryptophan depletion blocks morphineanalgesia in the cold-pressor test in humans. Psychopharmacol (Berl). 1992; 108: 60-66.

74. Crean J, Richards JB, de Wit H. Effect of tryptophan depletion on impulsive behavior in men with or without a familyhistory of alcoholism. Behav Brain Res. 2002; 136: 349-357.

75. Koszycki D, Zacharko RM, Le Melledo JM, Young SN, Bradwejn J. Effect of acute tryptophan depletion on behavioral,cardiovascular, and hormonal sensitivity to cholecystokinin-tetrapeptide challenge in healthy volunteers. Biol Psychiatry.1996; 40: 648-655.

76. Miller HEJ, Deakin JFW, Anderson IM. Effect of acute tryptophan depletion on CO2-induced anxiety in patients with panicdisorder and normal volunteers. Br J Psychiatry. 2000; 176: 182-188.

77. Oldman AD, Walsh AE, Salkovskis P, Laver DA, Cowen PJ. Effect of acute tryptophan depletion on mood and appetite inhealthy female volunteers. J Psychopharmacol. 1994; 8: 8-13.

78. Richell RA, Deakin JF, Anderson IM. Effect of acute tryptophan depletion on the response to controllable anduncontrollable noise stress. Biol Psychiatry. 2005; 57: 295-300.

79. Schmeck K, Sadigorsky S, Englert E, Demisch L, Dierks T, Barta S et al. Mood changes following acute tryptophandepletion in healthy adults. Psychopathology. 2002; 35: 234-240.

80. Weltzin TE, Fernstrom MH, Fernstrom JD, Neuberger SK, Kaye WH. Acute tryptophan depletion and increased foodintake and irritability in bulimia nervosa. Am J Psychiatry. 1995; 152: 1668-1671.

81. Coupland N, Zedkova L, Sanghera G, Leyton M, Le Melledo JM. Response to pentagastrin after acute phenylalanine andtyrosine depletion in healthy men: a pilot study. J Psychiatry Neurosci. 2001; 26: 247-251.


6

135

82. McCann UD, Thorne D, Hall M, Popp K, Avery W, Sing H et al. The effects of L-dihydroxyphenylalanine on alertness andmood in alpha-methyl-para-tyrosine-treated healthy humans. Further evidence for the role of catecholamines in arousaland anxiety. Neuropsychopharmacology. 1995; 13: 41-52.

83. Booij L, Swenne CA, Brosschot JF, Haffmans PM, Thayer JF, Van der Does AJ. Tryptophan Depletion Affects Heart RateVariability and Impulsivity in Remitted Depressed Patients with a History of Suicidal Ideation. Biol Psychiatry. 2006; 60:507-514.

84. Danjou P, Hamon M, Lacomblez L, Warot D, Kecskemeti S, Puech AJ. Psychomotor, subjective and neuroendocrineeffects of acute tryptophan depletion in the healthy volunteer. Psychiatry Psychobiol. 1990; 5: 31-38.

85. Ravindran AV, Griffiths J, Merali Z, Knott VJ, Anisman H. Influence of acute tryptophan depletion on mood and immunemeasures in healthy males. Psychoneuroendocrinology. 1999; 24: 99-113.

86. Price LH, Malison RT, McDougle CJ, McCance-Katz EF, Owen KR, Heninger GR. Neurobiology of tryptophan depletion indepression: effects of m-chlorophenylpiperazine (mCPP). Neuropsychopharmacology. 1997; 17: 342-350.

87. Price LH, Malison RT, McDougle CJ, Pelton GH, Heninger GR. The neurobiology of tryptophan depletion in depression:effects of intravenous tryptophan infusion. Biol Psychiatry. 1998; 43: 339-347.

88. Krahn LE, Lin SC, Klee GG, Lu PY, Ory SJ, Zimmermann RC. The effect of presynaptic catecholamine depletion on 6-hydroxymelatonin sulfate: a double blind study of alpha-methyl-para-tyrosine. Eur Neuropsychopharmacol. 1999; 9: 61-66.

89. Zimmermann RC, Krahn L, Klee G, Lu PY, Ory SJ, Lin SC. The impact of gender on alpha-methyl-para-tyrosine mediatedchanges in prolactin secretion and 6-hydroxymelatonin sulfate excretion. Psychoneuroendocrinology. 1996; 21: 469-478.

90. Bhatti T, Gillin JC, Seifritz E, Moore P, Clark C, Golshan S et al. Effects of a tryptophan-free amino acid drink challenge onnormal human sleep electroencephalogram and mood. Biol Psychiatry. 1998; 43: 52-59.

91. Kahkonen S, Makinen V, Jaaskelainen IP, Pennanen S, Liesivuori J, Ahveninen J. Serotonergic modulation of mismatchnegativity. Psychiatry Res. 2005; 138: 61-74.

92. Knott VJ, Howson AL, Perugini M, Ravindran AV, Young SN. The effect of acute tryptophan depletion and fenfluramine onquantitative EEG and mood in healthy male subjects. Biol Psychiatry. 1999; 46: 229-238.

93. Haynes PL, McQuaid JR, Kelsoe J, Rapaport M, Gillin JC. Affective state and EEG sleep profile in response to rapidtryptophan depletion in recently recovered nonmedicated depressed individuals. J Affect Disord. 2004; 83: 253-262.

94. Evans L, Golshan S, Kelsoe J, Rapaport M, Resovsky K, Sutton L et al. Effects of rapid tryptophan depletion on sleepelectroencephalogram and mood in subjects with partially remitted depression on bupropion.Neuropsychopharmacology. 2002; 27: 1016-1026.

95. Landolt HP, Kelsoe JR, Rapaport MH, Gillin JC. Rapid tryptophan depletion reverses phenelzine-induced suppression ofREM sleep. J Sleep Res. 2003; 12: 13-18.

96. Moore P, Gillin C, Bhatti T, DeModena A, Seifritz E, Clark C et al. Rapid tryptophan depletion, sleep electroencephalogram,and mood in men with remitted depression on serotonin reuptake inhibitors. Arch Gen Psychiatry. 1998; 55: 534-539.

97. Debener S, Strobel A, Kurschner K, Kranczioch C, Hebenstreit J, Maercker A et al. Is auditory evoked potentialaugmenting/reducing affected by acute tryptophan depletion? Biol Psychol. 2002; 59: 121-133.

98. Hughes JH, Ashton CH, Matthews D, Young AH. Acute depletion of plasma tryptophan does not alter electrophysiologicalvariables in healthy males. Psychopharmacol (Berl). 2000; 152: 119-121.

99. Voderholzer U, Hornyak M, Thiel B, Huwig-Poppe C, Kiemen A, Konig A et al. Impact of experimentally induced serotonindeficiency by tryptophan depletion on sleep EEG in healthy subjects. Neuropsychopharmacology. 1998; 18: 112-124.

100. Praschak-Rieder N, Wilson AA, Hussey D, Carella A, Wei C, Ginovart N et al. Effects of tryptophan depletion on theserotonin transporter in healthy humans. Biol Psychiatry. 2005; 58: 825-830.

101. Bremner JD, Innis RB, Salomon RM, Staib LH, Ng CK, Miller HL et al. Positron emission tomography measurement ofcerebral metabolic correlates of tryptophan depletion-induced depressive relapse. Arch Gen Psychiatry. 1997; 54: 364-374.

102. Smith KA, Morris JS, Friston KJ, Cowen PJ, Dolan RJ. Brain mechanisms associated with depressive relapse and associatedcognitive impairment following acute tryptophan depletion. Br J Psychiatry. 1999; 174: 525-529.

103. Bremner JD, Vythilingam M, Ng CK, Vermetten E, Nazeer A, Oren DA et al. Regional brain metabolic correlates of alpha-methylparatyrosine-induced depressive symptoms: implications for the neural circuitry of depression. JAMA. 2003; 289:3125-3134.

104. Yatham LN, Liddle PF, Shiah IS, Lam RW, Adam MJ, Zis AP et al. Effects of rapid tryptophan depletion on brain 5-HT(2)receptors: a PET study. Br J Psychiatry. 2001; 178: 448-453.

105. Talbot PS, Cooper SJ. Anterior cingulate and subgenual prefrontal blood flow changes following tryptophan depletion inhealthy males. Neuropsychopharmacology. 2006; 31: 09.

106. Rubia K, Lee F, Cleare AJ, Tunstall N, Fu CH, Brammer M et al. Tryptophan depletion reduces right inferior prefrontalactivation during response inhibition in fast, event-related fMRI. Psychopharmacol (Berl). 2005; 179: 791-803.

107. Evers EA, Cools R, Clark L, van der Veen FM, Jolles J, Sahakian BJ et al. Serotonergic modulation of prefrontal cortexduring negative feedback in probabilistic reversal learning. Neuropsychopharmacology. 2005; 30: 1138-1147.

108. Cools R, Calder AJ, Lawrence AD, Clark L, Bullmore E, Robbins TW. Individual differences in threat sensitivity predictserotonergic modulation of amygdala response to fearful faces. Psychopharmacol (Berl). 2005; 180: 670-679.

109. Shopsin B, Friedman E, Gershon S. Parachlorophenylalanine reversal of tranylcypromine effects in depressed patients.Arch Gen Psychiatry. 1976; 33: 811-819.

110. McCann UD, Penetar DM, Shaham Y, Thorne DR, Sing HC, Thomas ML et al. Effects of catecholamine depletion onalertness and mood in rested and sleep deprived normal volunteers. Neuropsychopharmacology. 1993; 8: 345-356.

111. Aberg-Wistedt A, Hasselmark L, Stain-Malmgren R, Aperia B, Kjellman BF, Mathe AA. Serotonergic 'vulnerability' inaffective disorder: a study of the tryptophan depletion test and relationships between peripheral and central serotoninindexes in citalopram-responders. Acta Psychiatr Scand. 1998; 97: 374-380.

112. Smith SE, Pihl RO, Young SN, Ervin FR. A test of possible cognitive and environmental influences on the mood loweringeffect of tryptophan depletion in normal males. Psychopharmacol (Berl). 1987; 91: 451-457.

Chap

ter

Chapter 6

6

136

113. Weltzin TE, Fernstrom JD, McConaha C, Kaye WH. Acute tryptophan depletion in bulimia: effects on large neutral aminoacids. Biol Psychiatry. 1994; 35: 388-397.

114. Delgado PL, Miller HL, Salomon RM, Licinio J, Krystal JH, Moreno FA et al. Tryptophan-depletion challenge in depressedpatients treated with desipramine or fluoxetine: implications for the role of serotonin in the mechanism ofantidepressant action. Biol Psychiatry. 1999; 46: 212-220.

115. Miller HL, Delgado PL, Salomon RM, Heninger GR, Charney DS. Effects of alpha-methyl-para-tyrosine (AMPT) in drug-freedepressed patients. Neuropsychopharmacology. 1996; 14: 151-157.

116. Moreno FA, Gelenberg AJ, Heninger GR, Potter RL, McKnight KM, Allen J et al. Tryptophan depletion and depressivevulnerability. Biol Psychiatry. 1999; 46: 498-505.

117. Cassidy F, Murry E, Weiner RD, Carroll BJ. Lack of relapse with tryptophan depletion following successful treatment withECT. Am J Psychiatry. 1997; 154: 1151-1152.

118. Smith KA, Fairburn CG, Cowen PJ. Relapse of depression after rapid depletion of tryptophan. Lancet. 1997; 349: 915-919.119. Spillmann MK, Van der Does AJ, Rankin MA, Vuolo RD, Alpert JE, Nierenberg AA et al. Tryptophan depletion in SSRI-

recovered depressed outpatients. Psychopharmacol (Berl). 2001; 155: 123-127.120. Booij L, Van der Does AJ, Haffmans PM, Riedel WJ. Acute tryptophan depletion in depressed patients treated with a

selective serotonin-noradrenalin reuptake inhibitor: augmentation of antidepressant response? J Affect Disord. 2005; 86:305-311.

121. Delgado PL, Price LH, Miller HL, Salomon RM, Aghajanian GK, Heninger GR et al. Serotonin and the neurobiology ofdepression. Effects of tryptophan depletion in drug-free depressed patients. Arch Gen Psychiatry. 1994; 51: 865-874.

122. Berman RM, Narasimhan M, Miller HL, Anand A, Cappiello A, Oren DA et al. Transient depressive relapse induced bycatecholamine depletion: potential phenotypic vulnerability marker? Arch Gen Psychiatry. 1999; 56: 395-403.

123. Bell CJ, Hood SD, Nutt DJ. Acute tryptophan depletion. Part II: clinical effects and implications. Aust N Z J Psychiatry. 2005;39: 565-574.

124. Delgado PL. Monoamine depletion studies: implications for antidepressant discontinuation syndrome. J Clin Psychiatry.2006; 67 Suppl 4: 22-26.

125. Rosenbaum JF, Fava M, Hoog SL, Ascroft RC, Krebs WB. Selective serotonin reuptake inhibitor discontinuation syndrome:a randomized clinical trial. Biol Psychiatry. 1998; 44: 77-87.

126. Charney DS. Monoamine dysfunction and the pathophysiology and treatment of depression. J Clin Psychiatry. 1998; 59Suppl 14: 11-14.

127. Freemantle N, Anderson IM, Young P. Predictive value of pharmacological activity for the relative efficacy ofantidepressant drugs. Meta-regression analysis. Br J Psychiatry. 2000; 177: 292-302.


129. Nelson JC, Mazure CM, Jatlow PI, Bowers MB, Jr., Price LH. Combining norepinephrine and serotonin reuptake inhibitionmechanisms for treatment of depression: a double-blind, randomized study. Biol Psychiatry. 2004; 55: 296-300.

130. Moreno FA, Heninger GR, McGahuey CA, Delgado PL. Tryptophan depletion and risk of depression relapse: A prospectivestudy of tryptophan depletion as a potential predictor of depressive episodes. Biol Psychiatry. 2000; 48: 327-329.

131. Neumeister A, Habeler A, Praschak-Rieder N, Willeit M, Kasper S. Tryptophan depletion: a predictor of future depressiveepisodes in seasonal affective disorder? Int Clin Psychopharmacol. 1999; 14: 313-315.

132. Badawy AA. Acute tryptophan or tyrosine depletion test: time for reappraisal? J Psychopharmacol. 2005; 19: 429-430.133. McTavish SF, Callado L, Cowen PJ, Sharp T. Comparison of the effects of alpha-methyl-p-tyrosine and a tyrosine-free

amino acid load on extracellular noradrenaline in the rat hippocampus in vivo. J Psychopharmacol. 1999; 13: 379-384.134. Ellenbogen MA, Young SN, Dean P, Palmour RM, Benkelfat C. Mood response to acute tryptophan depletion in healthy

volunteers: sex differences and temporal stability. Neuropsychopharmacology. 1996; 15: 465-474.135. Ellenbogen MA, Young SN, Dean P, Palmour RM, Benkelfat C. Acute tryptophan depletion in healthy young women with a

family history of major affective disorder. Psychol Med. 1999; 29: 35-46.136. Riedel WJ. Cognitive changes after acute tryptophan depletion: what can they tell us? Psychol Med. 2004; 34: 3-8.137. Russo S, Nielen MMA, Boon JC, Kema IP, Willemse PHB, De Vries EGE et al. Neuropsychological investigation into the

carcinoid syndrome. Psychopharmacol (Berl). 2003; 168: 324-328.138. Berman RM, Sanacora G, Anand A, Roach LM, Fasula MK, Finkelstein CO et al. Monoamine depletion in unmedicated

depressed subjects. Biol Psychiatry. 2002; 51: 469-473.139. Hughes JM, Matrenza C, Kemp AH, Harrison BJ, Liley D, Nathan PJ. Selective effects of simultaneous monoamine

depletion on mood and emotional responsiveness. Int J Neuropsychopharmacol. 2004; 7: 9-17.140. Leyton M, Pun VK, Benkelfat C, Young SN. A new method for rapidly and simultaneously decreasing serotonin and

catecholamine synthesis in humans. J Psychiatry Neurosci. 2003; 28: 464-467.141. Salomon RM, Miller HL, Krystal JH, Heninger GR, Charney DS. Lack of behavioral effects of monoamine depletion in

healthy subjects. Biol Psychiatry. 1997; 41: 58-64.142. Curtin F, Altman DG, Elbourne D. Meta-analysis combining parallel and cross-over clinical trials. I: Continuous outcomes.

Stat Med. 2002; 21: 2131-2144.143. Deeks JJ, Altman DG, Bradburn MJ. Statistical methods for examining heterogeneity and combining results from several

studies in meta-analysis. In: Egger M, Smith GD, Altman DG, eds. Systematic reviews in healthcare. Meta-analysis in context.2 ed. London, UK: BMJ Publishers Group; 2001:285-312.

144. Altman DG. Comparing groups-categorical data. In: Altman DG, ed. Practical statistics for medical research. 1st ed. London:Chapman & Hall; 1991:229-276.

145. Barr LC, Heninger GR, Goodman W, Charney DS, Price LH. Effects of fluoxetine administration on mood response totryptophan depletion in healthy subjects. Biol Psychiatry. 1997; 41: 949-954.


6

137

146. Klaassen T, Riedel WJ, Deutz NE, van Someren A, Van Praag HM. Specificity of the tryptophan depletion method.Psychopharmacol (Berl). 1999; 141: 279-286.

147. Smith KA, Clifford EM, Hockney RA, Clark DM, Cowen PJ. Effect of tryptophan depletion on mood in male and femalevolunteers: A pilot study. Hum Psychopharmacol. 1997; 12: 111-117.

148. Morris JS, Smith KA, Cowen PJ, Friston KJ, Dolan RJ. Covariation of activity in habenula and dorsal raphe nuclei followingtryptophan depletion. Neuroimage. 1999; 10: 163-172.

Supplementary appendix: statistical formulas.

Differences between intervention and control measurementsIn depletion studies changes in mood scores typically represented mean mood-scores both before(pre) and after (post) the depletion/challenge (experimental intervention) and the placebo/sham/control-intervention. Because the mood scores were not necessarily identical at the start of theexperiment and the control, we first calculated the mean change in mood-score (pooleddifference) for the experimental and control condition separately per study. Some studies alsoprovided the standard deviation (SD) of the pooled differences. When the standard deviation wasnot reported, we calculated the standard deviation of the pooled difference for paired data:

In this formula, the correlation coefficient R was calculable in four studies only40;41;50;119 andranged between 0.42and 1.00 for the experimental and between 0.34and 0.95 for the controlcondition. To be able to calculate the standard deviation of the change between pre and post testmood scores for the rest of the studies we imputed a correlation coefficient R of 0.5. This valuewas considered to be a conservative assumption.

Statistics for studies with a within subjects designThe difference in the changes of mood scores between intervention and control were expressedas difference of change scores:

For this difference the SD of the difference was calculated by again applying formula (1), withan assumed R of 0.5.

To acknowledge the different mood scales to measure change in mood, the difference inchanges between experimental and control condition were standardized by calculating Hedges’adjusted g, which is similar to Cohen’s d, but includes an adjustment for small sample bias:33

In this formula nAB and nBA represent the number of subjects randomized to intervention orcontrol as first test in the study. If the numbers for nAB and nBA were not reported, we assumedthat the sample was split half for the two sequences. For Hedges’ ga SE was calculated:142;143

)***2(22postprepostprechange SDSDRSDSDSD −+= (1).

INTCONTchanges ChangeChangeDifference −= (2).

)9)(4

31(*SD

Hedges' CONT

−+−−=

BAABchangesDifference

INT

nnChangeChangeg (3).

( )⎟⎟⎠⎞

⎜⎜⎝

⎛−+

+∗

+=94.32

'4

2

BAABBAAB

BAAB

nngHedges

nnnnSE (4).

Chap

ter

Chapter 6

6

138

Statistics for studies with a between subjects designFor between subject studies comparable statistics were used to calculate the mean change inmood-scores. Because the between-subjects design is a parallel group design, Hedges’ g wascalculated with formula (3) in which for nAB and nBA nINT and nCONT were substituted. The formulafor the SE was slightly different to acknowledge the absence of paired data:

Statistics for Relapse ratesFor relapse rates of MDD after depletion provided in a within subjects design the difference inrelapse rates was calculated as:

in which N is the total number of patients included. The standard error then is:

in which b represents the number of patients with a relapse after the intervention but not thecontrol condition and c the number of patients with a relapse after the control but not theintervention.144 If the numbers of ‘pairs’ were not extractable from the paper, a conservativeapproach was used assuming the minimal number of patients relapsing both after theintervention and the control condition (maximal c), resulting in the largest SE.

( )⎟⎟⎠⎞

⎜⎜⎝

⎛−+

+∗+=

94.32' 2

CONTINTCONTINT

CONTINT

nngHedges

nnnnSE (5).

Nn

Nn

Difference CONTlapseINTlapse _Re_Rerate-lapseRe −= (6),

Ncbcb

NSE raterelapseDifference

2

_)(1 −−+=−

(7),

CHA

PTER

SEROTONIN TRANSPORTER BINDING WITH [123I]β-CIT SPECT IN MAJOR DEPRESSIVE DISORDER

VERSUS CONTROLS: EFFECT OF SEASON AND

GENDER.

Submitted

Henricus G. Ruhé1

Jan Booij2

Johannes B. Reitsma3

Aart H. Schene1

1 Program for Mood Disorders, Department of Psychiatry, Academic Medical Center, University ofAmsterdam, Amsterdam, the Netherlands2 Department of Nuclear Medicine, Academic Medical Center, University of Amsterdam, Amsterdam,the Netherlands3 Department of Clinical Epidemiology, Biostatistics & Bioinformatics, Academic Medical Center,University of Amsterdam, Amsterdam, the Netherlands

Chap

ter

Chapter 7

7

140

Abstract

BackgroundThe serotonin system is undoubtedly involved in the pathogenesis of major depressive disorder(MDD). More specifically the serotonin transporter (SERT) serves as a major target forantidepressant drugs. There are conflicting results about SERT availability in depressed patientsversus healthy controls (HC).

AimTo measure SERT availability and study effects of age, gender and season of scanning in MDDpatients versus HC.

MethodsWe included 49 depressed outpatients (42.3 ± 8.3 [SD] years) with a Hamilton Depression RatingScale above 18, who were drug-naïve or drug-free for ≥4 weeks, and 49 age- (±2 years) and sex-matched HC. Subjects were scanned with single photon emission computed tomography (SPECT)using [123I]β-CIT. SERT availability was expressed as specific to non-specific binding ratios (BPND) inmidbrain (MID) and diencephalon (DIENC) with cerebellar binding as a reference.

ResultsIn crude comparisons between patients and HC, we found no significant differences in MID orDIENC SERT availability. In subgroup analyses, depressed males had numerically lower MID SERTavailability than HC, whereas in women SERT availability was not different (significantdiagnosis*gender interaction; p= 0.048). In DIENC we found a comparable diagnosis*gender (p=0.002) and an additional smoking*gender (p= 0.036) interaction. In MID the season of scanningshowed a significant main-effect (p= 0.018) with higher SERT availability in winter.

ConclusionsDifferences in MID and DIENC SERT availability in MDD patients compared with HC are modifiedby gender. The season of scanning is a covariate in MID. The diagnosis*gender andgender*smoking interactions on SERT availability should be considered in future studies of thepathogenesis of MDD.

Serotonin transporter availability in MDD

141

Chapter7

Introduction

Major depressive disorder (MDD) is a highly prevalent and disabling disease, often treated byselective serotonin reuptake inhibitors (SSRIs).1 SSRIs block the serotonin transporter (SERT),which lowers the reuptake of serotonin (5-HT) from the synaptic cleft and increasesneurotransmission. Despite the fact that the working-mechanism of antidepressants supports themonoamine deficiency theory, the pathogenesis of MDD remains unclear.2;3 Therefore,differences in SERT availability in patients and healthy controls (HC) have been studied previously.

Postmortem studies have shown reduced or unchanged concentrations of SERTs in MDD-patients compared with HC, but these studies may be biased by retrospective data collection,previous antidepressant use, suicidal behavior apart from MDD or non-selective ligands (reviewedby Stockmeier).4

Cerebral SERTs in humans can be quantified in-vivo with single photon emission computedtomography (SPECT) and positron emission tomography (PET). The first SERT radioligand, iodine-123-labeled 2β-carbomethoxy-3β-(4-iodophenyl)-tropane ([123I]β-CIT) binds both to SERTs anddopamine transporters (DATs).5 Diencephalon (DIENC) and brainstem (MID) [123I]β-CIT bindingpredominantly reflect SERT, while striatal [123I]β-CIT uptake reflects DAT.6

Studies comparing depressed patients with HC have reported either decreased,7-13

unchanged,14-18 or increased19 SERT availability in MDD patients (Table S7.1). A negative correlationbetween SERT availability and severity of depression (measured by Hamilton Depression RatingScale (HDRS) scores) was reported in patients with primary MDD9 or Wilson’s disease.20

Discrepant results among studies may be explained by differences in scanning techniques,analytic methods, and subject sampling, although the effects of additional variables and theirinteraction might also explain conflicting results. Staley et al. reported lower SERT availability inthe DIENC of MDD female patients versus HC,11 and suggested that this interaction accounted forthe contradictory results between studies. Furthermore, in HC significant effects on SERTavailability have been reported for gender,21 smoking behavior,21 aging22;23 and season ofscanning.24

Our objectives were to quantify SERT availability in MDD patients versus HC while accountingfor these potential covariates and possible interactions, and to correlate SERT availability withdepression severity. Therefore, we compared [123I]β-CIT SPECT scans of drug-free MDD patientswith age- and sex-matched HC.

Methods

SubjectsAfter approval by the institutional ethical committee and written informed consent, we recruiteddepressed patients from primary care, and our outpatient department (October 2003-August2006). Patients were eligible if they were 25-55 years old, had a diagnosis of MDD (diagnosed bystructured clinical interview for DSM-IV (SCID Patient Version)), had a HDRS score >18, and weredrug-free or used no more than one antidepressant (stopped for >4 weeks and ≥5 half-lives of thisantidepressant before scanning) for the present MDD-episode. Exclusion criteria were pregnancy(or desire to become pregnant), bipolar disorder, psychotic features, primary anxiety and/orsubstance abuse disorders and acute, severe suicidal ideation. We allowed secondary co-morbidanxiety and/or substance abuse.

We individually matched each patient by gender and age (±2 years). HC were in good physicalhealth and had never used psychotropic medication. Exclusion criteria were current or lifetimepsychiatric disorder(s) according to the SCID (including abuse or addiction disorders), a BeckDepression Inventory (BDI) score >9, alcohol use >4 units per day (last month) or a 1st-degreerelative with psychiatric disorder(s). We allowed HC to have incidentally used illicit drugs unless

Chap

ter

Chapter 7

7

142

criteria for a DSM-IV disorder were met, but prohibited illicit drug use the month prior to scanning.Patients and HC received €50 and €40, respectively. No restrictions were made with respect tosmoking behavior.

Procedure & SPECT-imagingWe performed all scans 230 ±18 (SD) minutes after intravenous injection of approximately 100MBq [123I]β-CIT, when the radioligand was at equilibrium for SERT binding in brain areasexpressing high densities of SERTs.22 Radiosynthesis of [123I]β-CIT and image acquisition weredescribed earlier.25 We performed SPECT imaging using a 12-detector single slice brain-dedicatedscanner (Neurofocus 810, Strichmann Medical Equipment; Cleveland, OH) with a full-width at half-maximum resolution of 6.5 mm, throughout the 20 cm field-of-view (http://www.neurophysics.com).

Image AnalysisAfter attenuation correction and reconstruction in 3D mode (http://www.neurophysics.com),

we selected regions of interest (ROIs) for midbrain (MID), DIENC and cerebellum (CER) by usingvalidated templates (see Figure 2.3).25 One examiner (HGR), blinded for diagnosis positioned allROIs in two series. Intra-class correlation coefficients were ≥0.98 for all ROIs. If the two seriesdiffered by >5%, scans were re-evaluated by a second investigator (JB). In the analyses weaveraged the counts for the two series.

We assumed activity in CER to represent non-displaceable activity (non-specific binding andfree radioactivity).26 We calculated the binding potential (BP) as the rate of specific to non-displaceable (ND) binding (BPND = ) for MID and DIENC.27 BPND is proportional totransporter number under equilibrium conditions.

StatisticsGeneral linear models were used to analyze differences in BPND in MID and DIENC betweendepressed patients and HC using the following modeling strategy.

We first compared mean BPND in MDD patients versus HC in univariable (‘crude’) models, onlycontaining the main effect of diagnosis (categorical: MDD/HC). We then fitted multivariablemodels by adding variables to the model, which in the literature have been reported to influenceBPND. These variables included: gender (categorical: male/female), age (continuous), smoking(categorical: yes/no), season of scanning (categorical: “winter” October-March/ “summer” April-September).24 In addition, a number of specific two-way interactions were examined, againbecause they have been reported as significant in previous studies, which included:diagnosis*gender, diagnosis*smoking, gender*smoking (‘full multivariable models’). Three-wayinteractions were not examined because of the relatively small sample-size (from a statisticalperspective). The Akaike’s Information Criterion (AIC) was used to judge whether the two-wayinteractions improved the model. If a two-way interaction did not improve the fit of the model, itwas removed from the model in order to facilitate the interpretation of the model (‘reducedmultivariable models’). Main effects always remained in the model, irrespective of theirsignificance, in order to report their (lack of) impact. Diencephalon and midbrain data wereanalyzed separately, but the same set of variables and interaction terms were examined using thesame modeling approach. If significant interactions were present, we performed post-hocanalyses in order to report the absolute differences in SERT availability in the involved subgroups.Differences between subgroups were analyzed within the framework of the multivariable modeland tested for significance using the residual variance estimate of the model. For explorativeanalyses see the supplementary appendix.

We examined the association of BPND with HDRS scores using linear regression models inpatients, correcting for covariates used in the multivariable models. We used SPSS (version 15.0.1)for statistical procedures (www.spss.com). We expressed all means ±SD, except in Figures 7.1 andS7.1, where ±SEM was used for legibility.

( )CER

CERROI

ActivityActivityActivity −


143

Chapter7

Results

We studied 17 male and 32 female patients with MDD, versus 17 male and 32 female HC (Table 7.1).Patients smoked significantly more (n= 27; 55.1%) than HC (n= 11; 22.4%; χ2= 11.2; df= 2; p= 0.004).Among HC we included more Caucasians (n= 44; 89.8%) than among patients (n= 31; 63.3%; χ2=11.9; df= 3; p= 0.008). We scanned 67% of patients and 55% of HC in winter (χ2= 1.55, df= 1, p= 0.213).In one female patient insufficient CER was scanned as a reference, and in three patients and onecontrol MID slices were insufficient; these were omitted in the analyses. Of 15 patients who usedantidepressants during their lives, 3 used antidepressants for the current episode. One patientused mirtazapine until 4 weeks before scanning, all others stopped antidepressants 6-132 monthsbefore scanning. Illicit drug abuse mainly involved cannabis. Lifetime MDMA-use occurred in noneof the patients and in one female control (<10 tablets; last use 8 months before scanning).

SERT availability in MDD patients versus healthy controls (‘crude models’)Mean MID BPND in patients (0.62 ±0.22 [SD]) was not significantly different from HC (0.63 ±0.19;F1,94= 0.118; p= 0.733). BPND in DIENC was 1.15 (±0.24) in patients and 1.09 (±0.26) in HC (F1,97=1.209; p= 0.274).

Multivariable models of SERT availability in MDD patients versus healthycontrolsMidbrainFor MID, the full multivariable model (including diagnosis, gender, age, smoking, season ofscanning, diagnosis*gender, diagnosis*smoking, and gender*smoking) was subsequentlyreduced by removing the non-significant interaction terms diagnosis*smoking, and

Table 7.1. Baseline characteristics of MDD patients and healthy controls (stratified by gender).

MDD patients Healthy controls (HC)

Male (n= 17) Female (n= 32) Male (n= 17) Female (n= 32)

Age (years) 43.2 ±7.8 41.8 ±8.6 43.0 ±8.1 41.9 ±8.9

Current cigarette smokers; n (%) *

Alcohol use >8 Units/ week; n (%) †12 (70.6) 3 (17.6)

15 (46.9) 4 (12.5)

3 (17.6) 7 (41.2)

8 (25.0) 8 (25.0)

Race: n (%) Caucasian Creole Asian

12 (70.6) 2 (11.8) 2 (11.8)

19 (59.4) 7 (21.9) 5 (15.6)

15 (88.2) 2 (11.8) 0

29 (90.6) 3 (9.4) 0

MDD Severity‡

First episode; n (%) Drug-naïve; n (%) Melancholic; n (%) Atypical; n (%) Suicidal thoughts, plan or attempt; n (%) Duration of episode: n (%) <5 months 5 months – 2 years > 2 years Age of first episode (years)

23.2 ±4.9 9 (52.9) 12 (70.6) 12 (70.6) 0 7 (41.2)

5 (29.4) 8 (47.1) 4 (23.5) 35.9 ±10.3

25.4 ±4.7 18 (56.3) 22 (68.8) 24 (75.0) 3 (9.4) 10 (31.3)

8 (25.0) 21 (65.6) 3 (9.4) 35.8 ±10.4

1.3 ±1.4 N/A

2.0 ±2.1 N/A

Co-morbidity: n (%) Anxiety disorder Dysthymia Alcohol dependence Drug (alcohol, cannabis, benzodiazepines) abuse

4 (23.5) 1 (5.9) 1 (5.9) 4 (23.5)

7 (21.9) 1 (3.1) 3 (9.4)3 (9.4)

N/A N/A

SPECT scan in winter; n (%) 9 (52.9) 24 (75.0%) 7 (41.2) 20 (62.5)

Numbers represent means ± SD* Significantly more smoking in MDD patients than in HC (χ2 = 11.0; 1 df; p< 0.01) † Trend of lower alcohol use in MDD patients than in HC (χ2 = 3.75; 1 df; p= 0.053)‡ Severity measured by Hamilton Depression Rating Scale (17-items) in MDD patients and by Beck Depression Inventory in HC

Chap

ter

Chapter 7

7

144

gender*smoking (AIC-decrease= 2.478; Figure 7.1A). This reduced multivariable model had asignificant diagnosis*gender interaction (F1,87= 4.039; p= 0.048). In post-hoc comparisons, MDDmales showed a trend of lower BPND compared with male HC (difference -0.124; t87= -1.718; p=0.089), while female patients and HC did not differ (difference 0.047; t87= 0.898; p= 0.372).Furthermore, the main effect of season of scanning was significant (F1,87= 5.814; p= 0.018). Scansperformed in winter showed on average 18% higher BPND than scans made in summer (F2,87=3.248; p= 0.044). The main effects of age and smoking were also included in this reduced model,but were not significant (p= 0.227 and p= 0.582, respectively; Figure 7.1A).

Diencephalon For DIENC, the full multivariable model (including the same variables as MID) could not bereduced as all three interactions terms (diagnosis*smoking, diagnosis*gender andsmoking*gender) improved the fit of the model as evaluated by the AIC (Figure 2B). This fullmultivariable model showed significant diagnosis*gender (F1,88= 10.227; p= 0.002) andsmoking*gender (F1,88= 4.541; p= 0.036) interactions. The diagnosis*smoking interaction was notsignificant (p= 0.127). In post-hoc comparisons, smoking MDD males had significant lower BPNDcompared with smoking male HC (difference -0.304; t88= -2.643; p= 0.010). In non-smoking MDDmales BPND was numerically lower than in non-smoking male HC (difference -0.140; t88= -1.340; p=0.184). Contrary, non-smoking female patients had higher BPND than non-smoking female HC(difference 0.221; t88= 3.064; p= 0.003) with almost no difference in BPND between MDD and HC insmoking females (difference 0.057; t88= 0.616; p= 0.539). Furthermore, male smoking HC hadsignificantly higher BPND than non-smoking male HC (difference 0.279; t88= -2.384; p= 0.019),while in female HC BPND was not affected by smoking (difference 0.032; t88= -0.371; p= 0.712). Themain effects of season of scanning and age were also included in this model, but were notsignificant (p= 0.679 and p= 0.247, respectively; Figure 7.1B).

Relation of SERT availability and severity of MDD With linear regression models, we found no significant relation between HDRS-scores and BPND,neither in MID, nor in DIENC when accounting for gender, age, smoking and season of scanning.

Figure 7.1. BPND values for midbrain and diencephalon.

Multivariable models for BPND in MDD patients (MDD) versus HC stratified for gender. Values represent estimated means±SEM. A. Midbrain (n= 94), corrected for main effects of diagnosis (p= 0.414), gender (p= 0.128), season of scanning (F1,87= 5.814; p=0.018), smoking (p= 0.582), age (p= 0.227) and diagnosis*gender interaction (F1,87= 4.039;p= 0.048). * Post-hoc differences t87= -1.718;p= 0.089 between: MDD males versus HC males. B. Diencephalon (n= 97), corrected for main effects of diagnosis (p= 0.476), gender (p= 0.277), season of scanning (p= 0.679),smoking (p= 0.223), age (p= 0.247) and diagnosis*gender interaction (F1,88= 10.227; p= 0.002), gender*smoking interaction(F1,88= 4.541; p= 0.036) and diagnosis*smoking interaction (p= 0.127). * Post-hoc difference t88= -2.384; p= 0.019 between smoking HC males versus non-smoking HC males. ** Post-hoc differencest88> 2.643; p< 0.01 between: smoking MDD males versus smoking HC males, non-smoking MDD females versus non-smokingHC females.

��

��

��

��

��

��

�

�

��

� ��

��

��

��

��

��

��

��

��

��

��

��

��

� ��

��

��

��

� �


145

Chapter7

Discussion

In the present – until now largest – study of MDD patients versus HC, we aimed to quantify SERTavailability in MDD patients versus HC while accounting for covariates and interactions, and tocorrelate SERT availability with depression severity. We did not find significant differences in SERTavailability in MID or DIENC in crude comparisons. However, a significant diagnosis*genderinteraction existed in MID and DIENC, combined with a significant gender*smoking interaction inDIENC only. Depressed males, but not females, had lower MID SERT availability compared withHC. In DIENC depressed smoking males had significantly lower MID SERT availability comparedwith smoking male HC, while non-smoking female patients had higher SERT availability thanfemale HC. Furthermore, the season of scanning influenced SERT availability in MID, with higherSERT availability in winter. We found no clinically relevant correlation of HDRS-scores with SERTavailability.

Comparison with previous studiesOur results replicate earlier reports of similar SERT availability in MID and DIENC in MDD patientsand HC.14-18 Other studies reported increased19 or decreased7-13 SERT availability in MDD patientsversus HC. However, none of these studies – except two11;13 – investigated the effect of genderand no study corrected for season. Furthermore, we replicated a significant contribution ofseason of scanning on MID BPND.24

The diagnosis*gender interaction in MID and DIENC is our most important finding. Contrary toStaley and colleagues,11;21 we found a different direction of this interaction in DIENC: significantlylower BPND in MDD males in MID (-17%) and DIENC (-18%), and higher BPND in MDD females in MID(+9%, non-significant) and DIENC (+13%, significant) than in HC. Staley et al. found 1% lower DIENCSERT availability in MDD males, and 22% lower DIENC SERT availability in MDD females. We found agender*smoking interaction in DIENC (with highest BPND in smoking healthy males), while Staleyfound higher SERT availability in the brainstem (attributable to males). Our findings suggest that afailure to stratify for a diagnosis*gender interaction may obscure differences between patientsand HC.

Methodological explanations for inconsistent findingsDespite technical differences between studies (scanning protocols, radioligands, image analyses),variation in the selection of HC (e.g. having relatives with psychiatric diagnoses) or patients (fromdifferent source populations) is the most probable explanation for inconsistent findings. Previousstudies recruited patients in general psychiatric outpatient and university clinics.7;8;10;12;18;19;28 Werecruited 65% of our patients from primary care settings. We adequately diagnosed patients bySCID, and required an HDRS >18 for inclusion. Thus, we recruited severe and often melancholicpatients that were drug-free, with 69% of the patients drug-naïve. Three studies7;13;29 includedlarger proportions of drug-naïve patients. Like Parsey et al.,10 we observed (non-significant -15%)lower MID SERT availability in drug-naïve patients (results available on request). Additionally,some studies suggested that anxiety disorders influence SERT availability,30 and MDD with co-morbid anxiety may differ from ‘pure MDD’. However, this was not observed in our sample(results available on request).

Role of SERT in the pathogenesis of MDDSERTs evacuate extracellular 5-HT from the synapse. Observed differences in SERTs betweenpatients and HC may represent differences in the number of SERT containing neurons, in thenumber of SERTs per neuron or a combination of both.

Two major mechanisms for the role of SERT in MDD are hypothesized.30 First, increased SERTavailability reduces 5-HT from the synapse more easily, which might lower 5-HT transmission,possibly leading to MDD. Second, as the brain might apply compensation mechanisms to retainhomeostasis, a decreased 5-HT transmission by MDD may result in downregulation (decrease) of

Chap

ter

Chapter 7

7

146

SERT in order to increase 5-HT transmission. A sequential occurrence of these two mechanismscould also be hypothesized: an initially increased SERT availability destabilizes (with or without anadditional factor) and leads to MDD, which is followed by a decrease in SERT to compensate fordecreased 5-HT transmission.

Differential effects of MDD on SERT availability between sexes may be explained via sex-hormones. Estrogen replacement after ovariectomy increased SERT mRNA and SERT availabilityin female rats31 and in hypothalamic regions of female macaques.32 Depressed women may havesignificantly higher 24h mean levels of diurnal estradiol rhythms, and may have highertestosterone levels compared with HC.33 Testosterone may increase SERT availability byconversion to estrogen by aromatase, which is especially available in DIENC. This could explain ourfinding of increased SERT availability in DIENC in females. In depressed men, the sex steroidtestosterone is decreased,33 with 34-61% biochemical hypogonadism in depressed malescompared with 6-14% in HC.34 This lack of testosterone in MDD may reduce SERT availability byreduced conversion to estrogen. Replacement of testosterone in castrated male rats increasedSERT mRNA and SERT availability.35 Since we did not measure sex hormones, and Best et al.36

found no relation between menstrual cycle or sex hormones and SERT availability in HC, theseexplanations remain speculative and should be examined further.

We replicated a main effect of season demonstrated previously in DIENC in 12 healthywomen37 and mesencephalon in 29 HC.24 Neumeister et al. observed decreased SERT availability inwinter.37 In contrast, Buchert et al.24 found increased SERT availability in winter, which was alsofound in our study. Serotonin modulates the effects of photic input in the suprachiasmaticnucleus (SCN). The SCN imposes a circadian rhythm by affecting hormonal and autonomic output(reviewed by Buijs).38 Serotonin release in the SCN is highest during waking and activity.39 Becauseraphe neurons show regular high firing rates during waking and decreased firing during sleep, itcould be hypothesized that during winter, with decreased daylight, more serotonergic activity isneeded, which may be mediated via the raphe input into the SCN.40 Increased serotonergicactivity (increased free synaptic serotonin) may result in a compensatory increase in SERTs.Nevertheless, the small size of the SCN (~0.27 mm3) by itself cannot explain higher SERTavailability found by SPECT or PET.

Limitations of the present studyThe cerebellum (especially the vermis) contains small amounts of SERT,41;42 which could result inan underestimation of BPND in patients and HC, expected to be 7% at most.43 Although asystematic underestimation of SERT due to DAT- or NET-rich areas in MID (substantia nigra andlocus coeruleus, respectively) cannot be ruled out, we think this does not differentially affectpatients or HC.

Second, lower levels of endogenous 5-HT (e.g. in MDD) could result in less competition withradioligands, increasing the specific binding measured. This was demonstrated in rhesus monkeyswith [123I]β-CIT SPECT44 but not in humans. After tryptophan depletion (artificially reducingendogenous 5-HT) no differences in SERT availability were observed45;46 but the radioligand([11C]DASB) in that study does not bind to the 5-HT recognition/translocation site, and may not besuitable to image such changes in extracellular 5-HT.47

Third, we used previously validated ROIs25 instead of Magnetic Resonance Imaging forcoregistration. Because these templates cover larger brain areas, BPND in small regions (raphenuclei) cannot be determined. This potential measurement error (non-differential for patients andHC), might have increased variance in our measurements, despite very good intra-ratercorrelation coefficients. Additionally, we did not correct for non-uniform photon attenuation andpartial volume effects in our gender-analyses. Greater skull thickness might underestimate BPNDin males, and smaller MID and DIENC in females might suppress BPND compared with males.However, these factors are unlikely to explain the observed interactions.


147

Chapter7

Fourth, the 4-week washout of antidepressants (binding to SERT) may be too short.48 Becauseall but one patient stopped antidepressants ≥6 months before scanning, we think no substantialbias of the BPND assessment was introduced by competitive binding by traces of previousantidepressants.

Fifth, we allowed previous incidental use of illicit drugs (marijuana/cannabis n= 10, or MDMAn= 1) in our HC. Because heavy use of MDMA (>50 MDMA tablets) can damage serotoninneurons,49 we performed an additional analysis in which we excluded data from the MDMA-userin the HC group. However, this exclusion did not affect our results (results available on request).Sixth, we did not check personal or family history of psychiatric illnesses in HC, nor did we test foralcohol or drug abuse.

ConclusionWe showed lower SERT availability in MID and DIENC in depressed males and higher SERTavailability in DIENC in depressed (non-smoking) females compared with HC. We replicated aseasonal influence on MID SERT availability, and found a gender*smoking interaction on DIENCSERT availability. This study points to complex effects of gender, smoking and season on theserotonergic system in the pathogenesis of MDD.

Acknowledgements

We wish to thank patients and healthy controls for their participation in the study. We thankgeneral practitioners and psychiatric residents for appropriate referrals of patients. Staffmembers and technicians of the department of Nuclear Medicine provided indispensable help inmaking the SPECT-scans. We would like to acknowledge Dr. Maartje M. de Win, MD PhD for herhelp to recruit HC. Dr. Jules Lavalaye, MD PhD helped in the initial design of this study. Mrs.Michelle L. Miller revised the text linguistically. Dr. Ruud M. Buijs, PhD commented constructivelyon an earlier version of the manuscript. Henricus G. Ruhé received a grant from the NetherlandsOrganisation for Health Research and Development (ZonMw), program Mental Health, educationof investigators in mental health (OOG; #100-002-002).


None

References


2. Ruhe HG, Mason NS, Schene AH. Mood is indirectly related to serotonin, norepinephrine and dopamine levels in humans:a meta-analysis of monoamine depletion studies. Mol Psychiatry. 2007; 12: 331-359.


4. Stockmeier CA. Involvement of serotonin in depression: evidence from postmortem and imaging studies of serotoninreceptors and the serotonin transporter. J Psychiatr Res. 2003; 37: 357-373.

5. Innis R, Baldwin R, Sybirska E, Zea Y, Laruelle M, al Tikriti M et al. Single photon emission computed tomography imagingof monoamine reuptake sites in primate brain with [123I]CIT. Eur J Pharmacol. 1991; 200: 369-370.

6. Laruelle M, Baldwin RM, Malison RT, Zea-Ponce Y, Zoghbi SS, al Tikriti MS et al. SPECT imaging of dopamine andserotonin transporters with [123I]β-CIT: pharmacological characterization of brain uptake in nonhuman primates.Synapse. 1993; 13: 295-309.

Chap

ter

Chapter 7

7

148

7. Lehto S, Tolmunen T, Joensuu M, Saarinen PI, Vanninen R, Ahola P et al. Midbrain binding of [123I]nor-β-CIT in atypicaldepression. Prog Neuropsychopharmacol Biol Psychiatry. 2006; 30: 1251-1255.

8. Malison RT, Price LH, Berman R, van Dyck CH, Pelton GH, Carpenter L et al. Reduced brain serotonin transporteravailability in major depression as measured by [123I]-2β-carbomethoxy-3β-(4-iodophenyl)tropane and single photonemission computed tomography. Biol Psychiatry. 1998; 44: 1090-1098.

9. Newberg AB, Amsterdam JD, Wintering N, Ploessl K, Swanson RL, Shults J et al. [123I]-ADAM binding to serotonintransporters in patients with major depression and healthy controls: a preliminary study. J Nucl Med. 2005; 46: 973-977.

10. Parsey RV, Hastings RS, Oquendo MA, Huang YY, Simpson N, Arcement J et al. Lower serotonin transporter bindingpotential in the human brain during major depressive episodes. Am J Psychiatry. 2006; 163: 52-58.

11. Staley JK, Sanacora G, Tamagnan G, Maciejewski PK, Malison RT, Berman RM et al. Sex differences in diencephalonserotonin transporter availability in major depression. Biol Psychiatry. 2006; 59: 40-47.

12. Willeit M, Praschak-Rieder N, Neumeister A, Pirker W, Asenbaum S, Vitouch O et al. [123I]β-CIT SPECT imaging showsreduced brain serotonin transporter availability in drug-free depressed patients with seasonal affective disorder. BiolPsychiatry. 2000; 47: 482-489.

13. Joensuu M, Tolmunen T, Saarinen PI, Tiihonen J, Kuikka J, Ahola P et al. Reduced midbrain serotonin transporteravailability in drug-naive patients with depression measured by SERT-specific [123I] nor-β-CIT SPECT imaging. PsychiatryRes. 2007; 154: 125-131.

14. Ahonen A, Heikman P, Kauppinen T, Koskela A, Bergström K. Serotonin transporter availability in drug free depressionpatients using a novel SERT ligand [Abstract #106]. Eur J Nucl Med. 2004; 31 (suppl 1): S227.

15. Catafau AM, Perez V, Plaza P, Pascual JC, Bullich S, Suarez M et al. Serotonin transporter occupancy induced byparoxetine in patients with major depression disorder: a [123I]-ADAM SPECT study. Psychopharmacology (Berl). 2006; 189:145-153.

16. Herold N, Uebelhack K, Franke L, Amthauer H, Luedemann L, Bruhn H et al. Imaging of serotonin transporters and itsblockade by citalopram in patients with major depression using a novel SPECT ligand [123I]-ADAM. J Neural Transm. 2006;113: 659-670.

17. Meyer JH, Houle S, Sagrati S, Carella A, Hussey DF, Ginovart N et al. Brain serotonin transporter binding potentialmeasured with carbon 11-labeled DASB positron emission tomography: effects of major depressive episodes and severityof dysfunctional attitudes. Arch Gen Psychiatry. 2004; 61: 1271-1279.

18. Reivich M, Amsterdam JD, Brunswick DJ, Shiue CY. PET brain imaging with [11C](+)McN5652 shows increased serotonintransporter availability in major depression. J Affect Disord. 2004; 82: 321-327.

19. Ichimiya T, Suhara T, Sudo Y, Okubo Y, Nakayama K, Nankai M et al. Serotonin transporter binding in patients with mooddisorders: a PET study with [11C]McN5652. Biol Psychiatry. 2002; 51: 715-722.

20. Eggers B, Hermann W, Barthel H, Sabri O, Wagner A, Hesse S. The degree of depression in Hamilton rating scale iscorrelated with the density of presynaptic serotonin transporters in 23 patients with Wilson's disease. J Neurol. 2003; 250:576-580.

21. Staley JK, Krishnan-Sarin S, Zoghbi S, Tamagnan G, Fujita M, Seibyl JP et al. Sex differences in [123I]β-CIT SPECT measuresof dopamine and serotonin transporter availability in healthy smokers and nonsmokers. Synapse. 2001; 41: 275-284.


23. van Dyck CH, Malison RT, Seibyl JP, Laruelle M, Klumpp H, Zoghbi SS et al. Age-related decline in central serotonintransporter availability with [123I]β-CIT SPECT. Neurobiol Aging. 2000; 21: 497-501.

24. Buchert R, Schulze O, Wilke F, Berding G, Thomasius R, Petersen K et al. Is correction for age necessary in SPECT or PET ofthe central serotonin transporter in young, healthy adults? J Nucl Med. 2006; 47: 38-42.

25. de Win MM, Habraken JB, Reneman L, Brink W vd, den Heeten GJ, Booij J. Validation of [123Iβ-CIT SPECT to assessserotonin transporters in vivo in humans: a double-blind, placebo-controlled, crossover study with the selective serotoninreuptake inhibitor citalopram. Neuropsychopharmacology. 2005; 30: 996-1005.

26. Laruelle M, Vanisberg MA, Maloteaux JM. Regional and subcellular localization in human brain of [3H]paroxetine binding,a marker of serotonin uptake sites. Biol Psychiatry. 1988; 24: 299-309.

27. Innis RB, Cunningham VJ, Delforge J, Fujita M, Gjedde A, Gunn RN et al. Consensus nomenclature for in vivo imaging ofreversibly binding radioligands. J Cereb Blood Flow Metab. 2007; 27: 1533-1539.

28. Dahlstrom M, Ahonen A, Ebeling H, Torniainen P, Heikkila J, Moilanen I. Elevated hypothalamic/midbrain serotonin(monoamine) transporter availability in depressive drug-naive children and adolescents. Mol Psychiatry. 2000; 5: 514-522.

29. Meyer JH, Wilson AA, Ginovart N, Goulding V, Hussey D, Hood K et al. Occupancy of serotonin transporters by paroxetineand citalopram during treatment of depression: a [11C]DASB PET imaging study. Am J Psychiatry. 2001; 158: 1843-1849.

30. Meyer JH. Imaging the serotonin transporter during major depressive disorder and antidepressant treatment. JPsychiatry Neurosci. 2007; 32: 86-102.

31. Fink G, Sumner BE, McQueen JK, Wilson H, Rosie R. Sex steroid control of mood, mental state and memory. Clin ExpPharmacol Physiol. 1998; 25: 764-775.

32. Lu NZ, Eshleman AJ, Janowsky A, Bethea CL. Ovarian steroid regulation of serotonin reuptake transporter (SERT)binding, distribution, and function in female macaques. Mol Psychiatry. 2003; 8: 353-360.

33. Swaab DF, Bao AM, Lucassen PJ. The stress system in the human brain in depression and neurodegeneration. Ageing ResRev. 2005; 4: 141-194.

34. McIntyre RS, Mancini D, Eisfeld BS, Soczynska JK, Grupp L, Konarski JZ et al. Calculated bioavailable testosterone levelsand depression in middle-aged men. Psychoneuroendocrinology. 2006; 31: 1029-1035.

35. Fink G, Sumner B, Rosie R, Wilson H, McQueen J. Androgen actions on central serotonin neurotransmission: relevance formood, mental state and memory. Behav Brain Res. 1999; 105: 53-68.


149

Chapter7

36. Best SE, Sarrel PM, Malison RT, Laruelle M, Zoghbi SS, Baldwin RM et al. Striatal dopamine transporter availability with[123I]β-CIT SPECT is unrelated to gender or menstrual cycle. Psychopharmacology (Berl). 2005; 183: 181-189.

37. Neumeister A, Pirker W, Willeit M, Praschak-Rieder N, Asenbaum S, Brucke T et al. Seasonal variation of availability ofserotonin transporter binding sites in healthy female subjects as measured by [123I]-2β-carbomethoxy-3β-(4-iodophenyl)tropane and single photon emission computed tomography. Biol Psychiatry. 2000; 47: 158-160.

38. Buijs RM, Scheer FA, Kreier F, Yi C, Bos N, Goncharuk VD et al. Organization of circadian functions: interaction with thebody. Prog Brain Res. 2006; 153: 341-360.

39. Pickard GE, Weber ET, Scott PA, Riberdy AF, Rea MA. 5HT1B receptor agonists inhibit light-induced phase shifts ofbehavioral circadian rhythms and expression of the immediate-early gene c-fos in the suprachiasmatic nucleus. J Neurosci.1996; 16: 8208-8220.

40. Moore RY, Speh JC. Serotonin innervation of the primate suprachiasmatic nucleus. Brain Res. 2004; 1010: 169-173.41. Kent JM, Coplan JD, Lombardo I, Hwang DR, Huang Y, Mawlawi O et al. Occupancy of brain serotonin transporters during

treatment with paroxetine in patients with social phobia: a positron emission tomography study with [11C]McN 5652.Psychopharmacology (Berl). 2002; 164: 341-348.

42. Parsey RV, Kent JM, Oquendo MA, Richards MC, Pratap M, Cooper TB et al. Acute occupancy of brain serotonintransporter by sertraline as measured by [11C]DASB and positron emission tomography. Biol Psychiatry. 2006; 59: 821-828.

43. Kish SJ, Furukawa Y, Chang LJ, Tong J, Ginovart N, Wilson A et al. Regional distribution of serotonin transporter protein inpostmortem human brain: is the cerebellum a SERT-free brain region? Nucl Med Biol. 2005; 32: 123-128.

44. Heinz A, Jones DW, Zajicek K, Gorey JG, Juckel G, Higley JD et al. Depletion and restoration of endogenous monoaminesaffects β-CIT binding to serotonin but not dopamine transporters in non-human primates. J Neural Transm Suppl. 2004;29-38.

45. Praschak-Rieder N, Wilson AA, Hussey D, Carella A, Wei C, Ginovart N et al. Effects of tryptophan depletion on theserotonin transporter in healthy humans. Biol Psychiatry. 2005; 58: 825-830.

46. Talbot PS, Frankle WG, Hwang DR, Huang Y, Suckow RF, Slifstein M et al. Effects of reduced endogenous 5-HT on the invivo binding of the serotonin transporter radioligand [11C]DASB in healthy humans. Synapse. 2005; 55: 164-175.

47. Hummerich R, Reischl G, Ehrlichmann W, Machulla HJ, Heinz A, Schloss P. DASB -in vitro binding characteristics on humanrecombinant monoamine transporters with regard to its potential as positron emission tomography (PET) tracer. JNeurochem. 2004; 90: 1218-1226.

48. Strauss WL, Layton ME, Dager SR. Brain elimination half-life of fluvoxamine measured by 19F magnetic resonancespectroscopy. Am J Psychiatry. 1998; 155: 380-384.

49. Reneman L, Booij J, de Bruin K, Reitsma JB, de Wolff FA, Gunning WB et al. Effects of dose, sex, and long-term abstentionfrom use on toxic effects of MDMA (ecstasy) on brain serotonin neurons. Lancet. 2001; 358: 1864-1869.

50. Heinz A, Ragan P, Jones DW, Hommer D, Williams W, Knable MB et al. Reduced central serotonin transporters inalcoholism. Am J Psychiatry. 1998; 155: 1544-1549.

51. Kugaya A, Sanacora G, Staley JK, Malison RT, Bozkurt A, Khan S et al. Brain serotonin transporter availability predictstreatment response to selective serotonin reuptake inhibitors. Biol Psychiatry. 2004; 56: 497-502.

52. Meyer JH, Wilson AA, Sagrati S, Hussey D, Carella A, Potter WZ et al. Serotonin Transporter Occupancy of Five SelectiveSerotonin Reuptake Inhibitors at Different Doses: An [11C]DASB Positron Emission Tomography Study. Am J Psychiatry.2004; 161: 826-835.

Chap

ter

Chapter 7

7

150

Supplementary appendix: explorative analyses

METHODSIn addition to the analyses described in the main text, in exploratory analyses we examinedwhether potentially relevant other factors (alcohol use,50 lifetime antidepressant use10 andduration of depression episode) had an impact on BPND by adding them as main effects to thefinal multivariable model. Because our exploratory models used small subgroups for assessmentof further confounding in multivariable models, we only present these analyses as hypothesisgenerating, exploratory models.

RESULTSMidbrainFor MID differentiation of patients who never took antidepressants (drug-naïve) versus thosewho did improved the reduced model (AIC-decrease= 1.399). Drug-naïve male and female patientshad non-significantly lower BPND than patients with a history of antidepressant use (data notshown).

Differentiation of the duration of the current episode, stratified as ≤2 years or >2 years(‘chronic’ depression) improved the reduced model (AIC-decrease= 7.289). Men depressed formore than 2 years had significantly higher BPND than men depressed ≤2 years (difference 0.333;t85= -3.102; p= 0.003), men depressed ≤2 years had significantly lower BPND than HC (difference -0.204; t85= -2.795; p= 0.006; Figure S7.1A). Differences between females were not significant.

Diencephalon Differentiation between patients with or without a secondary diagnosis of alcohol abuse/dependence improved the model for DIENC (AIC-decrease= 0.602). Smoking, alcohol-abusingmale patients had numerically higher BPND than smoking male patients not abusing alcohol(difference 0.717; t85= -1.175; p= 0.064). Non-smoking, female, alcohol-abusing patients hadsignificantly lower BPND than non-smoking, female patients not abusing alcohol (difference -0.388; t85= 2.182; p= 0.032; Figure S7.1B).

Figure S7.1. BPND values for midbrain and diencephalon (exploratory models)

Explorative models for BPND in MDD patients (MDD) versus HC stratified for gender. Values represent estimated means±SEM. Alc = Alcohol abuse/dependence.A. Midbrain (n= 94), corrected for main effects of duration of episode (F2,85= 4.777;p= 0.011), gender (p= 0.294), age (p=0.380), smoking (p= 0.636), season of scanning (F1,85= 6.194;p= 0.015), and duration of episode*gender interaction (F2,85=3.953;p= 0.023). ** Post-hoc differences for MDD males between: episodes ≤2 years versus >2 years (t85= 3.102; p= 0.003), episodes ≤2 yearsversus HC (t85= 2.795; p= 0.006). B. Diencephalon (n= 97), corrected for main effects of alcohol abuse/dependence (p= 0.521), gender (p= 0.489), season ofscanning (p= 0.511), age (p= 0.486), smoking (p= 0.067) and alcohol abuse/dependence*gender interaction (F2,85= 7.236;p=0.001), gender*smoking interaction (F1,85= 7.112;p= 0.009) and alcohol abuse/dependence*smoking interaction (p= 0.099). * Post-hoc differences t85= 2.182; p= 0.032 between non-smoking MDD females without, versus with alcohol abuse/dependence.

��

��"�

��

��

" � ��"�

��

��

" ��

� �

� �

� �

� �

�

� �

� � ��

��

��

��

��#$��%#� ��&��#$��%#

�#%�&

�%��

�&��

'��

�

��

��

��

��

��

�

��

��

��

��

��

��

��

�

��#$��%#� ��&��#$��%#

�#%�&

�%��

�&��

'��

��

151


Chapter7

Tabl

e S7

.1. P

revi

ous

stud

ies

mea

surin

g SE

RT a

vaila

bilit

y in

MD

D p

atie

nts

vers

us h

ealth

y co

ntro

ls.

Aut

hor &

Dat

e (r

efer

ence

)Po

pula

tion

(mea

n ag

e)Se

xN

SER

T Im

agin

g an

d R

OI*

Res

ults

Rem

arks

Aho

nen

2004

14D

rug

free

M

DD

pts

(34

±?)

Cont

rols

(35

±?)

M +

F ??

10 14

[123 I]

-AD

AM

SPE

CT, 1

0min

, 5h

, 7h

Mid

Br, T

hal,

Caud

, Put

am,

Pons

, Cer

Non

sign

. Diff

eren

ces

of V

3’’ i

n M

idBr

(7%

high

er in

pat

ient

s);

larg

e in

terin

divi

dual

var

iatio

n in

V3’

’ in

Mid

Br. O

ther

RO

Is n

o si

gn. d

iffer

ence

s.

Abs

trac

t of p

relim

inar

y re

sults

onl

y; li

ttle

info

rmat

ion

prov

ided

. G

ende

r not

repo

rted

. App

aren

tly n

o m

atch

ing

of c

ontr

ols

and

patie

nts.

Rec

ruitm

ent o

f pat

ient

s un

spec

ified

.

Cata

fau

2006

15D

rug-

free

M

DD

pts

(36

±11)

Cont

rols

(36

±11)

M +

F

10 10

[123 I]

-AD

AM

SPE

CT, 4

hM

idBr

, Tha

l, St

riatu

m, C

erSl

ight

ly lo

wer

but

non

sign

. diff

eren

ces

in S

ERT

avai

labi

lity

in

Mid

Br (-

4%),

Tha

l (-11

%) in

MD

D p

atie

nts

vs c

ontr

ols.

Patie

nts

wer

e dr

ug-fr

ee fo

r >6m

.A

ge m

atch

ed h

isto

ric c

ontr

ol g

roup

, no

mat

chin

g fo

r sex

. U

nequ

al m

ale-

fem

ale

dist

ribut

ion

in M

DD

vs.

con

trol

s. N

o in

form

atio

n on

fam

ily h

isto

ry o

f con

trol

s. R

ecru

itmen

t of

patie

nts

unsp

ecif

ied.

Dah

lstr

öm

2000

28D

rug-

naiv

e M

DD

pts

(13.

5 ±2

.5)

Dru

g-na

ive

Non

-MD

D p

ts (1

2.2

±2.9

)M

+ F

31 10

[123 I]β-

CIT

SPEC

T, 1,

4, 2

4h

Stria

tum

(24h

), PF

C, T

hal

(4h)

, Mid

Br (1

h), O

cc

Mid

Br V

3’’ i

n pa

tient

s si

gn. h

ighe

r (8%

) at 1

hr c

ompa

red

to

cont

rols

(diff

eren

ce n

ot s

ign.

at 4

h, a

gain

sig

n. a

t 24h

).

Pref

ront

al a

nd T

hal d

iffer

ence

s no

t sig

n. d

iffer

ent

Cont

rol-g

roup

is d

ue to

eth

ical

con

side

ratio

ns n

ot d

epre

ssed

bu

t a p

sych

iatr

ical

ly a

ffec

ted

grou

p of

ado

lesc

ents

. MD

D

patie

nts

and

cont

rols

are

not

mat

ched

for a

ge a

nd s

ex. N

o in

form

atio

n on

fam

ily h

isto

ry o

f con

trol

s. N

o ef

fect

of a

ge,

gend

er o

r co-

mor

bidi

ty fo

und.

Rec

ruitm

ent o

f out

patie

nts

in

univ

ersi

ty c

linic

.

Her

old

2006

16D

rug-

free

M

DD

pts

(42

±12)

Cont

rols

(36

±13)

M +

F

21 12

[123 I]

-AD

AM

SPE

CT, 4

hM

idBr

, Cer

Mid

Br V

3’’ i

n pa

tient

s 21

% hi

gher

(non

sign

.). M

DD

mal

es

nons

ign.

low

er V

3’’ t

han

MD

D fe

mal

es.

No

corr

elat

ion

with

SER

T an

d M

DD

sev

erity

ratin

g.

All

patie

nts

wer

e dr

ug-fr

ee fo

r >2m

. No

mat

chin

g fo

r age

or

sex.

Une

qual

mal

e-fe

mal

e di

strib

utio

n in

MD

D v

s. c

ontr

ols.

Stud

y al

so re

port

s on

occ

upan

cy a

fter

1 w

eek

of tr

eatm

ent

with

CIT

10 m

g (n

=13)

. Rec

ruitm

ent o

f pat

ient

s un

clea

r.

Ichi

miy

a 20

0219

Dru

g-fr

ee

MD

D p

ts (4

4.1 ±

13.5

) or

BD (4

1.7 ±

8.8)

Cont

rols

(42.

3 ±1

4.5)

M

13 21

[11C]

(+)M

cN56

52 P

ETM

idBr

, Tha

l, Ce

rSi

gn. h

ighe

r (22

%) in

Tha

l BP

in c

ompl

ete

patie

nt g

roup

vs

cont

rols

(in M

DD

onl

y (2

3%),

no

chan

ge in

Mid

Br6

patie

nts

had

BD; 1

0 pa

tient

s ha

d be

en tr

eate

d w

ith

antid

epre

ssan

ts ³6

w b

efor

e sc

anni

ng (1

pat

ient

with

BD

2w

). A

ge m

atch

ed (g

roup

wis

e??)

con

trol

s, n

ot fo

r sex

. Rec

ruitm

ent

of o

utpa

tient

s in

uni

vers

ity a

nd g

ener

al p

sych

iatr

ic c

linic

s.

Joen

suu

2007

13D

rug-

naïv

e M

DD

pts

(28.

8 ±8

.6)

Cont

rols

(30.

6 ±8

.9)

M +

F

29 19

[123 I]

nor β

-CIT

SPE

CT, 6

, 24

h M

idBr

, Cer

Sign

. low

er V

3’’ i

n pa

tient

s (1

0%) i

n M

idBr

vs

cont

rols

. N

o co

rrel

atio

n w

ith S

ERT

and

MD

D s

ever

ity ra

ting.

Alth

ough

men

tione

d as

suc

h, in

com

plet

e m

atch

ing

on g

ende

r an

d ag

e. 14

/29

pts

had

HD

RS ≤

18. R

ecru

itmen

t of o

utpa

tient

s fo

r psy

chod

ynam

ic p

sych

othe

rapy

in u

nive

rsity

clin

ic. N

o si

gn.

effe

cts

of g

ende

r (in

clud

ed p

atie

nts

mos

tly F

) or s

easo

n.

Leht

o 20

067

Dru

g-na

ïve

MD

D p

ts (2

8.3

±?)

Cont

rols

(30.

6 ±9

.2)

M +

F

29 18

[123 I]

nor-β

-CIT

SPE

CT, 5

m

in, 6

h, 2

4h

Mid

Br, S

tria

tum

, Cer

Sign

. low

er S

ERT

(-10

%) in

Mid

Br in

MD

D p

atie

nts v

s con

trol

s.

No

diff

eren

ces

betw

een

mel

anch

olic

, aty

pica

l, no

ndiff

eren

tiate

d M

DD

sub

type

s. L

inea

r inv

erse

cor

rela

tion

betw

een

SER

T an

d at

ypic

al s

core

. No

corr

elat

ion

betw

een

SERT

and

HD

RS.

Patie

nts

and

cont

rols

wer

e (g

roup

wis

e??)

mat

ched

for a

ge a

nd

sex.

No

subs

tant

ial r

atio

nale

giv

en fo

r rel

atio

n be

twee

n SE

RT

avai

labi

lity

and

atyp

ical

dim

ensi

on o

f MD

D. R

ecru

itmen

t of

outp

atie

nts

in u

nive

rsity

clin

ic.

152

Chapter 7

Chap

ter

7

Mal

ison

1998

8 ; Ku

gaya

200

451D

rug-

free

M

DD

pts

(44

±10)

Cont

rols

(45

±11)

M +

F

15 15

[123 I]β-

CIT

SPEC

T, 2

4h

Brai

nste

m, S

tria

tum

, Occ

Sign

. red

uctio

n (-1

9%) i

n br

ains

tem

V3’

’ of M

DD

vs.

con

trol

s,

but n

ot (-

11%)

in s

tria

tum

.N

o co

rrel

atio

n w

ith S

ERT

and

MD

D s

ever

ity ra

tings

Gro

up-m

atch

ing

of c

ontr

ols

for a

ge a

nd s

ex. N

ine

subj

ects

had

be

en tr

eate

d w

ith a

ntid

epre

ssan

ts ³3

w-7

y be

fore

sca

nnin

gO

ccip

ital c

orte

x w

as c

hose

n as

refe

renc

e.

Kuga

ya s

tudy

mai

nly

repo

rts

on o

ccup

ancy

of S

ERT

in T

hal a

nd

Mid

Br a

fter

2*

8 da

ys tr

eatm

ent w

ith C

IT 4

0 m

g ±

BUP

200

mg

or B

UP

200

mg

in c

ontr

ols

(n=

17) a

nd P

AR

20

mg

in M

DD

pa

tient

s (n

= 10

). Re

crui

tmen

t of o

utpa

tient

s in

uni

vers

ity c

linic

.

Mey

er 2

00129

Dru

g-fr

ee

MD

D p

ts (3

7 ±8

)Co

ntro

ls (?

±?)

M +

F

13 13

[11C]

DA

SP P

ETst

riatu

m, C

erN

o si

gn d

iffer

ence

in s

tria

tum

V3’

’ of M

DD

vs

cont

rols

. Sig

n.

effe

ct o

f age

.Ex

cept

2 (t

reat

ed >

2m b

efor

e) p

atie

nts

wer

e al

l dru

g-na

ïve;

in

divi

dual

age

mat

chin

g (±

2yea

rs),

no

sex-

mat

chin

g re

port

ed;

no s

epar

ate

mea

n ag

e pr

ovid

ed fo

r con

trol

s. S

tudy

mai

nly

repo

rts o

n oc

cupa

ncy

afte

r 4w

of t

reat

men

t with

PA

R 10

-20

mg

(n=

8) a

nd C

IT 2

0 m

g (n

=4).

Rec

ruitm

ent o

f pat

ient

s un

clea

r.

Mey

er 2

004a

52D

rug-

free

M

DD

pts

(? ±

?)Co

ntro

ls (?

±?)

M +

F

37 35

[11C]

DA

SP P

ETst

riatu

m, C

erN

o si

gn. d

iffer

ence

s in

str

iatu

m V

3’’ o

f MD

D v

s co

ntro

ls.

12 s

ubje

cts

(pat

ient

s/co

ntro

ls) w

ere

also

ana

lyze

d in

Mey

er e

t al

.29

Patie

nts

wer

e dr

ug-fr

ee fo

r >1m

. No

mat

chin

gfor

age

or

sex.

Rec

ruitm

ent o

f pat

ient

s un

spec

ifie

d.

Mey

er 2

004b

17

Dru

g fr

ee

MD

D p

ts (3

5 ±1

1)Co

ntro

ls (3

5 ±1

1)M

+F

20 20

[11C]

DA

SP P

ET

Bila

tera

l ant

erom

edia

l PF

C, d

orso

late

ral P

FC,

Ant

. Cin

g, C

auda

te,

Puta

m, T

hal,

Mid

Br, C

er

No

diff

eren

ce in

SER

T av

aila

bilit

y be

twee

n M

DD

and

co

ntro

ls fo

r any

regi

on.

With

in p

atie

nts

(but

not

in c

ontr

ols)

sig

n. c

orre

latio

ns

betw

een

incr

ease

d D

ysfu

nctio

nal A

ttitu

de S

cale

(DA

S)

scor

es a

nd in

crea

sed

SER

T av

ail.

Sign

ific

ant i

ncre

ase

in S

ERT

in p

atie

nts

vs. c

ontr

ols

in a

ll re

gion

s fo

r 8 p

atie

nts

with

hig

h D

AS

scor

es (>

190)

vs

cont

rols

.

MD

D p

atie

nts

wer

e/ d

rug

free

for ≥

3m; c

ontr

ols

wer

e in

divi

dual

ly a

ge-m

atch

ed (±

3 ye

ars)

but

not

for s

ex. A

ll pa

rtic

ipan

ts w

ere

non-

smok

ers;

no

info

rmat

ion

on fa

mily

hi

stor

y of

con

trol

s. N

o ef

fect

s of

age

or s

ex in

this

sam

ple.

Re

crui

tmen

t of p

atie

nts

unsp

ecifi

ed.

New

berg

200

59D

rug-

free

M

DD

pts

(38

±?)

Cont

rols

(37

±?)

M +

F

7 6

[123 I]

-AD

AM

SPE

CT, 4

hM

idBr

, Cer

Mid

Br S

ERT

sign

. low

er (-

7%) i

n M

DD

pat

ient

s vs

. con

trol

s.

SER

T M

idBr

cor

rela

ted

sign

ific

antly

with

sco

res

on H

DR

S2

drug

-nai

ve p

atie

nts,

2 >

3w a

nd 2

>6m

dru

g-fr

ee. N

o m

atch

ing

for a

ge o

r sex

. No

info

rmat

ion

on fa

mily

his

tory

of c

ontr

ols.

Re

crui

tmen

t of p

atie

nts

unsp

ecifi

ed.

Pars

ey 2

00610

Dru

g fr

ee

MD

D p

ts (3

8.0

±13.

4)Co

ntro

ls (3

8.8

±15.

9)M

+F

25 43

[11C]

McN

5652

PET

Mid

Br, P

utam

, Am

yg,

Thal

, Hip

poc,

Ant

. Cin

g,

Cer

Sign

. low

er (-

20%)

BP’

in a

ll re

gion

s in

MD

D p

atie

nts

vs

cont

rols

. Pos

t-ho

c as

sign

ed to

am

ygda

la a

nd m

idbr

ain

regi

ons.

Dru

g-na

ive

patie

nts

had

sign

. low

er B

P’ in

am

ygda

la a

nd

mid

brai

n vs

non

dru

g-na

ive

patie

nts.

No

corr

elat

ion

with

BP

and

depr

essi

on s

ever

ity ra

tings

MD

D p

atie

nts

wer

e dr

ug-fr

ee fo

r ≥2w

, 48%

dru

g-na

ive,

56%

in

patie

nts,

72%

fem

ales

. Con

trol

s ha

d no

1st d

egre

e po

sitiv

e fa

mily

his

tory

for p

sych

opat

holo

gyCo

ntro

ls n

ot m

atch

ed fo

r sex

, mat

chin

g fo

r age

unc

lear

, no

sex-

diff

eren

ces

obse

rved

. Rec

ruitm

ent o

f inp

atie

nts

and

outp

atie

nts.

Tabl

e S7

.1. P

revi

ous

stud

ies

mea

surin

g SE

RT a

vaila

bilit

y in

MD

D p

atie

nts

vers

us h

ealth

y co

ntro

ls.

(Con

tinue

d)

Aut

hor &

Dat

e (r

efer

ence

)Po

pula

tion

(mea

n ag

e)Se

xN

SER

T Im

agin

g an

d R

OI*

Res

ults

Rem

arks

153


Chapter7

Rei

vich

200

418D

rug

free

M

DD

pts

(22-

56)

Cont

rols

(23-

59)

M+F

4 4

[11C]

(+)M

cN56

52 P

ETBi

late

ral F

ront

al, C

ing,

Th

al, P

ons,

Cer

Sign

. lar

ger B

P in

left

fron

tal (

+17%

) and

righ

t cin

gula

te

(+24

%) in

pat

ient

s vs

con

trol

s. T

hala

mus

and

pon

s no

sig

n.

diff

eren

ce

MD

D p

atie

nts

wer

e dr

ug fr

ee fo

r ≥5

half-

lives

of t

he d

rug,

≥2w

fo

r MA

O-I

and ≥3

w fo

r flu

oxet

ine.

Pat

ient

s w

ere

not m

atch

ed

for a

ge o

r sex

; no

info

rmat

ion

on fa

mily

his

tory

of c

ontr

ols.

W

ide

age-

rang

e w

ithou

t pos

sibi

lity

to c

orre

ct fo

r age

-eff

ects

on

SER

T av

aila

bilit

y. L

arge

SD

s fo

r tha

lam

us o

r pon

s R

OIs

. Re

crui

tmen

t of o

utpa

tient

s.

Stal

ey 2

00611

Dru

g-fr

ee

MD

D p

ts (3

8.8

±9.7

)Co

ntro

ls (3

8.9

±10)

M+F

32 32

[123 I]β-

CIT

SPEC

T, 2

4h

Brai

nste

m, D

ienc

epha

lon,

St

riatu

m, C

er

Sign

. red

uctio

n (-1

2%) i

n di

ence

phal

ons

V3’’

of M

DD

vs.

co

ntro

ls, b

ut n

ot in

bra

inst

em.

Inte

ract

ion

MD

D b

y se

x on

die

ncep

halo

n V3

’’: w

omen

-22%

in

MD

D, m

en -1

%.

Subj

ects

wer

e (g

roup

??) m

atch

ed fo

r sex

, age

and

sm

okin

g st

atus

. No

info

rmat

ion

on fa

mily

his

tory

of c

ontr

ols.

Re

crui

tmen

t of p

atie

nts

unsp

ecifi

ed.

Will

eit 2

00012

Dru

g-fr

ee p

atie

nts

with

SA

D (3

0.5

±8)

Cont

rols

(29.

0 ±5

.5)

M +

F

11 11

[123 I]β-

CIT

SPEC

T, 4

, 24h

Th

al, H

ypoT

hal,

Mid

Br,

Pons

, Cer

Sign

. red

uctio

n (-

15%)

in T

hala

mus

and

Hyp

otha

lam

us V

3’’ o

f M

DD

vs.

con

trol

s, b

ut n

ot (-

8%) i

n M

idBr

-Pon

s.N

o co

rrel

atio

n w

ith S

ERT

and

depr

essi

on s

ever

ity ra

tings

5 su

bjec

ts h

ad b

een

trea

ted

with

ant

idep

ress

ants

³6m

bef

ore

scan

ning

. Gro

upw

ise

mat

chin

g fo

r age

and

sex

.A

ll di

ffer

ence

s in

SPE

CT a

cqui

sitio

ns 2

4h p

.i., a

t 4h

p.i.

no

sign

ific

ant d

iffer

ence

s fo

und.

Rec

ruitm

ent o

f out

patie

nts

of

univ

ersi

ty c

linic

.

* N

amin

g of

iden

tical

regi

ons

as p

rovi

ded

in th

e st

udie

s. R

efer

ence

regi

on in

ital

ics.

Abb

revi

atio

ns: A

myg

= A

myg

dala

, Ant

. Cin

g =

Ant

erio

r Cin

gula

te, B

D =

Bip

olar

Dis

orde

r, B

P =

Bind

ing

Pote

ntia

l, Ca

ud =

Cau

datu

s, C

er =

cer

ebel

lum

, CIT

= c

italo

pram

, h =

hou

r, H

ippo

c =

Hip

poca

mpu

s,

Hyp

oTha

l = H

ypot

hala

mus

, min

= m

inut

e, m

= m

onth

, MD

D =

Maj

or D

epre

ssiv

e D

isor

der,

Mid

Br =

Mid

brai

n, O

cc =

Occ

ipita

l lob

e, P

AR

= p

arox

etin

e, P

FC =

Pre

fron

tal C

orte

x, P

utam

= P

utam

en, S

AD

=

Seas

onal

Aff

ectiv

e D

isor

der,

Tha

l = T

hala

mus

, V3’

’ = B

P, w

= w

eek

Tabl

e S7

.1. P

revi

ous

stud

ies

mea

surin

g SE

RT a

vaila

bilit

y in

MD

D p

atie

nts

vers

us h

ealth

y co

ntro

ls.

(Con

tinue

d)

Aut

hor &

Dat

e (r

efer

ence

)Po

pula

tion

(mea

n ag

e)Se

xN

SER

T Im

agin

g an

d R

OI*

Res

ults

Rem

arks

NEUROBIOLOGICAL EFFECTS OF PAROXETINE IN THE TREATMENT OF MAJOR DEPRESSIVE DISORDER

PART

Neurobiological effects of paroxetine in the treatment of Major Depressive Disorder

CHA

PTER

SEROTONIN TRANSPORTER GENE PROMOTER

POLYMORPHISMS MODIFY THE ASSOCIATION

BETWEEN PAROXETINE SEROTONIN TRANSPORTER

OCCUPANCY AND CLINICAL RESPONSE IN MAJOR

DEPRESSIVE DISORDER

Pharmacogenetics and Genomics 2008; in press

Henricus G. Ruhé1

Wendy Ooteman1

Jan Booij2

Martin C. Michel3

Martina Moeton4

Frank Baas4

Aart H. Schene1

1 Program for Mood Disorders, Department of Psychiatry, Academic Medical Center, University ofAmsterdam, Amsterdam, the Netherlands2 Department of Nuclear Medicine, Academic Medical Center, University of Amsterdam, Amsterdam,the Netherlands 3Department of Pharmacology and Pharmacotherapy, Academic Medical Center, University ofAmsterdam, Amsterdam, the Netherlands4 Department of Neurogenetics, Academic Medical Center, University of Amsterdam, Amsterdam, theNetherlands

Chapter 8

158

Chap

ter

8

Abstract

BackgroundIn major depressive disorder (MDD), selective serotonin reuptake inhibitors (SSRIs) target theserotonin transporter (SERT). Their 30-50% response-rates are modified by SERT promotorpolymorphisms (5-HTTLPR).

AimTo quantify the relation between SERT occupancy and response, and whether 5-HTTLPR is amodifier.

MethodsDrug-free depressed outpatients (n= 49; both sexes; 25-55 years), received paroxetine 20 mg/day.We quantified SERT occupancy with [123I]β-CIT SPECT imaging at baseline and after 6 weeks; wegenotyped 5-HTTLPR (S, LG, LA). Primary outcomes: percentage decrease in Hamilton DepressionRating Scale (HDRS17) and response (≥50% decrease of HDRS17).

ResultsA significant positive relation between SERT occupancy and clinical response existed only in theLA/LA genotype (p< 0.002). Relative to paroxetine serum concentrations midbrain SERToccupancy was numerically higher for LA/LA compared with other genotypes, but this differencewas not significant (p= 0.188).

ConclusionsHigher SERT occupancy is only associated with more clinical improvement in the LA/LA-genotype.We hypothesize that the LA/LA carriers have a more dynamic serotonergic system, which appearsmore responsive to SSRIs. ISRCTN register ISRCTN44111488

ChapterSERT-occupancy, SSRI response and 5-HTTLPR

8

159

Introduction

Major depressive disorder (MDD) is often treated with antidepressants, most commonly selectiveserotonin reuptake inhibitors (SSRIs). SSRIs occupy the serotonin transporter (SERT) with highselectivity and affinity to block presynaptic serotonin reuptake. Unfortunately, clinical responserates for SSRIs are modest (30-50%) and difficult to predict.1 Unraveling the factors that modifyclinical response may contribute to more effective SSRI treatment.

SERT occupancy can be estimated in-vivo using radioligands and SPECT or PET-imaging.2 SERToccupancy reaches a plateau at relatively low SSRI serum concentrations in healthy subjects,3-6

social phobia patients,7 and MDD patients,4;8-11 explaining why SSRI-serum levels do not predictclinical response.12 Previous studies postulated that 80% SERT occupancy by SSRIs would berequired for clinical response.4;5;8 However, when SERT occupancy-rates and clinical responsewere correlated, the same studies found no clear relationship between striatal SERT occupancyand clinical response after 4 weeks of SSRI treatment in MDD patients.4;8 Contrary, Kugaya et al.found that higher pretreatment SERT availability and greater SERT occupancy in diencephalonpredicted better treatment response after 6 weeks of SSRIs.11 Recently, Zitterl et al. found acorrelation between clomipramine diencephalon SERT occupancy and severity-scores inobsessive compulsive disorder.13 However, in these two studies SERT occupancies rangedbetween 23-61%.

Polymorphisms of the SERT gene promoter region (5-HTTLPR) are associated with itstranscriptional activity and influences the rate of serotonin uptake.14 In human lymphoblasts, cellshomozygous for the long (L) allele produce higher concentrations of SERT mRNA than cellscontaining one or two copies of the short (S) allele. Furthermore, serotonin uptake by thetransporter is >2-fold higher in cells homozygous for the L-allele. In-vivo imaging inconsistentlyshowed lower midbrain SERT availability for S-allele carrying healthy subjects compared with L-allele carriers.15-18 Individuals with an S-allele have increased anxiety related traits,19;20 elevated riskof depression after stress,21;22 increased amygdala reactivity,23;24 and increased adverse eventsafter SSRI treatment,25 compared with subjects with an L-allele. An elegant, large MRI-study inhealthy volunteers showed associations of the 5-HTTLPR S/S polymorphism with unfavorablealterations in anatomy and function of the amygdala-cingulate feedback circuit.26 This studypoints to an important role of the 5-HTTLPR-polymorphism in the development and functioning ofemotional networks involved in MDD. A meta-analysis showed associations between the 5-HTTLPR-polymorphism and clinical SSRI efficacy within 4 weeks of treatment.27 Depressedpatients without S-alleles had higher response rates to SSRIs. Thus, apart from developmental,functional, and stress reactivity effects on emotional networks, the 5-HTTLPR polymorphism alsoappears to influence the response to SSRIs.

Nowadays, the 5-HTTLPR-polymorphism is considered tri-allelic.28 The L-allele is subdivided asLG and LA by a common single nucleotide polymorphism (SNP; rs25531). The LG SNP creates afunctional AP2 transcription factor binding site, behaving like an S-allele. The tri-allelic S:LA:LG ratiois approximately 4:5:1,28 and can be reclassified into a modified bi-allelic genotype.29 In-vivo studiespredominantly found higher SERT availability for LA/LA carriers in healthy and MDD subjects.29-31

Recently, a significant association between the LA/LA allele and citalopram adverse effects32 butnot with response33 was found in MDD patients.

In summary, the relation between SERT occupancy and clinical response to SSRIs is unclear, aSERT polymorphism likely influences the development and function of the serotonergic system(e.g. SERT availability) and affects clinical response to SSRIs. Since it was not investigatedwhether the relation between SERT occupancy and clinical response is modified by the 5-HTTLPR-polymorphism, we aimed to quantify (1) the relation between SERT occupancy and clinicalresponse, (2) pre-treatment SERT availability for different tri-allelic 5-HTTLPR genotypes and (3)the modification by 5-HTTLPR with respect to the relation between SSRI-occupancy and clinicalresponse.

Chap

ter

Chapter 8

8

160

We hypothesized that (1) the relation between occupancy and clinical response is nonlinearwith increased response-rates above 80% occupancy after 6 weeks of treatment, (2) we wouldfind lower pretreatment SERT availability for the tri-allelic S’/S’ 5-HTTLPR genotype and (3)because of the higher response-rates in patients with a L-allele, the association betweenoccupancy and clinical response would be distinct for tri-allelic 5-HTTLPR genotypes. We studiedthese questions in a 6 week open trial of paroxetine 20 mg/day, in subjects who participated in thefirst phase of a randomized dose-escalation trial with [123I]β-CIT SPECT assessment of SERToccupancy.34

Methods

ParticipantsFollowing approval by the institutional medical ethical committee and written informed consent,we recruited drug-free outpatients (25-55 years)35 from primary care, our outpatient department,and public psychiatric settings (October 2003 - August 2006) and included them in the studybefore antidepressants were started. Inclusion criteria were: MDD determined by the StructuredClinical Interview for DSM-IV (SCID),36 and a Hamilton Depression Rating Scale (HDRS17)37 scoreabove 18. Patients were drug-naïve or drug-free (>4 weeks and ≥ 5 half-lives of a previousantidepressant, when treated previously) and had used no more than one antidepressanttreatment (other than paroxetine) at an effective dose for ≥ 6 weeks for the present MDD-episode. Exclusion criteria were pregnancy (or wish), bipolar disorder, psychotic features,neurological cognitive impairments (i.e. dementia), primary anxiety and/or substance abusedisorders and acute, severe suicidal ideation. We allowed secondary co-morbid anxiety and/orsubstance abuse to increase applicability of the findings of our main study.34

SSRI treatmentAfter baseline assessment, patients were treated open-label with paroxetine 20 mg/day for sixweeks. When severe adverse effects occurred, dosages were reduced to 10 mg/day and againincreased to 20 mg/day after one week. We supplied paroxetine in pill-boxes to improvetreatment adherence. We checked adherence by pill-counts and medical history.38

Benzodiazepines (temazepam 10-20 mg/day or oxazepam 10-30 mg/day) were allowed ifnecessary.

Questionnaires and measurementsPrimary clinical outcomes were the percentage decrease in HDRS17-score, and the proportion ofpatients achieving clinical response (≥50% decrease in HDRS17). We administered questionnaires atstudy-entry and after 6 weeks of treatment. In addition, depressive symptoms were monitored atweek 2 and 4 using the Maier and Bech subscales of the HDRS17,39;40 and the self-rated Inventoryfor Depressive Symptomatology (IDS-SR30) scores.41 Three trained investigators whoadministered the clinician-rated questionnaires (HDRS17 and subscales) had good inter-rateragreement (intra-class correlation coefficient = 0.98).

SPECT imaging and analysisWe performed single photon emission computed tomography (SPECT) imaging for in-vivoassessment of SERT availability at study-entry and after 6 weeks between 2 to 10 pm as describedpreviously.42 The acquisition started 230 ±18 (SD) minutes after intravenous injection of 100 MBqiodine-123-labeled 2β-carbomethoxy-3β-(4-iodophenyl)-tropane ([123I]β-CIT), when the radioligandis at equilibrium for SERT binding in brain areas expressing high densities of SERTs (i.e. midbrainand diencephalon).43 To prevent thyroid uptake of [123I], all subjects received oral potassium-


8

161

iodide solution. We performed SPECT imaging using a 12-detector single slice brain-dedicatedscanner (Neurofocus 810, Strichmann Medical Equipment; Cleveland, OH) with a full-width at half-maximum resolution of 6.5 mm, throughout the 20 cm field-of-view (www.neurophysics.com).

After attenuation correction and reconstruction in 3D mode (www.neurophysics.com), wedefined regions of interest (RoIs) for midbrain, diencephalon and cerebellum by using validatedtemplates (see Figure 2.3).34;42 One investigator, blinded for scan session (study-entry/6 weeks),positioned all RoIs in two series. Intra-class correlation coefficients between series were >0.97 forall RoIs. If the two series differed by >5%, scans were re-evaluated by a second investigator. In theanalyses the counts were averaged for the two series.

Using activity in the cerebellum (CER) as indicator of non-displaceable activity (non-specificbinding and free radioactivity),44 we estimated the non-displaceable binding potential (BPND) ofthe radioligand to SERTby calculating the ratio of specific to non-specific binding per scan asBPND = . BPND is proportional to SERT availability under equilibrium conditions.45 In adifferent study, we found high reproducibility of SERT imaging with [123I]β-CIT SPECT afterrepeated scanning of subjects, using the same camera and scanning-protocol (de Win et al.,submitted). We calculated SERT occupancy at 6 weeks relative to the untreated SERT BPND(study-entry): OCC6 weeks = .

Paroxetine serum concentrationsWe collected blood for paroxetine serum concentrations (PSC; therapeutic range 10-75 µg/L) after6 weeks, immediately before SPECT scanning. Serum was stored at -20° C until analysis. PSCs weredetermined using a validated High Pressure Liquid Chromatography-MS/MS method (available onrequest). The lower limit of quantification was 5µg/L, the lower limit of detection was 0.3 µg/L.

Genotyping Procedures and AnalysisGenomic deoxyribonucleic acid (DNA) was isolated out of blood using a filter-based method

(QIAamp DNA Mini Kit, Qiagen Ltd, United Kingdom). The length of 5-HTTLPR28 was determinedby gel electrophoresis. The region around the polymorphism was amplified by PCR using forwardprimer tgtaaaacgacggccagtgccagcacctaacccctaat and reverse primercaggaaacagctatgaccaggggagatcctgggaga (M13 primer sequence in italics). The PCR reaction wasperformed in 10µl 1.5mM MgCl2, 0.2µM forward and reverse primer, 0.1mM dNTP’s, 0.5 UnitsHotfire Polymerase (Solis Biodyne, Estonia), Buffer B (Solis Biodyne, Estonia) and 20ng genomicDNA. The lengths of the different alleles were short ≈ 250 bp and long ≥298 bp. Genotyping of thers25531 SNP was done by sequencing (Sanger) using Big Dye Terminators (Applied Biosystems).The M13 forward primer tgtaaaacgacggccagt was used for sequencing. Reactions were performedcontaining 5ng of a forward primer, 5µl PCR product, BDT mix (Applied Biosystems) and 2.5xBDTbuffer (Applied Biosystems). The length of the 5-HTTLPRpolymorphism was confirmed by lookingat the length of the sequenced PCR product. We reclassified the genotypes as S’/S’ (S/S, LG/S, LG/LG), S’/LA (S/LA, LG/LA) and LA/LA.29 In post-hoc analyses we grouped S’/S’ and S’/LA genotypes tocontrast these with the ‘high-expression’ LA/LA genotype.

Statistical AnalysisFor cross tabulations and differences between groups we used χ2 tests, Fisher’s exact test andANOVA in SPSS for Windows v15.0.1.1 (www.spss.com). To investigate the relation between SERToccupancy and PSC, we modeled SERT occupancy (OCC) after 6 weeks in an Emax model asOCC= , in which a represents maximal SERT occupancy in the model (OCCmax) and b thePSC with 50% SERT occupancy (EC50).4-7;9 We calculated a and b by fitting a nonlinear regressionmodel that minimizes the sum of squares of the residuals (GraphPad Prism v5.00,www.graphpad.com). To assess whether PSC-occupancy or occupancy-response curves wereimproved by sub-grouping (e.g. genetic subgroups or responders versus non-responders), wefitted either one curve, or separate curves and determined whether separate curves decreased

( )CER

CERROI


( )entrystudyND

weeksNDentrystudyND

BPBPBP

−

− − _6

)(*

PSCbPSCa+

Chap

ter

Chapter 8

8

162

the Akaike Information Criterion (AIC; lower is better), which expresses the -2 log-likelihood of the(nested) model penalized for the number of independent variables in the model. We furthermoreverified these analyses by regression models including interaction terms for genotype.

To predict clinical response by SERT occupancy, we performed χ2 and ANOVA tests to estimategroup differences in clinical response-rate and proportional decrease in HDRS17, respectively, withSERT occupancy stratified as poor <50%, intermediate 51-79%, and high ≥80%, and linear andlogistic regression analyses for continuous outcomes (proportional decrease in HDRS17) anddichotomous outcomes (response). In these models occupancy was entered as a continuous anddummy-coded group variable (stratified SERT occupancy), while the models were corrected forage, baseline midbrain /diencephalon SERT availability and sex.

One responder and one non-responder were potentially non-adherent after 6 weeks (PSC<5µg/L), but were included in the analyses for three reasons. First, this estimation of medicationadherence was only based on a single measurement of PSC over a 6-week treatment period,second, patients had empty pill-boxes and had claimed adherence, and third, potential non-adherence in other patients before the measurement of PSC could not be excluded either.

Table 8.1. Characteristics of patients, stratified by clinical response after 6 weeks of paroxetine 20 mg/day.

Responders*(n= 10) Non-responders*(n= 32) p†

Age at baseline (years) 43.5 ±10.8 41.5 ±7.6 0.51

Female sex - n (%) 6 (60.0) 21 (65.6) 1.00

Current smoking - n (%) Alcohol use: n (%) ≤ 7 Units/week >7 Units/ week

5 (50.0)

7 (70.0) 3 (30.0)

17 (53.1)

30 (93.8) 2 (6.2)

1.000.08

MDD HDRS17 at baseline IDS-SR30 baseline First episode - n (%) No of episodes Drug-naïve - n (%) Used AD in current episode - n (%) Melancholic - n (%) Duration of episode: n (%) <5 months duration 5 months – 2 years duration ≥ 2 years Age of first episode (years)

23.8 ±3.7 41.0 ±7.2 4 (40.0) 2.0 ±1.1 7 (70.0) 0 (0.0) 7 (87.5)

3 (30.0) 6 (60.0) 1 (10.0) 33.3 ±11.2

25.0 ±4.7 44.7 ±9.118 (58.1)1.9 ±1.8 22 (68.8) 3 (9.4) 24 (96.0)

10 (31.3)18 (56.3) 4 (12.5) 36.4 ±9.6

0.470.280.480.921.001.000.430.97

0.40

Co-morbidity - n (%) Anxiety disorder Dysthymia Drug abuse / dependence‡

Alcohol abuse / dependence

3 (30.0) 0 (0.0) 3 (30.0) 2 (20.0)

3 (9.4) 1 (3.1)1 (3.1)1 (3.1)

0.141.000.040.14

SERT genotype S’/S’ (S/S, S/LG, LG/LG) S’/L’ (S/LA, LG/LA) L’/L’ (LA/ LA)

4 (40.0) 5 (50.0) 1 (10.0)

8 (25.0)17 (53.1)7 (21.9)

0.56

SERT availability Bsl-scan (BPND) Midbrain Diencephalon

0.60 (0.24) 1.16 (0.29)

0.61 (0.18) 1.15 (0.23 )

0.930.89

* Responders defined as patients with ≥50% decrease in baseline HDRS17-score† p values for χ2 for categorical, Fisher’s exact test for dichotomous, and independent T-tests for continuous data‡ cannabis, benzodiazepines


8

163

Results

PatientsFrom 177 patients assessed, 53 patients were excluded and 73 refused participation. Of the 51patients included in the study, 7 patients dropped out due to paroxetine adverse effects (n= 5) orrefusal of a second scan (n= 2), leaving 44 patients who were scanned at baseline and after sixweeks of treatment. Of these 44 patients, 12 patients were treatment-responder after 6 weeks. Ofresponders, two SPECT scans could not be used for analysis due to technical reasons, leaving 42patients for the final analysis. Benzodiazepines were prescribed in 11/42 (26%) patients.

At study-entry, no significant differences were found between responders and non-responders except fordrug (cannabis/benzodiazepine) abuse or dependence with higherprevalence in responders compared with non-responders (Fisher’s Exact; p= 0.036; Table 8.1).None of the patients reported lifetime use of 3,4-methylenedioxy methamphe tamine (MDMA).

SERT occupancy and clinical response When SERT occupancy was stratified as poor (<50%), intermediate (51-79%) and high (≥80%), wefound no significant relation between SERT occupancy and response-rate or the percentagedecrease in HDRS17 after 6 weeks of paroxetine treatment, neither in midbrain nor indiencephalon (p>0.05) (Table 8.2). Neither in linear nor in logistic regression models occupancysignificantly predicted clinical response (p>0.05). The minimum occupancy rate at which responseoccurred in at least one patient was 29% in midbrain and 49% in diencephalon. We found noincrease in response-rate for SERT occupancy >80%. Only midbrain SERT-occupancy wasassociated with PSC (p=0.02). PSC significantly predicted response (OR= 1.04 (95% CI= 1.00-1.09).After exclusion of 2 potential non-adherent patients, the association of PSC with responsebecame non-significant, other results did not change.

Modification of the relation between PSC, occupancy and response bygenotype

PSC and SERT occupancy PSC-occupancy curves in both the midbrain and diencephalon were curvilinear, with significantlydifferent (non-stratified) curves for midbrain and diencephalon (determined by a decrease in AIC;figure available on request). The curves for responders versus non-responders were notsignificantly different in both brain regions. Stratification of the PSC-occupancy curves by SERTgenotype indicated numerically higher OCCmax and lower EC50 in the LA/LA genotype (n= 8/42;19%) compared with the S’/S’ (n= 12/42; 29%) and S’/LA genotypes (n= 22/42; 52%), especially inmidbrain (OCCmax for LA/LA=99.6; p=0.188) but not in diencephalon (OCCmax for LA/LA=79.3; p=1.0;

Table 8.2. SERT occupancy by paroxetine and clinical response after 6 weeks of paroxetine 20 mg/day.

Occupancy after 6 weeks paroxetine

<50% 51-79.9% ≥80% p*

Diencephalon n= 9 25 8

Mean occupancy (%) Mean PSC (µg/L) Responders n (%) % decrease HDRS17

†

28.9 ±7.43 22.2 ±10.823 (33.3) 23.9 ±10.72

65.4 ±1.72 46.4 ±7.25 4 (16.0) 27.5 ±4.96

85.0 ±1.29 46.0 ±8.83 3 (37.5) 37.7 ±8.38

0.176 0.347 0.530

Midbrain n= 9 12 18

Mean occupancy (%) Mean PSC (µg/L) Responders n (%) % decrease HDRS17

†

30.4 ±5.40 18.2 ±6.27 4 (44.4) 30.2 ±10.89

68.6 ±2.46 38.5 ±7.90 2 (16.7) 26.1 ±6.88

93.2 ±2.18 56.3 ±9.18 3 (16.7) 30.6 ±4.29

0.020 0.149 0.896

Values represent means (±SEM) unless indicated different. N= 42, including 2 non-adherent patients. PSC = Paroxetine serum concentration* χ2 for linear trend for response rates; ANOVA for % decrease† after 6 weeks of treatment

Chap

ter

Chapter 8

8

164

Figures 8.1A and B). The difference in AIC between one or separate curves indicated no significantimprovement by including genotype in the models, both in midbrain (AIC increase 2.359) anddiencephalon (AIC increase 4.037).

SERT availability and SERT occupancy by genotypeMean pretreatment SERT availability and mean SERT occupancies after 6 weeks in midbrain ordiencephalon did not significantly differ between genotypes (Table 8.3). Six of 8 patients (75%)with the LA/LA genotype reached midbrain occupancies ≥ 80% after 6 weeks compared with 12 of31 patients (39%) with other genotypes (Fisher’s exact; p= 0.112).

After exclusion of 2 potentially non-adherent patients, mean SERT occupancies in midbrainwere significantly different between genotypes (S’/S’ 80.9% ±5.80, S’/LA 65.8% ±5.17, LA/LA 91.6%±6.07; p= 0.017). Moreover, 6 of 7 patients (86%) with the LA/LA genotype reached midbrainoccupancies ≥80% after 6 weekscompared with 12 of 30 patients (40%) with the other genotypes(Fisher’s exact; p= 0.042).

Figure 8.1. Paroxetine serum concentration and SERT occupancy by paroxetine, stratified by SERT genotype.

Panels represent PSC-occupancy curves for S’/S’ vs. S’/LAvs. LA/LA in midbrain (A) and diencephalon (B). Equotation fitted: OCC 6 weeks = . Differences between curves were not significant (lower AIC for 1 fitted curve vs. three fitted curves)PSC= Paroxetine Serum Concentration, OCC6 weeks = SERT occupancy after 6 weeks of paroxetine treatment.

Table 8.3. Ethnicity, clinical response and SERT occupancy by paroxetine stratified for SERT genotype.

SERT genotype

S/S, S/LG, LG/LG S/LA, LG/LA LA/ LA p

Ethnicity - n (%) Caucasian Creole Asian

7 (24.1) 3 (42.9) 2 (33.3)

15 (51.7) 3 (42.9) 4 (66.7)

7 (24.1) 1 (14.3) 0

0.605

Clinical n= 12 n= 22 n= 8

Week 6 % decrease HDRS17 39.4±8.18 22.2±5.34 30.4±7.97 0.183

Occupancy Diencephalon n= 12 n= 22 n= 8

Baseline BPND Mean occupancy (%) Occupancy ≥80%n (%)

101.7±3.94 58.3±6.94 2 (16.7)

116.2±6.0162.7±4.25 4 (18.2)

122.9±8.8462.0±9.76 2 (25.0)

0.353 0.857 0.888

Occupancy Midbrain n= 11 n= 20 n= 8

Baseline BPND Mean occupancy (%) Occupancy ≥80%n (%)

60.1±5.20 73.6±9.04 5 (45.5)

61.9±5.0365.8±5.17 7 (35.0)

57.8±4.52 81.3±11.56 6 (75.0)

0.879 0.381 0.159*

Continuous values represent means (±SEM). N= 42, including 2 non-adherent patients.* Post-hoc LA/LAvs.S/S, S/LG, LG/LG, S/LA, LG/LA Fisher’s exact p= 0.112

��

� �

�

�

�

�

��

��

��

��

��(��

�� )

��!�

"#$%&�#�

� �

�

�

�

�

��

��

��

��

��

��(��

�� )

��!�

� %

)(*

PSCbPSCa+


8

165

SERT occupancy and clinical response by genotype Despite no significant differences in mean proportional HDRS17-decrease across genotypes (Table8.3), we found a significant genotype*SERT occupancy interaction in diencephalon and midbrain(Figure 8.2). Higher diencephalon SERT occupancy was associated with larger proportionalHDRS17-decreases in LA/LA carriers (ANOVA; F2,5= 27.64; p= 0.002) (Figure 8.2B; corrected for age),and LA/LA and LA/S’ carriers grouped together (ANOVA; F1,28= 8.380; p= 0.007) (Figure 8.2D). Thegenotype*SERT occupancy interaction in diencephalon was significant (ANOVA; F2,36= 3.273; p=0.049, with correction for age) (Figure 2B). Higher midbrain SERT occupancy in LA/LA carriers wasassociated with larger proportional HDRS17-decreases when corrected for age (Figure 8.2C;ANOVA; F2,5= 56.918; p< 0.001). Introduction of baseline SERT availability in diencephalon (ormidbrain) showed no significant improvement of the association between proportional HDRS17decrease and SERT occupancy (p>0.05). These results remained unchanged with absolutedecrease in HDRS17-scores instead of proportional HDRS17-decreases. Performing the sameanalyses with the bi-allelic 5-HTTLPR classification found identical directions of the effects, butexplained less of the variances (available on request).

Figure 8.2. SERT occupancy by paroxetine and proportional decrease in HDRS17, stratified by SERT genotype.

Panels A and C: Midbrain, panels B and D: Diencephalon. In panel A and C (midbrain) a significant relation between SERT occupancy and % decrease in HDRS17 (determined by theslopes of regression lines) existed for the LA/LA genotype when confounding by age was accounted for (F2,5= 56.92; p<0.001).In panels B (diencephalon) this relation was significant for the LA/LA genotype (F1,6= 15.36; p= 0.008), and confounded by age(F2,5= 27.64; p= 0.002); for panel D this relation was significant for the S/LA, LG/LA, LA/LA genotypes (F1,28= 8.380; p= 0.007).In panel A and C there was no significant interaction between the different genotype groups. In panel B and D a significantinteraction existed between the genotype groups (F2,36=3.273; p= 0.049 and F1,38= 6.293; p= 0.017 respectively).

��

��

�

��

�

��

�

�

��

��

��

��

��

��'��

��

��

��

�'��

��

��

��

��

�

��

�

��

�

�

��

��

��

��

��

��'��

��

��

��

�'��

��

��

� !"#!$��"

��

�

��

�

��

�

�

��

��

��

��

��

��

��'��

��

��

��

�'��

��

��

% �&�� "

��

�

��

�

��

�

�

��

��

��

��

��

��

��'��

��

��

��

�'��

��

��

� &

# �

Chap

ter

Chapter 8

8

166

Discussion

We quantified the relation between SERT occupancy in midbrain and diencephalon and clinicalresponse after 6 weeks of paroxetine 20 mg/day. We found no significant relation between theseoccupancies and clinical response. Nor did we find significant differences in pretreatment SERTavailability in patients with different 5-HTTLPR-genotypes. We found a significant modifying effectof the 5-HTTLPR genotype with respect to the association between SERT occupancy and thedecrease in HDRS17-score (proportional and absolute). In diencephalon, increased SERToccupancy was associated with larger decreases in HDRS17 for LA/LA and S’/LA carriers. Inmidbrain, increased SERT occupancy was associated with larger proportional decreases in HDRS17for LA/LA carriers only. Age was a covariate in these associations. Unexpectedly, we found a trendof a modifying effect of the 5-HTTLPR genotype with respect to the association betweenparoxetine serum concentrations and occupancy.

Critique of methodsSome limitations might be addressed first. We observed a low 6 week response rate (24%), whichmay be attributable to 1) the inclusion of some patients with secondary co-morbid anxiety and/orsubstance abuse, who might show less and slower recovery,46 and 2) the relatively highpretreatment initial HDRS-scores. Nevertheless, the recent Sequential Treatment Alternatives torelieve Depression trial measured only 30% response rate after 6 weeks of citalopram as well.1

Second, we included patients who were ethnically heterogeneous (Table 8.3). This raises thepossibility of spurious results secondary to population stratification, which can be a problemwhen analyzing non-functional markers which are in linkage disequilibrium with the functionalvariants. Since we have analyzed the relation between a functional DNA sequence variant andtransporter occupancy we do not expect effects by the heterogeneity of our sample.

Third, as recruitment of MDD patients willing to undergo two SPECT scans is difficult, wecould only successfully scan 42 patients twice. This led to small genotype-subgroups, andreplication in larger samples will be necessary. Nevertheless, these 42 patients scanned twicewhile treated with the same antidepressant (and dose) is the largest patient sample to date, andallowed us to address the relation between SERT occupancy and clinical response, and itsmodification by the 5-HTTLPR-genotype for the first time. Furthermore, we had low dropout ratesand good adherence.

Fourth, in this open label study we did not control for placebo response (estimated to be 30%).Placebo-responses may potentially obscure the relation between occupancy and response.Nevertheless, our major finding of a significant relation between SERT occupancy andproportional response in LA/LA-carriers was found despite such potential placebo-effects.Additionally, a placebo comparator would have estimated test-retest variability in SERTavailability, but this variability is known to be much lower than the observed 60-70% SERToccupancy.

Fifth, we used [123I]β-CIT for SPECT imaging, which is a non-selective radioligand, and also bindsto dopamine transporters (DAT; e.g. midbrain substantia nigra).47 Nevertheless, uptake inmidbrain and diencephalon is considered to reflect predominantly SERT,48 as these structures arerich of SERT relative to DAT. Therefore, although some additional DAT-binding might haveconcealed SERT occupancy, we think the change in [123I]β-CIT-binding in diencephalon andmidbrain mainly reflects SERT occupancy. Nevertheless, it would be challenging to replicate ourstudy with selective radioligands like [11C]DASB or [123I]ADAM.

Finally, we did not further explore whether 5-HTTLPR (or other polymorphisms) alsoinfluenced transporter site kinetics. Rausch et al. showed that initial serotonin kinetic conditionsof SERT in platelets could predict treatment response in MDD patients, which was modified bySERT polymorphism.49 Patients with an L-allele were more responsive to fluoxetine than patientswith an S-allele, with the initial affinity constant Km and dose being significant covariates fortreatment outcome. It is unlikely that Km is directly affected by the SERT promoter polymorphismthat primarily regulates expression. Because the S and L variants actually represent 10 subtypes,50


8

167

other subtypes or an unknown other polymorphism could very well be associated withdifferences in Km, which may better explain the inter-individual variation between PSC and SERToccupancy.49 Recently, Smeraldi et al.51 investigated the relation between clinical response tofluvoxamine and different L and S subtypes (identified by Nakamura et al.50). They confirmedbetter responses in depressed patients bearing the L-allele, but also found significant differencesin response among L-allele carriers according to the subtype of the L-allele, accounting for 0.6% ofthe variance. Because these subtypes are relatively uncommon, it would be very interesting tospecifically study these L-allele subtypes in future neuroimaging studies using a case-controldesign.

SERT occupancy and clinical responsePrevious studies postulated that a SERT occupancy of at least 80% would be associated withclinical response to SSRIs.4;5;8 Our data did not show significant associations betweenpretreatment SERT availability or SERT occupancy and clinical response to paroxetine, and neithersuggested an 80% SERT occupancy threshold with increased response rates. We could notreplicate a relation between SERT occupancy and decrease of symptoms.11;13 Receiver operatingcharacteristic curves relating SERT occupancy to response (available on request) showed thatSERT occupancy had no diagnostic value in the prediction of response.

The relation between SERT occupancy and clinical response is modified by 5-HTTLPR genotypeWe found that 5-HTTLPR-polymorphisms modify the relation between SERT occupancy andclinical response. Previous clinical studies and meta-analyses found superior treatment effects inL-allele carriers. Despite a low response rate in our sample, we found that in LA/LA carriers higherSERT occupancy was associated with more improvement. We put forward two possibleexplanations for this finding.

First, the LA/LA genotype is associated with a more than two-fold higher serotonin uptake inhuman lymphoblasts.14 If we assume that serotonergic cells maintain a certain tonus inneurotransmission, in LA/LA carriers more serotonin must be released as this is more effectivelyevacuated. When SERTs are blocked with paroxetine, this may increase serotoninneurotransmission more in LA/LA carriers than in other genotypes, which may result in largerpostsynaptic effects.

Second, 5-HTTLPR also modifies the development and synaptic plasticity of neural networkscritically involved in MDD.14 Our findings might thus point to differences in postsynaptic effects(evoked by increased serotonin neurotransmission) in neuronal networks that have beendeveloped differently as a result of different SERT genotypes. Pezawas et al. showed that healthysubjects with an S/S polymorphism demonstrated relative uncoupling of the amygdala from theanterior perigenual cingulate.26 S/S carriers were also associated with increased anxiety traits,increased amygdala reactivity,23 decreased mood after tryptophan depletion52 and increased riskfor MDD.22 In summary, the limbic-cortical network and the serotonergic innervations appear tobe more flexible in non S/S carriers. Our results may be supportive of the hypothesis that in LA/LAcarriers the significant association between higher SERT occupancy and increased reduction ofsymptoms could be indicative for a broader range for regulation of the serotonergic system.Higher SERT occupancy might then result in more effects of serotonergic antidepressants in LA/LAcarriers.

Interestingly, Pollock et al. found quicker response in L/L carriers when treated withparoxetine for 12 weeks, but no difference when treated with the noradrenergic antidepressantnortriptyline.53 Additionally better treatment outcomes and fewer adverse events in S/Sgenotypes were found when treated with mirtazapine (a postsynaptic 5-HT2A, 5-HT2Cantagonistand noradrenergic agonist),54;55 which were contrary to the effects of paroxetine.55 Futurepharmacogenomic studies should investigate whether non S’/S’ carriers will benefit more fromserotonergic drugs while S’/S’ carriers may benefit from noradrenergic antidepressants.Furthermore, since the mechanisms of action of antidepressants may be influenced bypolymorphisms of other genes, future research should investigate additive effects of multiplecandidate gene polymorphism combinations.56

Chap

ter

Chapter 8

8

168

Are paroxetine serum concentrations and SERT occupancy modified by 5-HTTLPR genotype?Six of 8 patients (75%) with the LA/LA genotypereached midbrain occupancies ≥80% after 6weekscompared with 12 of 31 patients (39%) with other genotypes, although this was statisticallynot significant (p= 0.112, Fisher’s exact). Which factor determines the maximum SERT occupancyremains unclear. 5-HTTLPR-polymorphisms primarily affect gene transcription, and as aconsequence affect SERT expression. Therefore, we a priori did not expect differences in SERToccupancy for different genotypes, as occupancy is expected to be independent of availableSERTs. As such, our results could represent an epiphenomenon: e.g. a different SNP in the SERTgene, in linkage disequilibrium with the 5-HTTLPR-polymorphism (e.g. rs2228673;www.ncbi.nlm.nih.gov/SNP/snp_ref.cgi?rs=2228673) may have affected SERT occupancy.Contrary, the suggested modifying properties of the 5-HTTLPR-genotype for the relation betweenPSC and occupancy may also lack power, since the PSC-occupancy curves in diencephalon show aclear trend of modification by 5-HTTLPR. Therefore, this needs further exploration in largersamples.

A rationale for a modifying role of the 5-HTTLPR polymorphism in the occupancy of the SERTby paroxetine and clinical response could be postulated as follows. SERT imaging assesses the in-vivo binding to SERTs. After 6 weeks of treatment with paroxetine, the availability of SERTs tobind to the radiotracer is lower, presumably due to SERT blockade by the SSRI. However, it is notpossible to assess whether in addition to blockade, secondary effects occur such as down-regulation of the SERT. This might also result in lowering of the availability of SERTs to bind withthe radiotracer. Indeed, in rats down-regulation of SERTs after prolonged exposure to paroxetinehas been reported.57;58 Said differently, we are unable to discriminate between direct (i.e.,blockade of SERTs by paroxetine) and potential indirect pharmacological effects (i.e., down-regulation). Importantly, Benmansour et al., suggested that SSRI-induced down-regulation of theSERT may be a key component for the clinical response to SSRIs.57;58 Down-regulation of SERTs inrats was not caused by decreased gene transcription, but is presumably caused by increased(posttranslational) internalization of SERTs.57 Although speculative, the level of down-regulationof SERTs might be diverse for different 5-HTTLPR-genotypes, with faster or greater down-regulation in LA/LA versus other genotypes. This hypothesis could be studied in future imagingstudies with antidepressant-exposed primates (having 5-HTTLPR-polymorphisms) in which abruptdiscontinuation is possible.

ConclusionIn conclusion, we find that SERT genotype modifies the relation between SERT occupancy andclinical response, with more improvement at higher midbrain and diencephalon SERT occupancyin LA/LA carriers. Although this SERT promoter-polymorphism presumably does not influenceSERT occupancy at given PSCs, our data may point to a more flexible serotonergic system in LA/LAcarriers, which is more easily influenced by serotonergic antidepressants. This hypothesis couldserve as a starting point for future pharmacogenetic studies. This might provide the necessarydatato guide the choice of antidepressant treatment for individual patients, reducing thepatient's suffering, and lowering healthcare costs.

Acknowledgements

The authors wish to thank the patients in this study for their participation, and willingness toparticipate in the SPECT-study. We thank all participating general practitioners in the area ofAmsterdam Oost and Zuidoost, Hoofddorp, Nieuw Vennep, and Abcoude for their inclusions andreferrals for the study. Mrs. E. Miedema M.D. and Mrs. M.C. ten Doesschate M.D. wereindispensable for their help in rating questionnaires. This study was financed by a grant from theNetherlands Organisation for Health Research and Development (ZonMw), program MentalHealth, education of investigators in mental health to H.G. Ruhé (OOG; #100-002-002).


8

169


None

References



3. Erlandsson K, Sivananthan T, Lui D, Spezzi A, Townsend CE, Mu S et al. Measuring SSRI occupancy of SERT using the noveltracer [123I]ADAM: a SPECT validation study. Eur J Nucl Med Mol Imaging. 2005; 32: 1329-1336.


5. Suhara T, Takano A, Sudo Y, Ichimiya T, Inoue M, Yasuno F et al. High levels of serotonin transporter occupancy with low-dose clomipramine in comparative occupancy study with fluvoxamine using positron emission tomography. Arch GenPsychiatry. 2003; 60: 386-391.

6. Takano A, Suzuki K, Kosaka J, Ota M, Nozaki S, Ikoma Y et al. A dose-finding study of duloxetine based on serotonintransporter occupancy. Psychopharmacology (Berl). 2006; 185: 395-399.

7. Kent JM, Coplan JD, Lombardo I, Hwang DR, Huang Y, Mawlawi O et al. Occupancy of brain serotonin transporters duringtreatment with paroxetine in patients with social phobia: a positron emission tomography study with [11C]McN 5652.Psychopharmacology (Berl). 2002; 164: 341-348.



10. Kugaya A, Seneca NM, Snyder PJ, Williams SA, Malison RT, Baldwin RM et al. Changes in human in vivo serotonin anddopamine transporter availabilities during chronic antidepressant administration. Neuropsychopharmacology. 2003; 28:413-420.

11. Kugaya A, Sanacora G, Staley JK, Malison RT, Bozkurt A, Khan S et al. Brain serotonin transporter availability predictstreatment response to selective serotonin reuptake inhibitors. Biol Psychiatry. 2004; 56: 497-502.

12. Mitchell PB. Therapeutic drug monitoring of non-tricyclic antidepressant drugs. Clin Chem Lab Med. 2004; 42: 1212-1218.13. Zitterl W, Aigner M, Stompe T, Zitterl-Eglseer K, Gutierrez-Lobos K, Wenzel T et al. Changes in Thalamus-Hypothalamus

serotonin transporter availability during clomipramine administration in patients with Obsessive-Compulsive Disorder.Neuropsychopharmacology. 2008; E-publication doi:10.1038/npp.2008.35.


15. Heinz A, Jones DW, Mazzanti C, Goldman D, Ragan P, Hommer D et al. A relationship between serotonin transportergenotype and in vivo protein expression and alcohol neurotoxicity. Biol Psychiatry. 2000; 47: 643-649.

16. Shioe K, Ichimiya T, Suhara T, Takano A, Sudo Y, Yasuno F et al. No association between genotype of the promoter regionof serotonin transporter gene and serotonin transporter binding in human brain measured by PET. Synapse. 2003; 48:184-188.

17. Van Dyck CH, Malison RT, Staley JK, Jacobsen LK, Seibyl JP, Laruelle M et al. Central serotonin transporter availabilitymeasured with [123I]β-CIT SPECT in relation to serotonin transporter genotype. Am J Psychiatry. 2004; 161: 525-531.

18. Willeit M, Stastny J, Pirker W, Praschak-Rieder N, Neumeister A, Asenbaum S et al. No evidence for in vivo regulation ofmidbrain serotonin transporter availability by serotonin transporter promoter gene polymorphism. Biol Psychiatry. 2001;50: 8-12.

19. Schinka JA, Busch RM, Robichaux-Keene N. A meta-analysis of the association between the serotonin transporter genepolymorphism (5-HTTLPR) and trait anxiety. Mol Psychiatry. 2004; 9: 197-202.

20. Sen S, Burmeister M, Ghosh D. Meta-analysis of the association between a serotonin transporter promoterpolymorphism (5-HTTLPR) and anxiety-related personality traits. Am J Med Genet B Neuropsychiatr Genet. 2004; 127: 85-89.

21. Caspi A, Sugden K, Moffitt TE, Taylor A, Craig IW, Harrington H et al. Influence of life stress on depression: moderation bya polymorphism in the 5-HTT gene. Science. 2003; 301: 386-389.

22. Lotrich FE, Pollock BG. Meta-analysis of serotonin transporter polymorphisms and affective disorders. Psychiatr Genet.2004; 14: 121-129.

23. Hariri AR, Mattay VS, Tessitore A, Kolachana B, Fera F, Goldman D et al. Serotonin transporter genetic variation and theresponse of the human amygdala. Science. 2002; 297: 400-403.

24. Hariri AR, Drabant EM, Munoz KE, Kolachana BS, Mattay VS, Egan MF et al. A susceptibility gene for affective disordersand the response of the human amygdala. Arch Gen Psychiatry. 2005; 62: 146-152.

Chap

ter

Chapter 8

8

170





29. Parsey RV, Hastings RS, Oquendo MA, Hu X, Goldman D, Huang YY et al. Effect of a triallelic functional polymorphism ofthe serotonin-transporter-linked promoter region on expression of serotonin transporter in the human brain. Am JPsychiatry. 2006; 163: 48-51.

30. Praschak-Rieder N, Kennedy J, Wilson AA, Hussey D, Boovariwala A, Willeit M et al. Novel 5-HTTLPR allele associates withhigher serotonin transporter binding in putamen: a [11C] DASB positron emission tomography study. Biol Psychiatry. 2007;62: 327-331.

31. Reimold M, Smolka MN, Schumann G, Zimmer A, Wrase J, Mann K et al. Midbrain serotonin transporter binding potentialmeasured with [11C]DASB is affected by serotonin transporter genotype. J Neural Transm. 2007; 114: 635-639.


33. Kraft JB, Peters EJ, Slager SL, Jenkins GD, Reinalda MS, McGrath PJ et al. Analysis of association between the serotonintransporter and antidepressant response in a large clinical sample. Biol Psychiatry. 2007; 61: 734-742.

34. Ruhe HG, Booij J, Weert HC van, Reitsma JB, Fransen EJF, Michel MC et al. Evidence why dose-escalation of paroxetine inmajor depressive disorder is not effective: A 6-week, randomized-controlled trial with assessment of serotonintransporter occupancy. Neuropsychopharmacology. 2008. [In press].

35. van Dyck CH, Malison RT, Seibyl JP, Laruelle M, Klumpp H, Zoghbi SS et al. Age-related decline in central serotonintransporter availability with [123I]β -CIT SPECT. Neurobiol Aging. 2000; 21: 497-501.


37. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 23: 56-61.38. Saunders K, Simon G, Bush T, Grothaus L. Assessing the feasibility of using computerized pharmacy refill data to monitor



40. Ruhe HG, Dekker JJ, Peen J, Holman R, de Jonghe F. Clinical use of the Hamilton Depression Rating Scale: is increasedefficiency possible? A post hoc comparison of Hamilton Depression Rating Scale, Maier and Bech subscales, ClinicalGlobal Impression, and Symptom Checklist-90 scores. Compr Psychiatry. 2005; 46: 417-427.

41. Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH. The Inventory of Depressive Symptomatology (IDS):psychometric properties. Psychol Med. 1996; 26: 477-486.

42. de Win MM, Habraken JB, Reneman L, van den Brink W, den Heeten GJ, Booij J. Validation of [123I]β-CIT SPECT to assessserotonin transporters in vivo in humans: a double-blind, placebo-controlled, crossover study with the selective serotoninreuptake inhibitor citalopram. Neuropsychopharmacology. 2005; 30: 996-1005.




46. Fava M, Rush AJ, Alpert JE, Balasubramani GK, Wisniewski SR, Carmin CN et al. Difference in Treatment Outcome inOutpatients With Anxious Versus Nonanxious Depression: A STAR*D Report. Am J Psychiatry. 2008; 165: 342-351.



49. Rausch JL, Johnson ME, Fei YJ, Li JQ, Shendarkar N, Hobby HM et al. Initial conditions of serotonin transporter kineticsand genotype: influence on SSRI treatment trial outcome. Biol Psychiatry. 2002; 51: 723-732.

50. Nakamura M, Ueno S, Sano A, Tanabe H. The human serotonin transporter gene linked polymorphism (5-HTTLPR) showsten novel allelic variants. Mol Psychiatry. 2000; 5: 32-38.

51. Smeraldi E, Serretti A, Artioli P, Lorenzi C, Catalano M. Serotonin transporter gene-linked polymorphic region: possiblepharmacogenetic implications of rare variants. Psychiatr Genet. 2006; 16: 153-158.

52. Neumeister A, Hu XZ, Luckenbaugh DA, Schwarz M, Nugent AC, Bonne O et al. Differential effects of 5-HTTLPR genotypeson the behavioral and neural responses to tryptophan depletion in patients with major depression and controls. Arch GenPsychiatry. 2006; 63: 978-986.


8

171

53. Pollock BG, Ferrell RE, Mulsant BH, Mazumdar S, Miller M, Sweet RA et al. Allelic variation in the serotonin transporterpromoter affects onset of paroxetine treatment response in late-life depression. Neuropsychopharmacology. 2000; 23:587-590.

54. Kang RH, Wong ML, Choi MJ, Paik JW, Lee MS. Association study of the serotonin transporter promoter polymorphismand mirtazapine antidepressant response in major depressive disorder. Prog Neuropsychopharmacol Biol Psychiatry. 2007;31: 1317-1321.

55. Murphy GM, Jr., Hollander SB, Rodrigues HE, Kremer C, Schatzberg AF. Effects of the serotonin transporter genepromoter polymorphism on mirtazapine and paroxetine efficacy and adverse events in geriatric major depression. ArchGen Psychiatry. 2004; 61: 1163-1169.

56. Serretti A, Benedetti F, Zanardi R, Smeraldi E. The influence of Serotonin Transporter Promoter Polymorphism (SERTPR)and other polymorphisms of the serotonin pathway on the efficacy of antidepressant treatments. ProgNeuropsychopharmacol Biol Psychiatry. 2005; 29: 1074-1084.

57. Benmansour S, Cecchi M, Morilak DA, Gerhardt GA, Javors MA, Gould GG et al. Effects of chronic antidepressanttreatments on serotonin transporter function, density, and mRNA level. J Neurosci. 1999; 19: 10494-10501.


Chap

ter

Chapter 8

8

172

CHA

PTER

EVIDENCE WHY PAROXETINE DOSE-ESCALATION

IS NOT EFFECTIVE IN MAJOR DEPRESSIVE

DISORDER: A RANDOMIZED-CONTROLLED TRIAL

WITH ASSESSMENT OF SEROTONIN TRANSPORTER

OCCUPANCY

Neuropsychopharmacology 2008; in press

Henricus G. Ruhé1

Jan Booij2

Henk C. v Weert3

Johannes B. Reitsma4

Eric J.F. Fransen5

Martin C. Michel6

Aart H. Schene1

1 Program for Mood Disorders, Department of Psychiatry, Academic Medical Center, University ofAmsterdam, Amsterdam, the Netherlands2 Department of Nuclear Medicine, Academic Medical Center, University of Amsterdam, Amsterdam,the Netherlands 3 Department of General Practice, Academic Medical Center, University of Amsterdam, Amsterdam,the Netherlands 4 Department of Clinical Epidemiology, Biostatistics & Bioinformatics, Academic Medical Center,University of Amsterdam, Amsterdam, the Netherlands 5 Department of Pharmacy and Clinical Pharmacology, Onze Lieve Vrouwe Gasthuis, Amsterdam, theNetherlands6 Department of Pharmacology and Pharmacotherapy, Academic Medical Center, University ofAmsterdam, Amsterdam, the Netherlands

Chap

ter

Chapter 9

9

174

Abstract

BackgroundDose-escalation is often used in depressed patients who fail to respond to standard doses ofSSRIs, but clinical efficacy is equivocal.

AimTo reassess the efficacy of paroxetine dose-escalation and quantify whether paroxetine dose-escalation increases occupancy of the serotonin transporter (SERT) more than placebo dose-escalation, in a randomized controlled trial.

MethodsWe recruited 107 non-psychotic, unipolar depressed outpatients (18-70 yrs; Hamilton DepressionRating Scale (HDRS17) >18) from primary care and psychiatric outpatient departments. After 6weeks open-label paroxetine 20 mg/day (T0), non-responding patients (HDRS17 decrease <50%; n=60) were randomized to double-blind paroxetine (30-50 mg/day as tolerable) or placebo dose-escalation (paroxetine 20 mg/day + placebo). Patients were followed until 6 weeks afterrandomization (T1). Forty-nine patients, drug free at study-entry, underwent SPECT-scanningbefore treatment and were scanned repeatedly at T0 and T1. Paroxetine serum concentrationsand SERT occupancy were determined at T0 and T1 (n= 32).

ResultsWe terminated the dose-escalation trial after an interim analysis. Thirty non-responding patientswere randomized to paroxetine (46.7 ±5.5 mg/day), 27 to placebo dose-escalation. Response-rates were 10/30 (33.3%) and 10/27 (37.0%), respectively. Repeated measurement analyses showedno significant effect for treatment (p= 0.88, exceeding a priori stopping rules for futility [p>0.5]).Overall dropout was higher for placebo (26.7%) than paroxetine (3.3%; p= 0.03). Paroxetine dose-escalation increased paroxetine serum concentrations (p<0.001). SPECT measurements (12patients randomized to paroxetine (46.9 ±4.8 mg) and 14 to placebo dose-escalation) showed nosignificant increase of midbrain SERT occupancy (2.5 ±26.4%, paroxetine; 3.1 ±25.8% placebo; p=0.687) nor in diencephalon (p= 0.529).

ConclusionsParoxetine dose-escalation in depressed patients has no clinical benefit over placebo dose-escalation. This is explained by the absence of significant increases of SERT occupancy byparoxetine dose-escalation, despite increased paroxetine serum concentrations. (ISRCTN register nr. ISRCTN44111488)

ChapterSERT occupancy by paroxetine dose-escalation in MDD

9

175

Introduction

Major depressive disorder (MDD) is often treated with antidepressants, particularly selectiveserotonin reuptake inhibitors (SSRIs). Unfortunately, response and remission rates are modest(30-50%), which require additional strategies to gain remission.1;2 Switching3 and augmentation4

have recently been evaluated. A third, and frequently applied option is dose-escalation,recommended in treatment guidelines1;2 and frequently used preceding other strategies. Only therecent NICE guideline is more reluctant in recommending dose-escalation.5 Although an individualpatient may improve after dose-escalation, this could also represent a delayed drug response orreflect the natural course of the disease. Prolonged (up to 10 weeks), unaltered treatment withfluoxetine 20 mg/day improved the response rates of initial week 6 non-responders.6

Theoretically, the concept of dose-escalation assumes linear dose-response relationships whichhave not been proven for SSRIs.7;8 Therefore, the efficacy of dose-escalation of SSRIs has beenquestioned.5;8-10

Previous studies did not show improved clinical effectiveness of dose-escalation, but hadserious methodological weaknesses.8-17 All previous studies increased dosages probably too early(mostly after 3-4 weeks) and too abruptly, which may have obscured true dose-escalation effectsby delayed effect of the standard doses and selective early drop-out of patients receiving truedose-escalation.9;10 Moreover, no study provided a rationale why dose-escalation was ineffective.

The primary molecular target of SSRIs is the serotonin transporter (SERT). Imaging techniquessuch as positron emission tomography (PET) and single photon emission computed tomography(SPECT) allow in-vivo labeling of SERT in the brain, which can be used to study their occupancy. Todate, several imaging studies measured SERT occupancy after short or prolonged treatment withSSRIs (Table S9.1).18-31 Particularly, Meyer et al.20 showed 60-80% SERT occupancy after standardclinical doses of SSRIs, and demonstrated curvilinear dose-response relationships for SERToccupancy by SSRIs. However, high doses of SSRIs were rarely studied,30 and dose-escalation wasnever studied.

Taking into account previous methodological criticisms, and considering the molecular targetof SSRIs, we have tested whether paroxetine dose-escalation increases SERT occupancy, andimproves depressive symptoms more than placebo dose-escalation. We performed a 6 week,multicenter, randomized study in depressed patients not responding to 6 weeks of paroxetine at20 mg/day. As a novel extension to previous clinical trials, and in order to elucidate theneurobiological basis for an expected lack of benefit of dose-escalation, we included a SPECTimaging approach. Herewith, we quantified whether paroxetine dose-escalation increased SERToccupancy more than placebo dose-escalation. This enabled us to relate clinical findings to theneurobiological correlate of SERT occupancy.

Methods

ParticipantsFollowing approval by the institutional ethical committee and written informed consent, werecruited outpatients (18-70 years) from primary care, our outpatient department, and publicpsychiatric settings between October 2003 and February 2007. Inclusion criteria were: MDDdetermined by the structured clinical interview for DSM-IV (SCID),32 and a Hamilton DepressionRating Scale (17 items; HDRS17)33 score above 18. All participants were drug-free or had undergoneno more than one antidepressant treatment (other than paroxetine) at an effective dose for ≥ 6weeks for the present MDD-episode. By the latter criterion, we avoided treatment resistance aspotential bias for inefficacy of dose-escalation. Exclusion criteria, apart from pregnancy (or wish),were bipolar disorder, psychotic features, neurological cognitive impairments (i.e. dementia),

Chap

ter

Chapter 9

9

176

primary anxiety and/or substance abuse disorders and acute, severe suicidal ideation. Contrary,we allowed secondary co-morbid anxiety and/or substance abuse to increase applicability of ourfindings.

InterventionsPatients were treated by their referring physician or were referred to our outpatient department.After assessment at study-entry, all patients were treated open-label with paroxetine 20 mg/dayfor 6 weeks (see Figure 2.1). When severe adverse effects occurred, dosages were reduced to 10mg/day and again increased to 20 mg/day after one week. We randomized all patients who did notachieve ≥50% decrease in HDRS17-score after 6 weeks, relative to study entry. They received a trueparoxetine or a placebo dose-escalation added to paroxetine 20 mg/day. Dose-escalation wasprovided in blue capsules containing 10 mg paroxetine or placebo. Randomization was stratifiedfor treatment setting (SPECT-group, outpatient department AMC, primary care, publicpsychiatry), gender and age. Within strata, we applied a minimization method to achieve abalanced distribution. We concealed allocation by using an independently operated computerprogram.

Dose-escalation consisted of incremental steps of one capsule every 5 days towards amaximum of 50 mg/day (20 mg + 3 capsules). Patients were allowed to increase at a slower pace(e.g. by 7 days) or stop further escalation (e.g. 20 mg + 2 capsules) according to adverse effects.34

No dosage adjustments were allowed during the last 3 weeks of the study. We checkedadherence by pill-counts and anamnesis.35

Outcomes and measurementsPrimary clinical outcomes were HDRS17-scores, and the proportion of patients achieving response(≥50% decrease in HDRS17) or remission (HDRS17 ≤7). Secondary outcomes were total and specific(adverse effects / inefficacy) dropout rates, the Maier and Bech 6 item subscales of theHDRS17,36;37 the Inventory for Depressive Symptomatology self-rated (IDS-SR30)38 scores, theoccurrence of adverse effects and health-related quality of life (MOS-SF36; physical and mentalcomponent scales standardized to a general Dutch population).39

We administered questionnaires at study-entry, randomization (T0), and 6 weeks afterrandomization (T1). Depressive symptoms were also monitored at week 1, 2 and 4 using the Maierand Bech subscales and IDS-SR30 (see Figure 2.1). Three trained investigators administeredclinician-rated questionnaires. Agreement between raters was good (intraclass correlationcoefficient = 0.98). Raters and patients were blinded for treatment.

Subgroup for SPECT imagingFrom all patients who entered the trial, we recruited patients who were drug-free (>4 weeks and≥5 half-lives of a previous antidepressant) as potential candidates for SPECT imaging. Thesepatients were asked to participate in the SPECT sub-study if their age was between 25-55 years toreduce variability in SERT measurements by age.40 Forty-nine patients could thus be recruited fora first SPECT scan. None of these patients reported past or present use of 3,4-methylenedioxymethamphetamine. We made a second scan in those patients who completed 6weeks of paroxetine treatment (n= 44; including 12 responders), while only randomized non-responders (n= 32) were invited for a third scan at the end of the study. We treated SPECTpatients at the AMC outpatient department. Medication was supplied in pillboxes.

SPECT imaging and analysisWe performed SPECT imaging at study-entry (baseline-scan), T0 and T1 (see Figure 2.1) between 2to 10 pm according to previously described procedures.41 We made all scans 230 ±18 (SD) minutesafter intravenous injection of approximately 100 MBq iodine-123-labeled 2β-carbomethoxy-3β-(4-iodophenyl)-tropane ([123I]β-CIT), when the radioligand is at equilibrium for SERT binding in brainareas expressing high densities of SERTs.42 To prevent thyroid uptake of [123I], all subjects received


9

177

oral potassium-iodide solution. We performed SPECT imaging using a 12-detector single slicebrain-dedicated scanner (Neurofocus 810, Strichmann Medical Equipment; Cleveland, OH) with afull-width at half-maximum resolution of 6.5 mm, throughout the 20 cm field-of-view (http://www.neurophysics.com). Blood for paroxetine serum concentrations (PSC) was collected at T0and T1 immediately before scanning. Serum was stored at -20° C until analysis. PSC weredetermined in May 2007 using a validated HPLC-MS/MS method (therapeutic range 10-75 µg/L; seeappendix). The lower limit of quantification was 5 µg/L, the lower limit of detection was 0.3 µg/L.

After attenuation correction and reconstruction in 3D mode (http://www.neurophysics.com),we defined regions of interest (RoIs) for midbrain, diencephalon and cerebellum by usingvalidated templates (see Figure 2.3).41 One examiner, blinded for scan session (baseline-scan/T0/T1), positioned all RoIs in two series. Intraclass correlation coefficients were >0.97 for all RoIs. Ifthe two series differed by >5%, scans were re-evaluated by a second investigator. In the analyseswe averaged the counts for the two series.

Using activity in cerebellum as indicator of non-displaceable activity (non-specific binding andfree radioactivity),43 we calculated specific to non-specific binding ratios per scan as BPND = . BPND is proportional to transporter number under equilibriumconditions.44 In a different study, we found high reproducibility of SERT imaging with [123I]β-CITSPECT after repeated scanning of subjects, using the same camera and scanning-protocol (de Winet al., submitted). As primary outcomes, we calculated SERT occupancies at T0 or T1 relative tountreated baseline-scan SERT availability: OCCT0 or T1 = .

Power and interim analysis We performed a-priori power-calculations for two co-primary endpoints: (a) To detect a differenceof ≥5-points in HDRS17 scores between paroxetine and placebo dose-escalation, while assuming acommon standard deviation of 7 and using a one-tailed α = 0.025 and β = 0.05, sample sizes of 60per group were required; (b) For response rates (assumed to be 50% and 30% for paroxetine vs.placebo dose-escalation) a two-tailed α = 0.05 and β = 0.20 required 110 participants per group.Because previous dose-escalation studies indicated no benefits relative to placebo dose-escalation, we planned an interim analysis after SPECT data for had been collected on at least 30randomized patients in the SPECT subgroup. Stopping criteria, using the most informativecontinuous scores in a mixed model, were predetermined using the O'Brien and Fleming45

approach, were undisclosed while performing the interim analysis, and were p<0.0026 in case ofsuperiority and p>0.50 for futility.

Data analysesAnalyses were performed while blinded for treatment allocation. We based endpoint analyses onintention to treat (ITT), with last-observation carried forward (LOCF). To examine theeffectiveness of paroxetine vs. placebo dose-escalation, we compared the proportion of patientswith response, remission and drop-outs at the end of study using χ2 or Fisher’s exact test. Weexamined differences in mean continuous endpoints by ANCOVA with treatment as factor andvalue at randomization (T0) as covariate.

We used linear mixed models to assess differences in trends over time between groups inMaier, Bech and IDS-SR30 scores. Mean scores for these questionnaires were modeled as afunction of the randomized group (paroxetine vs. placebo dose-escalation), score atrandomization, and time since randomization (categorical, four levels). The interaction betweentime*group was added to the model to test whether trends over time were different between thetwo treatment groups. We used the Akaike Information Criteria to choose the best fittingvariance/covariance structure (unstructured, compound symmetry or first-order auto-regressive)for each outcome parameter.

To examine changes in SERT occupancy between T0 and T1, we used ANCOVAs with treatmentas factor, and SERT occupancy at T0 and age as covariates. In order to obtain maximuminformation of dose-escalation in these analyses, we excluded patients that were likely non-adherent to paroxetine at T0 or T1 (PSC < 5 μg/l). Thereafter, we plotted SERT occupancy against

CER

CERROI


( )BslND

TorTNDBslND

BPBPBP 10−

Chap

ter

Chapter 9

9

178

dose and PSC. We modeled dose-response in an Emax model as OCC= , in which arepresents maximal SERT occupancy and b the PSC with 50% SERT occupancy.18;20-22;27;28 Wecalculated a and b by fitting a nonlinear regression model that minimizes the sum of squares ofthe residuals. For quantification of differences in SERT occupancy between final responders andnon-responders, we used ANCOVA models corrected for differences at T0, age and baseline-scanSERT availability in diencephalon.46 We performed all analyses in SPSS v15.0.1.1 (www.spss.com ).

Results

Patient dispositionOne-hundred and seven patients (mean age 43.8 ±9.8) started open-label paroxetine (Figure 9.1).The response-rate in the open phase was 27/107 (25.2%), and 60 non-responding patients wererandomized for the double-blind phase. Randomization over the 2 treatment-arms resulted incomparable groups (Table 9.1). Fifty-one patients completed the 6 week randomization phaseincluding 31 from the SPECT study. We obtained at least 1 post-randomization HDRS17-score for 57patients.

HDRS17 scores at study-entry (-6 weeks) were comparable in T0 responders vs. non-responders (ANOVA, F1,85= 1.972, p= 0.164). At T0, HDRS17 scores (±SD) were 7.8 ±3.62 (66.6 ±14.3%decrease) in responders vs. 20.5 ±6.25 (17.7 ±20.3% decrease) in non-responders.

Figure 9.1. Recruitment and flow of participants.

* Three patients who refused dose-escalation after randomization, never ingested study drugs and refused furtherquestionnaires, were excluded for endpoint analysis.† One SPECT patient dropped out early due to inefficacy, but for all SPECT-patients clinical data could be obtained. For SPECTanalyses, 6 patients were excluded: 1 patient missed the T1 scan (placebo dose-escalation), 3 patients were likely non-adherent at T0 (paroxetine serum concentration <5μg/l; all paroxetine dose-escalation), and 2 were likely non-adherent at T1(1 paroxetine dose-escalation, 1 placebo dose-escalation).

)(*

PSCbPSCa+

��

5�� 5��

'�!�� "�� " 5��

'�� 5��

#!�� 5��$��

�� 5��+�

%�� 5��

�� &�� '�� #�� 5��(�

3�� 5��

�� 5��

� !�� 5��

�� 5��)�

"�� "� �� 5��

3�� * �� 5��

�� 5��+�

� !�� 5��$�

�� 5��

��&�� 5�� .��

� �� .��

��

��

��

��

��

�� .�� +�

� �� .��

�� .�� !�"��! ��#�

��

$� �� "%��&�

3��! ��'�(� � ��

��))��

��*�� ))��

� (��

�� .�� !�"��! ��#�


9

179

Table 9.1. Characteristics of non-responding MDD patients after 6 weeks of open treatment with paroxetine 20 mg/day (trial population).

All patients (n= 60) SPECT subgroup (n= 32)

Paroxetine DE (n= 30) Placebo DE (n= 30) Paroxetine DE (n= 16) Placebo DE (n= 16)

Age at baseline (years) 41.9 ±9.1 42.9 ±10.3 42.5 ±7.7 40.4 ±7.7

Female sex - n (%) 21 (70.0) 19 (63.3) 11 (68.8) 10 (62.5)

Marital status - n (%) Single (never married) Married Divorced WidowedEducational level - n (%) Low Intermediate HighUnemployed - n (%) Income €/month (median; 25 & 75 quartiles)

15 (51.7) 8 (27.6) 4 (13.8) 2 (6.9)

6 (20.0) 19 (63.3) 5 (16.7) 6 (20.0) 1485 (875-1769)

12 (40.0) 6 (20.0) 11 (36.7) 1 (3.3)

11 (36.7) 15 (50.0) 4 (13.3) 11 (36.7) 1197 (767-1820)

6 (37.5) 7 (43.8) 2 (12.5) 1 (6.3)

2 (12.5) 10 (62.5) 4 (25.0) 4 (25.0) 1177 (715-1530)

5 (31.3) 4 (25.0) 7 (43.8) 0

3 (18.8) 10 (62.5) 3 (18.8) 4 (25.0) 1185 (610-2277)

Current smoking - n (%) Alcohol use: n (%) ≤ 2 Units/week 3-7 Units/ week 8-21 Units/ week >22 Units/ week

13 (43.3)

20 (66.7) 6 (20.0) 2 (6.7) 2 (6.7)

16 (53.3)

21 (70.0) 7 (23.3) 1 (3.3) 1 (3.3)

6 (37.5)

10 (62.5) 4 (25.0) 1 (6.3) 1 (6.3)

11 (68.8)

11 (68.8) 5 (31.3) 0 0

Race - n (%) Caucasian Creole Asian Other

17 (56.7) 4 (13.3) 6 (20.0) 3 (10.0)

19 (63.3) 7 (23.3) 1 (3.3) 3 (10.0)

9 (56.2) 2 (12.5) 3 (18.8) 2 (12.5)

13 (81.2) 3 (18.8) 0 0

MDD HDRS17 at study-entry (-6 weeks) HDRS17 (T0)*

Maier at (T0)*

Bech (T0)*

IDS-SR30 (T0)*

First episode - n (%) No of episodes Drug-naïve - n (%) Used AD in current episode - n (%)Melancholic - n (%) Duration of episode: n (%) <5 months duration 5months – 2 years duration > 2 yearsAge of first episode (years)

24.5 ±4.7 20.1 ±6.6 10.0 ±3.0 10.5 ±2.9 38.1 ±12.3 17 (56.7) 1.6 ±0.8 24 (80.0) 3 (10.0) 23 (76.7)

9 (30.0) 19 (63.3) 2 (6.7) 37.1 ±9.1

25.5 ±5.0 21.0 ±5.9 10.5 ±3.1 11.3 ±3.0 40.8 ±12.2 21 (70.0) 1.7 ±1.8 19 (63.3) 5 (16.7) 19 (63.3)

5 (16.7) 22 (73.3) 3 (10.0) 38.1 ±11.8

25.6 ±5.0 21.3 ±7.4 10.4 ±3.6 11.0 ±3.4 40.3 ±14.6 9 (56.3) 1.6 ±0.8 14 (87.5)†

1 (6.3) 12 (75.0)

7 (43.8) 7 (43.8) 2 (12.5) 38.4 ±8.8

24.4 ±4.6 19.3 ±5.1 10.1 ±2.9 10.8 ±3.1 39.9 ±11.8 9 (56.3) 2.3 ±2.4 8 (50.0)†

2 (12.5) 12 (75.0)

3 (18.8) 11 (68.8) 2 (12.5) 34.4 ±10.2

Co-morbidity – n (%) Anxiety disorder Dysthymia Drug (alcohol, cannabis, benzodiazepines) abuse / dependence

5 (16.7) 2 (6.7) 2 (6.7)

7 (23.3) 0 1 (3.3)

0 1 (6.3) 1 (6.3)

4 (25.0) 0 0

MOS-SF36 Physical*

Mental*41.2 ±9.0 26.1 ±8.4

40.3 ±10.7 25.7 ±10.2

41.2 ±8.6 25.3 ±5.9

43.9 ±9.9 22.3 ±6.7

SERT availability baseline-scan (BPND) Midbrain DiencephalonSERT occupancy (% of BPND in Bsl-scan)* ‡

Midbrain Diencephalon

N/A N/A

N/A N/A

N/A N/A

N/A N/A

0.553 ±0.119 1.157 ±0.226

73.7±17.1 63.8 ±15.4

0.657 ±0.217 1.134 ±0.247

82.2 ±18.5 70.3 ±12.1

Numbers represent means (±standard deviation) unless specified otherwise.* at randomization (T0).† sign. difference between SPECT patients randomized to conditions (Fisher’s exact p = 0.026).‡ n = 29; excluding three patients who were likely nonadherent at T0 (PSC <5μg/l). BPND = Binding Potential (non-displaceable), DE = dose-escalation, MDD = Major Depressive Disorder, PSC = Paroxetine serum concentration

Chap

ter

Chapter 9

9

180

Clinical effectiveness of paroxetine vs. placebo dose-escalationDuring dose-escalation (T0-T1), 1, 8 and 21 patients reached final doses of 30, 40 and 50 mg/dayrespectively. The placebo group escalated to a comparable number of capsules (χ2= 0.895, df= 2,p= 0.639). Adherence based upon pill-counts was comparable between both groups (Fisher’sexact, p= 0.492).

Paroxetine dose-escalation did not yield better outcomes in depression severity and health-related quality of life compared to placebo dose-escalation (ITT; Table 9.2). The robustness of thisfinding was confirmed in the longitudinal analysis (mixed model). Changes over time in the Maiersubscale and IDS-SR30-scores (Figure 9.2), and Bech subscale-scores (available on request), werecomparable between the two groups. Overall dropout was higher with placebo (26.7%) than withparoxetine dose-escalation (3.3%; p= 0.03). Paroxetine dose-escalation had significantly moreadverse effects than placebo dose-escalation, but this did not result in higher discontinuationrates due to adverse effects (Table S9.2). Instead, adverse effects by paroxetine dose-escalationmoderately decreased over time, suggestive of habituation.

Table 9.2. Depression and health-related quality of life scores after 6 weeks paroxetine vs. placebo dose-escalation (T1); all patients and SPECT subgroup.

All patients (n = 57) SPECT subgroup (n = 32)

Paroxetine DE (n = 30) Placebo DE (n = 27) p Paroxetine DE (n = 16) Placebo DE (n = 16) p

Mean dosage mg/day

46.7 ±1.00 NA - 46.9 ±1.20 NA -

HDRS17 16.1 ±1.22 15.3 ±1.28 0.650 15.8 ±1.39 14.5 ±1.39 0.519

Maier subscale 7.5 ±0.61 7.5 ±0.64 1.000 7.6 ±0.69 7.4 ±0.69 0.868

Bech subscale 8.1 ±0.63 8.1 ±0.66 0.996 8.2 ±0.75 8.0 ±0.75 0.832

Response* - n (%) 10 (33.3) 10 (37.0) 0.788 4 (25.0) 8 (50%) 0.273

Remission† - n (%) 4 (13.3) 2 (7.4) 0.673 1 (6.3) 2 (12.5) 1.000

IDS-SR30‡ 34.8 ±1.83 32.5 ±2.05 0.406 39.2 ±2.38 32.0 ±2.38 0.042

MOS-SF36§

Physical 41.8 ±0.93 42.4 ±1.06 0.679 42.5 ±1.33 44.1 ±1.33 0.399

Mental 29.6 ±1.37 27.3 ±1.57 0.264 27.2 ±1.78 25.9 ±1.78 0.620

Scores at endpoint of the study are based on intention to treat, with last observation carried forward for early drop-outs. Values are means ±standard error, corrected for (mean) scores at randomization (T0) (ANCOVA).* ≥50% decrease in HDRS17 with baseline score (-6 weeks) as reference; Fisher’s exact test.† HDRS17 ≤7; Fisher’s exact test.‡ Due to missing values: n = 29 paroxetine DE and n = 23 placebo DE for all patients.§ Due to missing values: n = 29 paroxetine DE and n = 22 placebo DE for all patients. DE = dose-escalation.

Figure 9.2. Changes over time in Maier and IDS-SR scores after randomization.

Points represent mean Maier (A) and IDS-SR (B) scores (±SEM) adjusted for scores at randomization (T0) for paroxetine (n=30) and placebo (n= 27) dose-escalation. Mixed model analysis (Maier: n= 57, IDS-SR: n= 53): overall difference betweenparoxetine vs. placebo dose-escalation for Maier scores F4,101,400 = 0.295, p= 0.880, and for IDS-SR F4,45,519= 1,516; p= 0.213. * Mean IDS-SR score at week 4 differed significantly in favor of placebo dose-escalation (t49,161= 2.11; p= 0.040).DE = dose-escalation.

� � ��

��

��

��

��

��

�

�

��

��

��

� � � ��

�

�

��

��

��

��

��

� �


9

181

SERT occupancy and clinical responseOf 32 randomized patients in the SPECT subgroup, only 3 (9%) previously used mirtazapine orfluoxetine in the current episode (1 patient stopped mirtazapine 4 weeks prior to scanning, othersstopped <2 months before scanning). One patient missed the T1 scan. Based upon PSC atrandomization or T1, five patients in the SPECT-subgroup (4 with paroxetine and 1 with placebodose-escalation; Fisher’s Exact: p= 0.172; see Figure 9.1) with PSC <5 μg/l were considered non-adherent, despite adherence according to pill-counts. We excluded these five patients foranalyses of changes in SERT occupancy after dose-escalation, leaving 26 T1 scans analyzable fordiencephalon occupancies after true or placebo dose-escalation. Paroxetine dose-escalationincreased mean PSC from 36.2 to 154.3 μg/l (paired t-test; p<0.001), while mean PSC in placebodose-escalation remained unchanged (Table S9.3). At randomization (T0), mean SERToccupancies (±SEM) for the paroxetine dose-escalation group were 76.2 ±4.70% in midbrain and64.3 ±4.60% in diencephalon. For the placebo dose-escalation group these were 84.6 ±4.95% and72.2 ±3.08%, respectively. Neither paroxetine nor placebo dose-escalation significantly increasedSERT occupancy further (Figure 9.3A). Plotting PSC vs. SERT occupancy showed that PSCs >50 μg/l were not associated with further increases of SERT occupancy in midbrain or diencephalon(Figure 9.3BC). Furthermore, individual changes in PSC (T0 to T1) were not significantly associatedwith changes in occupancy in midbrain (F1,24= 0.101; p= 0.754) and diencephalon (F1,27= 1.332; p=0.259; Figure S9.1).

Figure 9.3. SERT occupancy during randomized dose-escalation of paroxetine. Changes over time and relation with paroxetine serum concentration.

A. Mean SERT occupancy (±SEM) for paroxetine dose-escalation and placebo dose-escalation at randomization (T0) and after6 weeks of dose-escalation (T1). SERT occupancy was calculated as percentage of initial available SERTs (expressed as BPND)at baseline (-6 weeks) scans (see text). Changes in SERT occupancy between T0 and T1 for paroxetine dose-escalation andplacebo dose-escalation were non-significant in ANCOVA models correcting for age and differences in T0 SERT occupancy(MIDBRAIN: F1,21= 0.167; p= 0.687; DIENCEPHALON: F1,22 = 0.409; p= 0.529). For 1 patient insufficient midbrain was scanned atstudy entry to compute subsequent SERT occupancies.B & C. Data for randomized patients used from both T0 (open circles) and T1 (diamonds). Dose - occupancy relationships aremodeled as OCC= . For Midbrain (B): a = 86.0 ±4.03 (SE), b= 2.65 ±1.39 (n= 30 at T0 and n= 27 at T1); for Diencephalon(C) a= 73.3 ±2.28, b = 3.97±1.07 (n= 32 at T0 and n= 29 at T1). Dashed lines represent 95% confidence interval of fitting.DE = dose-escalation, OCC = occupancy, PAR = paroxetine, PLAC = placebo. PSC = paroxetine serum concentration.

��

�

�

��

��

��

��

��

��'�*

��+��

��'��'

�� !

� "� �� "� �� "� #��

�

�

��

��

��

�

��

��,��'��$��&��'��'��'�� %&�!

��'�*

�� !

� "� �� "� �� "� #��

�

�

��

��

��

�

��

��,��'��$��&��'��'��'�� %&�!

��'�*

�� !

�

� �

)(*

PSCbPSCa+

Chap

ter

Chapter 9

9

182

We explored whether SERT occupancy was related to clinical response at T1 irrespective ofparoxetine or placebo dose-escalation. Responders at T1 (n= 12) had numerically higher SERToccupancy (±SEM) in midbrain (91.2 ±5.8%) and diencephalon (69.2 ±2.8%) than non-responders(n= 14; 77.8 ±5.1% and 63.8 ±2.6%, respectively; ANCOVA: p= 0.107 and p= 0.178). These models usedbaseline-scan diencephalon SERT availability as covariate as this accounted for a major part ofvariance (F1,21= 4.831, p= 0.039 and F1,22= 10.407, p= 0.004 respectively). Contrary, T1 SERToccupancy in midbrain or diencephalon did not significantly predict the percentage decrease inHDRS17 in linear regression, or response status in logistic regression (neither when corrected forbaseline-scan SERT availability in diencephalon or age).

Discussion

In this randomized trial we examined clinical effectiveness of dose-escalation in MDD patients,who were non-responders to 6 weeks of 20 mg/day paroxetine, and explored potential underlyingmechanisms. Despite markedly increased drug exposure, paroxetine dose-escalation to 30-50 mg/day did not improve depressive symptoms more than placebo dose-escalation, but wasassociated with more adverse effects. Concomitantly, increased paroxetine serum concentrationswere not associated with substantially greater SERT occupancy, indicating that standardparoxetine doses (20 mg/day) already resulted in maximum SERT occupancy.

Clinical outcomes of dose-escalation in non-respondersThe dose-response relationship for paroxetine was previously examined in fixed dose, parallelgroup designs. Twenty mg/day and higher paroxetine doses yielded similar clinicalimprovements.7 Similar findings were reported from parallel-group, fixed dose studies of otherSSRIs.8 Accordingly, the usefulness of dose-escalation in non-responders to paroxetine and otherSSRIs has been questioned.8-10 However, the underlying studies had methodological short-comings, and dose-escalation remains a recommended standard approach for non-responders.1;2

If dose-escalation is applied too early (before week 6), randomization of ‘late responders’ willlikely dilute the difference between true and placebo dose-escalation, resulting in potentially falsenegative findings. While one study reported randomized dose-escalation of sertraline after 6weeks of treatment, this was compromised by a non-randomized dose-increase 2 weeks prior torandomization.17

The benefits of dose-escalation might be under-estimated if actively-treated patients drop-outearly due to adverse effects34 or become more non-adherent. Our schedule for dose-escalationdid not increase drop-out. Hypothetically, patients receiving a paroxetine dose-escalation mighthave interpreted the increased level of adverse effects as subjective clue of greater drug effects,encouraging them to persevere in the trial. At first sight a misbalance in adherence is suggestedwith 4 patients with paroxetine vs. 1 patient with placebo dose-escalation having low PSCs.However, this was not due to dose-escalation. As mentioned in Figure 9.1, in three patients, lowPSCs at T0 already classified them as likely non-adherent, with only 1 paroxetine vs 1 placebo dose-escalation patient becoming likely non-adherent during dose-escalation (T0-T1). Therefore, wethink that neither adverse effects nor non-adherence account for the observed inefficacy of dose-escalation.

Thus, the present study overcomes methodological limitations of previous SSRI dose-escalation studies by using a randomized, placebo-controlled, double-blind dose-escalation innon-responders to 6 weeks treatment with a standard dose of paroxetine. We also avoidedtreatment resistance as a factor for inefficacy of dose-escalation by inclusion of patients whoreceived no more than one effective antidepressant trial for the current episode. Under theseconditions paroxetine was not superior to placebo in dose-escalation. Moreover, our study was


9

183

more inclusive than most previous ones and thus may be more applicable to ‘real-world’ first-lineantidepressant treatment. This may also explain why we observed lower response and remissionrates than previous studies.

Neurobiological effects of dose-escalationOur pharmacokinetic and imaging measurements were designed to explore why paroxetine dose-escalation would or would not improve treatment outcomes. Our imaging of SERT occupancybypasses potential bias by inclusion of patients with ultrarapid drug-metabolism, which is oftencausally linked to non-response. Hypothetically the clinical selection of non-responders eligible fordose-escalation, might represent a selection of patients not reaching high levels of SERToccupancy.

Comparing SERT occupancies of different SSRI doses faces several methodologicalchallenges. Firstly, the assessment of occupancy requires knowledge on the available number ofSERTs. Due to inter-individual differences in available SERTs, only assessments with individualdrug-free baseline-scans yield reliable data. Secondly, a given drug dose may yield a range ofserum concentrations due to inter-patient pharmacokinetic differences. Hence, associationsbased upon serum concentrations are more reliable than those based upon administered dose.Finally, intra-individual comparisons of occupancy changes following dose-escalation are morepowerful than those with historic data.

Against this background, Voineskos et al.30 recently reported high SERT occupancies instriatum (~85%), thalamus (~79%), and midbrain (~98%) in 12 depressed patients exposed to > 4weeks of venlafaxine 225-450 mg/day, sertraline 150-200 mg/day or citalopram 60-80 mg/day in a[11C]DASB PET study. They concluded that high doses significantly increased occupancy comparedto an average of 80% SERT occupancy determined in previous studies with standard SSRIdoses,20;47 which would favor the concept of dose-escalation. However, they did not determineoccupancy relative to baseline scans of the same patients without medication, nor at standarddoses. We performed drug-free study-entry scans, in addition, >90% of patients did not useantidepressants for the current episode of MDD. Furthermore, low therapeutic dosages of severalSSRIs also yielded high SERT occupancy in most studies.18;19;21-29;48 The present study resolves thiscontroversy by showing that 4-fold increases of PSC upon paroxetine dose-escalation did notsignificantly increase SERT occupancy. This offers an explanation for our findings: SERToccupancy is limited by a ceiling effect. A PSC achieved with a 20 mg/day paroxetine dose issufficient to yield maximum SERT occupancy (Figures 9.3B,C). If low doses already yield maximumSERT occupancy, dose-escalation cannot be expected to increase treatment efficacy, which is inline with our clinical findings. Our results do not necessarily challenge the relationships betweendose, SERT occupancy and clinical response but rather suggest that these relationships existmainly at low and sub-therapeutic doses. Furthermore, the relationship between SERT occupancyand response might be confounded by other factors such as SERT gene polymorphisms.

In a recent study Owens et al.49 showed increased SERT occupancy with increasing paroxetineCR doses (12.5-75 mg/day) in an ex-vivo model using human transporter transfected cells, whichmight be at odds with our findings. However, validation of this ex-vivo method (in cultured cells)with concomitant in-vivo SPECT or PET SERT occupancy (the gold standard) is not yet available.Additionally, Zitterl et al.50 found a significant relation between SERT occupancy and clinicalresponse in obsessive compulsive disorder treated with clomipramine (150 mg/day), but did notstudy the effects of dose-escalation in their study. Therefore, our study optimally quantifies theneurobiological effects of dose-escalation of antidepressants in patients.

Critique of methodsFor logistic reasons, we used [123I]β-CIT for SPECT imaging, which is a non-selective radioligand,and also binds to dopamine transporters (DAT; e.g. substantia nigra) and norepinephrinetransporter (NET; e.g. locus coeruleus).51-53 Furthermore, imaging studies indicated increasedstriatal DATbinding after treatment with paroxetine,48;54 especially when the occipital cortex wasused as a reference.54 Nevertheless, uptake in midbrain and diencephalon is considered to reflect

Chap

ter

Chapter 9

9

184

predominantly SERT,55 as these structures are rich of SERT relative to DAT and NET. Therefore,although this non-selectivity might have concealed changes in SERT occupancies due to additionalDAT or NETbinding, we think our findings in diencephalon and midbrain mainly reflect SERToccupancy. Nevertheless, it would be challenging to replicate our study using a selective ligandfor SERT like [11C]DASB for PET or [123I]ADAM for SPECT imaging.

Our study did not investigate secondary effects of paroxetine. Many adaptive pre- and post-synaptic effects of chronic administration of SSRIs have been documented, includingneuroadaptive alterations in serotonin receptors and intracellular signalling pathways,56-61 as wellas time-dependent effects on neurogenesis.62-64 Thesehypothetical additional effects of dose-escalation remain to be investigated. Nevertheless, neither the results of our trial, nor the findingsin previous randomized controlled trials indicate that dose-escalation is an efficacious strategy forSSRI non-responders in MDD.8-10

The fourfold increase of PSC after dose-escalation from 20 to 50 mg/day may questionadherence of patients in the open phase of the study. However, paroxetine inhibits thecytochrome P450 enzyme 2D6, also responsible for the metabolism of paroxetine. Thereforenonlinear increases of PSC reflect normal paroxetine pharmacokinetics.65

Our study was discontinued after an interim analysis with relatively small patient numbers.However, the criteria for premature trial termination regarding futility had been pre-specified,making it highly unlikely that we overlooked clinically relevant differences. Moreover, theneurobiological parts of our study provide a rationale why even with much larger patient numbersno substantially different outcome can be expected. On the other hand, premature stoppingreduced the power to examine whether subgroups of patients were more responsive to dose-escalation.

The present study was not designed to test the efficacy of paroxetine per se, as this is wellestablished in patients with severe MDD (HDRS17>18).1;2;5 Therefore, we did not include a pureplacebo arm. Rather we investigated dose-escalation, and accordingly only included placeboduring dose-escalation. This approach is similar to e.g. the STAR*D project, which interestinglyreported similar response rates for open treatment with citalopram.66

ConclusionPrevious studies had failed to demonstrate a clinical benefit of dose-escalation by SSRIs, but hadmethodological limitations.8-10 Addressing those limitations, our trial replicates that dose-escalation of paroxetine above the 20 mg/day standard dose has no additional clinical benefit. Asa novel extension, we revealed the underlying neurobiological mechanism for this inefficacy:maximum SERT occupancy was already reached with the standard dose. Similarly high SERToccupancies reported with low doses of other SSRIs suggest that our conclusion may beapplicable to the entire drug class. However, this does not exclude that dose-escalation hasclinical benefits for antidepressants with additional molecular targets, e.g. the norepinephrinetransporter, such as venlafaxine. This drug has shown dose-dependency of the clinical response infixed-dose studies.67;68

If dose-escalation is not promising for paroxetine and presumably other SSRIs, two clinicaloptions remain for the treatment of non-responders to standard doses. These are eithercontinuation of treatment until 10 weeks while waiting for a potential delayed response, or achange to a different and potentially more effective treatment strategy. Both strategies willfurther improve response-rates, but studies directly comparing these strategies have not yet beenperformed.


9

185

Acknowledgements

The authors wish to thank the patients in this study for their participation, and especially thankthe patients that were willing to participate in the SPECT-study. We thank all participating generalpractitioners in the area of Amsterdam Oost and Zuidoost, Hoofddorp, Nieuw Vennep, andAbcoude for their inclusions and referrals for the study. Mrs. E. Miedema M.D. and Mrs. M.C. tenDoesschate M.D. were indispensable for their help in rating questionnaires. Mrs M. Haagesmanaged randomization and maintained blinding. This study was financed by a grant from theNetherlands Organisation for Health Research and Development (ZonMw), program MentalHealth, education of investigators in mental health to H.G. Ruhé (OOG; #100-002-002). Weespecially thank Professor M.E. Thase M.D. for his constructive comments on a previous versionof the manuscript.


All authors reported no biomedical financial interests or potential conflicts of interest.

References

1. Kennedy SH, Lam RW, Cohen NL, Ravindran AV. Clinical guidelines for the treatment of depressive disorders. IV.Medications and other biological treatments. Can J Psychiatry. 2001; 46 Suppl 1: 38S-58S.



4. Trivedi MH, Fava M, Wisniewski SR, Thase ME, Quitkin F, Warden D et al. Medication augmentation after the failure ofSSRIs for depression. N Engl J Med. 2006; 354: 1243-1252.



7. Dunner DL, Dunbar GC. Optimal dose regimen for paroxetine. J Clin Psychiatry. 1992; 53 Suppl: 21-26.8. Adli M, Baethge C, Heinz A, Langlitz N, Bauer M. Is dose escalation of antidepressants a rational strategy after a medium-

dose treatment has failed? A systematic review. Eur Arch Psychiatry Clin Neurosci. 2005; 255: 387-400.9. Ruhe HG, Huyser J, Swinkels JA, Schene AH. Dose escalation for insufficient response to standard-dose selective

serotonin reuptake inhibitors in major depressive disorder: Systematic review. Br J Psychiatry. 2006; 189: 309-316.10. Baker CB, Tweedie R, Duval S, Woods SW. Evidence that the SSRI dose response in treating major depression should be

reassessed: a meta-analysis. Depress Anxiety. 2003; 17: 1-9.11. Dornseif BE, Dunlop SR, Potvin JH, Wernicke JF. Effect of dose escalation after low-dose fluoxetine therapy.

Psychopharmacol Bull. 1989; 25: 71-79.12. Schweizer E, Rickels K, Amsterdam JD, Fox I, Puzzuoli G, Weise C. What constitutes an adequate antidepressant trial for

fluoxetine? J Clin Psychiatry. 1990; 51: 8-11.13. Fava M, Rosenbaum JF, McGrath PJ, Stewart JW, Amsterdam JD, Quitkin et al. Lithium and tricyclic augmentation of

fluoxetine treatment for resistant major depression: a double-blind, controlled study. Am J Psychiatry. 1994; 151: 1372-1374.



16. Fava M, Alpert J, Nierenberg A, Lagomasino I, Sonawalla S, Tedlow J et al. Double-blind study of high-dose fluoxetineversus lithium or desipramine augmentation of fluoxetine in partial responders and nonresponders to fluoxetine. Journalof Clinical Psychopharmacology. 2002; 22: 379-387.


Chap

ter

Chapter 9

9

186



20. Meyer JH, Wilson AA, Sagrati S, Hussey D, Carella A, Potter WZ et al. Serotonin Transporter Occupancy of Five SelectiveSerotonin Reuptake Inhibitors at Different Doses: An [11C]DASB Positron Emission Tomography Study. American Journalof Psychiatry. 2004; 161: 826-835.




24. Hiltunen J, Akerman KK, Kuikka JT, Bergstrom KA, Halldin C, Nikula T et al. Iodine-123 labeled nor-β-CIT as a potentialtracer for serotonin transporter imaging in the human brain with single-photon emission tomography. Eur J Nucl Med.1998; 25: 19-23.

25. Klein N, Sacher J, Geiss-Granadia T, Mossaheb N, Attarbaschi T, Lanzenberger R et al. Higher serotonin transporteroccupancy after multiple dose administration of escitalopram compared to citalopram: an [123I]ADAM SPECT study.Psychopharmacology (Berl). 2007; 191: 333-339.

26. Klein N, Sacher J, Geiss-Granadia T, Attarbaschi T, Mossaheb N, Lanzenberger R et al. In vivo imaging of serotonintransporter occupancy by means of SPECT and [123I]ADAM in healthy subjects administered different doses ofescitalopram or citalopram. Psychopharmacology (Berl). 2006; 188: 263-272.




30. Voineskos AN, Wilson AA, Boovariwala A, Sagrati S, Houle S, Rusjan P et al. Serotonin transporter occupancy of high-doseselective serotonin reuptake inhibitors during major depressive disorder measured with [11C]DASB positron emissiontomography. Psychopharmacology (Berl). 2007; 193: 539-545.

31. Cavanagh J, Patterson J, Pimlott S, Dewar D, Eersels J, Dempsey MF et al. Serotonin transporter residual availabilityduring long-term antidepressant therapy does not differentiate responder and nonresponder unipolar patients. BiolPsychiatry. 2006; 59: 301-308.

32. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV Axis I Disorders Patient Edition(SCID-I/P version 2.0). Translated in Dutch by Groenestijn MAC, Akkerhuis GW, Kupka RW, Schneider N, and Nolen WA.1999. Lisse, the Netherlands, Swets & Zeitlinger B.V.

33. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 23: 56-61.34. Baker CB, Woods SW. Is there a SSRI dose response in treating major depression? The case for re-analysis of current data






39. Aaronson NK, Muller M, Cohen PD, Essink-Bot ML, Fekkes M, Sanderman R et al. Translation, validation, and norming ofthe Dutch language version of the SF-36 Health Survey in community and chronic disease populations. J Clin Epidemiol.1998; 51: 1055-1068.

40. van Dyck CH, Malison RT, Seibyl JP, Laruelle M, Klumpp H, Zoghbi SS et al. Age-related decline in central serotonintransporter availability with [123I]β-CIT SPECT. Neurobiol Aging. 2000; 21: 497-501.

41. de Win MM, Habraken JB, Reneman L, van den Brink W, den Heeten GJ, Booij J. Validation of [123I]β-CIT SPECT to assessserotonin transporters in vivo in humans: a double-blind, placebo-controlled, crossover study with the selective serotoninreuptake inhibitor citalopram. Neuropsychopharmacology. 2005; 30: 996-1005.





9

187

45. O'Brien PC, Fleming TR. A multiple testing procedure for clinical trials. Biometrics. 1979; 35: 549-556.46. Kugaya A, Sanacora G, Staley JK, Malison RT, Bozkurt A, Khan S et al. Brain serotonin transporter availability predicts

treatment response to selective serotonin reuptake inhibitors. Biol Psychiatry. 2004; 56: 497-502.47. Meyer JH. Imaging the serotonin transporter during major depressive disorder and antidepressant treatment. J

Psychiatry Neurosci. 2007; 32: 86-102.48. Kugaya A, Seneca NM, Snyder PJ, Williams SA, Malison RT, Baldwin RM et al. Changes in human in vivo serotonin and

dopamine transporter availabilities during chronic antidepressant administration. Neuropsychopharmacology. 2003; 28:413-420.

49. Owens MJ, Krulewicz S, Simon JS, Sheehan DV, Thase ME, Carpenter DJ et al. Estimates of Serotonin and NorepinephrineTransporter Inhibition in Depressed Patients Treated with Paroxetine or Venlafaxine. Neuropsychopharmacology. 2008;Advance online publication; doi:10.1038/npp.2008.47.

50. Zitterl W, Aigner M, Stompe T, Zitterl-Eglseer K, Gutierrez-Lobos K, Wenzel T et al. Changes in Thalamus-HypothalamusSerotonin Transporter Availability during Clomipramine Administration in Patients with Obsessive-Compulsive Disorder.Neuropsychopharmacology. 2008; Advance online publication; doi:10.1038/npp.2008.35.


52. Neumeyer JL, Wang SY, Milius RA, Baldwin RM, Zea-Ponce Y, Hoffer PB et al. [123I]-2 beta-carbomethoxy-3 beta-(4-iodophenyl)tropane: high-affinity SPECT radiotracer of monoamine reuptake sites in brain. J Med Chem. 1991; 34: 3144-3146.

53. Neumeyer JL, Tamagnan G, Wang S, Gao Y, Milius RA, Kula NS et al. N-substituted analogs of 2 beta-carbomethoxy-3 beta-(4'-iodophenyl)tropane (beta-CIT) with selective affinity to dopamine or serotonin transporters in rat forebrain. J MedChem. 1996; 39: 543-548.

54. Booij J, de Jong J, de Bruin K, Knol R, de Win MM, van Eck-Smit BL. Quantification of striatal dopamine transporters with[123I]-FP-CIT SPECT is influenced by the selective serotonin reuptake inhibitor paroxetine: a double-blind, placebo-controlled, crossover study in healthy control subjects. J Nucl Med. 2007; 48: 359-366.


56. Meyer JH, Kapur S, Eisfeld B, Brown GM, Houle S, DaSilva J et al. The effect of paroxetine on 5-HT2A receptors indepression: an [18F]setoperone PET imaging study. Am J Psychiatry. 2001; 158: 78-85.


58. Blendy JA. The role of CREB in depression and antidepressant treatment. Biol Psychiatry. 2006; 59: 1144-1150.59. Blier P, De Montigny C. Current advances and trends in the treatment of depression. Trends Pharmacol Sci. 1994; 15: 220-

226.60. Davidson C, Stamford JA. Effect of chronic paroxetine treatment on 5-HT1B and 5-HT1D autoreceptors in rat dorsal raphe

nucleus. Neurochem Int. 2000; 36: 91-96.61. Kim SW, Park SY, Hwang O. Up-regulation of tryptophan hydroxylase expression and serotonin synthesis by sertraline.

Mol Pharmacol. 2002; 61: 778-785.62. Djavadian RL. Serotonin and neurogenesis in the hippocampal dentate gyrus of adult mammals. Acta Neurobiol Exp

(Wars). 2004; 64: 189-200.63. Malberg JE, Eisch AJ, Nestler EJ, Duman RS. Chronic antidepressant treatment increases neurogenesis in adult rat

hippocampus. J Neurosci. 2000; 20: 9104-9110.64. Martinowich K, Lu B. Interaction between BDNF and serotonin: role in mood disorders. Neuropsychopharmacology. 2008;

33: 73-83.65. Paxil® Prescribing information and medication guide PXL:47PI (available at http://us.gsk.com/products/assets/

us_paxil.pdf). 1-44. 2008. Research Triangle Park, North Carolina, GlaxoSmithKline. 66. Trivedi MH, Rush AJ, Wisniewski SR, Nierenberg AA, Warden D, Ritz L et al. Evaluation of outcomes with citalopram for

depression using measurement-based care in STAR*D: implications for clinical practice. Am J Psychiatry. 2006; 163: 28-40.67. Thase ME, Shelton RC, Khan A. Treatment with venlafaxine extended release after SSRI nonresponse or intolerance: a

randomized comparison of standard- and higher-dosing strategies. J Clin Psychopharmacol. 2006; 26: 250-258.68. Rudolph RL, Fabre LF, Feighner JP, Rickels K, Entsuah R, Derivan AT. A randomized, placebo-controlled, dose-response

trial of venlafaxine hydrochloride in the treatment of major depression. J Clin Psychiatry. 1998; 59: 116-122.69. Parsey RV, Kegeles LS, Hwang DR, Simpson N, Abi-Dargham A, Mawlawi O et al. In vivo quantification of brain serotonin

transporters in humans using [11C]McN 5652. J Nucl Med. 2000; 41: 1465-1477.70. Laasonen-Balk T, Viinamaki H, Kuikka JT, Husso-Saastamoinen M, Lehtonen J, Tiihonen J. 123I-beta-CIT binding and

recovery from depression. A six-month follow-up study. Eur Arch Psychiatry Clin Neurosci. 2004; 254: 152-155.71. Pirker W, Asenbaum S, Kasper S, Walter H, Angelberger P, Koch G et al. β-CIT SPECT demonstrates blockade of 5HT-

uptake sites by citalopram in the human brain in vivo. J Neural Transm Gen Sect. 1995; 100: 247-256.72. Shang Y, Gibbs MA, Marek GJ, Stiger T, Burstein AH, Marek K et al. Displacement of serotonin and dopamine transporters

by venlafaxine extended release capsule at steady state: a [123I]2β-carbomethoxy-3β-(4-iodophenyl)-tropane singlephoton emission computed tomography imaging study. J Clin Psychopharmacol. 2007; 27: 71-75.

73. Takano A, Suhara T, Ichimiya T, Yasuno F, Suzuki K. Time course of in vivo 5-HTT transporter occupancy by fluvoxamine. JClin Psychopharmacol. 2006; 26: 188-191.

74. Tauscher J, Pirker W, de Zwaan M, Asenbaum S, Brucke T, Kasper S. In vivo visualization of serotonin transporters in thehuman brain during fluoxetine treatment. Eur Neuropsychopharmacol. 1999; 9: 177-179.

75. Viinamäki H, Kuikka JT, Tiihonen J, Lehtonen J. Change in monoamine transporter density related to clinical recovery: acase-control study. Nord J Psychiatry. 1998; 52: 39-44.

Chap

ter

Chapter 9

9

188

Supplementary appendix: methods of paroxetine serum concentrationsand results.

Paroxetine serum concentrationsPSC were determined in May 2007 using a previously validated HPLC-MS/MS method.A ThermoFinnigan (San Jose, CA, USA) Surveyor HPLC system was used, equipped with a Surveyorautosampler. Separation was carried out on a Agilent Eclipse XDB-CN column, 100 x 2.1 mm,particle size 3,5 µm (Agilent, Amstelveen, The Netherlands). A mixture of 60% (v/v) 2 mMammonium acetate (pH 3.2) and 40% (v/v) acetonitrile was used as mobile phase.Paroxetine wasextracted from serumby liquid-liquid extraction. Briefly, 500μL of serum samples were mixed withsodium carbonate solution (0.5M). An internal standard solution (disopyramide 20 µg/l) wasadded and the mixture was vortexed for 5 seconds. Thereafter, 1-chlorobutane/ethylacetatemixture (60/40; v/v) was added and vortexed for 10 minutes. After centrifugation for 10 minutes(3000g),the upper layer was transferred to a second vial, and evaporated under a nitrogenatmosphere at 40 °C. After addition of methanol the solution was vortexed and transferred ontothe HPLC-MS/MS system.Selected reaction monitoring (SRM) was used for drug quantification(SRM paroxetine 330.1/192.0 and SRM disopyramide 340.2/239.0).Calibration curves wereconstructed by adding known amounts of paroxetine to blank serum (5-75 µg/L range). The lowerlimit of quantification was 5 µg/L, which was 50% below the lower end of the therapeutic range ofparoxetine in serum (10-75 µg/L range). The lower limit of detection was 0.3 µg/L.

189


9

Tabl

e S9

.1. S

tudi

es o

f SER

T oc

cupa

ncy

afte

r tre

atm

ent w

ith s

erot

oner

gic

antid

epre

ssan

ts

Aut

hor &

Dat

e (r

efer

ence

) Re

sults

Popu

latio

n (m

ean

age)

Sex

N

Follo

w-u

p D

esig

n In

terv

entio

n / c

ontr

ol

SER

T Im

ag

ing

and

RoI

Resu

ltsRe

mar

ks

Cata

fau

2006

28D

rug-

free

M

DD

pts

(36

±11)

M

+ F

10Bs

l, 4-

6 w

Pa

roxe

tine

20 m

g [12

3 I]-A

DA

M

SPEC

T, 4

h M

ID, T

hal,

Stria

tum

, Ce

r

In 9

pat

ient

s a

2nd s

can

was

mad

e.

Occ

upan

cies

wer

e: M

ID 6

6.4

±9.5

%, T

hal

63.0

±9.

6%, S

tria

t 61.3

±10

.8%

Cons

ider

able

var

iatio

n in

occ

upan

cy o

bser

ved.

A

utho

rs s

ugge

st th

at S

ERT

occu

panc

y is

pro

babl

y no

t rel

ated

to S

SRI r

espo

nse.

Cava

nagh

20

0631

MD

D p

ts tr

eate

d 24

N

/A m

ean

AD

use

26

.3-12

.8m

Varie

ty o

f dru

gs u

sed:

mon

othe

rapy

(n=

17):

ven

lafa

xine

(75-

300

mg)

, SSR

Is (2

0-60

mg)

, tric

yclic

(150

mg)

, Mirt

azap

ine

(30

mg)

; com

bina

tion(

n= 7

) co

mbi

natio

ns o

f 2 a

ntid

epre

ssan

ts,

addi

tion

of li

thiu

m, v

alpr

oate

, ca

rnba

maz

epin

e, T

3 or

ant

ipsy

chot

ics

; do

sage

s un

chan

ged

for ≥

2wee

ks

[123 I]β-

CIT

SPEC

T, 3

.3h

Br

St/D

ienc

, O

cc

No

occu

panc

y pe

rcen

tage

s av

aila

ble.

No

sign

ific

ant d

ifere

nce

in S

ERT

resi

dual

ac

tivity

bet

wee

n re

spon

ders

and

non

-re

spon

ders

. Wid

e ra

nge

of S

ERT

avai

labi

lity

Patie

nts

wer

e no

t sca

nned

bef

ore

trea

tmen

t (bs

l),

inst

ead

a co

mpa

rison

of b

indi

ng p

oten

tials

was

m

ade.

Hig

hly

hete

roge

neou

s gr

oup

cons

ider

ing

drug

trea

tmen

t.

Erla

ndss

on

2005

23H

ealth

y vo

lunt

eers

(3

2; 2

5-52

) M

16

Bsl,

2-7d

Ci

talo

pram

at d

iffer

ent d

osag

es (1

0-60

m

g) fo

r diff

eren

t dur

atio

ns (2

-7 d

ays)

. Vo

lunt

eers

wer

e ra

ndom

ized

to o

ne o

f 7

diff

eren

t dos

ing

regi

men

s

[123 I]

-AD

AM

SP

ECT,

4h

MID

, Tha

l, St

riat,

Cer

No

mea

n oc

cupa

ncy

for s

epar

ate

dosi

ng

regi

men

s gi

ven.

Max

imum

MID

oc

cupa

ncy

84%

Stud

y ai

ms

at fi

ndin

g th

e be

st ti

me

fram

e (s

igle

sc

an p

roto

col)

for s

cann

ing

with

[123 I]

-AD

AM

. Co

nsid

erab

le v

aria

tion

in re

latio

n oc

cupa

ncy

and

bloo

d-le

vels

of c

italo

pram

obs

erve

d.

Her

old

2006

29D

rug-

free

M

DD

pts

(42

±12)

M

+ F

21

Bsl,

7d

Cita

lopr

am 10

mg,

sca

ns w

ere

mad

e 6-

7 ho

urs

afte

r las

t app

licat

ion

of o

ral d

ose

[123 I]

-AD

AM

SP

ECT,

4h

MID

, Cer

In 13

pat

ient

s a

2nd s

can

was

mad

e.M

ean

MID

occ

upan

cy w

as 6

1% (r

ange

37-

88%)

Co

nsid

erab

le v

aria

tion

in o

ccup

ancy

obs

erve

d. N

o co

rrel

atio

n of

occ

upan

cy a

nd d

ecre

ase

in

depr

essi

on s

ever

ity.

Hilt

unen

1998

24H

ealth

y vo

lunt

eers

(3

0; 2

5-35

) M

+ F

5 Bs

l 1 p

atie

nt re

ceiv

ed c

italo

pram

30

mg

3hr

prio

r to

inje

ctio

n; 1

patie

nt re

ceiv

ed

cita

lopr

am 2

0 m

g, 1

patie

nt v

enla

faxi

ne

37.5

mg1

hr a

fter

inje

ctio

n.

[123 I]

nor-β

-CI

T SP

ECT,

0-

24h

a.o

. aC

G, M

ID,

Thal

, Ba

sGan

g,

Cer

Aft

er c

italo

pram

30

mg

3hr p

rior t

o in

ject

ion

spec

ific

bind

ing

in th

e m

idbr

ain

was

52%

less

than

in u

ntre

ated

sub

ject

s.

For v

enla

faxi

ne a

nd c

italo

pram

20

mg

no

data

giv

en.

Stud

y ai

ms

at e

stab

lishi

ng tr

acer

-pro

pert

ies

of

[123 I]

nor-β-

CIT.

Spe

cific

bin

ding

of [

123 I]

nor-β

-CIT

is

33%

high

er c

ompa

red

to [12

3 I]β-

CIT

(in o

ther

su

bjec

ts).

Kent

200

218Pa

tient

s w

ith

Soci

al F

obia

(3

5 ±1

1)

M +

F

5 Bs

l, 3-

6m

Paro

xetin

e 20

-40

mg

(titr

ated

by

clin

ical

re

spon

se) (

3-6m

) [11

C]M

cN56

52 P

ET M

ID,

Thal

, Str

iat,

H

ip, A

myg

, Ci

ngA

, Cer

Occ

by

paro

xetin

e: M

ID 9

8±3%

, Tha

l 81

±6%,

Str

iat 7

5±7%

, Hip

92±

12%,

Am

yg

94±7

%, C

ingA

81±

22%

Hig

her o

ccup

ancy

of S

ERT

than

aft

er a

cute

ch

alle

nge

with

hig

her d

oses

(see

69).

Une

xpec

ted

decr

ease

in V

T in

Cer

ebel

lum

and

whi

te m

atte

r,

indi

cativ

e of

dec

reas

e in

non

-spe

cifi

c di

strib

utio

n vo

lum

e of

liga

nd a

fter

chr

onic

trea

tmen

t with

pa

roxe

tine

190

Chapter 9

Chap

ter

9

Klei

n 20

0626

Hea

lthy

volu

ntee

rs

(26.

8 ±4

.3)

M

25

Bsl,

6h

Cita

lopr

am (1

0 m

g or

20

mg)

, es

cita

lopr

am (5

mg,

10 m

g, 2

0 m

g).

Volu

ntee

rs w

ere

rand

omiz

ed to

one

of

5 di

ffer

ent d

rug/

dosi

ng re

gim

ens

[123 I]

-AD

AM

SP

ECT,

4h

MID

/Tha

l, Ce

r

Occ

upan

cy o

f 60

±6%,

64

±6%

and

75 ±

5%

for 5

, 10

and

20 m

g es

cita

lopr

am. F

or 10

an

d 20

mg

cita

lopr

am o

ccup

ancy

65

±10%

an

d 70

±6%

Stud

y ai

ms

to d

eter

min

e ac

ute

occu

panc

y of

SER

T af

ter s

ingl

e do

se d

rug

adm

inis

trat

ion

of c

itaol

pram

or

esc

italo

pram

Klei

n 20

0725

Hea

lthy

volu

ntee

rs(2

8 ±7

) M

15

Bsl,

10d

Cita

lopr

am 2

0 m

g, e

scita

lopr

am 10

mg.

Vo

lunt

eers

wer

e ra

ndom

ized

to o

ne o

f th

e do

sing

gro

ups

[123 I]

-AD

AM

SP

ECT,

4h,

52

h M

ID/

Thal

, Cer

Occ

upan

cy e

scita

lopr

am 8

1.5 ±

5.4%

, ci

talo

pram

64.

0 ±1

2.7%

In

crea

sed

occu

panc

y ob

serv

ed fo

r esc

itaol

pram

, al

so re

pres

ente

d in

Em

ax c

urve

s. E

max

cur

ves

cons

truc

ted

with

dat

a fo

r 4h

and

52h

post

inje

ctio

n sc

ans

in 1

mod

el

Kuga

ya 2

003,

20

0446

;48

1. Vo

lunt

eers

(3

7.4

±14.

3)

M +

F

2. P

atie

nts

with

M

DD

(4

6.4

±7.6

) M

+ F

9 10

Bsl,

8d,

16d

Bsl,

1-3w

, 6w

Cita

lopr

am 4

0 m

g (8

d), C

ita lo

pra

m 4

0 m

g +

Bupr

o pi

on 10

0 m

g (8

-16d)

Pa

roxe

tine

20 m

g (6

w)

[123 I]β-

CIT

SPEC

T, 2

4h

BrSt

, Die

nc,

Cer

Occ

by

cita

lopr

am a

t 8d:

BrS

t 51.

4%, D

ienc

39

.4%;

no

sign

. cha

nge

ther

eaft

er O

cc b

y pa

roxe

tine

at 1-

3w: B

rSt 3

6.5%

;, D

ienc

29

.1% &

at 6

w B

rSt 3

2.6%

, Die

nc 2

3.4%

Incr

ease

d D

AT-b

indi

ng in

Str

iatu

m d

urin

g pr

olon

ged

SSR

I tre

atm

ent.

Bup

ropr

ion

(100

-200

m

g) o

nly

has

no in

fluen

ce o

n D

AT-b

indi

ng in

st

riatu

m In

stud

y 2

high

er B

sl D

ienc

SER

T av

aila

bilit

y pr

edic

ted

bett

er tr

eatm

ent r

espo

nse

at w

eek

6.

Laas

onen

-Bal

k 20

0470

Patie

nts

with

M

DD

(4

2.7

±?)

M +

F

18

Bsl,

6m

(≥1m

dru

g-fr

ee)

Ope

n tr

eatm

ent w

ith v

ario

us

antid

epre

ssan

ts (n

= 3)

and

/or

benz

odia

zepi

nes

or s

uppo

rtiv

e co

unse

lling

[123 I]β-

CIT

SPEC

T, 1h

&

21-2

4h M

ID,

Cer

Bind

ing

in M

ID w

as s

ign.

incr

ease

d in

re

spon

ders

to tr

eatm

ent (

corr

ecte

d fo

r ag

e an

d se

x)

Low

dep

ress

ion

seve

rity

at B

sl (H

DRS

13.9

±6.

7);

only

thre

e pa

tient

s re

ceiv

ed a

ntid

epre

ssan

ts, a

nd

six

benz

odia

zepi

nes o

nly.

Incr

ease

d SE

RT b

indi

ng in

re

pond

ers

sugg

ests

dec

reas

ed m

idbr

ain

dens

ity o

f SE

RT d

urin

g M

DD

Mey

er 2

00119

1. Pa

tient

s w

ith

MD

D ,

HD

RS ≥1

6 R

x-fr

ee (3

2 ±8

) M

+ F

2.

Hea

lthy

volu

ntee

rs

(age

-mat

ched

) M

+ F

12

17

Bsl,

4w

Bsl,

4w

1. O

pen

trea

tmen

t with

par

oxet

ine

20

mg

(n=

7), 1

0 m

g (n

= 1)

or c

italo

pram

20

mg

(n=

4) 2

. No

trea

tmen

t

[11C]

DA

SB

PET,

0-1

½ h

St

riat,

Tha

l, Ci

ngA

, PFC

, Ce

r

1. O

cc in

Str

iat a

fter

par

oxet

ine

20 m

g:

83%

±5, a

nd a

fter

cita

lopr

am 2

0 m

g: 7

7%

±10.

Occ

in T

hal a

fter

par

oxet

ine

20 m

g:

75-7

8% ±

8, a

nd a

fter

cita

lopr

am 2

0 m

g:

65-7

0% ±

7-15

. O

cc in

Cin

gA a

fter

pa

roxe

tine

20 m

g: 7

6-77

% ±1

5, a

nd a

fter

ci

talo

pram

20

mg:

77-

79%

±29.

No

rela

tions

hip

was

foun

d be

twee

n H

DRS

-sco

re

and

occu

panc

y le

vel i

n an

y R

oI.

Stria

tal O

cc

incr

ease

d w

ith h

ighe

r ser

um-le

vels

of p

arox

etin

e,

with

app

. 85%

occ

at s

erum

leve

ls o

f 28 μg

/l.

Ref

eren

ces

to o

ther

stu

dies

that

des

crib

e a

hype

rbol

ic re

latio

nshi

p of

ser

um le

vel a

nd

occu

panc

y. T

est-

rete

st in

hea

lthy

subj

ects

show

ed a

m

ean

diff

eren

ce o

f -3.

7% ±

3.7

(ran

ge -8

– 2

%) o

ver 4

w

eeks

. Mea

n ab

solu

te d

iffer

ence

was

10.9

% ±2

.9.

Tabl

e S9

.1. S

tudi

es o

f SER

T oc

cupa

ncy

afte

r tre

atm

ent w

ith s

erot

oner

gic

antid

epre

ssan

ts (C

ontin

ued)

Aut

hor &

Dat

e (r

efer

ence

) Re

sults

Popu

latio

n (m

ean

age)

Sex

N

Follo

w-u

p D

esig

n In

terv

entio

n / c

ontr

ol

SER

T Im

ag

ing

and

RoI

Resu

ltsRe

mar

ks

191


9

Mey

er 2

00420

1. Pa

tient

s w

ith

MD

D

2. P

atie

nts

with

M

DD

and

co-

mor

bid

anxi

ety

diso

rder

3.

Hea

lthy

subj

ects

(2

0-50

) M

+ F

29

16

37

Bsl,

4w

MD

D: c

italo

pram

20-

40 m

g, fl

uoxe

tine

20 m

g, s

ertr

alin

e 50

-100

mg,

par

oxet

ine

20 m

g, v

enla

faxi

ne 7

5 m

g

MD

D+a

nxie

ty: c

italo

pram

40-

60 m

g,

fluox

etin

e 40

-60

mg,

ser

tral

ine

150-

200

mg,

par

oxet

ine

40-6

0 m

g, v

enla

faxi

ne

150-

225

mg

Hea

lthy

subj

ects

: cita

lopr

am

1-10

mg,

fluo

xetin

e 1-1

0 m

g, s

ertr

alin

e 5-

25 m

g, p

arox

etin

e 5-

10 m

g, v

enla

faxi

ne

2.4-

37.5

mg

[11C]

DA

SB

PET,

0-1

½ h

St

riat,

Tha

l, Ci

ngA

, PFC

, Cu

neus

, M

ID, C

er

Mea

n st

riata

l occ

was

81.

4 ±7

.2%

for

cita

lopr

am (2

0-40

mg)

,76.

2 ±7

.5 fo

r flu

oxet

ine

(20

mg)

, 85.

0 ±7

.0%

for

sert

ralin

e (5

0-10

0), 8

4.5

±6.0

% fo

r pa

roxe

tine

and

83.7

±2.

4 fo

r ven

lafa

xine

(7

5 m

g).

Occ

for T

hal:

72.3

±7.

6%, 6

9.1

±9.4

%, 7

6.8

±4.3

%, 7

4.7

±15.

7% a

nd 7

1.3

±10.

2% re

sp. O

cc fo

r MID

: 87.

5 ±7

.7%,

82.

3 ±9

.3%,

91.8

±4.

0%, 9

3.4

±7.7

% an

d 91

.0

±7.7

% re

sp.

12 p

atie

nst a

lso

desc

ribed

in M

eyer

et a

l. 20

0119

2

MD

D-p

atie

nts

rece

ived

sub

ther

apeu

tical

dos

es in

‘H

ealth

y’ g

roup

; in

5 pa

tient

s of

MD

D g

roup

do

sage

s w

ere

seco

ndar

ily in

crea

sed

and

also

an

alys

ed (t

wic

e) in

MD

D +

anx

iety

gro

up. N

o re

latio

n be

twee

n st

riata

l occ

upan

cy a

nd c

linic

al

rem

issi

on o

r per

cent

age

chan

ge in

Ham

ilton

de

pres

sion

sco

res.

Pars

ey 2

00069

Hea

lthy

volu

ntee

rs

(38

±12)

M

2 Bs

l 2

of 6

vol

unte

ers

rece

ived

par

oxet

ine

60m

or8

0 m

g pr

ior t

o sc

anni

ng

(blo

ckin

g st

udy)

; blo

ckin

g co

mpa

red

to

PET

befo

re u

sing

par

oxet

ine

[11C]

McN

5652

PET

, 0-2

½

h M

ID, T

hal,

Hip

, Am

yg,

Cing

A, m

TL,

Occ

by

60 m

g pa

roxe

tine

‘pre

trea

tmen

t’

(n=

1): A

myg

64.

8%, H

ip 4

6.0%

, Tha

l 38.

4%,

Mid

br 8

3.9%

, Cin

gA 2

6.4%

Mai

nstu

dy (n

= 6)

aim

s to

qua

ntify

cha

ract

eris

tics

and

prot

ocol

for [

11C]

McN

5652

.

Pars

ey 2

00621

Hea

lthy

volu

ntee

rs17

Bs

l, 4-

6d

Sert

ralin

e 25

mg,

50

mg

and

100

mg

(4

days

at d

esig

nate

d do

sage

) [11

C]D

ASB

PE

T, 0

-1½

h

Stria

t, T

hal,

Cing

, DL/

O/

M P

FC,

Am

yg,H

ip,

Insu

la,

MID

,Occ

, Ce

r

Aver

age

acro

ss 15

RoI

s: m

ax o

ccup

ancy

10

6.8

±8.3

% ra

nge

126.

9 ±3

5.9

in O

PFC

to

79.3

±10

.4%

in T

hal

Wid

e ra

nge

of R

oIs

take

n. S

tudy

als

o ex

amin

es

SERT

ava

ilabi

lity

in c

ereb

ellu

m a

nd e

xam

ines

its

role

as

refe

renc

e. S

peci

fic S

ERT

bind

ing

was

foun

d hi

gher

in th

e ce

rebe

llar v

erm

is c

ompa

red

to th

e ce

rebe

llar g

rey.

In c

oncl

usio

n no

idea

l ref

eren

ce

regi

on fo

r in

vivo

SER

T st

udie

s ca

n be

re

com

men

ded.

Pirk

er 19

9571

1. Pa

tient

s w

ith

MD

D

(44

26-7

1)

M +

F2.

‘Hea

lthy’

co

ntro

ls

(42.

3 24

-70)

M

+ F

13

11

1 sca

n so

me

whe

re in

tr

eatm

ent

Cita

lopr

am 2

0 m

g (n

= 5)

, 40

mg

(n=

6)

and

60 m

g (n

= 1)

for a

t lea

st o

ne w

eek;

on

e un

trea

ted

patie

nt.

[123 I]β-

CIT

SPEC

T, 2

, 4,

16, 2

0, 2

4h

Thal

, hy

poTh

al,

MID

, Pon

s,

Cer

No

diff

eren

ce in

bin

ding

in S

tria

t be

twee

n pa

tient

s an

d co

ntro

ls.

Cita

lopr

am p

atie

nts

show

ed s

ign.

de

crea

se in

Tha

l, hy

poTh

al, M

ID

com

pare

d to

con

trol

s. N

o di

ffer

ence

in

bind

ing

betw

een

patie

nts

with

ci

talo

pram

20

mg

or 4

0 m

g.

Stud

y in

clud

ed p

atie

nts

with

bip

olar

dis

orde

r (n=

1)

and

conv

ersi

on d

isor

der (

n= 1)

. Con

trol

gro

up

incl

uded

4 p

atie

nts

with

per

iphe

ral n

euro

logi

cal

dise

ase.

Cita

lopr

am-d

osin

g w

as n

ot b

linde

d an

d hi

gher

dos

es m

ay b

e th

e re

sult

of a

n un

clea

r se

lect

ion

proc

ess.

Spe

cific

bou

nd in

Tha

l, hy

poTh

al

and

Mid

br re

gion

pea

ked

at 4

h; in

Str

iat t

his

was

re

ache

d af

ter 1

6-24

h.

Suha

ra 2

00322

1. H

ealth

y vo

lunt

eers

(2

2.0

±2.3

) M

2.

Pat

ient

s w

ith

MD

D(3

5.7

±12.

1)

27

10

1 sca

n 1.

Clo

mip

ram

ine

5-50

mg,

fluv

oxam

ine

12.5

-50

mg

as s

ingl

e ad

min

istr

atio

n 2.

Cl

omip

ram

ine

20-2

50 m

g, fl

uvox

amin

e 25

-200

mg

as lo

ng-t

erm

adm

inis

trat

ion

[11C]

McN

5652

PET

, 0-1

½

h Th

al

1. Cl

omip

ram

ine

occ ≥8

3.9-

100%

at

dosa

ges

5-50

mg;

fluv

oxam

ine

occ ≥7

.7-

87.7

% at

dos

ages

12.5

-50

mg.

2.

Clom

ipra

min

e oc

c ≥6

1.3-10

0% a

t dos

ages

20

-250

mg;

fluv

oxam

ine

occ ≥7

6.6-

93.6

% at

dos

ages

25-

200

mg.

Hyp

erbo

lic re

latio

nshi

p be

twee

n or

al d

ose,

pla

sma

conc

entr

atio

n an

d oc

c. fo

r clo

mip

ram

ine

and

fluvo

xam

ine.

Pat

ient

dat

a su

gges

t tha

t the

du

ratio

n of

trea

tmen

t with

fluv

oxam

ine

may

be

corr

elat

ed w

ith o

cc. T

his

does

not

hol

d fo

r cl

omip

ram

ine.

Tabl

e S9

.1. S

tudi

es o

f SER

T oc

cupa

ncy

afte

r tre

atm

ent w

ith s

erot

oner

gic

antid

epre

ssan

ts (C

ontin

ued)

Aut

hor &

Dat

e (r

efer

ence

) Re

sults

Popu

latio

n (m

ean

age)

Sex

N

Follo

w-u

p D

esig

n In

terv

entio

n / c

ontr

ol

SER

T Im

ag

ing

and

RoI

Resu

ltsRe

mar

ks

192

Chapter 9

Chap

ter

9

Shan

g 20

0772

Hea

lthy

cont

rols

(2

3.6

± 6.

3)

M +

F ?

8 Bs

l, 9d

Ve

nlaf

axin

e 15

0 m

g (4

day

s st

able

dos

e)

[123 I]β-

CIT

SPEC

T, 2

3h

Die

nc, M

ID,

Stria

t, C

er

Occ

in D

ienc

: 52.

5 ±

4.7%

, in

MID

: 55.

7 ±4

.5%

Stud

y al

so in

vest

igat

es in

fluen

ce o

f ven

lafa

xine

on

stria

tal D

AT a

vaila

bilit

y, w

hich

incr

ease

s 10

.7 ±

3.0

%

Taka

no 2

006a

73H

ealth

y co

ntro

ls

(24.

3 ±4

.8)

M

6 Bs

l, 5h

, 26

h, 5

3h

Fluv

oxam

ine

50 m

g on

ce

[11C]

DA

SB

PET,

0-1

½ h

PF

C, T

hal,

Stria

t,

Am

yg, H

ip,

Cer

Mea

n oc

c of

the

5 re

gion

s of

6 s

ubje

cts

wer

e 72

.9%

± 4.

9% a

t 5 h

ours

, 50.

3% ±

11.0

% at

26

hour

s, a

nd 2

4.7%

± 15

.3%

at 5

3 ho

urs.

5h

occ

: Tha

l 71.8

±3.

5%, A

myg

71.

6 ±1

2.8%

; 53

h oc

c: T

hal 2

4.2

±12.

3%, A

myg

27.

4 ±1

8.2%

Stud

y in

vest

igat

es ti

me-

cour

se o

f occ

upan

cy o

f SE

RT

afte

r one

(low

) dos

e of

fluv

oxam

ine.

In E

max

m

odel

dep

ende

nt m

easu

rem

ents

at 3

tim

epoi

nts

are

used

as

inde

pend

ent d

ata-

poin

ts

Taka

no 2

006b

27H

ealth

y co

ntro

ls

(24.

1 ±2.

4)

M

15

Bsl,

1. 6

h,

25h,

53h

2.

7d, 9

d, 10

d

1. si

ngle

dul

oxet

ine

dose

(5, 2

0, 4

0, 6

0 m

g) (n

= 12

) 2.

dul

oxet

ine

60 m

g fo

r 7 d

ays

then

st

oppe

d (n

= 3)

[11C]

DA

SB

PET,

0-1

½ h

Th

al, C

er

1. Th

al o

cc 3

5.3-

86.5

% at

incr

easi

ng

dosa

ges:

at 6

hrs

occ:

43.

6±8.

8% (5

mg)

, 71

.3±5

.3%

(20

mg)

, 80.

6±4.

8% (4

0 m

g),

81.8

±4.3

% (6

0 m

g) 2

. Tha

l occ

: 84.

3+2.

8%

(7d)

, 71.9

+2.6

% (9

d), 4

7.1+

3.7%

(11d

)

Stud

y co

nclu

des

that

mor

e th

an 4

0 m

g is

requ

ired

to a

ttai

n 80

% oc

cupa

ncy.

Inst

ead

of p

lasm

a co

ncen

trat

ions

that

dec

reas

e in

a o

ne-e

xpon

entia

l w

ay a

fter

stop

ping

, occ

upan

cy a

ppea

rs to

dec

reas

e lin

early

and

slo

wer

than

pla

sma

conc

entr

atio

ns.

Taus

cher

1999

74Pa

tient

with

BN

an

d M

DD

(1

8)

F

1 1 s

can

Fluo

xetin

e 60

mg.

Unt

il 1 w

eek

befo

re

scan

add

ition

with

dox

epin

e 75

mg.

[12

3 I]β-

CIT

SPEC

T, 4

, 24

h Th

al,

hypo

Thal

, St

riat

Com

pare

d to

con

trol

s an

occ

in T

hal a

nd

hypo

Thal

of a

pp. 4

1% w

as e

xpec

ted

by

fluox

etin

e. H

ighe

r Str

iat D

AT th

an in

co

ntro

ls; i

n ac

cord

ance

with

oth

er M

DD

-pa

tient

s

Met

hodo

logi

cally

poo

r. R

esul

ts c

ompa

red

to a

hi

stor

ic c

ontr

ol g

roup

(n=

13, M

+ F

; age

46

±20;

age

ra

nge

24-7

1).

Viin

amäk

i 19

9875

1. Pa

tient

s w

ith

Pers

Dis

(25

&

34) M

2.

Hea

lthy

con

trol

s (2

5±3

& 3

4±1)

M

2 10

Bsl,

1-1.5

y Ps

ycho

dyna

mic

psy

chot

hera

py o

r no

trea

tmen

t

No

trea

tmen

t

[123 I]β-

CIT

SPEC

T, 1h

&

21h

mPF

C, O

C,

MID

, Tha

l, St

riat

(Suc

cesf

ully

) tre

ated

pat

ient

1: in

crea

se in

bi

ndin

g in

MID

: 34%

, mPF

C 43

%, T

hal 3

1%;

with

no

chan

ges

in u

ntre

ated

pat

ient

2;

leve

ls o

f bin

ding

pos

t-tr

eatm

ent

appr

oach

leve

ls o

f con

trol

s

Patie

nts

wer

e bo

th a

busi

ng a

lcoh

ol; p

atie

nt 1

abst

aine

d du

ring

trea

tmen

t; p

atie

nt 2

did

not

. Pa

tient

1 ha

d lo

w, a

nd p

atie

nt 2

had

hig

h H

DRS

-sc

ores

at b

asel

ine

and

follo

w-u

p. C

ontr

ols

used

al

coho

l occ

asio

nally

, no

data

on

HD

RS

are

give

n.

Met

hodo

logi

cally

poo

r, s

tudy

is d

escr

ibed

poo

rly.

Voin

esko

s 20

0730

Patie

nts

with

M

DD

(36

±9)

Hea

lthy

cont

rols

(3

2 ±9

) M+F

12

>4w

Bsl

Ve

nlaf

axin

e 22

5 m

g, 4

50 m

g, s

ertr

alin

e 15

0 m

g, 2

00 m

g, c

italo

pram

60

mg,

80

mg

[11C]

DA

SB

PET,

0-1

½ h

St

riat,

MID

, Th

al,

PFC,

ACg

, Ce

r

Occ

Str

iat:

85.

8 ±3

.4%

(ven

lafa

xine

), 8

5.8

±2.3

(ser

tral

ine)

and

85.

4±2.

0 (c

italo

pram

) Occ

MID

: 99.

5 ±4

.1%

(ven

lafa

xine

), 9

8.2

±3.2

(ser

tral

ine)

and

95

.7±1

.4 (c

italo

pram

) Occ

Tha

l: 77

.6 ±

4.7%

(v

enla

faxi

ne),

76.

3 ±7

.9 (s

ertr

alin

e) a

nd

82.2

±3.

6 (c

italo

pram

)

No

base

line

scan

s w

ere

perf

orm

ed in

pat

ient

s.

Inst

ead

as re

fere

nce

base

line

scan

s of

con

trol

s w

ere

used

to d

eter

min

e oc

cupa

ncy.

Abb

revi

atio

ns: A

myg

= A

myg

dala

, aCG

= a

nter

ior C

ingu

late

Gyr

us, B

asG

ang

= Ba

sal G

angl

ia, B

N =

Bul

imia

Ner

vosa

, BrS

t = B

rain

Stem

, Bsl

= B

asel

ine,

Cin

gA =

Cin

gula

te A

nter

ior,

d =

day

s, D

AT =

Dop

amin

e Tr

ansp

orte

r, D

ienc

= D

ienc

epha

lon,

h =

hou

rs, H

ip =

Hip

poca

mpu

s, H

DRS

= H

amilt

on D

epre

ssio

n R

atin

g Sc

ale,

hyp

oTha

l = H

ypot

hala

mus

, m =

mon

ths,

MD

D =

Maj

or d

epre

ssiv

e di

sord

er, m

TL =

med

ial

tem

pora

l lob

e, M

ID =

Mid

brai

n, m

PFC

= m

edia

l Pre

Fron

tal C

orte

x, O

C =

Occ

ipita

l Cor

tex,

Occ

= O

ccup

ancy

,Per

sDis

= P

erso

nalit

y D

isor

der,

PFC=

Pre

Fron

tal C

orte

x, R

oI =

Reg

ions

of I

nter

est,

Str

iat =

St

riatu

m, T

hal =

Tha

lam

us, w

= w

eeks

, y =

yea

rs.* m

ost s

tudi

es u

sed

antid

epre

ssan

tsSE

RT

occ.

= S

ERT

occu

panc

y re

lativ

e to

non

-spe

cifi

c bi

ndin

g (c

ereb

ellu

m)

Tabl

e S9

.1. S

tudi

es o

f SER

T oc

cupa

ncy

afte

r tre

atm

ent w

ith s

erot

oner

gic

antid

epre

ssan

ts (C

ontin

ued)

Aut

hor &

Dat

e (r

efer

ence

) Re

sults

Popu

latio

n (m

ean

age)

Sex

N

Follo

w-u

p D

esig

n In

terv

entio

n / c

ontr

ol

SER

T Im

ag

ing

and

RoI

Resu

ltsRe

mar

ks


9

193

Table S9.2. Drop-outs and emerging adverse effects reported more than once between T0 and T1 (ITT).

Paroxetine DE (n = 30) Placebo DE (n = 30) p

Drop-outs

Overall 1/30 (3.3%) 8/30 (26.7%) 0.026

Inefficacy 0/30 (0.0%) 3/30 (10.0%) 0.237

Adverse effects 0/30 (0.0%) 4/30 (13.3%) 0.112

Adverse effect rates Paroxetine DE (n = 30) Placebo DE (n = 27)

% all* T0 ≥2x RP End RP Rec. AE† T0 ≥2x RP End RP Rec. AE†

Headache 69.0 46.7 70.0 46.7 5/10 (50.0) 50.0 67.9 50.0 5/6 (83.3)

Sleepiness 67.2 63.3 70.0 56.7 6/7 (85.7) 46.4 64.3 46.4 7/8 (87.5)

Sweating 63.8 56.7 80.0‡ 63.3 3/5 (60.0) 39.3 46.4‡ 42.9 3/5 (60.0)

Dry mouth 63.8 66.7 63.4 50 1/5 (20.0) 64.3 64.3 46.4 2/4 (50.0)

Orthostatic hypotension 46.6 50.0 53.3 50.0 3/5 (60.0) 35.7 39.3 28.6 6/8 (75.0)

Constipation 44.8 36.7 50.0 43.3 2/3 (66.7) 28.6 39.3 21.4 4/7 (57.1)

Libido 41.4 36.7 56.7‡ 56.7 2/2 (100.0) 28.6 25.0‡ 39.3 1/4 (25.0)

Agitation 41.4 33.3 50.0 33.3 6/9 (66.7) 21.4 32.1 17.9 5/10 (50.0

Appetite 39.7 46.7 46.7 43.3‡ 3/7 (42.9) 35.7 32.1 14.3‡ 3/8 (37.5)

Blurry vision 37.9 40.0 43.3 43.3 0/4 (0.0) 14.3 32.1 25.0 2/4 (50.0)

Tremor 36.2 36.7 46.7 40.0‡ 4/7 (57.1) 32.1 25.0 14.3‡ 3/5 (60.0)

Dizziness 34.5 36.7 46.7 26.7 8/11 (72.7) 28.6 21.4 17.9 2/7 (28.6)

Concentration 34.5 26.7 43.3 30.0 5/8 (62.5) 28.6 25.0 25.0 4/5 (80.0)

Paraesthesia 32.8 23.3 36.7 36.7 4/5 (80.0) 32.1 28.6 25.0 2/3 (66.7)

Sleep disturbances 31.0 23.3 43.3‡ 43.3§ 8/9 (88.9) 35.7 17.9‡ 10.7§ 3/8 (37.5)

Nausea 29.3 33.3 33.3 33.3 5/11 (45.5) 28.6 25.0 14.3 4/7 (57.1)

Weakness 29.3 16.7 36.7 26.7 3/6 (50.0) 17.9 21.4 17.9 2/7 (28.6)

Diarrhea 24.1 16.7 23.3 13.3 4/5 (80.0) 25.0 25.0 17.9 3/4 (75.0)

Urinary retention 24.1 13.3 36.7‡ 23.3 2/3 (66.7) 14.3 10.7‡ 7.1 0/1 (0.0)

Other sexual dysfunction

22.8 16.7 26.7 30.0 3/3 (100) 17.9 18.5 17.9 0/0

Palpitations 22.4 16.7 26.7 16.7 3/7 (42.9) 10.7 17.9 17.9 4/9 (44.4)

Rash 10.3 6.7 6.7 3.3 1/4 (25.0) 7.1 14.3 10.7 0/2 (0.0) * Adverse effects are listed in order of overall frequency (mentioned more than once in Randomization Phase).† Recurrence of adverse effect was determined if an adverse effect was initially present during the open phase, disappeared at the end of the open phase, and reemerged during the randomization phase. Columns represent reemerging cases (nominater) of all subjects with the adverse effect initially present, but in whom it disappeared at end of the open phase (denominater). Values in brackets represent this rate as percentages.‡ Sign. difference paroxetine DE vs. placebo DE (p<0.05; Fisher’s exact test).§ Sign. difference paroxetine DE vs. placebo DE (p<0.01; Fisher’s exact test). DE = dose-escalation, T0 = time of randomization at end of open phase, ≥2x RP = mentioned more than once in randomization phase, Rec. AE = Recurrence of adverse effect.

Table S9.3.Changes in paroxetine serum concentrations (μg/l) in the SPECT subgroup due to paroxetine or placebo dose-escalation.

T0 T1 p*

Paroxetine DE (n= 11) 36.2 (22.7-52.8) 154.3 (112.4-202.7) <0.001

Placebo DE (n= 15) 60.3 (41.0-83.3) 52.2 (32.9-75.8) 0.268

Values represent means with 95% confidence intervals, computed on square-root transformed paroxetine serum concentrations. Probable adherentpatients only; in 1 patient (paroxetine dose-escalation) the measurement of paroxetine concentration was missing. The change in paroxetine serum concentrations after paroxetine dose-escalation vs. placebo dose-escalation was significant (ANCOVA correcting for paroxetine serum concentration at T0: F1,23= 59.938; p <0.001). *significance of difference between T0 and T1; paired T-test. DE = dose-escalation.

Chap

ter

Chapter 9

9

194

Figure S9.1. Changes in paroxetine serum concentration versus changes in occupancy (randomization to endpoint).

Regression lines show linear relationship (with 95% CI) between increase in paroxetine serum concentration (PSC) andchange in occupancy of midbrain (circles; beta= 0.03 ±0.08; F1,24= 0.101; p= 0.754) and diencephalon (diamonds; beta= 0.07±0.06; F1,27= 1.332; p= 0.259)

��

��

��

��

��

��

��

��

��

��

�+�,��

��

� �

��

��

��

��

��

��

��!

�

CHA

PTER

SUCCESSFUL PHARMACOLOGICAL TREATMENT OF

MAJOR DEPRESSIVE DISORDER ATTENUATES

AMYGDALA ACTIVATION TO NEGATIVE FACIAL

EXPRESSIONS. AN FMRI STUDY

Submitted

Henricus G. Ruhé1*

Jan Booij2

Dick J.Veltman3,4

Martin C. Michel5

Aart H. Schene1

1 Program for Mood Disorders, Department of Psychiatry, Academic Medical Center, University ofAmsterdam, Amsterdam, the Netherlands2 Department of Nuclear Medicine, Academic Medical Center, University of Amsterdam, Amsterdam,the Netherlands 3 Department of Addiction Research, Academic Medical Center, University of Amsterdam,Amsterdam, the Netherlands 4 Department of Psychiatry, Vrije Universiteit, Amsterdam, the Netherlands 5Department of Pharmacology and Pharmacotherapy, Academic Medical Center, University ofAmsterdam, Amsterdam, the Netherlands

Chap

ter

Chapter 10

10

196

Abstract

BackgroundDespite modest response and remission rates to selective serotonin reuptake inhibitors (SSRIs),underlying neurobiological mechanisms of pharmacological treatment of major depressivedisorder (MDD) are scarcely investigated in vivo. Amygdala function might be a biomarker forMDD treatment-response.

Aimto quantify the changes in activation by emotional faces in amygdala and other limbic-subcortical-prefrontal structures after paroxetine treatment in MDD.

MethodsWe performed an event-related functional magnetic resonance imaging study in SCID-positive,non-psychotic, unipolar depressed patients (n= 22; M+F; 25-55 years, HDRS17>18) who werescanned at study-entry, and at 6 weeks (T0) and 12 weeks (T1) of paroxetine treatment. After 6weeks open-label paroxetine 20mg/day (T0), 15 non-responding patients (HDRS17 decrease <50%)were randomized to double-blind paroxetine (30-50mg/day) or placebo dose-escalation(paroxetine 20mg/day + placebo). The paradigm consisted of sex-judgments of fearful, angry,happy and neutral faces, relative to scrambled (baseline) faces. Twenty-one age- (±2.5 yrs) andsex-matched healthy controls (HC; without 1st degree relative with, or lifetime psychiatricdisorder) were scanned once as reference.

ResultsAt study-entry, MDD-patients showed increased ventral/limbic (amygdala and insula) anddecreased dorsal cortical activations relative to HC. At T0 and T1, 5/20 and 13/20 patients wereresponders, respectively. Treatment response was associated with decreased activations inamygdala, right insula and orbitofrontal cortex (OFC), and increased activation in rightdorsolateral prefrontal cortex (DLPFC) and nucleus accumbens. Amygdala activation wasinversely correlated with increased pregenual cingulate and DLPFC activation as well as withimprovement using the HDRS17. Paroxetine dose-escalation was associated with decreasedactivity in ventral (limbic) and dorsal cortical regions, whereas activity of right hippocampus andleft subgenual cingulate increased relative to placebo dose-escalation.

ConclusionsDecreased amygdala activation by negative faces is a biomarker for clinical response toantidepressants, but needs further exploration in other MDD-treatments. Paroxetine response-effects appear to be mediated by increased fronto-limbic control.(ISRCTN register nr. ISRCTN44111488)

Chapter10

197

Amygdala deactivation in responders to paroxetine

Introduction

Major depressive disorder (MDD) is a highly prevalent and disabling disease,1 often treated withantidepressants, particularly selective serotonin reuptake inhibitors (SSRIs). Unfortunately,response and remission rates are only modest (30-50%). An additional strategy to achieveremission2-4 is dose-escalation2;3 of which the underlying neurobiological mechanism has scarcelybeen investigated in vivo. The development of non-invasive neuroimaging techniques may aid inclarifying this issue. For example, we recently showed that a true dose-escalation of paroxetinedid not increase serotonin transporter occupancy (the molecular target of SSRIs) more thanplacebo-dose-escalation.5

Previous studies of the etiopathogenesis of MDD have provided evidence for a dysfunctionallimbic-subcortical-prefrontal network in MDD.6;7 A processing bias to perceive negative emotionswas associated with increased activation of the ‘ventral’ compartment of this network (consistingof amygdala, insula, ventral striatum, ventral anterior cingulate gyrus and prefrontal cortex),responsible for the identification of the emotional significance of stimuli and the production ofaffective states.7 Activation of the (left) amygdala to negative facial expressions (sad, angry orfearful) has been consistently found increased in MDD compared with healthy subjects, and maybe considered a biomarker for MDD.8-11

Treatment with SSRIs may not fully reverse these differences between MDD and healthysubjects. Four studies have investigated changes in brain activation after 8 weeks of treatmentwith antidepressants in 8-19 patients per study.10;12-14 Another study in patients who remitted had aduration of 22 weeks (Table S10.1).15 These studies found an increase in activation in neocorticalregions, including the cingulate gyrus (dorsal and anterior parts) after treatment withantidepressants.10;13 A decrease in amygdala activation, as originally found by Sheline et al.,12 wasreplicated in one study with sertraline (sad faces task),10 and in one study with bupropion(emotional oddball task).14 In addition, two randomized placebo-controlled studies in healthycontrols found decreased activation in right16 or bilateral17 amygdala after 7 days of treatmentwith reboxetine and citalopram, respectively. However, as most of these treatment studiesreported high response rates and one study explicitly excluded non-responders,13 it is unclearwhether the observed changes over time are drug effects, or due to clinical response. This couldbe addressed in analyses of responders vs. non-responders. Furthermore, although dose-escalation is a recommended strategy for non-responders, studies employing fMRI to investigatedose-response relationships have not yet been published.

Therefore, the aim of this study was to evaluate changes within the limbic-subcortical-prefrontal network, including the amygdala, to (negative) facial expressions after paroxetinetreatment of MDD. We specifically aimed to investigate clinicalresponse versus drug effects,including dose-escalation effects. To this end, we performed a functional MRI study in 22depressed patients who were scanned at study-entry, 6 weeks and 12 weeks of paroxetinetreatment,and 22 matched controls scanned once.

We expected to observe increased activation of the amygdala in patients versus healthycontrols at baseline, followed by an attenuation of amygdala hyperactivity together with anincrease of activation in the neocortical areas (anterior cingulate and dorsomedial and -lateralprefrontal cortex (DMPFC/DLPFC)) after 12 weeks of treatment. We hypothesized that thesechanges would correlate with clinical response, but not with dose-escalation of paroxetinecompared with placebo.

Chap

ter

Chapter 10

10

198

Methods

ParticipantsFollowing approval by the institutional ethical committee and written informed consent, werecruited 22 outpatients (25-55 years) from our outpatient department (August 2005 - August2006).5 Inclusion criteria were: MDD determined by the structured clinical interview for DSM-IV(SCID18), and a Hamilton Depression Rating Scale (17 items; HDRS17)19 score above 18. Patientswere drug-free (>4 weeks and ≥5 half-lives of a previous antidepressant) and had not used morethan one antidepressant treatment (other than paroxetine) at an effective dose for ≥6 weeks forthe present MDD-episode. Exclusion criteria, apart from pregnancy (or wish) and standard fMRIcontraindications, were bipolar disorder, psychotic features, neurological impairments, primaryanxiety and/or substance abuse disorders and acute, severe suicidal ideation.

We individually matched each patient by gender and age (±2.5 years) with a healthy control(HC), who had good physical health and had never used psychotropic medication. Exclusioncriteria for HC were any lifetime psychiatric disorder according to the SCID (including abuse oraddiction disorders), a Beck Depression Inventory (BDI)20 score >9, average alcohol use >4 unitsper day (preceding month) and a 1st-degree relative with psychiatric disorder(s).

Treatment scheduleAfter assessment at study-entry patients were treated with paroxetine 20 mg/day (supplied in pill-boxes) for 6 weeks. All patients who did not achieve ≥50% decrease in HDRS17-score after 6 weeks(T0; see Figure 2.1) were randomized (stratified for gender and age). Patients received a trueparoxetine or a placebo dose-escalation added to paroxetine 20 mg/day. Methods of dose-escalation were described previously,5 we applied incremental steps of one capsule every 5 days(reaching 30-50 mg/day according to adverse effects).21 Dosages remained unchanged the last 3weeks of the study. Adherence was checked by self-report,22 pill-counts and measured paroxetineserum concentrations.5

MeasurementsPrimary clinical outcome was the proportion of patients achieving response (≥50% decrease inHDRS17). Depression severity was measured at study-entry, randomization (T0), and 6 weeks afterrandomization (T1), also using the Inventory for Depressive Symptomatology self-rated (IDS-SR30).23 At these time-points we planned fMRI-sessions of 50 minutes, each including a cognitivetask (reported elsewhere), a structural scan and the facial expression task reported here.Agreement between trained raters was good (intra-class correlation coefficient = 0.98). Ratersand patients were blinded for the randomized dose-escalation.

Facial expression task paradigmWe used an event-related emotional faces paradigm, which reliably activates the anterior medialtemporal lobe including the amygdala.24;25 We presented four human face stimuli: angry, fearful,happy, and neutral human faces,26 and scrambled faces (with two arrows in the centre of thescreen) as baseline condition. Each face stimulus condition consisted of 10 pictures; each picturewas presented three times. Stimuli were randomized once and presented in identical order to allsubjects, using the same task for each session. Stimuli were displayed for 2500ms with a variableinterstimulus interval (400-600ms), to increase experimental power and to decrease expectancyeffects. To control for overflow effects, we displayed a baseline stimulus after each one or twoface pictures. Subjects were instructed to make gender judgements during presentation of facestimuli to control for attention differences. To familiarize participants, the task was first explainedoutside the scanner. No feedback was provided during the task.

Chapter10

199


fMRI imaging We acquired fMRI scans in the afternoon/early evening using a 3Tesla Intera MRI scanner (Philips,Eindhoven, Netherlands). A 6-channel head-coil was used for radiofrequency reception, with thehead fixated in the coil by foam pads to prevent movement-artefacts. Stimuli were generated by aPentium PC and projected on a screen at the patient’s feet, visible through a mirror mounted onthe headcoil. Stimulus onset was triggered by a pulse from the scanner. Two magnet compatibleresponse boxes were used to record subject’s performance and reaction times (RTs).

At each session, we obtained a volumetric T1-weighted coronal scan (TE/TR= 4.6/9.63 msec,field of view= 24×24 cm, flip angle= 8°, number of excitations= 1, matrix= 256×256, 182 slices, slicethickness= 1.2 mm, interslice gap= 0 mm, scan time= 7 min) covering the entire brain volume,and260 T2*-weighted axial echoplanar imaging (EPI) images sensitive to blood oxygen leveldependent (BOLD) contrast (TE/TR= 35/2530.4 msec, field of view= 24×24 cm, flip angle= 90°,number of excitations= 1, matrix= 128×128, 36 ascending slices, slice thickness= 3 mm, interslicegap= 0.3 mm, scan time=10 min).

fMRI data analysis

Individual analysisFor all fMRI data-analyses we used SPM5 (Statistical Parametric Mapping; Wellcome Departmentof Cognitive Neurology, London, UK; http://www.fil.ion.ucl.ac.uk/spm/), operated under Matlabversion 7.3.0.267 (2006b; the Mathworks,Natick, Massachusetts, USA). Standard preprocessing ofscans consisted of correcting for slice-timing differences, head movements, coregistration to thestructural scan, normalization to SPM/MNI standard space, and smoothing (8 mm full-width half-maximum Gaussian filter). Next, BOLD responses were modeled to affective facial expressionsand baseline conditions for each voxel. For each subject, weighted contrasts were computed forsimple main effects across all stimulus types combined (angry/anxious, happy, and neutral facesvs. baseline = ‘all faces’), and within stimulus type contrasts (angry/anxious vs. baseline =‘negative faces’; happy vs. baseline = ‘happy faces’).

Group analysisThe contrast images obtained in individual analyses were entered into second level (randomeffects) analyses for between-group and within subject (i.e. T0-T1) comparisons (withinterdependent measures and assuming equal variances).Main effects were identified at p <0.01and FDR-corrected for multiple comparisons, with an extent threshold of three voxels.27 Plannedcontrasts were changes over time (study-entry – T0 – T1), responders versus non-responders, andeffect of paroxetine dosage. Interactive effects of time/response*stimulus type (all faces,negative faces, happy faces; masked with the relevant main effect) were identified at p<0.001 (z>3.09), uncorrected, with an extent threshold of three voxels, but in the amygdala region atp<0.005 (z>2.57).

Statistics We compared study-entry characteristics of patients and controls with independent samples t-tests for continuous and χ2 or Fisher’s exact test for categorical variables. We used linear mixedmodels (compound symmetry variance/covariance structure) to assess differences in scanperformance (reaction times and errors); between patients and healthy controls, and in patientsin trend over time. Parameter estimates were extracted for bilateral amygdala from SPM5, andthese values were used for graphical representation. With regression models, we quantified theassociation of parameter estimates with HDRS17-scores. We performed additional analyses inSPSS v15.0.1.1 (www.spss.com) and GraphPad Prism v5.00 (www.graphpad.com).

Chap

ter

Chapter 10

10

200

Results

Patients, controls and behavioural dataTable 10.1 displays demographic data and MDD ratings for patients and HC. HC had significantlyhigher education-levels. MDD-patients reported significantly higher state and trait anxiety thancontrols (p<0.001), and showed slower RTs p= 0.045) and more gender-judgement errors (p=0.032). Three male patients withdrew from the study after the study-entry scan; 2 withdrewcompletely. One female HC did not attend her visit and could not be replaced. Due to excessivemovements, two study-entry scans (one male, one female patient), two T0 scans (one male, onefemale patient) and three T1 scans (one male, two female patients) could not be analyzed. Thus,41 study-entry scans (20 patients, 21 controls), 17 T0 and 16 T1 scans were available for analyses.

Twenty patients completed the treatment protocol, 5/20 (25%) and 13/20 (65%) had a clinicalresponse at week 6 and 12 respectively. No week 6 responders deteriorated afterwards (Table10.2). A randomized dose-escalation was provided to 15 non-responding patients after 6 weeks ofparoxetine treatment; for 11 of these repeated scans were analyzable. Clinical outcomes wereneither numerically, nor statistically significantly different between true and placebo dose-escalation. RTs and errors did not change over time, and were not significantly different forresponders versus non-responders nor associated with HDRS17-scores.

Main effects at study-entry in patients and healthy controlsCombining the study-entry scans of patients and HCs (all faces contrast) showed robust activationof bilateral amygdala, fusiform gyrus, DLPFC, (anterior) insula, occipital cortices, and rightorbitofrontal cortex (OFC; extending into the right anterior insula), parietal cortex and DMPFC.These effects were also found for negative faces, except for the right amygdala, left insula, andleft DLPFC, which were not activated above threshold. With the happy faces contrast, we foundmain effects for bilateral fusiform gyrus, insula, occipital cortices, and right DLPFC, OFC (extendedfrom insula), thalamus and parietal cortex (Table S10.2).

Table 10.1. Characteristics of study-sample.

Patients (n= 22) Healthy Controls (n= 22) p

Age ±SD 43.3 ±7.93 43.7 ±7.99

Males n (%) 14 (63.6) 14 (63.6)

Handedness R/L 20/2 21/0* 0.488

Educational level n (%) Low Intermediate High

5 (22.7) 12 (54.5) 5 (22.7)

0 6 (27.3) 16 (72.7)

0.002

MDD-severity HDRS17 IDS-SR30 STAI-I STAI-II

23.1 ±3.61 42.9 ±7.69 58.9 ±8.84 62.7 ±9.25

N/A 4.6 ±3.87* 29.4 ±7.74* 29.8 ±11.14*

<0.001 <0.001 <0.001

Co-morbidity n (%) Anxiety Alcohol dependance Cannabis dependance

3 (13.6) 2 (9.1) 1 (4.5)

N/A

Scan task performance Reaction Time (ms) Correct response (%)

954.7 ±238.3 95.5 ±3.24

* 830.0 ±143.3 97.5 ±2.66

0.045 0.032

* 1 female HC did not attend; no questionnaire and scan performance data available

Chapter10

201


Activations in the amygdala in MDD-patients versus HCThe all faces contrast showed higher activations in bilateral (extended) amygdala in MDD-patientsthan in HC. Higher (right) amygdala activities in patients were also found when we contrastednegative faces, but not when contrasting happy faces (Table 10.3A). In post-hoc analyses, baselineamygdala activations were not different between final treatment responders and non-responders(all contrasts).

Other activations of the limbic-subcortical-prefrontal network in MDD-patientsversus HCContrasting all faces versus baseline showed higher activations in the left insula in MDD-patientscompared with HC (Table 10.3A). For happy faces MDD-patients showed higher activation in theleft subthalamic nucleus.

With the all faces contrast, we found lower activations in MDD-patients relative to HC inbilateral ventrolateral prefrontal cortex (VLPFC), posterior and medial dorsal cingulate, leftDMPFC, bilateral DLPFC and fusiform gyrus (Table 10.3B). For negative faces, we found loweractivations in MDD-patients in bilateral VLPFC, left posterior cingulate gyrus, left DMFPC, rightDLPFC, and bilateral fusiform gyrus. With happy faces, we found lower activations in right VLPFC,right premotor cortex, and left fusiform gyrus in MDD-patients relative to HC.

In post-hoc analyses, final treatment responders showed higher activations at study-entry inthe right pregenual (rostral) cingulate (MNI 14, 44, 3; k= 4; Z= 2.77; p= 0.003; negative faces),relative to final non-responders. Instead, non-responders showed higher study-entry activationsin the subgenual cingulate (MNI 0, 26, -3; k= 11; Z= 3.90; p<0.001; negative faces).

Table 10.2. Clinical response and task performance of patients during follow-up (n= 20).

T0 (Randomization) T1 (Endpoint)

HDRS17All (n= 20) T0-Responders (n= 5) True DE (n= 8) Placebo DE (n= 7)

15.7 ±5.14 9.8 ±3.11 18.0 ±5.13 17.3 ±2.75

11.1 ±5.31 7.2 ±1.48 13.5 ±6.70 11.0 ±3.87

% decrease in HDRS17All (n= 20) T0-Responders (n= 5) True DE (n= 8) Placebo DE (n= 7)

30.9 ±26.67 59.9 ±9.53 20.5 ±26.36 22.0 ±21.01

51.6 ±25.66 70.5 ±2.24 39.4 ±34.07 52.0 ±15.09

Responders (HDRS17 ≥50%)All (n= 20) T0-Responders (n= 5) True DE (n= 8) Placebo DE (n= 7)

5 (25%) 5 (100%) - -

13 (65%) 5 (100%) 4 (50%) 4 (57%)

Remission (HDRS17 ≤7)All (n= 20) T0-Responders (n= 5) True DE (n= 8) Placebo DE (n= 7)

2 (10%) 2 (40%) - -

5 (25%) 3 (60%) 1 (13%) 1 (14%)

IDS-SR30All (n= 20) T0-Responders (n= 5) True DE (n= 8) Placebo DE (n= 7)

30.6 ±8.80 19.8 ±6.46 35.3 ±6.71 33.0 ±5.66

27.2 ±11.46 21.6 ±4.72 33.5 ±14.92 23.9 ±7.03

Scan task performance* Reaction Time (ms) Correct response (%)

936.8 ±242.8 96.3 ±3.74

921.7 ±252.2 95.6 ±6.26

Values represent means ±SD. DE= dose-escalation. All differences between True DE and Placebo DE non-significant (ANOVA; p>0.14) * n= 18 at T0 and n= 17 at T1

Chap

ter

Chapter 10

10

202

Changes in amygdala activations after 6 and 12 weeks of paroxetine treatmentAfter 6 weeks of treatment (T0), for the all faces and negative faces contrasts, the activations inbilateral amygdala regions had not significantly decreased relative to the study-entry scan (TableS10.3A). In contrast, the right lateral amygdala showed a significant increase in activation after 6weeks of treatment (all faces, negative faces and happy faces). Further exploration showed thatweek 6 treatment non-responders (n= 12) had increased amygdala activations for all faces (left z=2.33; p= 0.01; right z= 2.52; p= 0.006) and negative faces (left z= 3.22; p= 0.001; right z= 2.87; p=0.002) relative to week 6 responders.

After 12 weeks of treatment (T1), when we contrasted all faces, the right amygdala showeddecreased activations (although subtreshold), relative to study-entry (Table 10.4A). Othercontrasts did not reveal amygdala deactivations over time.

Table 10.3. Study-entry scans: MDD-patients versus HC.

Contrast Brain region L/R x,y,z (MNI mm) Cluster size (k) Max. voxel Z p

A. MDD > HC

All Faces Amygdala R 18 2 -9 12 3.04 0.001

L -16 -4 -9 39 2.63 0.004

Insula L -26 4 12 30 3.13 0.001

Neg. Faces Amygdala R 18 2 -9 13 2.99 0.001

L -16 -4 -9 16 2.46 0.007

Insula L -28 6 15 32 3.26 0.001

Hap. Faces Subthalamic nucleus L -12 -6 -6 14 3.35 <0.001

B. HC > MDD

All Faces VLPFC R 50 20 -6 222 3.95 <0.001

L -34 22 -6 23 3.20 <0.001

DLPFC R 42 16 27 61 3.73 <0.001

R 48 -2 51 16 3.31 <0.001

L -54 16 0 25 3.48 <0.001

DMFPC L -4 10 63 76 3.49 <0.001

Fusiform gyrus L -42 -54 -21 61 3.61 <0.001

Cingulate, medial L -8 26 42 58 3.36 <0.001

posterior L -6 -20 51 23 3.59 <0.001

Neg. Faces DMFPC L -4 10 63 383 4.37 <0.001

VLPFC R 50 20 -6 241 3.74 <0.001

L -34 22 -6 41 3.09 0.001

DLPFC R 42 16 27 76 4.25 <0.001

Fusiform gyrus L -42 -54 -21 67 3.78 <0.001

R 46 -40 -24 49 3.10 0.001

Cerebellum L -16 -38 -21 13 3.28 0.001

Sup. Temporal gyrus R 58 -6 -12 37 3.45 <0.001

Cingulate, posterior L -12 -22 48 39 3.41 <0.001

Hap. Faces VLPFC R 54 32 3 81 3.51 <0.001

Premotor cortex R 44 2 48 27 3.37 <0.001

Sup. Temporal gyrus R 54 -48 12 34 3.51 <0.001

R 48 -36 6 27 3.20 0.001

Fusiform gyrus L -40 -54 -18 50 3.13 0.001

DLPFC= dorsolateral prefrontal cortex; HC = healthy control; MDD= major depressive disorder; MNI= Montreal neurological institute; VLPFC= ventrolateral prefrontal cortex.

Chapter10

203


Changes in activations in other brain areas after 6 and 12 weeks of paroxetinetreatmentAfter 6 weeks of treatment (T0), relative to study-entry, we additionally found decreasedactivations in the left posterior hippocampus (all faces; Table S10.3A) and bilateral cuneus (allfaces and negative faces). In contrast, increased activations were found in the right dorsalhippocampus, left posterior and medial cingulate gyrus, right pregenual cingulate gyrus, and leftDMPFC (all faces contrast; Table S10.3B). For negative faces, activations of right dorsalhippocampus, right anterior and pregenual cingulate gyrus, bilateral cingulate gyrus, bilateralDLPFC, left DMPFC and left nucleus accumbens were increased. With the happy faces contrast,after 6 weeks no activations in any of these regions were found.

After 12 weeks of treatment (T1), relative to study-entry, we found decreased activations forthe all faces contrastin left OFC, right posterior hippocampus, and right posterior cingulate gyrus(Table 10.4A). After 12 weeks, relative to study-entry, we found no significant decreases inactivation for the negative faces. For happy faces we found a decrease in activation in the leftinsula, right dorsal hippocampus and right parietal cortex. Furthermore, after 12 weeks, relative to

Table 10.4. Changes after 12 weeks of treatment (T1) relative to study entry.


A. Study entry > T1


Sup.Temporal Gyrus R 50 -48 18 101 3.67 <0.001

Hippocampus posterior R 22 -36 0 34 3.30 <0.001

OFC L -38 32 -12 9 3.15 0.001

Cingulate, posterior R 4 -36 48 4 3.09 0.001

Neg. Faces -

Hap. Faces OFC L -34 30 -9 140 3.91 <0.001

Hippocampus, dorsal R 20 -36 0 34 3.80 <0.001

Parietal cortex* R 46 -34 57 8 3.46 <0.001

R 46 -50 33 6 3.35 <0.001

Insula L 36 12 -3 12 3.33 <0.001

B. T1 > Study entry

All Faces Premotor Cortex R 28 6 48 194 6.02 <0.001

L -28 4 51 21 4.15 <0.001

DLPFC L -38 2 48 23 3.62 <0.001

L -48 22 18 32 3.58 <0.001

L -36 34 30 7 3.10 0.0001

Motor cortex R 10 -12 63 16 3.25 0.001

L -4 0 60 97 3.19 0.001

Neg. Faces Premotor Cortex R 28 6 48 487 5.14 <0.001

L -28 4 51 107 4.82 <0.001

DLPFC R 42 20 21 30 3.66 <0.001

R 60 -20 30 65 3.63 <0.001

L -46 20 24 114 3.82 <0.001

L -50 8 0 7 3.45 <0.001

L -46 8 30 35 3.44 <0.001

Motor cortex L -26 -16 57 39 3.67 <0.001

R 12 -4 66 20 3.31 0.001

Cingulate, posterior R 6 -12 30 106 3.53 <0.001

Hap. Faces DLPFC L -26 -6 51 30 3.84 <0.001

Hippocampus R 24 -16 -18 19 3.23 0.001

* Intraparietal sulcus and supramarginal gyrus respectively Abbreviations: see also Table 10.3. OFC= orbitofrontal cortex.

Chap

ter

Chapter 10

10

204

study-entry, we found increased activations in left DLPFC, and bilateral premotor and motorcortices for the all faces and negative faces contrasts, and in the right hippocampus and leftDLFPC for the happy faces contrast (Table 10.4B).

Changes in amygdala activations in relation to clinical responseWhen we compared responders and non-responders after 6 weeks and 12 weeks (full factorialmodel), we found higher right bilateral amygdala activations in non-responders relative toresponders for the all faces contrast (Table 10.5A). With the negative faces contrast, higheractivations in non-responders were significant in bilateral amygdala (p= 0.001; Figure 10.1).Moreover, the amygdala negative faces contrast estimates were significantly associated withHDRS17-scores, with higher activations in more depressed patients. No differential effectsbetween left and right amygdala were found.

In post-hoc analyses, left amygdala activations correlated positively with the contralateralamygdala, bilateral OFC and right subgenual anterior cingulate (Table S10.3). Right amygdalaactivation correlated positively with the contralateral amygdala, bilateral OFC and right DLPFC.Left amygdala signal was inversely correlated with the pregenual anterior cingulate, rightamygdala signal was inversely correlated with the left DLPFC and the left nucleus accumbens.

Table 10.5. Full-factorial model of 6-weeks and 12-weeks responders versus non-responders.


A. Non-Responders > Responders

All Faces Amygdala R 18 -2 -18 30 3.24 0.001

L -18 -10 -21 24 2.22 0.013

OFC lateral R 28 16 -18 98 3.80 <0.001

posterior R 10 12 -18 98 3.03 0.001

R 40 38 -12 39 3.56 <0.001

Insula R 42 -10 3 60 3.62 <0.001

Brainstem (L. coeruleus) R 6 -16 -12 30 3.45 <0.001

Neg. Faces Amygdala R 18 -2 -18 30 3.20 0.001

L -22 2 -18 24 3.11 0.001

OFC R 40 38 -12 33 3.48 <0.001

lateral R 26 16 -18 46* 3.01 0.001

posterior R 10 12 -18 * 2.90 0.002

Insula R 42 -10 3 49 3.33 <0.001

Hap. Faces -

B. Responders > Non-Responders

All Faces DLPFC R 20 20 63 5 3.71 <0.001

N.Accumbens L -10 10 -9 22 3.21 0.001

Neg. Faces N.Accumbens L -8 12 -9 40 3.69 <0.001

Parietal cortex L -46 -52 45 21 3.49 <0.001

Hap. Faces Cerebellum R 26 -66 -24 11 3.67 <0.001

R 8 -74 -15 78 3.64 <0.001

Hippocampus, dorsal L -24 -32 -6 15 3.38 <0.001

Cingulate, medial LR 4 16 42 41 3.29 0.001

Insula L -34 24 6 30 3.17 0.001

Hypothalamus R 8 -10 3 15 3.11 0.001 * in one OFC cluster. Abbreviations: see Table 10.3 and 10.4.

Chapter10

205


Changes in activations in other brain areas in relation to clinical response Non-responders also showed significant (p≤0.001) higher activations in right OFC and right insula(all faces and negative faces), brainstem (all faces), relative to non-responders after 6 and 12weeks of treatment (Table 10.5A). In contrast, treatment responders showed higher activations inright DLPFC, and left nucleus accumbens (all faces and negative faces; Table 10.5B). Furthermore,with the happy faces contrast, responders had higher activations in the left dorsal hippocampus,bilateral medial cingulate gyrus, left insula and right hypothalamus.

Effects of true dose-escalation versus placebo dose-escalationWe investigated the interaction of randomized dose-escalation by time in full factorial models.Relative to placebo dose-escalation, 6 weeks of true dose-escalation decreased activations in leftOFC, bilateral insula, right VMPFC, left ventrolateral thalamus, right anterior cingulate, left DMPFCand motor cortex, right parietal cortex, right putamen and left globus pallidus (all faces; TableS10.5A). With negative faces decreased activations after dose-escalation were found in rightinsula, bilateral OFC, left ventrolateral thalamus, right anterior cingulate, left DMPFC and bilateral

Figure 10.1. Amygdala response relative to clinical response (HDRS17 ≥50% decrease) after 6 and 12 weeks of treatment with paroxetine.

Negative faces contrast. Week 6 & 12 scans and response status were combined in a full factorial model. For color figure see page 284A. SPM-results showing design matrix, contrast and activation in right amygdala (MNI 18,-2,-18).B. Activations of left (MNI -22,2,-18; dark) and right (MNI 18,-2,-18; white) amygdala stratified for non-response and responseat 6 and 12 weeks. Significant difference between non-responders and responders (F1,32= 4.591; p<0.0001); no significantlateralization (F1,32= 0.1517; p= 0.70).C. Plot of contrast estimate by HDRS17-score (centralized to mean). Significant positive correlations for left (0.33 ±0.12 [SE];F1,31= 7.076; p= 0.0123; r2= 0.19; line) and right (0.38 ±0.14 [SE]; F1,31= 6.929; p= 0.0131; r2= 0.18; dashed line) amygdala.Regression lines for left and right amygdala are not significantly different (F1,62= 0.0683; p= 0.79).

��

��

��

�

�

��

��

��

��

��

��

��

��

�

�

�

�

��

��

��

��

��

� �

�

Chap

ter

Chapter 10

10

206

motor cortex. With happy faces decreased activations were found in right insula, bilateral OFC,right VMPFC, bilateral anterior cingulate, left pregenual cingulate, bilateral parietal cortex andright putamen.

Relative to placebo dose-escalation, 6 weeks of true dose-escalation increased activations inright hippocampus, left subgenual cingulate, bilateral medial temporal gyrus, left parietal andDLPFC and right cerebellum (all faces; Table S10.5B). With the negative faces contrast, dose-escalation increased activations in right hippocampus, left parahippocampal gyrus, left subgenualcingulate, left medial temporal gyrus and bilateral parietal cortex and cerebellum. With happyfaces 6 weeks of dose-escalation increased activations in left parahippocampal gyrus, left medialtemporal gyrus and left DLPFC.

Discussion

This fMRI study evaluated the changes in amygdala activation to (negative) facial expressionsafter paroxetine treatment of MDD, as well as changes in other brain regions within the limbic-subcortical-prefrontal network. Depressed patients were scanned at study-entry, 6 weeks and 12weeks of paroxetine treatment. To our knowledge, this is the first fMRI study investigating SSRI-treatment with a randomized, placebo-controlled dose-escalation.

MDD-patients showed higher ventral (amygdala and insula) and decreased dorsal activationsrelative to HC at study entry. After 6 weeks of paroxetine treatment, when only 5/20 patientswere classified as responders, we found increased ventral (right amygdala) and dorsal activations(posterior, medial and pregenual cingulate cortex, left DMPFC and DLPFC). In contrast, after 12weeks of treatment (with 13/20 responders), we found decreased activations in ventral (left OFC)and increased activations in the dorsal compartment, all relative to study-entry. Treatmentresponse was associated with lower activations in amygdala, right OFC and insula, and higheractivations in right DLPFC and nucleus accumbens. The placebo-controlled randomized dose-escalation resulted in decreased activations in a number of ventral regions, but not amygdala,while right hippocampus and left subgenual cingulate activations increased, relative to placebodose-escalation. These results suggest that the attenuation of amygdala activation during SSRI-treatment is not a direct pharmacological response, but appears to be a biomarker for treatmentresponse.

Brain activations in MDD and changes following antidepressant treatmentOur study replicates an increased reactivity of the ventral ‘limbic’ structures in MDD-patients (e.g.amygdala and insula), with decreased activations in dorsal prefrontalareas (PFC and cingulate).8-11

Furthermore, like previous treatment studies10;13-15 paroxetine treatment over 12 weeks increaseddorsal cortical function, and reduced activations in the ventral structures. However, these effectswere more prominent when we distinguished treatment responders and non-responders.

In the present study, we investigated the role of the amygdala as a biomarker forantidepressant treatment. Previous reports suggested that the SSRIs sertraline and fluoxetinedecreased abnormal amygdala activations in response to negative stimuli in MDD-patients,10;12 aswas found for bupropion.14 Two studies in healthy volunteers, treated for 7 days with citalopramand the noradrenaline uptake inhibitor reboxetine also found an attenuation of amygdalaresponse to negative faces.16;17 Although we replicate this attenuation, our data suggest animportant difference regarding the interpretation previous findings. Specifically, our resultsindicate that abnormal amygdala activation is decreased in association with treatment response,but is not a direct pharmacological effect, as was also indicated by the absence of amygdala-effects after dose-escalation. Because the response-rates in previous MDD fMRI-studies werehigh; 10/11 (91%),12 13/19 (68%),10;28 and 6/8 (75%)14 this response-related effect may have beenmissed in previous group analyses.

Chapter10

207


As expected, in our study amygdala activation at week 6 and 12 was correlated with otherventral structures (contralateral amygdala, OFC, subgenual cingulate) and inversely correlated topregenual (rostral) cingulate cortex and DLPFCactivity. This highlights an inhibitive connectionbetween the amygdala and the pregenual cingulate, a region that plays a crucial role inintegrating the function of dorsal and ventral prefrontal compartments.6;29-31 Previous pre-treatment, resting state, fluorodeoxyglucose positron emission tomography (PET) studies haveshown that increased metabolism in the pregenual cingulate predicted better treatmentresponse (reviewed by Dougherty and Rauch).29 This finding has been replicated in two fMRIstudies with negative faces/pictures,10;13;31 and in our post-hoc comparisons of study-entry scans.Moreover, we also found increased activations in the subgenual cingulate in final treatment non-responders,32;33 suggesting that activity of pregenual and subgenual cingulate areas may serve asadditional biomarkers for (future) treatment response.

The present study also included a paroxetine versus placebo dose-escalation comparison in asecondary randomization of treatment non-responders after 6 weeks. Relative to placebo, higherparoxetine doses was associated with decreased activity in several regions, including left OFC,bilateral insula, right VMPFC and left ventrolateral thalamus. In contrast, dose-escalation vs.placebo resulted in increased activity of the (para)hippocampal cortex and, unexpectedly, thesubgenual cingulate cortex. Two explanations can be put forward to account for these results.First, higher doses of paroxetine may further decrease ventral (limbic) hyperactivation andincrease dorsal prefrontal activation, in line with the effects seen at standard doses.10;12;14

However, the observed increased activity of the subgenual cingulate after dose-escalation is atodds with this hypothesis, because persistent or higher activation in the subgenual cingulate waspreviously associated with non-response.32;33 Also within this hypothesis, the increased activationsin the hippocampus might represent increased hippocampal neurogenesis after prolongedantidepressant administration, as has been found in rodents,34;35 and which is presumablyassociated with second-messenger effects of SSRIs resulting in increased levels of brain derivedneurotrophic factor (BDNF)36 In contrast, Mayberg et al. previously reported decreasedhippocampal metabolism in fluoxetine-responders, and no change in non-responders.33 However,these studies used fluorodeoxyglucose resting-state PET, comparing changes between week 1and week 6, so that direct comparisons are problematic.

An alternative explanation might be that dose-escalation does not lead to furtherimprovement MDD in patients, nor affect its neurobiological substrate. This would be in line withthe absence of unequivocal improvement after dose-escalation found in previous clinicaltrials.5;37-39 Following this line of argument, the increased activations in the subgenual cingulatemay be interpreted as supportive to the finding of Licht et al.,40 who reported that patients withincreased sertraline doses (200 mg/day) had a poorer response-rate relative to placebo dose-escalation (sertraline 100 mg/day).

In our study, correlation analyses showed that decreased amygdala activity in responders wasassociated with increased rostral cingulate and DLPFC activity, suggesting that antidepressantresponse is associated with increased cognitive control by the dorsal compartment as well as areduction of the hyperactivation in the ventral compartment. Recently, Chen et al. investigatedfunctional connectivity of the amygdala, and reported increased functional coupling with thefrontal and (pregenual) cingulate cortex (and striatum and thalamus) after 8 weeks of sertralinetreatment.31 Although assessed differently, we also found associations of the amygdala withthese regions. Therefore, it may be hypothesized that SSRI-exposure over time improves dorsalprefrontal regulation of abnormal limbic activity.31;41;42 However, the SSRI-effects on pregenualcingulate-amygdala coupling remain controversial,31;41 and may be modified by patient selectionand other variables (e.g. polymorphisms of the serotonin transporter promoter region andBDNF).31;43;44

In treatment responders, we found increased activation not only of dorsal prefrontal areasbut also of the left nucleus accumbens. Activation of the nucleus accumbens is associated withanticipation of reward.45 In contrast, expected loss might deactivate the nucleus accumbens.Although a recent pilot-study did not reveal differences in nucleus accumbens activation betweenMDD patients and HC during a monetary reward task,46 it might still be that during MDD either the

Chap

ter

Chapter 10

10

208

anticipation of reward is reduced, or the responses to loss are increased compared with a lessdepressed state. This might hypothetically explain the increased activations (or reduceddeactivations) seen in the nucleus accumbens in patients who showed a clinical response. Thisfinding merits further investigation of reward-processing in MDD.

LimitationsSimilar to previous studies,10;12;13 we refrained from a placebo-treated group, but instead included arandomized placebo-controlled dose-escalation. A full placebo-group, as in previous PET studies,47

might have enabled a clearer differentiation between true drug-effects and response effects.Therefore, as proposed recently, future fMRI-studies should preferably include a full placebo-group.31

Second, we scanned healthy controls once, so that differential habituation effects could notbe assessed.10;13 Nevertheless, our patient sample showed a biphasic response, the initial increasein amygdala activity being associated with a high non-response ratio at T0. In addition, we foundan overall clear distinction between treatment responders and non-responders. Both findingssuggest that changes in responsiveness during our fMRI paradigm are related to clinical staterather than time-effects. We therefore consider our results valid.

Third, the findings for the randomized dose-escalation in our modest sample size need to beinterpreted with caution, given the small sample size. Therefore, replications in larger samples arewarranted.

Fourth, the inclusion of patients with co-morbid secondary anxiety disorders (3/22) orsubstance abuse disorders (3/22), might theoretically have influenced our findings. We considerthis unlikely; first, one patient with co-morbid anxiety and alcohol dependence could not beincluded in the analyses because of excessive movements in the scans, and second, the alcoholand cannabis use was stopped during the paroxetine treatment in the remaining 2 subjects.

ConclusionIn this fMRI-study in MDD patients, we investigated changes in amygdala activation and otherbrain regions within the limbic-subcortical-cortical network after paroxetine treatment, andapplied a secondary randomized, placebo-controlled dose-escalation in non-responders after 6weeks of paroxetine at 20mg/day. We found that amygdala activation by negative faces washigher in depressed patients than in HC and that bilateral amygdala activation decreased intreatment responders only. Decreased amygdala activations were correlated with increased rightpregenual cingulate and DLPFC activations. These results support the use of amygdala activitymeasured by fMRI as a biomarker for treatment response, and indicate increased fronto-limbiccontrol as a mechanism of paroxetine drug-response effects. Future studies need to clarifywhether spontaneous, psychotherapy and/or placebo responders show similar attenuation inamygdala activation by negative faces.

Acknowledgements

The authors wish to thank the patients who participated in this fMRI-study. Mrs. E. Miedema M.D.was indispensable for clinical treatment and assisted in scanning. Mr. A. Nederveen, PhD, and Mr.M. de Ruiter, PhD and Mrs. T. Dekker are acknowledged for their contributions to scanning andanalyses. Mrs. M. Haages managed randomization and maintained blinding. This study wasfinanced by a grant from the Netherlands Organisation for Health Research and Development(ZonMw), program Mental Health, education of investigators in mental health (OOG; #100-002-002) to H.G. Ruhé, and a grant from the Dutch Brain Foundation (14F06.45).

Chapter10

209



None

References

1. Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med. 2006; 3: e442.2. Kennedy SH, Lam RW, Cohen NL, Ravindran AV. Clinical guidelines for the treatment of depressive disorders. IV.

Medications and other biological treatments. Can J Psychiatry. 2001; 46 Suppl 1: 38S-58S.3. American Psychiatric Association. Practice guideline for the treatment of patients with major depressive disorder

(revision). American Psychiatric Association. Am J Psychiatry. 2000; 157: 1-45.4. NICE. Clinical Guideline 23. Depression: management of depression in primary and secondary care. December 2004. 23, 1-

63. 2004. London, National Institute for Clinical Excellence. 5. Ruhe HG, Booij J, Weert HCv, Reitsma JB, Fransen EJF, Michel MC et al. Evidence why dose-escalation of paroxetine in

major depressive disorder is not effective: A 6-week, randomized-controlled trial with assessment of serotonintransporter occupancy. Neuropsychopharmacology. 2008; [In Press].

6. Mayberg HS. Modulating dysfunctional limbic-cortical circuits in depression: towards development of brain-basedalgorithms for diagnosis and optimised treatment. Br Med Bull. 2003; 65: 193-207.

7. Phillips ML, Drevets WC, Rauch SL, Lane R. Neurobiology of emotion perception II: Implications for major psychiatricdisorders. Biol Psychiatry. 2003; 54: 515-528.

8. Abler B, Erk S, Herwig U, Walter H. Anticipation of aversive stimuli activates extended amygdala in unipolar depression. JPsychiatr Res. 2007; 41: 511-522.

9. Anand A, Li Y, Wang Y, Wu J, Gao S, Bukhari L et al. Activity and connectivity of brain mood regulating circuit indepression: a functional magnetic resonance study. Biol Psychiatry. 2005; 57: 1079-1088.


11. Surguladze S, Brammer MJ, Keedwell P, Giampietro V, Young AW, Travis MJ et al. A differential pattern of neuralresponse toward sad versus happy facial expressions in major depressive disorder. Biol Psychiatry. 2005; 57: 201-209.





16. Norbury R, Mackay CE, Cowen PJ, Goodwin GM, Harmer CJ. Short-term antidepressant treatment and facial processing.Functional magnetic resonance imaging study. Br J Psychiatry. 2007; 190: 531-532.

17. Harmer CJ, Mackay CE, Reid CB, Cowen PJ, Goodwin GM. Antidepressant drug treatment modifies the neural processingof nonconscious threat cues. Biol Psychiatry. 2006; 59: 816-820.


19. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 23: 56-61.20. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry. 1961; 4:

561-571.21. Baker CB, Woods SW. Is there a SSRI dose response in treating major depression? The case for re-analysis of current data




24. Hariri AR, Bookheimer SY, Mazziotta JC. Modulating emotional responses: effects of a neocortical network on the limbicsystem. Neuroreport. 2000; 11: 43-48.

25. Wolfensberger SPA, Veltman DJ, Hoogendijk WJG, Boomsma DI, de Geus EJC. Amygdala responses to emotional faces intwins discordant or concordant for the risk for anxiety and depression. Neuroimage. 2008; E-pub: doi: 10.1016/j.neuroimage.2008.01.053.

26. Ekman P, Friesen W. Pictures of facial affect. Palo Alto, CA: Consulting Psychologists; 1976.

Chap

ter

Chapter 10

10

210

27. Genovese CR, Lazar NA, Nichols T. Thresholding of statistical maps in functional neuroimaging using the false discoveryrate. Neuroimage. 2002; 15: 870-878.

28. Walsh ND, Williams SC, Brammer MJ, Bullmore ET, Kim J, Suckling J et al. A longitudinal functional magnetic resonanceimaging study of verbal working memory in depression after antidepressant therapy. Biol Psychiatry. 2007; 62: 1236-1243.

29. Dougherty DD, Rauch SL. Brain correlates of antidepressant treatment outcome from neuroimaging studies indepression. Psychiatr Clin North Am. 2007; 30: 91-103.

30. Mayberg HS, Brannan SK, Mahurin RK, Jerabek PA, Brickman JS, Tekell JL et al. Cingulate function in depression: apotential predictor of treatment response. Neuroreport. 1997; 8: 1057-1061.

31. Chen CH, Suckling J, Ooi C, Fu CH, Williams SC, Walsh ND et al. Functional Coupling of the Amygdala in Depressed PatientsTreated with Antidepressant Medication. Neuropsychopharmacology. 2008; 33: 1909-1918.

32. Mayberg HS, Liotti M, Brannan SK, McGinnis S, Mahurin RK, Jerabek PA et al. Reciprocal limbic-cortical function andnegative mood: converging PET findings in depression and normal sadness. Am J Psychiatry. 1999; 156: 675-682.

33. Mayberg HS, Brannan SK, Tekell JL, Silva JA, Mahurin RK, McGinnis S et al. Regional metabolic effects of fluoxetine inmajor depression: serial changes and relationship to clinical response. Biol Psychiatry. 2000; 48: 830-843.

34. Malberg JE, Eisch AJ, Nestler EJ, Duman RS. Chronic antidepressant treatment increases neurogenesis in adult rathippocampus. J Neurosci. 2000; 20: 9104-9110.

35. Djavadian RL. Serotonin and neurogenesis in the hippocampal dentate gyrus of adult mammals. Acta Neurobiol Exp(Wars). 2004; 64: 189-200.

36. Martinowich K, Lu B. Interaction between BDNF and serotonin: role in mood disorders. Neuropsychopharmacology. 2008;33: 73-83.

37. Adli M, Baethge C, Heinz A, Langlitz N, Bauer M. Is dose escalation of antidepressants a rational strategy after a medium-dose treatment has failed? A systematic review. Eur Arch Psychiatry Clin Neurosci. 2005; 255: 387-400.

38. Baker CB, Tweedie R, Duval S, Woods SW. Evidence that the SSRI dose response in treating major depression should bereassessed: a meta-analysis. Depress Anxiety. 2003; 17: 1-9.



41. Anand A, Li Y, Wang Y, Wu J, Gao S, Bukhari L et al. Antidepressant effect on connectivity of the mood-regulating circuit:an FMRI study. Neuropsychopharmacology. 2005; 30: 1334-1344.

42. Mayberg HS. Modulating limbic-cortical circuits in depression: targets of antidepressant treatments. Semin ClinNeuropsychiatry. 2002; 7: 255-268.


44. Pezawas L, Meyer-Lindenberg A, Goldman AL, Verchinski BA, Chen G, Kolachana BS et al. Evidence of biologic epistasisbetween BDNF and SLC6A4 and implications for depression. Mol Psychiatry. 2008; 13: 709-716.

45. Knutson B, Adams CM, Fong GW, Hommer D. Anticipation of increasing monetary reward selectively recruits nucleusaccumbens. J Neurosci. 2001; 21: RC159.

46. Knutson B, Bhanji JP, Cooney RE, Atlas LY, Gotlib IH. Neural responses to monetary incentives in major depression. BiolPsychiatry. 2008; 63: 686-692.

47. Mayberg HS, Silva JA, Brannan SK, Tekell JL, Mahurin RK, McGinnis S et al. The functional neuroanatomy of the placeboeffect. Am J Psychiatry. 2002; 159: 728-737.

48. Fu CH, Williams SC, Brammer MJ, Suckling J, Kim J, Cleare AJ et al. Neural responses to happy facial expressions in majordepression following antidepressant treatment. Am J Psychiatry. 2007; 164: 599-607.

Chapter10

211


Tabl

e S1

0.1

Prev

ious

fMRI

stu

dies

of a

ffec

tive

stim

uli (

face

s/IA

PS) i

n M

DD

dur

ing

trea

tmen

t with

ant

idep

ress

ants

Aut

hor &

Dat

e (r

efer

ence

) Re

sult

s

Popu

latio

n (m

ean

age)

Sex

N

Des

ign

fMRI

par

adig

mR

oIs

stud

ied

/re

port

edRe

sults

Rem

arks

Dav

idso

n et

al.

(200

3)13

MD

D P

ts R

x-fr

ee ≥

? W

eeks

, HD

RS ≥

? (3

8.2

±9.3

) H

C (2

7.8

±10.

4)

M+F

12

5

8 w

eek,

VLX

225

mg

(ope

n),

3 sc

ans

Bsl,

2wk,

8w

k; 1.

5T

Neg

ativ

e, p

ositi

ve, n

eutr

al

IAPS

evo

king

hig

h ar

ousa

l.

L.In

s, L

.CG

(A),

Vi

sual

Cor

tex,

Am

yg

Act

ivat

ion

in L

.Ins

decr

ease

d in

Pts

vs

HC

(neg

. vs

neut

ral);

aft

er 2

wee

ks P

ts =

HC;

aft

er 8

wee

ks P

ts

incr

ease

d ac

tivat

ion

vs H

C (n

.s.)

Act

ivat

ion

L.CG

(A) d

ecre

ased

in P

ts v

s H

C (n

eg. v

s ne

urtr

al);

af

ter 8

wee

ks P

ts in

crea

sed

activ

atio

n vs

HC

(n.s

.)

Bsl A

myg

act

ivity

in P

ts ≈

HC;

Bsl

DLP

FC a

ctiv

ity in

Pt

s <

HC

(L) a

nd P

ts ≈

HC

(R).

Recr

uitm

ent b

y ad

vert

isem

ent,

ass

essm

ent

not s

peci

fied

. Hig

her p

re-t

reat

men

t CG

(A)

activ

atio

n pr

edic

ted

impr

ovem

ent.

Sta

rt

stim

ulus

and

sca

nnin

g m

anua

lly. N

o pl

aceb

o.

Res

pond

ers

only

: of o

rigin

ally

14 M

DD

Pts

2

non-

resp

onde

rs w

ere

excl

uded

.

Fu e

t al.

(200

4)10

MD

D P

ts R

x-fr

ee ≥

4 w

ks, H

DRS

≥18

(4

2.8

±6.7

) M

atch

ed H

C (4

3.2

±8.8

) M

+F

19

19

8 w

eekF

LX 2

0 m

g (o

pen)

, 2

scan

s Bs

l, 8w

k; 1.

5T S

ad

Ekm

an F

aces

of v

aria

ble

inte

nsity

: stu

dy o

f act

ivat

ion

and

dyna

mic

rang

e

HCM

, PH

C, A

myg

, H

ypot

hal,

Thal

, Ins

, CG

(A+D

), VS

tria

t N

C, In

fPC,

CER

Act

ivat

ion

in a

ll re

gion

s in

crea

sed

in p

ts v

s. H

C at

bs

l. Li

mbi

c an

d su

bcor

tical

act

ivat

ions

dec

reas

ed,

dyna

mic

rang

e in

neo

cort

ical

regi

ons

incr

ease

d,

in p

ts a

fter

8 w

eeks

. Dec

reas

e in

HD

RS

~ re

duct

ion

dyna

mic

rang

e in

CG

(A) a

nd C

ER.

Patie

nts

recr

uite

d by

adv

ertis

emen

ts, a

sses

s ed

by

SCID

, no

axis

1 co

-mor

bidi

ty. N

o pl

aceb

o,

13/1

9 w

eek

8 re

spon

ders

; 9 /1

9 re

mis

sion

; HD

RS

resp

onde

rs 5

.8 ±

2.7;

non

-res

pond

ers1

5.0

±4.9

28

Fu e

t al.

(200

7)48

MD

D P

ts R

x-fr

ee ≥

4 w

ks, H

DRS

≥18

(4

2.8

±6.7

) M

atch

ed H

C (4

3.2

±8.8

) M

+F

19

19

8 w

eekF

LX 2

0 m

g (o

pen)

, 2

scan

s Bs

l, 8w

k; 1.

5T H

appy

Ek

man

Fac

es o

f var

iabl

e in

tens

ity: s

tudy

of a

ctiv

atio

n an

d dy

nam

ic ra

nge

extr

astr

iata

l cor

tex,

Li

ngG

, CER

, Cun

, CG

(P),

Put

, Tha

l, H

CM, P

reC

Act

ivat

ion

in e

xtra

stria

tal c

orte

x, C

ER, L

ingG

, Cun

, CG

(P) l

ower

in p

ts v

s H

C at

bsl

. A

ctiv

atio

n in

crea

sed

in p

ts a

fter

8w

ks. P

ts h

ave

smal

ler

dyna

mic

rang

e in

Put

, Tha

l, Pr

eC, C

G(P

) vs

HC

at

base

line

and

wk

8.

See

unde

r Fu

et a

l. (2

004)

Har

mer

et a

l. (2

006)

17H

C w

ithou

t (lif

etim

e)

psyc

hiat

ric d

isor

der

24

7 da

ys d

oubl

e bl

ind

rand

omis

ed: P

L or

CIT

20

mg

(mat

ched

on

gend

er, a

ge a

nd

IQ),

1 sc

an a

t 7 d

ays;

1.5T

M

aske

d (1

7ms

+ 16

7ms

neut

ral)

fear

ful,

happ

y or

ne

utra

l Ekm

an fa

ces

Am

yg (L

+R)

Am

yg(L

+R) a

nd M

ed.P

FC a

ctiv

ity s

igni

fica

ntly

lo

wer

aft

er fe

arfu

l fac

es u

nder

CIT

vs

PL. N

o si

gn.

effe

cts

of C

IT v

s PL

on

happ

y fa

ces.

Recr

uitm

ent n

ot s

peci

fied,

ass

essm

ent n

ot

spec

ified

. No

base

line

scan

s. N

o pa

tient

s.

Dec

reas

ed M

ed.P

FC a

ctiv

ity m

ay b

e co

ntra

dict

ory

to d

ecre

ased

Med

.PFC

act

ivity

in

MD

D. E

xpla

ined

as

decr

ease

in H

C du

e to

CIT

in

hibi

tion

of A

myg

Nor

bury

et a

l. (2

007)

16H

C w

ithou

t (lif

etim

e)

psyc

hiat

ric d

isor

der

24

7 da

ys d

oubl

e bl

ind

rand

omis

ed: P

L or

REB

8m

g,

1 sca

n at

7 d

ays;

1.5T

Mas

ked

(17m

s +

183m

s ne

utra

l) an

d ov

ert (

200m

s) fe

arfu

l, ha

ppy

or n

eutr

al E

kman

face

s

Am

yg (L

+R)

Mas

ked

Am

yg(R

) act

ivity

sig

nifi

cant

ly lo

wer

aft

er

fear

ful f

aces

und

er R

EB v

s PL

; no

sign

diff

eren

ces

in o

vert

stim

ulus

pre

sent

atio

n or

Am

yg(L

).

Incr

ease

d ac

tivity

und

er R

EB v

s PL

in F

G(R

) aft

er

happ

y fa

ces

Recr

uitm

ent n

ot s

peci

fied,

ass

essm

ent n

ot

spec

ified

. No

base

line

scan

s. N

o pa

tient

s.

Resp

onse

to m

aske

d fa

ces

mig

ht re

pres

ent

effe

ct o

f REB

on

auto

mat

ic a

spec

ts o

f pr

oces

sing

.

212

Chap

ter

Chapter 10

10

Robe

rtso

n et

al.

(200

7)14

MD

D P

ts R

x-fr

ee ≥

? W

ks, H

DRS

≥15

(4

1.4 ±

7)

M+F

108

wee

k BU

P 15

0-45

0 m

g (o

pen)

, 2 sc

ans B

sl, 8

wk;

1.5

T N

eg. I

APS

(war

fare

/vio

lenc

e)

in E

mot

iona

l Odd

ball

Task

OFC

, VM

PFC,

CG

(A.P

), In

fPFC

, FG

, D

MPF

C, A

myg

, HC

Redu

ced

activ

atio

n in

OFC

(R),

VM

PFC(

R),

CG(A

.R),

InfP

FC(R

), H

C+A

myg

dala

(R),

NC(

R),

FG(R

), D

MPF

C(L)

, CG

(P.L

) aft

er 8

wee

ks o

f BU

P In

crea

sed

activ

atio

n in

InfP

FC(L

), F

G(L

). D

ecre

ase

in a

ctiv

atio

n in

OFC

(R),

InfP

FC(R

), FG

(R) a

nd

Am

yg (L

) cor

rela

ted

with

impr

ovem

ent i

n H

DR

S

Recr

uitm

ent n

ot s

peci

fied.

Sca

ns o

f 8 s

ubje

cts

com

plet

ing

follo

w-u

p us

ed fo

r ana

lyse

s. N

o pl

aceb

o, n

o co

ntro

ls. A

fter

BU

P, 2

/8 o

f sub

ject

s sh

owed

no

impr

ovem

ent,

4/8

reac

hed

HD

RS

≤7. 3

/8 p

ts w

ithdr

ew b

efor

e 8

wee

ks, s

cans

m

ade

at w

ithdr

awal

.

Scha

efer

et a

l. (2

006)

15M

DD

or d

ysth

ymia

Pts

R

x-fr

ee ≥

4 w

ks, H

DR

S ≥2

2 (3

5.9

±?)

HC

with

out 1

st d

egre

e re

lativ

es w

ith p

sych

. D

isor

d. (2

8.2

±?)

M+F

9 14

Aver

age

22 w

eek

VLX

150-

300

mg

(ope

n), 2

sca

ns B

sl, a

fter

re

mis

sion

(4 w

eeks

HD

RS

≤10)

; 1.5

T Po

s. IA

PS (s

ocia

l in

tera

ctio

n, s

ingl

e pe

rson

, hu

man

face

s, p

eopl

e, e

rotic

a,

nonh

uman

app

etiv

e) in

pa

ssiv

e vi

ewin

g

Inf.,

Med

., Su

p.PF

C,

Ins,

Med

.DN

T,

FG(R

)

For F

aces

vs

peop

le/n

onhu

man

app

etiti

ve: I

n pa

tient

s si

gn in

crea

sed

activ

ity a

t 2nd

sca

n in

In

f.,M

ed.,S

up.P

FC(L

), In

s(L,

R),

Med

.DN

T(R)

, FG

(R),

whi

le in

con

trol

s th

ese

regi

ons

show

ed

sign

, inc

reas

ed a

ctiv

ity a

t Bsl

sca

n on

ly.

Recr

uitm

ent b

y ad

vert

isem

ent,

ass

essm

ent b

y SC

ID. 2

nd s

can

afte

r rem

issi

on o

nly.

6/9

pa

tient

s re

ache

d re

mis

sion

. Tw

o ot

her

cont

rast

s: s

ocia

l int

erac

tion

and

erot

ica

also

pr

esen

ted

in m

anus

crip

t. In

exp

erim

ents

for

face

s ex

cept

for I

ns a

nd F

G(R

) a s

igni

fica

nt

late

ralit

y w

as o

bser

ved.

No

inte

ract

ion

with

cl

inic

al im

prov

emen

t or V

LX d

ose

foun

d.

Shel

ine

et a

l. (2

001)

12M

DD

Pts

Rx-

free

≥4

wks

, HD

RS ≥

18

(40.

3 ±

?)

Mat

ched

HC

HD

RS <

8 (3

9.8

±?)

M +

F

11

11

8 w

eek

SER

~10

0 m

g (o

pen)

, 2

scan

s Bs

l, 8w

k; 1.

5T M

aske

d (4

0ms)

fear

ful,

neut

ral o

r ha

ppy

follo

wed

by

neut

ral

(160

ms)

Ekm

an F

aces

Am

yg (L

+R)

Bsl:

L.A

myg

act

ivat

ion

Pts

> H

C; R

.Am

yg P

ts≥H

C (n

.s.)

(fea

rful

/hap

py v

s ne

utra

l) Pt

s: a

fter

SER

, L+

R A

myg

act

ivat

ion

afte

r all

or fe

arfu

l fac

es s

ign.

D

ecre

ased

. Aft

er S

ER L

+R A

myg

act

ivity

in P

ts =

H

C

Recr

uitm

ent b

y ad

vert

isem

ent,

ass

essm

ent b

y D

SM-IV

crit

eria

. Aft

er S

ER m

ean

HD

RS

decr

ease

d fo

rm 2

3.3

to 9

.7. O

nly

1 non

-re

spon

der a

nd 2

par

tial r

espo

nder

s. N

o co

rrel

atio

n be

twee

n L.

Am

yg s

igna

l int

ensi

ty

chan

ge a

nd H

DR

S or

anx

iety

Abb

revi

atio

ns: A

myg

= A

myg

dala

, BU

P= b

upro

pion

, CER

= Ce

rebe

llum

, CG

= Ci

ngul

ate

gyru

s (A

-ant

erio

r, D

-dor

sal,P

-pos

terio

r), C

IT=

Cita

lopr

am, C

un=

Cune

us, D

MPF

C= D

orsa

l Med

ial P

refr

onta

l Cor

tex,

FLX

= flu

oxet

ine,

FG

= Fu

sifo

rm G

yrus

, HC=

Hea

lthy

Cont

rol,

HCM

= H

ippo

cam

pus,

HD

RS=

Ham

ilton

Dep

ress

ion

Rat

ing

Scal

e, H

ypot

hal=

Hyp

otha

lam

us, I

nfFC

= In

fFro

ntal

Cor

tex,

InfP

C= In

ferio

r Par

ieta

l Cor

tex,

In

s= In

sula

, L=

left

, Lin

gG=

Ling

ual g

yrus

, MD

D=

Maj

or D

epre

ssiv

e D

isor

der,

Med

.DN

T= M

edia

l Dor

sal N

ucle

us o

f Tha

lam

us, N

C= N

ucle

us C

auda

tus,

PFC

= Pr

efro

ntal

Cor

tex,

PH

C= P

arah

ippo

cam

pal g

yrus

, PL

= pl

aceb

o, P

reC=

Pre

cune

us, P

ts=

patie

nts,

Put

= Pu

tam

en, R

= rig

ht, R

EB=

Reb

oxet

ine,

SER

= se

rtra

line,

Tha

l= T

hala

mus

, VLX

= ve

nlaf

axin

e, V

Stria

t= V

entr

al s

tria

tum

.

Tabl

e S1

0.1

Prev

ious

fMRI

stu

dies

of a

ffec

tive

stim

uli (

face

s/IA

PS) i

n M

DD

dur

ing

trea

tmen

t with

ant

idep

ress

ants

(Con

tinue

d)

Aut

hor &

Dat

e (r

efer

ence

) Re

sult

s

Popu

latio

n (m

ean

age)

Sex

N

Des

ign

fMRI

par

adig

mR

oIs

stud

ied

/re

port

edRe

sults

Rem

arks

Chapter10

213


Table S10.2 Main effects in study-entry scans: MDD-patients and HC combined.


MDD + HC

All Faces Amygdala R 20 -4 -12 12 3.73 <0.001

L -18 -4 -15 8 4.00 <0.001

Fusiform gyrus R 40 -52 -21 160 6.01 <0.001

L -42 -56 -24 99 5.34 <0.001

DLPFC R 46 16 30 405 5.94 <0.001

L -40 24 21 15 3.61 <0.001

Insula (Ant.) R 36 30 -3 249 5.21 <0.001

L -34 30 -6 119 4.49 <0.001

OFC R 32 34 -12 * 4.77 <0.001

Occipital R 12 -94 6 88 4.86 <0.001

L -14 -98 15 64 4.99 <0.001

Parietal R 48 -50 6 126 4.72 <0.001

R 48 -68 12 34 4.03 <0.001

DMPFC R 4 22 45 35 3.91 <0.001

Cuneus R 4 -64 33 15 3.84 <0.001

Neg. Faces Amygdala L -18 -4 -15 8 3.95 <0.001

Fusiform gyrus R 40 -52 -21 165 6.28 <0.001

L -42 -56 -24 82 4.61 <0.001

DLPFC R 42 16 27 339 5.65 <0.001

OFC R 34 36 -12 107 4.37 <0.001

Insula (Ant.) R 36 30 -3 * 4.35 <0.001

Occipital R 14 -96 9 45 4.31 <0.001

L -12 -98 15 56 5.24 <0.001

Parietal R 48 -50 6 61 4.25 <0.001

R 50 -68 12 9 3.73 <0.001

DMPFC R 4 22 45 11 3.66 <0.001

Cuneus R 4 -60 33 10 3.93 <0.001

Hap. Faces DLPFC R 46 16 30 173 5.60 <0.001

Fusiform gyrus R 38 -56 -21 90 5.36 <0.001

L -44 -56 -24 61 4.62 <0.001

Insula (Ant.) R 36 22 3 163 5.01 <0.001

L -34 24 0 72 4.44 <0.001

OFC R 32 34 -12 * 4.70 <0.001

Occipital R 16 -94 6 38 4.41 <0.001

L -10 -96 9 17 3.98 <0.001

Thalamus R 4 -14 6 20 4.04 <0.001

Parietal R 52 -50 6 8 3.80 <0.001

Clusters are FDR (p<0.05)-corrected with extend threshold of 3 voxels.* cluster size not reported, one region extended from right anterior insula to right OFC Abbreviations: see also Table 10.3 and 10.4. DMPFC= dorsomedial prefrontal cortex

Chap

ter

Chapter 10

10

214

Table S10.3 Changes after 6 weeks of treatment (T0) compared with study-entry.Contrast Brain region L/R x,y,z (MNI mm) Cluster size (k) Max. voxel Z p

A. Study-entry > T0


L -16 0 -15 12 1.97 0.024

Cuneus LR 0 -66 12 67 4.27 <0.001

Hippocampus, posterior L -18 -48 9 7 3.60 <0.001

R 28 -36 0 31 3.27 0.001

Neg. Faces Amygdala L -14 2 -12 11 2.25 0.012

Cuneus L -2 -66 12 51 3.97 <0.001

Hap. Faces Sup.Temporal Sulcus L -48 -46 6 11 3.48 <0.001

DLPFC L -50 0 15 6 3.37 <0.001

Insula L -36 16 -18 31 3.36 <0.001

B. T0 > study-entry

All Faces Amygdala R 28 -2 -15 28 3.22 0.001

Cingulate, posterior L -12 -18 48 21 4.13 <0.001

medial L 2 4 33 70 3.37 <0.001

pregenual R 4 36 3 15 3.41 <0.001

Sup.Temporal gyrus R 62 -8 -9 8 3.79 <0.001

Hippocampus, dorsal R 28 -28 -12 31 3.74 <0.001

DMFPC L -10 22 60 32 3.39 <0.001

Inf. Temporal Gyrus L -44 -4 -39 14 3.28 0.001

Neg. Faces Amygdala R 28 0 -15 47 3.47 <0.001

DMFPC L -4 22 60 288 4.56 <0.001

Cingulate, anterior R 2 30 33 133 3.21 0.001

medial LR 2 2 36 69 3.27 0.001

pregenual R 6 34 3 5 3.43 <0.001

VLPFC L -24 58 30 82 3.89 <0.001

DLPFC L -50 26 -3 16 3.65 <0.001

L -28 44 42 55 3.51 <0.001

L -30 22 54 14 3.28 0.001

R 32 -2 54 12 3.22 0.001

R 24 42 48 136 3.09 0.001

Hippocampus, dorsal R 30 -28 -9 48 3.53 <0.001

Cerebellum L -22 -34 -27 15 3.21 0.001

Sup. Temporal gyrus R 62 -8 -9 7 3.17 0.001

N.Accumbens L -4 8 -12 9 3.13 0.001

Hap. Faces Cerebellum L -24 -32 -24 23 3.78 <0.001

Sup.Temporal gyrus R 62 -8 -9 3 3.57 <0.001

Amygdala R 22 -12 -18 8 3.22 0.001

L -18 -10 -18 3 2.47 0.007

Abbreviations: see Table 10.3, 10.4 and S10.2.

Chapter10

215


Table S10.4 Positive and negative correlations with amygdala activation (negative faces).


A. Left Amygdala (MNI -22, 2, -18) activation

Positive* Hippocampus R 26 -18 -12 102 4.24 <0.001

Amygdala R 22 4 -18 127 5.09 <0.001

OFC R 24 14 -21 † 4.34 <0.001

L -38 18 -21 33 3.77 <0.001

Parahippocampal gyrus L -30 -34 -18 98 4.25 <0.001

Insula R 38 -6 12 18 3.60 <0.001

Cingulate, subgenual R 6 14 -12 10 3.59 <0.001

Parietal R 52 -54 -15 37 3.88 <0.001

Negative‡ Precuneus R 14 -58 21 4 2.78 0.003

Cingulate, pregenual R 4 50 0 14 2.73 0.001

B. Right Amygdala (MNI 18, -2, -18) activation

Positive* Hippocampus R 24 -22 -12 † 4.56 <0.001

R 30 -32 -6 17 4.51 <0.001

L -16 -30 -6 57 5.07 <0.001

Amygdala L -18 -4 -21 65 4.49 <0.001

L -18 -10 -12 42 4.20 <0.001

OFC L -38 16 -21 40 4.01 <0.001

L -34 34 -12 18 3.77 <0.001

R 40 14 -21 13 3.48 <0.001

DLPFC R 58 14 36 14 3.70 <0.001

Negative‡ DLPFC L -22 16 57 6 3.16 0.001

R 22 18 63 5 2.87 0.002

Accumbens L -10 12 -9 3 2.68 0.004 * activations for p<0.001 with extend voxel size 10† in one cluster with right amygdala ‡ activations for p<0.01 with extend voxel size 3 Abbreviations: see Table 10.3, 10.4 and S10.2.

Table S10.5 Changes between T0 and T1 for true dose-escalation (paroxetine 30-50 mg) relative to placebo dose-escalation (paroxetine 20 mg/day).


A. Decrease in activation by DE relative to placebo

All Faces DMPFC L -6 14 51 333 4.27 <0.001

DLPFC L -24 16 51 96 4.31 <0.001

R 26 14 63 7 3.10 0.001

Thalamus, ventrolateral L -8 -18 3 96 4.26 <0.001

Motor cortex R 26 -2 63 138 4.19 <0.001

L -4 -6 66 56 3.97 <0.001

OFC L -38 28 -3 226 3.93 <0.001

Insula R 44 -6 6 100 3.91 <0.001

R 36 22 3 110 3.71 <0.001

L -42 -10 3 42 3.65 <0.001

Putamen R 28 0 -3 24 3.84 <0.001

Parietal R 62 -36 39 86 3.69 <0.001

R 66 -32 18 60 3.64 <0.001

R 56 -52 39 99 3.48 <0.001

Cingulate, anterior R 14 32 18 36 3.30 <0.001

Precuneus R 12 -60 21 42 3.42 <0.001

VMPFC R 16 58 33 37 3.28 0.001

Sup.Temporal Gyrus L -42 -32 12 19 3.15 0.001

G. Pallidus L -10 0 -3 10 3.14 0.001

Chap

ter

Chapter 10

10

216

Neg. Faces Motor cortex LR -4 -6 66 62 4.42 <0.001

LR -8 -14 48 85 3.24 0.001

DLPFC L -18 6 57 101 4.22 <0.001

L -24 16 51 77 3.60 <0.001

DMFPC L -6 14 51 * 3.72 <0.001

Thalamus, ventrolateral L -6 -18 0 60 3.80 <0.001

Insula R 36 22 3 72 3.74 <0.001

R 44 -6 6 41 3.62 <0.001

OFC R 38 34 6 † 3.46 <0.001

L -36 30 -3 61 3.55 <0.001

Cingulate, anterior R 12 30 21 14 3.28 0.001

Precuneus R 12 -60 21 9 3.24 0.001

Hap. Faces Insula R 36 12 0 207 4.21 <0.001

R 32 16 -21 40 3.67 <0.001

Putamen R 22 12 3 134 3.20 0.001

Parietal L -60 -36 18 109 4.08 <0.001

R 62 -28 42 166 3.10 0.001

R 26 -10 60 10 3.55 <0.001

R 22 -26 66 67 3.21 0.001

Cingulate, anterior L -6 32 18 218 3.17 0.001

R 12 42 24 180‡ 3.59 <0.001

pregenual L -8 46 -6 127 3.90 <0.001

VMPFC R 22 58 18 180 3.91 <0.001

OFC L -28 42 -15 ¶ 3.00 0.001

R 22 46 -15 136 3.43 <0.001

R 16 16 -15 23 3.23 0.001

Premotor Cortx R 4 0 54 14 3.21 0.001

DLPFC R 26 38 45 19 3.15 0.001

L -24 20 54 67 3.10 0.001

B. Increase in activation by DE relative to placebo

All Faces Med.Temporal Gyrus L -52 -6 -12 32 4.02 <0.001

R 60 -34 -3 19 3.44 <0.001

Hippocampus R 24 -14 -18 69 3.85 <0.001

Cingulate, subgenual L -2 22 -6 48 3.65 <0.001

Parietal L -60 -42 -6 25 3.63 <0.001

DLPFC L -54 24 15 45 3.16 0.001

Cerebellum R 12 -46 -15 23 3.14 0.001

Neg. Faces Cingulate, subgenual L -2 22 -6 66 4.16 <0.001

Med.Temporal Gyrus L -52 -6 -12 65 4.12 <0.001

Hippocampus R 20 -24 -21 147 3.85 <0.001

Parietal R 8 -34 57 6 3.4 <0.001

L -40 -34 66 9 3.19 0.001

Cerebellum LR -6 -42 -15 79 3.39 <0.001

Parahippocampal gyrus L -26 -32 -21 19 3.10 0.001

Hap. Faces Med.Temporal Gyrus L -60 -42 -6 43 4.51 <0.001

Parahippocapal gyrus L -28 -32 -12 12 3.55 <0.001

DLPFC L -58 20 21 43 3.01 <0.001 * in one cluster with left DLPFC † in one cluster with right insula‡ in one cluster with right VMPFC ¶ in one cluster with left pregenual cingulate Abbreviations: see also Table 10.3, 10.4 and S10.2. VMPFC= ventromedial prefrontal cortex.

Table S10.5 Changes between T0 and T1 for true dose-escalation (paroxetine 30-50 mg) relative to placebo dose-escalation (paroxetine 20 mg/day). (Continued)


CHA

PTER

EFFECT OF THE SELECTIVE SEROTONIN REUPTAKE

INHIBITOR PAROXETINE ON PLATELET FUNCTION IS MODIFIED BY A SLC6A4 SEROTONIN

TRANSPORTER POLYMORPHISM

Journal of Thrombosis and Haemostasis 2008; in press

Nadia Abdelmalik1,2*

Henricus G. Ruhé1*

Kurdo Barwari2

Erik-Jan van den Dool3

Joost C.M. Meijers2

Saskia Middeldorp2,4,5

Harry R. Büller2

Aart H. Schene1

Pieter W. Kamphuisen2

* Both authors contributed equally to the manuscript1 Program for Mood Disorders, Department of Psychiatry, Academic Medical Center,University ofAmsterdam, Amsterdam, the Netherlands2 Department of Vascular Medicine, Academic Medical Center, University of Amsterdam, Amsterdam,the Netherlands 3 Department of Clinical Chemistry, Academic Medical Center, University of Amsterdam, Amsterdam,the Netherlands4 Department of Clinical Epidemiology, Leiden University Medical Center, Leiden, the Netherlands 5 Department of General Internal Medicine and Endocrinology, Leiden University Medical Center,Leiden, the Netherlands

Chap

ter

Chapter 11

11

218

Abstract

BackgroundSelective serotonin reuptake inhibitors (SSRIs) have been associated with an increased bleedingtendency.

AimTo prospectively quantify the dose-response effects of paroxetine and the influence of theserotonin transporter gene (SLC6A4) promoter polymorphism (5-HTTLPR) on platelet function.

MethodsNineteen drug-free psychiatric outpatients (44.5 ±10.8 years), were tested before and after 6weeks of paroxetine treatment (20 mg/day). Based on clinical symptoms, paroxetine dosageswere increased (40-50 mg/day) for 6 more weeks in 11 patients. Parameters related to plateletfunction were assessed by bleeding time, platelet function analyzer (PFA), platelet serotonin,platelet factor 4 (PF4), β-thromboglobulin (β-TG), and aggregation tests.

ResultsParoxetine 20 mg/day increased mean bleeding time by 1.2 minutes (95% confidence interval (95%CI) -0.2-2.7) and reduced median platelet serotonin level (463 ng/109 platelets; Inter QuartileRange (IQR) 361-666), and platelet ß-TG concentration (3.1 IU/106 platelets; IQR 0.3-6.0). Otherplatelet parameters did not change significantly. Serial platelet aggregation tests did not becomeabnormal. Paroxetine dose-escalation did not further influence platelet function. However, 5-HTTLPR polymorphisms modified these effects: in LA/LA-carriers, bleeding times did not change (-0.2 minutes; (95% CI -0.6 to 0.9)), while bleeding times significantly increased in <2LA-allele carriers(2.3 minutes (95% CI 0.5 to 4.07); p= 0.032). Platelet serotonin decreases were larger in patientswithout LA-alleles (868 ng/109 platelets; (IQR 585 to 1213)) than in ≥1 LA-allele carriers (457 ng/109

platelets; (IQR 392 to 598); p= 0.035). PFA closure time and PF4 increased significantly in patientswithout LA-alleles.

ConclusionsParoxetine 20 mg/day does not increase overall bleeding time, but impairs platelet function bydecreasing the levels of platelet serotonin and platelet ß-TG. These paroxetine effects appear tobe mediated by 5-HTTLPR, with most pronounced effects in patients without LA-alleles.

ChapterParoxetine induced platelet function changes

11

219

Introduction

Selective serotonin reuptake inhibitors (SSRIs) belong to the most widely prescribed classes ofdrugs and are often used for the treatment of depression and anxiety disorders.1 However,because of their extensive use, the impact of relatively infrequent side effects is of potentialclinical importance. One of these harmful side effects is an increased bleeding tendency.

Over the years, numerous case reports2-6 and case control studies7-11 have suggested that SSRIuse is associated with an increased risk of bleeding. The bleeding pattern varied from minorbleeding to spontaneous gastro-intestinal bleeding and excessive perioperative hemorrhage. Onestudy showed that SSRI users who underwent orthopedic surgery had a nearly four-fold increasedrisk to require blood transfusion.2 In addition, large population based case-control studies haveconsistently shown an increased incidence of upper gastrointestinal bleeding associated withSSRI use,12 especially when co-medicated with nonsteroidal anti-inflammatory drugs, aspirin8;11 orcoumarins.13

The underlying pathophysiological mechanisms of the increased bleeding tendency of SSRIswere previously studied, and significant changes in platelet function related to SSRI use wereconsistently reported. These changes included a decreased aggregation with epinephrine14;15 andcollagen,15;16 a prolongation of the PFA closure time,17 and a decrease in plasma beta-thromboglobulin (β-TG)18 and plasma platelet factor 4 (PF4).18;19

Furthermore, SSRIs block serotonin transporters (SERTs), which increases serotonergictransmission between neurons and results in antidepressant effects. However, SSRIs also blockthe SERTs of blood platelets, causing decreased platelet uptake of serotonin.20 Serotoninmediates vasoconstriction, platelet aggregation, and platelet activation after vessel injury, andsince platelets cannot synthesize serotonin, treatment with an SSRI could lead to depletion ofplatelet serotonin and impair hemostasis.20

Polymorphisms of the SERT gene (SLC6A4) promoter region (5-HTTLPR) are associated withthe transcriptional activity of the SERT gene and the rate of serotonin uptake.21 The 5-HTTLPRpolymorphism is located approximately 1 kb upstream of the transcription initiation site and iscomposed of 16 repeat elements. Human lymphoblasts homozygous for the long (L) 5-HTTLPRallele produce higher concentrations of SERT mRNA than cells containing one or two copies of theshort (S) allele. Furthermore, the rate of serotonin uptake by the transporter is more than twofoldhigher in cells homozygous for the L allele.21 In Caucasians the L allele is found more frequently(57%) than the S allele (43%), with a 5-HTTLPR genotype distribution of 32% LL, 49% LS and 19% SS.However, other populations show different prevalences, especially Asians, who have more thantwo-fold higher SS genotype frequencies.22 Nowadays, the 5-HTTLPR polymorphism is consideredtri-allelic.23 The L allele can be subdivided in an LG and an LA variant by a common SNP (rs25531),which creates a functional transcription factor binding site (LG), behaving like a short allele. Thistri-allelic classification is probably more reliable to find associations between 5-HTTLPRpolymorphisms and phenotypes (e.g. SSRI adverse effects).24 In Caucasians the S:LA:LG ratio isapproximately 8:10:2, while in black patients this is 5:10:5.23

Although an association between SSRI use and platelet dysfunction is undoubted,14;17;18;25;26 itremains uncertain whether a dose-response relationship exists between SSRI use and plateletfunction.20 In addition, since functional polymorphisms of the serotonin transporter influence therate of serotonin uptake, they may also mediate the effects on hemostasis.27

In this study, we evaluated the effect of standard and increasing dosages of the SSRIparoxetine on the platelet function of SSRI free patients. In addition, we assessed whether thiseffect is influenced by the 5-HTTLPR/rs25531 polymorphism.

Chap

ter

Chapter 11

11

220

Methods

ParticipantsAfter approval by the local ethics committee and written informed consent, we recruited 19 adultpsychiatric outpatients (18-75 years) from February 2006 until February 2007. The inclusioncriterion was indication of SSRI use (for mood or anxiety disorders) as determined by the treatingphysician. We excluded patients who used aspirin, NSAIDs (in the preceding 48 hours) orantidepressants (less than 4 weeks prior to inclusion), and patients who had a history of increasedbleeding tendency.

Treatment and timing of measurementsAfter baseline assessment, patients were treated with paroxetine 20 mg/day for 6 weeks. If at 6weeks the clinical symptoms of the patients had not improved by 50%, physicians could increasethe dose (dose-escalation) to a maximum of 50 mg/day over the next 6 weeks, as recommendedby current psychiatric treatment guidelines.1 As a measure to detect increased bleeding tendency,in all patients the bleeding severity score (Tosetto et al.)28 was assessed at baseline, after 6 weeks,and after 12 weeks of treatment (if a dose-escalation was prescribed). Platelet parameters wereassessed at the same time points by technicians that were unaware of the treatment regimen.

Measurements of platelet parametersBlood samplesBlood was collected in open 10 mL tubes containing 1 ml 3.2% sodium citrate by venapuncturefrom the antecubital vein using a 19-gauge needle. Within 60 minutes after collection, platelet richplasma (PRP) was prepared by centrifugation at 190 g at room temperature for 10 minuteswithout braking. PRP was separated from the packed cells; after 30 minutes aggregation studieswere started. PECT tubes for 4 ml blood containing 0.4 ml of a mixture of 94 nmol/L ProstaglandinE1, 90 mmol/L EDTA, 0.63 mmol/L sodium carbonate and 10 mmol/L theophilline were used for thedetermination of ß-TG and PF4 in plasma. Blood samples were collected in PECT tubes to preventin vitro release of ß-TG and PF4 from platelets. K3EDTA 7.5% anti-coagulant tubes were used forthe determination of platelet count.

Platelet parametersBleeding time was measured with a standard incision using the Surgicutt device (Surgicutt Adult,ITC Edison, USA). Platelet function was measured with the PFA-100 (Siemens HealthcareDiagnostics, Marburg Germany) using epinephrine (EPI) and ADP containing cartridges. Plateletaggregation in platelet-rich-plasma with ADP, ristocetin, collagen and arachidonic acid wasanalyzed qualitatively with a standardized aggregometer (model 540-vs, Chrono-Log Corporation,Havertown, USA).

The amount of ß-TG and PF4 in platelets and plasma was measured using enzyme-linkedimmunosorbent assays (Asserachrom β-TG and PF4, Diagnostica Stago, Roche). ß-TG and PF4 inplasma were determined in PECT samples. ß-TG and PF4 in platelets were determined afterpelleting the platelets from PRP. The platelets were then destroyed by a combination of Triton (2%Triton X-100) and sonication for 15 seconds on ice (microtip, Branson, amplitude 50%). After 5minutes of centrifugation at 13,000 rpm, the supernatant was used for analysis.

SerotoninIn the sonicated supernatant, the amount of serotonin was measured in an acidified (HClO4)environment by fluorimetry (Fluostar Galaxy, BMG Offenburg, Germany).


11

221

Paroxetine serum concentrationsWe collected blood samples for measurement of paroxetine serum concentrations (PSC) after 6and 12 weeks of paroxetine treatment from 19 and 11 patients, respectively. PSC were measuredusing a validated HPLC-MS/MS method (details available on request). The lower limit ofquantification was 5 µg/L, which was at the lower end of the therapeutic range of paroxetine inserum (5-75 µg/L). The lower limit of detection was 0.3 µg/L.

Genotyping procedures and analysisGenotyping was performed as described earlier.23 In brief, genomic deoxyribonucleic acid (DNA)was isolated and the length of the 5-HTTLPR polymorphism was determined by gelelectrophoresis. The region around the polymorphism was amplified by PCR. Genotyping of thers25531 SNP was done by sequencing. The length of the 5-HTTLPR polymorphism was confirmedby looking at the length of the sequenced PCR product. Taking into account this SNP, wereclassified the tri-allelic genotypes into a modified bi-allelic classification: S’/S’ (S/S, LG/S, LG/LG =non LA), S’/LA (S/LA, LG/LA = 1 LA) and LA/LA (= 2 LA). Because it is unknown which allele is dominant(heterozygosity), we either grouped S’/S’ and S’/ LA as <2 LA alleles versus 2 LA,or contrasted ≥1 LAvs. non LA genotypes in our analyses.

Statistical analyses We first computed means for the platelet parameters at baseline, after 6 and 12 weeks ofparoxetine administration. We compared patients who received a further dose-escalation fromweek 6 onwards with patients who remained at 20 mg/day with χ2 tests for categorical data andindependent T-tests for continuous data. We compared the changes in platelet parameters overtime with paired T-tests (baseline vs. 6 weeks and 6 vs. 12 weeks). In case of a non-normaldistribution (assessed by histograms and Shapiro Wilks’ test) we present medians withinterquartile ranges (IQR) and used nonparametric Wilcoxon signed ranks tests for paired data.For the platelet parameters that appeared to change significantly after 6 weeks of treatment, wefurther explored the changes in linear mixed models;29 we examined the effects of paroxetinedosage by the significance of a dose*time interaction.

In addition, we investigated the relation of PSC with changes in relevant platelet parametersin linear regression models using all observations (after 6 and 12 weeks) at once. We introducedthe 5-HTTLPR polymorphism to explore the genetic influence on these relations. Because of oursample size and non-normal distribution of bleeding times and platelet parameters, we testeddifferences between genotypes with nonparametric Mann-Whitney tests. All analyses wereperformed in SPSS v15.0.1.

Results

PatientsNineteen subjects who were diagnosed with either a mood disorder (n= 13), an anxiety disorder(n= 2) or both (n= 4) participated in this study. The mean age was 44.5 years (SD 10.8) and 11 (58%)were female. Co-medication consisted of depakine (for epilepsy; n= 1), levothyroxine (n= 1) andomeprazol (n= 1).

All subjects received a standard dose of paroxetine 20 mg/day for 6 weeks(mean PSC 25.6 µg/L (95% CI 14.3 to 36.9)). In 11 subjects who failed to show sufficient clinical improvement after thisperiod, paroxetine dosages were increased to 40 (n= 4) or 50 mg/day (n= 7) for another 6 weeks(mean PSC 107.2 µg/L (95% CI 42.2 to 172.2)). The group that only received standard dosage wascomparable with the increased dosage group in type of psychiatric disorder, age and sex (Table11.1).

Chap

ter

Chapter 11

11

222

Bleeding severity scoresAt baseline, the bleeding severity score was low (<3) in all subjects. Six weeks at paroxetine 20mg/day, or a subsequent dose-escalation for 6 weeks thereafter did not affect the bleeding score.

Platelet parametersSix weeks of treatment with paroxetine non-significantly increased the mean bleeding time by 1.2minutes (95% CI -0.2 to 2.7; Paired t-test; p= 0.083; Table 11.2). Furthermore, paroxetine resulted ina significant reduction of median platelet serotonin level of 463 ng/109 platelets (IQR 361 to 666;Wilcoxon signed ranks test; p<0.001), and median platelet ß-TG level of 3.1 international units (IU)/106 platelets (IQR 0.3 to 6.0; Wilcoxon signed ranks test; p= 0.016).Other changes in plateletparameters were not statistically significant, nor did the aggregation tests reveal qualitativeabnormalities.

Dose-escalation of paroxetine to 40-50 mg/day for another 6 weeks did not lead to furtherreduction of platelet serotonin or platelet ß-TG levels or to a further increase of the bleeding time(Table 11.3). The absence of further changes in platelet ß-TG and serotonin levels were confirmedby non-significant dose*time interactions in mixed models.

There was no correlation between plasma paroxetine levels and changes in platelet functiontests, including the change in platelet serotonin and platelet ß-TG levels. Platelet aggregationtests with ADP, ristocetin, collagen and arachidonic acid were not influenced by paroxetineadministration (data not shown).

Table 11.1. Baseline characteristics of patients starting treatment with paroxetine 20 mg/day.

All (n= 19)

Dose not increased after 6 weeks (n= 8)

Dose-escalation 40-50 mg after 6 weeks (n= 11)

Age ±SD 44.5 ±10.8 45.6 ± 13.2 43.6 ± 9.2

M/F ratio 8/11 4/4 4/7

Psychiatric diagnosisMood disorder*

Anxiety disorder†

Mood + anxiety disorder Substance abuse‡

Eating disorder

n (%) 13 (68.4%) 2 (10.5%) 4 (21.1%) 4 (21.1%) 1 (5.3%)

n (%) 5 (62.5%) 1 (12.5%) 2 (25.0%) 2 (25%) 1 (12.5%)

n (%) 8 (72.7%) 1 (9.1%) 2 (18.2%) 2 (18.2%) -

* Unipolar depression (n= 16; 89.5%) and/or dysthymia (n= 3; 15.8%)† Panic disorder with or without agoraphobia (n= 4; 21.1%), Post traumatic stress disorder (n= 2; 10.5%)‡ Cannabis abuse (n= 1; 5.3%), Alcohol and cannabis abuse (n= 2; 10.5%), Benzodiazepine abuse (n= 1; 5.3%)

Table 11.2. Changes in platelet function after 6 weeks of treatment with paroxetine 20 mg/day (total n= 19).

Baseline (mean ±SD)

6 weeks (mean ±SD)

n* Difference baseline vs. 6 weeks (mean 95% CI)

p†

Platelets (x109/l) 252 ±58 249 ±59 19 -3.2 (-14.0 – 7.6) 0.547

Bleeding time (min) 4.3 ±1.1 5.3 ±2.0 13 1.2 (-0.2 – 2.7) 0.083

PFA-ADP (sec) 85.7 ±18.4 88.9 ±18.9 19 3.2 (-7.2 – 13.5) 0.530

PFA-epinephrine (sec) 112.4 ±26.4 119.5 ±33.9 19 7.1 (-7.8 – 22.0) 0.329

(median IQR) (median IQR) (median IQR)

PF4 platelets‡ (IU/106 platelets) 15.9 (14.7 – 16.7) 15.9 (13.2 – 17.9) 17 -0.2 (-2.3 – 2.7) 0.925

PF4 plasma‡ (IU/mL) 6.9 (4.0 – 12.1) 5.8 (3.4 – 17.6) 18 -0.7 (-6.8 – 8.9) 0.983

ß-TG platelets‡(IU/106 platelets) 33.5 (30.4 – 36.4) 30.7 (28.1 – 34.2) 17 -3.1 (-6.0 - -0.3) 0.016

ß-TG plasma‡ (IU/mL) 35.8 (22.4 -39.7) 28.0 (20.3 – 64.7) 18 -2.2 (-16.4 – 22.0) 0.879

Platelet serotonin‡(ng/109 platelets) 667 (489 – 758) 128 (80 – 141) 16 -463 (-666 - -361) <0.001 * n for pairs without missing values† p-values for paired t-tests for normally distributed data and paired Wilcoxon signed ranks test for non-normally distributed data‡ due to non-normal distribution medians and interquartile ranges (IQR) are shown.


11

223

Effects of the 5-HTTLPR polymorphismGenotype was analyzed in eighteen patients, of whom 3 had the S’/S’, 9 the S’/LA, and 6 the LA/ LAgenotype, respectively. For these 18 patients, 26 observations after 6 and 12 weeks were availablewhen combining data from these time points. In patients with the 2LA genotype, paroxetinetreatment did not increase bleeding time (-0.2 minutes; 95% CI -0.6 to 0.9), while in patientscarrying the <2LA genotype, bleeding time increased by 2.3 minutes (95%CI 0.5 to 4.1, Mann-Whitney test; p= 0.032). In a linear regression model the <2LA genotype significantly predicted thechange in bleeding time (p<0.001). This difference in bleeding time between the two genotypegroups was not explained by a decrease in platelet serotonin, dosage or PSC (linear regression;pchange> 0.37).

The median decrease in platelet serotonin levels of patients with paroxetine treatment wasnot different between the <2LA and 2LA genotypes (<2LA: 587 ng/109 platelets; IQR 393 to 841; 2LA:457 ng/109 platelets; IQR 373 to 582; Mann-Whitney test; p= 0.244). However, in patients withoutLA alleles, we found significant decreases in serotonin, and increases in PFA-ADP, PFA-EPI andplatelet PF4 after 6 and 12 weeks of paroxetine treatment (Mann-Whitney test; Table 11.4; Figure11.1).

Table 11.3. Changes in platelet parameters in patients who received a secondary dose-escalation (40-50 mg) paroxetine after 6 weeks (n= 11).

Difference baseline vs. 6 weeks(mean 95% CI)

6 weeks (mean ±SD)

12 weeks (mean ±SD)

n* Difference 6 vs.12 weeks(mean 95% CI)

p†

Platelets (x109/l) -9.9 (-23.5 – 3.7) 242 ±61 249 ±53 11 7.0 (-13.8 – 27.8) 0.470

Bleeding time (min) 1.3 (-0.1 – 3.1) 5.6 ±2.3 5.8 ±2.3 9 0.4 (-1.1 – 2.0) 0.523

PFA-ADP (sec) -1.4 (-15.8 – 13.1) 88.0 ±16.4 91.7 ±15.3 11 3.7 (-8.2 – 15.6) 0.502

PFA-epinephrine (sec) 9.5 (-14.7 – 33.6) 122.6 ±39.7 124.5 ±30.0 10 0.9 (-19.5 – 21.3) 0.923

(median IQR) (median IQR) (median IQR) (median IQR)

PF4 platelets‡ (IU/106 platelets)

-0.2 (-2.3 – 2.7) 16.0 (12.5 – 18.4) 15.0 (14.3 – 17.0) 10 -0.8 (-2.2 – 2.2) 0.799

PF4 plasma‡ (IU/mL) 0.1 (-6.8 – 8.9) 6.4 (4.3 – 18.2) 5.2 (3.7 – 12.2) 10 -2.6 (-13.4 – 9.9) 0.575

ß-TG platelets‡(IU/106 platelets)

-3.7 (-9.4 - -0.3)§ 31.2 (27.4 – 35.3) 32.8 (28.6 – 35.4) 10 1. 0 (-2.2 – 2.5) 0.575

ß-TG plasma‡ (IU/mL) 0.9 (-15.5 – 22.0) 28.0 (20.9 – 65.2) 22.7 (19.3 – 38.9) 10 -5.6 (-39.6 - -0.1) 0.074

Platelet serotonin‡

(ng/109 platelets) -499 (-780 - -366)¶ 118 (74 – 143) 100 (86 – 145) 10 1.50 (-23.50 – 14.00) 0.799

* n for pairs week 6 – week 12 without missing values† p-values for paired t-tests for normally distributed data and paired Wilcoxon signed ranks test for non-normally distributed data ‡ due to non-normal distribution medians and interquartile ranges (IQR) are shown. § Wilcoxon signed ranks test p= 0.051 ¶ Wilcoxon signed ranks test p= 0.008

Table 11.4. Changes in platelet parameters after 6 and 12 weeks of paroxetine treatment, relative to baseline, stratified for genotype.

No LA/LA ≥1 LA p*

Platelet serotonin(ng/109 platelets) -868 (-1213 – -585) -457 (-598 – -392) 0.035

PFA-ADP (sec) 23.0 (17.0 – 48.0) -0.5 (-15.5 – 5.0) 0.001

PFA-epinephrine (sec) 47.0 (26.5 – 63.5) 4 (-16.0 – 19.0) 0.001

PF4 plateletsc (IU/106 platelets) 3.2 (0.7 – 4.7) -0.4 (-1.5 – 1.7) 0.003

Values represent changes of platelet parameters of combined observations after 6 and 12 weeks of paroxetine treatment (relative to baseline; 29 observations for 18 patients). Negative numbers indicate a decrease of the parameter. * Mann-Whitney non-parametric test

Chap

ter

Chapter 11

11

224

Discussion

This study confirms that treatment with the SSRI paroxetine at 20 mg/day decreases plateletserotonin and platelet β-TG levels, and increases bleeding time. Paroxetine treatment did notaffect other platelet parameters or bleeding tendency. Dose-escalation of paroxetine to 40-50mg/day did not further change platelet parameters. Additionally, the 5-HTTLPR polymorphism ofthe patients strongly affected the effect of paroxetine on bleeding time, platelet serotonin andPF4, PFA-ADP and PFA-EPI. In patients with the LA/LA genotype, paroxetine did not influencebleeding time, whereas patients with non-LA/LA genotypes had an increased bleeding time.Especially patients with the S’/S’ genotype had a greater reduction in platelet serotonin level,andincreasedPFA-ADP, PFA-EPI and platelet PF4.

Previous studies did not investigate dose-escalation. Because of the large inter-individualdifference of plasma SSRI levels at stable doses, proper investigation of a dose-response relationrequire quantification of within-subject changes, instead of comparisons of (small) groups.14;16 Bydoing so, paroxetine at 20 mg/day already maximally prolonged bleeding time and decreasedplatelet serotonin, which was not altered by consecutive dose-escalation (to 40-50 mg/day). Thisis in accordance with the recent finding that SERT occupancy in the brain is not increased by dose-escalation.30

Of interest are two in vitro studies that reported dose-response relations for sertraline17 andvenlafaxine.31 Serebruany et al. reported a significantly prolonged in vitro PFA-ADP and PFA-epinephrine with sertraline concentrations that were claimed to mimic the plasma levelsobserved in patients using sertraline 50, 100 or 200 mg/day.17 Increased platelet aggregation withADP, arachidonic acid, epinephrine and collagen was found in vitro, but with venlafaxine

Figure 11.1. Reduction of platelet serotonin levels after paroxetine exposure stratified by SERT genotype (n= 18).

Boxplots show mean (+), median (-), interquartile range, and minimum-maximum reductions in serotonin levels after 6 and 12weeks of paroxetine treatment (compared with baseline). For 18 patients observations after 6 weeks were available, for 8patients (secondary receiving higher doses) observations after 12 weeks were available. Data are stratified for genotype: S’/S’(S/S, S/LG, LG/LG; n= 3; 3 observations after 6, and 2 after 12 weeks), S’/LA (S/LA, LG/LA; n= 9; 9 observations after 6, and 1 after12 weeks) and LA/LA (n= 6; 6 observations after 6, and 5 after 12 weeks). There was a significant difference between thegenotype group S’/S’ vs. S’/LA + LA/LA (Mann-Whitney; p= 0.035).

��

��

��

��

��

��

��

��

� ��

��

��

��

� ��


11

225

concentrations 1000-fold higher than usual plasma concentrations of treated patients.31 However,these in vitro studies incubated samples with antidepressants immediately before assessingplatelet function. Normally platelet serotonin levels decrease in vivo only after >7 days, since thecirculating platelet pool must be renewed before effects of SSRIs on platelet function aremeasurable.31 This raises concern about the validity of in vitroexperiments.

The effect of the SERT polymorphism on platelet functionThe effect of paroxetine on platelet function highly depended on the SERT polymorphism of thepatients. In mice the influence of the serotonin transporter can be studied in strains without(knockout Slc6a4-/-), with 1 (Slc6a4+/-) or with 2 alleles (Slc6a4+/+) of the SERT gene. In knockoutmice the serotonin level in peripheral tissues (e.g. platelets) is <10% compared with non-knockouts. In Slc6a4+/--mice (best resembling humans with S/S genotypes), peripheral serotoninlevels were unchanged compared with Slc6a4+/+ mice. Thus, in Slc6a4+/--mice overall serotoninhomeostasis can be retained, despite a 4 to 5 fold decrease in serotonin reuptake.32 Furthermore,a recent platelet aggregation study in Slc6a4-/--mice showed 80% reduction in ADP-inducedaggregation, not observed in Slc6a4+/--mice, while results of this study also indicated a moreprominent role of the requirement of ongoing SERT transport of serotonin in maintaining plateletaggregation.33 In other words, apart from the direct effect of SSRIs reducing serotonin granules inplatelets, blockade of the SERT will also impair the functioning of the SERT at the platelet surface,which is required for the aggregation process itself. This might explain our findings: despitecompromised serotonin reuptake in S’/S’ subjects, platelets manage to achieve adequate, normalserotonin levels under normal circumstances. However, blocking the SERT by paroxetine mightaffect this compromised transporter system more than in patients with the most effective LA/LAgenotype, resulting in larger decreases of serotonin, reduced aggregation capacity (increasedPFA-ADP and PFA-EPI closure times), and longer bleeding times.

A recent study failed to find an effect of 5-HTTLPR polymorphisms on PFA-closure time.27

Consistent with our findings, this study reported no significant differences in bleeding tendency,but unfortunately platelet serotonin levels were not measured. Furthermore, in this study the bi-allelic variant of 5-HTTLPR was genotyped, while the reclassification of the ‘tri-allelic’ A to G SNP inthe long allele as an S’-variant is probably more precise to detect differences.

LimitationsDue to the small number of patients, the results of dose-escalation and genotype must beinterpreted with some restraint. However, despite the modest sample size of our study, theeffects of the genetic variants on platelet serotonin and bleeding time were demonstrated inconservative nonparametric tests. We furthermore consider our findings as valid, because ourbleeding time abnormalities and platelet serotonin findings for subgroups of SERT genotypesclearly coincide. We did not test changes in platelet parameters over time in control patientswithout SSRI, nor did we investigate dose-escalation in a randomized design. Additionally, oursample size did not allow further exploration of large versus small changes in platelet parameterswithin genotype groups.34

Clinical relevanceTwo potential clinical implications emerge. First, dose-escalation of paroxetine above 20 mg/daydoes not further impair platelet function, as it is already maximally impaired at a standard dose of20 mg/day. In other words, increased bleeding tendency associated with SSRI use will occurirrespective of the administered dose. Second, and most important, the impairment of plateletfunction appears to depend on the common functional SERT promoter polymorphism. Althoughwe only investigated paroxetine, we think this finding, if replicated, can be extrapolated to otherSSRIs.35

The absence of an increased bleeding tendency or abnormal bleeding severity score mayrepresent low sensitivity of the bleeding severity score to detect changes. However, this may alsosuggest that the observed decrease in platelet function is not clinically relevant in patients

Chap

ter

Chapter 11

11

226

without trauma or surgery, unless a preexisting platelet abnormality is present.20 However,because of their enhanced antiplatelet response to SSRIs, patients with a <2LA genotype may beat increased risk of bleeding complications during surgery or while using concomitantanticoagulants.

With a low prevalence of clinically important bleeding complications associated with SSRI use,determination of the SERT polymorphism in all patients who start with an SSRI will not be cost-effective. However, for patients with a previous severe bleeding episode (either with or withoutSSRI use) genotyping may be considered. When an S’/S’ genotype is found, switching to another,non-serotonergic antidepressant might be advisable. Generally, non-serotonergic antidepressants(i.e. mirtazapine, maprotiline, doxepin or bupropion) are associated with fewer hospitalizationsdue to bleeding complications than intermediate serotonin reuptake inhibitors (e.g. venlafaxineand amitriptyline; Odds ratio 1.9, 95% CI 1.1 to 3.5), and high degree serotonin reuptake inhibitors(fluoxetine, sertraline, clomipramine and paroxetine; Odds ratio 2.6, 95% CI 1.4-4.8), althoughcalculated odds ratios differed for individual antidepressants.9 The decision to switch must thendepend on the risks (e.g. oncoming surgery) and benefits (clinical response) for individualpatients.

ConclusionThis study shows that the SSRI paroxetine already maximally impairs platelet function at 20 mg/day, with no overall effect on bleeding time, but significant decreases in platelet serotonin and β-TG levels. Dose-escalation does not further influence hemostasis. In addition, the effect ofparoxetine on platelet function appears to be largely mediated by the common functional 5-HTTLPR SERT polymorphism, with patients with the S’/S’ polymorphism having the largesthemostatic impairment with larger decreases in serotonin, and increases in PFA closure time andplatelet PF4.

Acknowledgements

We wish to thank patients for their participation in the study. We thank general practitioners andpsychiatric residents for appropriate referrals of patients.

Conflicts of interests

All authors reported no biomedical financial interests or potential conflicts of interest. This studywas partly funded by a grant from the Netherlands Organization for Health Research andDevelopment (ZonMw), program Mental Health, education of investigators in mental health(OOG; #100-002-002) to Henricus G. Ruhé.


11

227

References


2. Movig KL, Janssen MW, de Waal MJ, Kabel PJ, Leufkens HG, Egberts AC. Relationship of serotonergic antidepressantsand need for blood transfusion in orthopedic surgical patients. Arch Intern Med. 2003; 163: 2354-2358.

3. Sewnath ME, van HR, Koopman MM, Levi MM, Gouma DJ. [Increased perioperative blood loss during treatment withparoxetine]. Ned Tijdschr Geneeskd. 2002; 146: 1800-1802.

4. Ceylan ME, psan-Omay MH. Bleeding induced by SSRIs. Eur Psychiatry. 2005; 20: 570-571.5. Linnebur SA, Saseen JJ, Pace WD. Venlafaxine-associated vaginal bleeding. Pharmacotherapy. 2002; 22: 652-655.6. Lake MB, Birmaher B, Wassick S, Mathos K, Yelovich AK. Bleeding and selective serotonin reuptake inhibitors in childhood

and adolescence. J Child Adolesc Psychopharmacol. 2000; 10: 35-38.7. de Abajo FJ, Montero D, Rodriguez LA, Madurga M. Antidepressants and risk of upper gastrointestinal bleeding. Basic

Clin Pharmacol Toxicol. 2006; 98: 304-310.8. Wessinger S, Kaplan M, Choi L, Williams M, Lau C, Sharp L et al. Increased use of selective serotonin reuptake inhibitors in

patients admitted with gastrointestinal haemorrhage: a multicentre retrospective analysis. Aliment Pharmacol Ther.2006; 23: 937-944.

9. Meijer WE, Heerdink ER, Nolen WA, Herings RM, Leufkens HG, Egberts AC. Association of risk of abnormal bleeding withdegree of serotonin reuptake inhibition by antidepressants. Arch Intern Med. 2004; 164: 2367-2370.

10. Dalton SO, Johansen C, Mellemkjaer L, Norgard B, Sorensen HT, Olsen JH. Use of selective serotonin reuptake inhibitorsand risk of upper gastrointestinal tract bleeding: a population-based cohort study. Arch Intern Med. 2003; 163: 59-64.

11. de Abajo FJ, Rodriguez LA, Montero D. Association between selective serotonin reuptake inhibitors and uppergastrointestinal bleeding: population based case-control study. BMJ. 1999; 319: 1106-1109.

12. Serebruany VL. Selective serotonin reuptake inhibitors and increased bleeding risk: are we missing something? Am J Med.2006; 119: 113-116.

13. Schalekamp T, Klungel OH, Souverein PC, de BA. Increased bleeding risk with concurrent use of selective serotoninreuptake inhibitors and coumarins. Arch Intern Med. 2008; 168: 180-185.

14. Laine-Cessac P, Shoaay I, Garre JB, Glaud V, Turcant A, Allain P. Study of haemostasis in depressive patients treated withfluoxetine. Pharmacoepidemiol Drug Saf. 1998; 7 Suppl 1: S54-S57.

15. Hergovich N, Aigner M, Eichler HG, Entlicher J, Drucker C, Jilma B. Paroxetine decreases platelet serotonin storage andplatelet function in human beings. Clin Pharmacol Ther. 2000; 68: 435-442.

16. Markovitz JH, Shuster JL, Chitwood WS, May RS, Tolbert LC. Platelet activation in depression and effects of sertralinetreatment: An open-label study. Am J Psychiatry. 2000; 157: 1006-1008.

17. Serebruany VL, Gurbel PA, O'Connor CM. Platelet inhibition by sertraline and N-desmethylsertraline: a possible missinglink between depression, coronary events, and mortality benefits of selective serotonin reuptake inhibitors. PharmacolRes. 2001; 43: 453-462.

18. Pollock BG, Laghrissi-Thode F, Wagner WR. Evaluation of platelet activation in depressed patients with ischemic heartdisease after paroxetine or nortriptyline treatment. J Clin Psychopharmacol. 2000; 20: 137-140.

19. Musselman DL, Marzec UM, Manatunga A, Penna S, Reemsnyder A, Knight BT et al. Platelet reactivity in depressedpatients treated with paroxetine: preliminary findings. Arch Gen Psychiatry. 2000; 57: 875-882.

20. Skop BP, Brown TM. Potential vascular and bleeding complications of treatment with selective serotonin reuptakeinhibitors. Psychosomatics. 1996; 37: 12-16.





25. Alderman CP, Seshadri P, Ben-Tovim DI. Effects of serotonin reuptake inhibitors on hemostasis. Ann Pharmacother. 1996;30: 1232-1234.

26. McCloskey DJ, Postolache TT, Vittone BJ, Nghiem KL, Monsale JL, Wesley RA et al. Selective serotonin reuptakeinhibitors: measurement of effect on platelet function. Transl Res. 2008; 151: 168-172.

27. Hougardy DM, Egberts TC, van der GF, Brenninkmeijer VJ, Derijks LJ. Serotonin transporter polymorphism and bleedingtime during SSRI therapy. Br J Clin Pharmacol. 2008; 65: 761-766.

28. Tosetto A, Rodeghiero F, Castaman G, Goodeve A, Federici AB, Batlle J et al. A quantitative analysis of bleeding symptomsin type 1 von Willebrand disease: results from a multicenter European study (MCMDM-1 VWD). J Thromb Haemost. 2006;4: 766-773.

29. Dean CB, Nielsen JD. Generalized linear mixed models: a review and some extensions. Lifetime Data Anal. 2007.30. Ruhe HG, Booij J, Weert HCv, Reitsma JB, Fransen EJF, Michel MC et al. Evidence why dose-escalation of paroxetine in

major depressive disorder is not effective: A 6-week, randomized-controlled trial with assessment of serotonintransporter occupancy. Neuropsychopharmacology. [In Press].

Chap

ter

Chapter 11

11

228

31. Sarma A, Horne MK, III. Venlafaxine-induced ecchymoses and impaired platelet aggregation. Eur J Haematol. 2006; 77:533-537.

32. Murphy DL, Lesch KP. Targeting the murine serotonin transporter: insights into human neurobiology. Nat Rev Neurosci.2008; 9: 85-96.

33. Carneiro AM, Cook EH, Murphy DL, Blakely RD. Interactions between integrin alphaIIbbeta3 and the serotonintransporter regulate serotonin transport and platelet aggregation in mice and humans. J Clin Invest. 2008; 118: 1544-1552.

34. Schreijer AJ, Cannegieter SC, Meijers JC, Middeldorp S, Buller HR, Rosendaal FR. Activation of coagulation system duringair travel: a crossover study. Lancet. 2006; 367: 832-838.

35. Owens MJ, Morgan WN, Plott SJ, Nemeroff CB. Neurotransmitter receptor and transporter binding profile ofantidepressants and their metabolites. J Pharmacol Exp Ther. 1997; 283: 1305-1322.


11

229

SUMMARY AND GENERAL DISCUSSION

PART

Summary and general discussion

CHA

PTER

SUMMARY, CONCLUSIONS AND GENERAL DISCUSSION

ChapterSummary and general discussion

12

235

Summary

In this chapter, a summary of the conclusions of our studies and the discussion of these findingswill be presented. Possible limitations and strengths of the thesis will be discussed and directionsfor future research will be presented.

Summary of this thesis and conclusions

This thesis addressed several questions:

1 Is a short, easy to use clinician rated questionnaires as effective and precise as the routineHamilton depression rating scale (HDRS)?In chapter 3, we reanalyzed the treatment-outcomes of two antidepressant-psychotherapytrials.1 In line with previous reports,2-5 we found that the effect-sizes of two 6-item sub-scalesof the HDRS (17-items) – the Maier6 and Bech7 sub-scales – were comparable to the originalHDRS in the measurement of depression severity, and the sensitivity to measure changes.Furthermore, this comparability was stable across the full range of response to treatment,across both pharmacotherapy and psychotherapy, and for patients with different baselineseverities of their depression. With an item response theory approach, we calculated aconversion table linking HDRS-scores and Maier and Bech scores, and determined cut-offpoints for remission for these subscales compared with conventional HDRS definitions.8-10

With these subscales clinicians can measure depression severity and clinical response moreefficient than with the original HDRS.

2 What is the evidence for dose-escalation as a strategy for non-response to a first SSRI?In chapter 4, we presented a systematic review of the evidence for the dose responserelationship of selective serotonin reuptake inhibitors (SSRIs) in major depressive disorder(MDD).11 In our literature-search, we identified 8 dose-escalation studies that increaseddosages after at least 3 weeks of a standard dosage.12-19 Furthermore, 3 systematic reviewswere published by then, which included three20;21 or four22 of the eight identified dose-escalation studies. Only one of the dose-escalation studies approached stringentmethodological criteria.19 We found no evidence for increased efficacy of dose-escalationwithin the first 4 weeks of treatment. Dose-escalation after 6 weeks appeared less effectivethan continuing the same dose. We found some, but limited evidence for efficacy of dose-escalation after 8 weeks, particularly in partial responders. This effect was seen within 4weeks after dose-escalation. Irrespective of efficacy, dose-escalation unequivocally increasedside-effects, but effects on drop-out rates due to side effects were less straightforward. Wetherefore concluded that the available evidence for dose-escalation neither unequivocallyconfirmed its efficacy, nor deemed it ineffective. This conclusion was the starting point for theDELPHI-study, described in the chapters 2 and 7 to 10.

3 What is the evidence for switching antidepressants as a strategy for non-response to a firstSSRI?In chapter 5, we presented a systematic review of the evidence for switching after failure of afirst SSRI in MDD.23 In addition to our literature-search we included four studies released afterthese searches, of which three studies from the Sequenced Treatment Alternatives to RelieveDepression (STAR*D) trial.24-27 We identified eight randomized controlled trials (RCTs),24-31 and23 open switch studies. Studies were of variable methodological quality and carried out inheterogeneous populations. The STAR*D results largely increased the amount, and quality ofthe available evidence, but did not show differential class effects to guide switching. Threerandomized switch-studies investigating the switch from SSRI to another SSRI or SNRI(venlafaxine) were eligible for a meta-analysis.24;25;28 This meta-analysis showed that in themost favorable analysis, venlafaxine was slightly more effective as a second antidepressant,

Chap

ter

Chapter 12

12

236

compared with other SSRIs (number needed to treat (NNT) = 10 (95% confidence interval (95%CI) 6.3-33.3). We therefore concluded that after the failure of a first SSRI switching is open toall antidepressant classes (except irreversible MAO-inhibitors), without clearrecommendations other than those that apply for the selection of initial treatment.

4 Does the depletion of monoamine (5-HT and NA/DA) systems lower mood in humans, and isthis lowering of mood different across different populations?In chapter 6, we presented a systematic review of monoamine depletion studies reportingmood effects of depletion.32 As an extension of previous systematic reviews of monoaminedepletion studies,33-38 we pooled the results of 53 small-sized depletion studies with anadapted pooling technique (modified from conventional meta-analyses of RCTs to handle thestatistically paired cross-over designs, and including an adjustment for small sample bias).Pooling is important because it quantifies the balance of positive versus negative studies.

We pooled 45 acute tryptophan depletion (ATD) studies and 8 acute phenylalanine/tyrosine depletion (APTD) studies, acutely lowering serotonergic and dopaminergic/norepinephrinergic neurotransmission, respectively. Serotonin or dopamine/norepinephrinedepletion did not decrease mood in healthy controls, but slightly lowered mood in healthycontrols with a family history of MDD. In drug-free patients with MDD in remission, amoderate mood decrease was found for ATD, without an effect of APTD. ATD induced relapsein patients with MDD in remission who used serotonergic antidepressants. MDD patients didnot experience mood deterioration after ATD while depressed.

We concluded that simple, direct correlations of serotonin or dopamine/norepinephrinelevels in the brain and mood do not exist. Because the serotonin or dopamine/norepinephrinedepletion induced by ATD or APTD most clearly decreases mood in vulnerable individuals(patients who are in remission from MDD and healthy controls with a positive family historyfor MDD), we concluded that the monoamine systems are important systems in thevulnerability to become depressed. Moreover, the changes in brain metabolism in remittedpatients who relapse after ATD or AMPT suggest that the serotonergic andnorepinephrinergic systems give input to a final common pathway which needs furtherresearch to be clarified.

5 Do MDD-patients and healthy controls differ in the number of central serotonin transporters,and is the amount of available SERTs correlated with depression severity?In chapter 7,we measured the availabilities of the central serotonin transporters (SERTs) inbaseline SPECT scans of the DELPHI-SPECT participants, and compared these with SERTavailabilities in age- and sex-matched healthy controls.39 Because of earlier reports, weinvestigated potential covariates as smoking behaviour,40 age,41;42 season of scanning43 andpossible interactions with gender40;44 in multivariate models. We found interaction effects ofsmoking and diagnosis with gender for diencephalon SERT availability. For midbrain SERTavailability, an interaction of gender with diagnosis was found, while season of scanning was acovariate. These interaction effects were expressed as lower SERT availability in midbrain anddiencephalon in depressed males, and higher SERT availability in diencephalon in depressed(non-smoking) females compared with healthy controls. These findings point to complexeffects of gender, smoking and season of scanning on the serotonergic system in healthycontrols and patients with MDD, which must be considered in future studies comparing SERTavailabilities in MDD patients versus healthy controls.

6 Does a common genetic polymorphism of the promoter region of the serotonin transportergene (SLC6A4) modify the association between the SERToccupancy by paroxetine and theclinical response?In chapter 8, we studied the relationship between clinical response after 6 weeks ofparoxetine treatment and SERT occupancy (which is the change in SERT availability after 6weeks of treatment relative to the pre-treatment SERT availability).45 Because clinicalresponse to SSRIs is likely associated with polymorphisms of the serotonin transporter genepromoter region (5-HTTLPR),46;47 we also aimed to investigate whether this relation of


12

237

response and SERT occupancy was modified by the 5-HTTLPR genotype. We obtained study-entry and week 6 SPECTscans for 44 patients treated with paroxetine 20 mg/day, of which 42scans were analyzable (10 responders and 32 non-responders).

We found that for all patients, SERT occupancy was not associated with clinical response(expressed as the proportional decrease in HDRS relative to study entry). However, when wegrouped patients by 5-HTTLPR polymorphism, we found that in patients with the(favorable)46;47 LA/LA variant a significant relation between SERT occupancy and clinicalresponse (absolute and proportional change in HDRS) existed. A previous study found abetter amygdala-cingulate coupling in subjects with an L-allele relative to the unfavorable S/Svariant,48 probably reflecting differences in the development of the serotonergic systeminduced by 5-HTTLPR.49 As an explanation of our findings, we therefore hypothesized thatpatients with the 5-HTTLPR LA/LA genotype have a more flexible serotonergic system, which ismore easily influenced by serotonergic antidepressants.

7 Is dose-escalation of paroxetine an effective clinical strategy for non-response in MDD?In chapter 9, we studied the strategy of dose-escalation which was found equivocallyefficacious in the systematic review,11 as described in chapter 4. In this study, we addressedthe methodological flaws found in previous dose-escalation trials,50 but moreover, we alsomeasured whether paroxetine dose-escalation increased SERT occupancy more than placebodose-escalation.51 After inclusion of 49 SPECT patients (with successful follow-up of 31patients participating in the randomized placebo-controlled dose-escalation SPECT study), weperformed an interim analysis including all randomized patients (n= 57), and testeddifferences between placebo and true dose-escalation with pre-defined cut-off values forfutility and superiority. We found no clinical benefits of true dose-escalation compared withplacebo, and so the trial was stopped because of futility.

Most important, our data also showed that true dose-escalation of paroxetine did notincrease SERT occupancy more than placebo dose-escalation, probably because a plateau wasreached at standard doses already. We concluded, that in line with previous, though moreequivocal evidence,52;53 dose-escalation is not beneficial in MDD. Therefore, two clinicaloptions remain for the treatment of MDD-patients who do not respond to standard doses:either continuation of treatment until 10 weeks while waiting for a potential delayedresponse, or a switch to a different and potentially more effective treatment strategy (e.g.another SSRI or SNRI or TCA, or psychotherapy, or a combination).

8 Does treatment with paroxetine normalize amygdala hyperactivation in MDD?In chapter 10, we describe the (negative) faces task of the DELPHI-fMRI sub-study.54 Westudied the effects of paroxetine treatment on the hyperactivation of the amygdala (andother brain areas) during a depressive episode. We specifically investigated: 1) whetheramygdala-activation by (negative) facial expressions differed from healthy controls, 2)whether amygdala activation changed after 6 and 12 weeks of treatment, 3) whether theamygdala activation merely changed by paroxetine treatment or in relation with clinicalresponse, and 4) whether dose-escalation of paroxetine in week 6 non-responders affectedactivations, compared with placebo-dose-escalation. Apart from the amygdala, we alsoinvestigated these questions for other brain areas. We performed fMRI scans in 22 MDDpatients and 21 age- and sex-matched healthy controls (controls were scanned once).

Compared with healthy controls, we found increased activations in bilateral (extended)amygdala and left insula in MDD-patients. In contrast with previous studies,55-57 we found anincrease in right amygdala activation after 6 weeks of treatment, which was attributed to thehigh number (12/20) of treatment non-responders. This increased amygdala-activation wasreduced thereafter until week 12. When we analyzed the differences in week 6 and 12treatment responders versus week 6 and 12 non-responders, bilateral amygdala activationwas reduced in responders versus non-responders. Amygdala activations in week 6 and 12were associated with HDRS-scores. Furthermore, non-responders had higher activations in

Chap

ter

Chapter 12

12

238

right orbitofrontal cortex (OFC) and insula, while treatment responders showed higheractivation in right dorsolateral prefrontal cortex (DLPFC) and left nucleus accumbens. Thesedifferences were non-existent at study-entry.

Although randomized groups became small, changes over time in the true dose-escalationgroup showed decreased activations in ventral and dorsal regions, while activations of theright hippocampus and left subgenual cingulate increased,when compared with changesobserved after placebo dose-escalation. We concluded that in line with previous treatmentstudies56-59 paroxetine treatment over 12 weeks reduced activations in the amygdala and other(ventral) emotion-generating structures, and increased cortical function (in the dorsalregulatory structures; see chapter 1, pages 18 and 21), especially in treatment responders.These findings may point to increased fronto-limbic control as a mechanism of paroxetinedrug-response effects, which was supported by findings of other groups.60;61

9 What are the changes in hemostasis and blood platelet parameters when patients are treatedwith paroxetine, and are these changes modified by dose-escalation or a geneticpolymorphism of the promoter region of the serotonin transporter gene?In chapter 11, we present the results of a study investigating an infrequent, but dangerousadverse effect of SSRIs: increased bleeding tendency.62 In this study, we measured plateletparameters and coagulation, while we applied a secondary (non-randomized) dose-escalationin paroxetine treatment non-responders. We additionally investigated whether 5-HTTLPRpolymorphisms modified these effects on platelet parameters. We found that a standard doseof paroxetine (20 mg/day) already significantly decreased platelet serotonin levels andplatelet β-thromboglobulin (β-TG), without further decreases after dose-escalation.Moreover, we found that 5-HTTLPR polymorphisms modified these effects: compared withLA/LA-carriers, bleeding time significantly increased in <2LA-allele carriers, and plateletserotonin decrease was larger in patients without LA-alleles. Furthermore, the plateletfunction analyzer closure time significantly increased in patients without LA-alleles. Althoughthe observed bleeding tendency presumably is not clinically relevant in patients withouttrauma or surgery (unless a preexisting platelet abnormality is present),63 these findings areapplicable to patients with a previous severe bleeding episode. For them genotyping may beconsidered; when an S’/S’ genotype is found, switching to another, non-serotonergicantidepressant might be advisable.

Clinical relevance

Antidepressants have comparable efficacy, however only 50% of the MDD-patients respond to thefirst antidepressant trial given, while fewer achieve full remission of symptoms.64;65 When themiracle doesn’t happen, there are 5 strategies for non-response: prolongation of the initial trial,dose-escalation, switching to another drug, augmentation with another drugora combination ofantidepressants. This thesis addresses switching strategies23 and dose-escalation.11;51

SwitchingDespite a small benefit of a switch to venlafaxine compared with a second SSRI (in the mostfavorable analysis only), we concluded that switching-options after a first SSRI are open to allantidepressant classes, without clear recommendations other than those that apply for theselection of initial treatment. One more RCT comparing venlafaxine and citalopram,66 and threeopen studies, one with venlafaxine (with randomization to different doses)67, one with avenlafaxine-mirtazapine combination strategy,68 and one with duloxetine69 appeared thereafter.

In addition, another meta-analysis was published, comparing a non-SSRI switch (venlafaxine(3 studies) or mirtazapine (1 study) or buproprion (1 study)) versus a second SSRI.70 This meta-analysis included 2 different studies (abstracts from symposia)71;72 compared with our meta-analysis, and excluded one study with questionable methodology, like we did in our most


12

239

favorable analysis.24 Their conclusion was that the NNT for remission was 22 (confidence intervalnot given), modestly in favor of a non-SSRI switch, which is far below the standard for clinicalrelevance (NNT ≤10) as suggested by the United Kingdom’s National Institute of ClinicalExcellence.73;74 When they considered response as an outcome, they found no significant benefit,comparable to our results.70 In the final manuscript of the previous abstract,71 Lenox-Smith andJiang reported increased efficacy of venlafaxine over citalopram (flexible doses) in a subgroup ofpatients with severe MDD (HDRS21>31), who participated in a multicenter RCT of SSRI non-responders.66 However, this difference was only significant for the continuous scores of theHDRS21 (difference approximately 4.7 points; SD not given), and not for the remission rates.Therefore, we think that only at the level of public health, the difference between switching-strategies might be interesting, but not at the level of individual patients.

Dose-escalationIn our systematic review of dose-escalation it was concluded that the available evidence neitherunequivocally confirmed its efficacy, nor deemed it ineffective. This was in agreement withanother review, performed by an independent group, and published just before ours.52 The resultsof the dose-escalation RCT in this thesis first of all addressed the methodological shortcomings ofprevious studies, and was stopped for futility after an interim analysis.51 Even though the numbersof patients randomized to placebo or true dose-escalation were modest (n= 27 and n= 30,respectively), the changes in HDRS-scores and subscales were far from significantly different,while the level of this significance was far above the a prioricut-off for futility(which wasundisclosed at the time of analysis). This finding aligns with the dose-escalation study by Licht etal.19 Unfortunately, the premature termination of the trial precluded planned secondary analysesto distinguish subgroups who still might benefit form dose-escalation (e.g. partial responders;defined as 25-50% reduction of HDRS).15;18 In our sample the subgroup of partial respondersconsisted of 24 patients (from 60 randomized; 40%), and did not show any indication of beneficialtreatment-effects for true dose-escalation. Nevertheless this observation cannot replace truesubgroup analyses in larger samples.

Our sub-study of the changes in SERT occupancy during the randomized, placebo-controlleddose-escalation phase further corroborates our clinical findings. We therefore conclude that oursystematic review, our dose-escalation study and our imaging study provide substantial evidencefor clinicians, that dose-escalation is not an effective strategy for depressed patients who do notrespond to 6 weeks of a first SSRI at a standard dose.

In a recent meta-analysis of 9 placebo-controlled, fixed-dose, dose-finding studies of SSRIs,75

Papakostas et al. found a modest but statistically significant difference of 4% higher responserates in patients who started with higher than standard doses of citalopram, escitalopram,fluoxetine, fluvoxamine, paroxetineand sertraline (p= 0.04). This 4% difference in response ratescorresponds with a NNT of 25, which is clinically not impressive, nor relevant. In their meta-analysis, Papakostas et al. could not investigate whether this difference was attributable to asingle SSRI. However, when inspecting their data, all paroxetine trials did not show benefits ofhigher paroxetine doses compared with standard 20 mg/day doses, with relative risks of 1.00 (95%CI 0.67-1.50), 1.02 (95% CI 0.69-1.52) and 0.94 (95% CI 0.65-1.35). In contrast, the rate ofdiscontinuation due to adverse events was also higher in the patients treated in the abovestandard doses group (16.5% compared with 9.8%). Therefore, to our opinion, it remains alsoquestionable whether higher doses of SSRIs at the initiation of treatment indeed areadvantageous to MDD-patients. Instead, they could give patients more adverse effects andpreclude prolonged exposure to tolerable standard doses. Finally, but most important for thisthesis, these findings presumably hardly apply to paroxetine.

Other findings of clinical relevanceBesides the switching and dose-escalation strategies, two other relevant points from this thesisemerge. First, we showed that the Maier and Bech subscales of the HDRS17 are equivalent to thefull HDRS17.1 Others even suggested that these subscales might measure the core-symptoms and

Chap

ter

Chapter 12

12

240

severity of MDD more specifically, and their use might even be more efficient,2;3 because of thereduction of noise by less responsive items.76 We therefore propose that the Maier or Bech-subscales should be used in the clinical assessment of (treatment of) MDD patients, instead of, orbesides self-rating scales like the Inventory for depressive symptoms self-rated (IDS-SR) or Beckdepression inventory (BDI). This will be of great importance in patients who cannot reliably fill outself-rated questionnaires, like in MDD with psychotic features or MDD with severe cognitivedysfunction. Of course, providing a shorter, easy to use clinician rated scale will not translate indirect implementation of one of these subscales, but may at least reduce the time burden ofrating scales for physicians.

Second, our findings regarding the increased bleeding tendency in patients merit attention.62

Increased bleeding tendency, especially in combination with non-steroidal anti-inflammatorydrugs (NSAIDs) was observed before.77-82 This even led to the recommendation to use TCAs whenconcomitant NSAIDs should be used in the guideline for Dutch general practitioners.83 However,the serotonergic TCA clomipramine, presumably has comparable effects. Because of the relativelylow incidence of dangerous gastro-intestinal bleeding complications in general (4.3/1000 SSRItreatment years; 14.5/1000 SSRI+NSAID/aspirin treatment years; relative to 1.2/1000/year inuntreated humans),80 the increased bleeding tendency by SSRIs is of no clinicalrelevance inpatients without trauma, surgery or a previous severe bleeding episode.63 However, in patientswith a previous severe bleeding episode, or in patients who need elective surgery, and maybe alsoin patients regularly using NSAIDs, genotyping of the 5-HTTLPR may be considered. When an S’/S’genotype is found, switching to another, non-serotonergic antidepressant might be advisable. Forwhich of the aforementioned subgroups this genotyping will be cost-effective, remains to beinvestigated.

Neurobiology of the treatment of major depressive disorder

With our studies we also aimed to investigate the neurobiology of paroxetine treatment effects inMDD in vivo. The enhancement of serotonergic neurotransmission by paroxetine is undoubted,caused by a reduced serotonin uptake after SERT occupancy as the primary mechanism.Therefore, the most important question which we addressed was whether dose-escalation ofparoxetine would increase SERT occupancy compared with placebo.

SERT occupancy after short or prolonged treatment with SSRIs was investigated before,84-96

even at variable and higher doses.86;96 However, without a pre-assessment of SERT availabilitybefore treatment96 and without randomization to a placebo-controlled dose-escalation,86 one cannever exclude the possibility that non-responders after 6 weeks of a standard dose consist of aselected sample who had lower SERT occupancy. This hypothetically lower SERT occupancy couldthen be considered explanatory for non-response, and might increase by subsequent dose-escalation compared with placebo dose-escalation.

Our SPECT measurements in the subgroup of the DELPHI study showed no overall relationbetween SERT occupancy and clinical response in the open phase of treatment with paroxetineover 6 weeks. Furthermore, we found no evidence for the postulated 80% SERT occupancy as arequisite for clinical response.45 In addition, the second, randomized, placebo-controlled dose-escalation phase of DELPHI showed no differences in changes in SERT occupancy betweenplacebo and true dose-escalation for 6 more weeks.51 With this latter finding, we are the firstgroup that studied sequential dose-escalation in non-responders, controlling for placebo-effectsand measurement-bias by repeated SPECT scans. Interestingly, we did find an associationbetween SERT occupancy and clinical response in the subgroup of patients with the (tri-allelic) LA/LA polymorphism of the 5-HTTLPR. Because the tri-allelic LA:LG ratio is approximately 5:1,97 ourfinding aligns with data from previous treatment studies which found increased response rates insubjects with the bi-allelic L/L 5-HTTLPR polymorphism.46;47


12

241

However, the mechanism why increased SERT occupancy is associated with increased clinicalresponse in LA/LA carrying patients, cannot be explained by higher SERT occupancies in LA/LApatients.45 Probably, the explanation must be sought in differences in the function of theserotonergic system (and its development) mediated by 5-HTTLPR polymorphisms. From previouslines of research, it may be concluded that the S-allele predisposes to a more reactive arousalsystem, and that the L-allele accounts for more flexibility and/or the possibility to react and adaptbetter to external changes imposed on the subject.98 Furthermore an elegant fMRI-study showedthat both the anatomy and functional connections between limbic, subcortical and corticalregions are mediated by 5-HTTLPR polymorphisms, with S/S carriers showing relative uncouplingof the amygdala-pregenual cingulate connectivity.48

The biochemical consequences of chronic administration of SSRIs are thought to include anincrease in extracellular levels of serotonin followed by neuroadaptive alterations in serotoninreceptors and postsynaptic intracellular signalling pathways (see figure 1.2, page 20), as well astime-dependent effects on neurogenesis.99;100 Not surprisingly, animal studies demonstrated thatin mice knocked out for the murine Slc6a4 SERT transporter (Slc6a4-/-), fluoxetine andfluvoxamine are ineffective on behavioural and other measures, while mice with at least one copyof the allele (Slc6a4+/- or Slc6a4+/+) showed the typical antidepressant effects.99 This is pivotalevidence for the working mechanism of SSRIs, for which SERTs are apparently required to bepresent. The Slc6a4+/- mice have ~50% less expression and transportfunction and can be viewed asa model for humans with the S/S genotype.101 Therefore, the findings of almost identical effectsof SSRIs in Slc6a4+/- or Slc6a4+/+ mice102 do not support the findings of differential effects for 5-HTTLPR polymorphisms from meta-analyses of human studies.46;47 However, the 5-HTTLPRpolymorphism is only present in humans and higher order primates (e.g. macaques). To ourknowledge, no studies in primates investigated the mediating effects of 5-HTTLPR polymorphismson the relation between SERT occupancy and behavioural effects. Furthermore, we are unawareof studies investigating dose-response relationships of SSRIs in animals. Finally, although a robustincrease in anxiety-like behaviourwas found in Slc6a4-/-mice compared with other genotypes, sofar no studies were performed to investigate the differences in brain development and functionalconnections for mice with a Slc6a4-/-, Slc6a4+/- or Slc6a4+/+ genotype.103

Serotonin has an important role in the regulation of neuronal information processing byconstraining (instead of inhibiting) the reactivity of the brain to internal and external stressors.104

In general, serotonin appears 1) to prevent an overshoot of other dynamic systems (e.g.constraining the release of dopamine after rewarding stimuli), and 2) to control the sensitivity ofthe system to perturbation by new elements entering the system (e.g. constraining a naturalpropensity to switch to alternate behaviours; this can be provoked by serotonin depletion). Thisconstraint can be experimentally challenged by acute tryptophan depletion (ATD), which resultsin increased perception of aversive stimuli, increased food intake, increased sexual, aggressive,depression-like and anxiety-like behaviours, but also decreased sensitivity to cues of punishment,and less flexibility to change behaviour in rats.104

Besides the cognitive effects,98 our meta-analysis quantified the mild mood lowering effectsby ATD in humans, which were especially seen in healthy controls with depressed relatives, andpatients who had remitted from a previous MDD-episode.32 Because mood-effects can beconsidered as more complex behaviour, these may be more difficult to provoke in relatively short-term ATD studies compared with longer-lasting studies in animals. Nevertheless, the large body ofresearch on the serotonergic system has now changed the categorical model –decreasedserotonin is specific for depression – into a model in which alterations in the serotonergicneurotransmission can be seen as a biological risk-factor, which interacts with innate and/orexternal factors. This so-called serotonergic vulnerability is now seen as neither a sufficient, nornecessary factor to develop and maintain MDD, but in combination with other factors increasesthe risk for MDD.98

Chap

ter

Chapter 12

12

242

Our fMRI-study,54 corroborates the hypothesis that paroxetine treatment interferes with the(maladaptive) functioning of various interconnected regions in the limbic-subcortical-corticalnetwork.105;106 Like others, we found increased ventral* activations (e.g. in the amygdala andinsula), and decreased dorsal* activations (anterior cingulate, DMPFC, DLPFC) in MDD-patientsrelative to healthy controls. We furthermore found that the ventral hyperactivation and dorsalhypoactivation was normalized in treatment responders, but not in non-responders. Recently,Chen et al. demonstrated increased functional coupling of the amygdala with the frontal and(pregenual) cingulate cortex (and striatum and thalamus) after 8 weeks of sertraline treatment.61

Although we did not investigate this coupling per se, we found that the amygdala responses in ourpatients were correlated with pregenual cingulate and frontal regions.

Furthermore, in responders we found increased activation in the nucleus accumbens, pointingto the involvement of the limbic-cortico-striato-pallido-thalamic network, which indirectlysuggests the involvement of changes in dopaminergic neurotransmission after paroxetinetreatment.107 Two coinciding explanations can be given for this increased activation of the nucleusaccumbens. First, in rodents, chronic treatment with antidepressants (among others SSRIs)potentiated dopamine transmission, due to an increased sensitivity of postsynaptic dopaminereceptors and possibly also due to a decreased sensitivity of presynaptic (inhibitive) dopamineautoreceptors, preferentially in the limbic system.108 Second, again in rodents, decreases inserotonin are associated with a release of normally inhibited behaviour, including impairedsuppression of behaviour inducing punishment.104 Activation of the nucleus accumbens isassociated with anticipation of reward,109 while the nucleus accumbens is deactivated by loss orpunishment. Therefore, it might be postulated that the increase in serotonergicneurotransmission by SSRIs, reduces a lack of constraint on dopaminergic neurotransmission inthe nucleus accumbens, with subsequently less punishment perceived. This would then result inless deactivation of the nucleus accumbens. The fact that this was demonstrated in treatmentresponders relative to non-responders might suggest that in non-responders this enhanceddopaminergic constraint might not be achieved.

As interesting, but contradictive finding, we did not find any indication that dose-escalationincreased SERT occupancy, but at the other hand, we found indications that a true dose-escalation decreased hyperactivations in the ventral compartment (OFC, insula, VMPFC,ventrolateral thalamus), and increased activation in the dorsal compartment (hippocampus,DLPFC, and parietal cortex). Also we counter intuitively found an increased activation in thesubgenual cingulate after dose-escalation.54 As explained, we could unfortunately not disentangleresponse- and drug-effects. If the decreases in the ventral, and the increases in the dorsalcompartment could have been attributed to responders to dose-escalation, while the increasedactivation in the subgenual cingulate was associated with non-responders (despite dose-escalation), this would have made our findings less surprising.

Limitations

Limitations of the studies in this thesis were discussed in each chapter. However, some must berecapitulated in this general discussion.

At the moment of publication of this thesis, the systematic reviews presented in the chapters4, 5, and 6 might have become outdated, since the literature searches were performed untilOctober 200632 or February 2005.11;23 Although we did not repeat the searches until July 2008, forthe dose-escalation and switching reviews, we commented on additional studies that weencountered thereafter in this general discussion. We did not review additional monoaminedepletion studies, as this should have resulted in new meta-analyses.

* For explanation of the ventral and dorsal compartments of the limbic-subcortical-cortical network, see chapter 1, page 18.


12

243

Furthermore, in contrast with other authors,53;75 we did not perform a meta-analysis for thedose-escalation studies, and only for the venlafaxine versus second SSRI switch studies. At thispoint, we considered the dose-escalation studies with various moments of randomization afterthe initiation of treatment too heterogeneous to pool. For switching, we decided not to pool (bythen unpublished) data on various dual acting antidepressants versus second SSRIs.

In the SPECT-studies, we used [123I]β-CIT for SPECT imaging, which is a non-selectiveradioligand, and also binds to the dopamine transporter (DAT; e.g. substantia nigra) andnorepinephrine transporter (NET; e.g. locus coeruleus).41;110-112 Nowadays, selective SERT ligandslike [11C]DASB for PET or [123I]ADAM for SPECT are available. However, previous imaging study inprimates, showed that [123I]β-CIT uptake in midbrain and diencephalon predominantly reflectsSERT,113 as these structures are rich of SERT relative to DAT and NET. Although this non-selectivitymight have slightly concealed changes in SERT occupancies due to additional DAT- or NET-binding,we think our findings in diencephalon and midbrain mainly reflect SERT occupancy. Furthermore,if one would consider non-selectivity as relevant, the use of SPECT-scans at randomization as areference for changes in occupancy will further reduce the magnitude of this non-selectivity bias,because non-selectivity will be measured systematically between T0 and T1. Despite this rationale,in future studies, a selective ligand for SERT ([11C]DASB or [123I]ADAM) should be considered.Furthermore, it would be a challenge to replicate our study using one of these selective ligands.

Although the region-of-interest (RoI) approach to determine non-displaceable bindingpotential (BPND) and SERT occupancies is a valid and widely accepted method,114 we did not co-register individual SPECT images with individual structural MRI scans. This might have reducedsignal-to-noise variance in our measurements. However, because we applied a randomized,double-blind dose-escalation, and analyzed the SPECT scans while blinded for the moment intreatment (T0 or T1) and intervention, we think that this RoI approach only affectedmeasurement-bias, and did not introduce differential bias between the intervention groups.

Scientific studies are powered to investigate major questions, but often additional questionswill be explored. In the case-control study of SERT availability of MDD-patients versus controls,only small subgroups remained for effect-modifying variables, which might have increased thechance of detecting spurious results. We therefore presented these differences as exploratoryanalyses only. For the SERT occupancy-response relation by genotype interaction, we alsocompared small groups, but nevertheless found significant results, which we consider valid.Nevertheless, replication in a larger sample would strengthen these results. Because of theinterim-analysis, we stopped our trial for futility. Thereafter, our modest sample size of therandomized dose-escalation for example precluded proper subgroup analyses. Our fMRI-study,which was relatively large for imaging studies found robust effects. However, the comparison ofthe dose-escalation versus placebo dose-escalation in our fMRI study was hindered by a modestnumber of patients in the true dose-escalation group. This precluded the investigation of the 3-way interaction intervention by time by response, which could have properly answered whicheffects were drug-induced. Therefore, as indicated, these results must be considered with somerestraint. Finally, in the hemostasis study, we studied 19 patients, but did so in a within-subjectdesign, which has more power to detect changes. Again the number of subjects receiving a dose-escalation was modest, which might have precluded changes in other platelet parameters withsmaller effect-sizes.

A general limitation is that our neurobiological investigations did not investigateadaptiveneuronal effects caused by paroxetine. We either investigated the pre-synaptic effects of dose-escalation on SERT occupancy, or the changes in the final common pathway: the changes inactivations of the limbic-subcortical-cortical network. Many adaptive pre- and post-synapticeffects of SSRIs have been documented,100;115-123 mostly in animal studies, but also in humans.124 Wedid not investigate any of these specific secondary effects of SSRIs. The radiation burden forscientific research did not allow additional radioligand scans of other receptors (e.g. 5-HT2A or5-HT1A) or other monoamines (e.g. endogenous dopamine release). Furthermore, additionalchallenges (e.g. gepirone to assess 5-HT1A receptor sensitivity125) were considered tooburdensome in these patients. We did not perform animal studies to investigate dose-escalation,

Chap

ter

Chapter 12

12

244

for example in microdialysis experiments or studies investigating SERT occupancy, SERT down-regulation, or other secondary effects of serotonergic neurotransmission on receptors and/orother pathways.

Nevertheless these limitations, one should bear in mind that our, and previous clinical studiesdid not support dose-escalation as a rational strategy for SSRI non-responders in MDD. Therefore,the merits of additional dose-escalation studies would be to provide fundamental insights inworking mechanisms.

Finally, this thesis might also make clinicians wonder why initial, high doses of SSRI have beenfound effective in (some) fixed-dose, dose-finding studies in anxiety disorders,126-129 and especiallyin obsessive compulsive disorder (OCD).130-134 Of course, anxiety disorders are beyond the scope ofthis thesis, but this question remains intriguing and might warrant an approach as we did forMDD.

Strengths

First of all, a major strength of this thesis is its clinical starting point. All psychiatrists recognize theproblem of non-response to a first SSRI, and most of them increase the dose as a first strategy,followed by switching.135-139 As such, this thesis covers an important clinical topic, which could bevery relevant for future treatment guidelines.

Our systematic reviews were conducted in accordance with the stringent guidelines of theCochrane Collaboration.140 We performed ‘sensitive’ literature searches in several databases,decreasing the chance of missing valuable studies. All studies were critically appraised andabstracted. We only performed meta-analyses when clinical heterogeneity allowed pooling, amethodological requirement which is often violated by referring to a random effects approach.141

In the obvious heterogenous populations investigated in the monoamine depletion review, weachieved better homogeneity by stratification. In this latter meta-analyses, we acknowledged thestatistical problems of pooling crossover studies (having dependent measurements) and adaptedthe standard statistical approach to solve this problem.32

The empirical studies in this thesis were designed carefully and the methodology of the dose-escalation trial was based on a thorough review of the previous dose-escalation studies.11;50 Wetherefore randomized patients after sufficient time to expect not many further delayed clinicalresponses. Also, we gradually increased doses which obviously prevented early drop-out due tointolerable adverse effects. Most importantly, our SERT occupancy SPECT protocol was unique,and provided the rationale why dose-escalation was ineffective.

With the fMRI-study, which could be started after a second grant was obtained, we lookedbeyond SERT occupancy to investigate the effects of paroxetine treatment on emotion-regulation. Although the dose-escalation results of this study are not unambiguous, this study isonly the sixth fMRI-study investigating the treatment-effects of SSRIs. The fact that we scannedpatients three times, and that many patients initially were treatment non-responders allowed usto investigate the distinction in brain activations between drug-effects and clinical response,which is an important addition to the existing literature.

Finally, we also looked at peripheral effects of SSRIs, far beyond the brain, but alsoconsidering the inhibition of SERTs by paroxetine in association with hemostasis. Increasedbleeding tendency is an infrequent but potentially dangerous adverse effect. We showed thateven standard doses of paroxetine might affect hemostasis, and suggested subgroups who areeligible for secondary prevention by genetic screening of the 5-HTTLPR.


12

245

Future research

Future studies with data from DELPHISeveral lines of research on data from the DELPHI-study still need to be worked out. We collectedsalivary samples at study-entry, T0 and T1 (see chapter 2, page 36). We will investigate thechanges in cortisol by treatment over time, and might be able to associate these changes withclinical response or other clinical or imaging (e.g. SERT) variables. We therefore will collaboratewith the group of Kirschbaum and Rohleder.142;143 In addition, we collected blood-samples atstudy-entry, T0 and T1, to determine ex-vivo SERT and norepinephrine inhibition byparoxetine.144;145 Finally, we determined omega-3 and omega-6 polyunsaturated fatty acids,146

again at study-entry, T0 and T1, which will be analyzed in relation with clinical response toparoxetine, and additionally in relation with SERT availability and occupancy.147

Furthermore, we are currently working on a sub-study which investigates the association ofmidbrain and amygdala SERT occupancy with the changes in amygdala-activation after (negative)facial expressions in the patients who underwent both SPECT and fMRI scans.

Future clinical studiesThe findings presented in this thesis may be the starting point of new research.- For better evidence-based recommendations about when to choose between switching,

augmentation, combination, or psychotherapeutic strategies as a next step, comparisons ofthese strategies relative to each other are necessary. This will require future algorithm-basedswitch studies, including psychotherapy and augmentation or combination strategies toimprove our knowledge to guide treatment for SSRI non-responders. In these studies, one armshould be to continue treatment (e.g. for another 6-8 weeks, depending on the timing ofswitching) as a ‘placebo’ control. Of course the somehow disappointing results fromSTAR*D148;149 should be reconsidered when such a major endeavor is set up again.

- As a smaller enterprise, the relative benefits of switching after 6 weeks versus a continuationof the same treatment for another 6 weeks would provide important guidance to clinicians.Quitkin et al. showed that MDD-patients treated with fluoxetine 20 mg/day who were non-responders at week 6, showed remission rates at week 12 between 31%-41%, indicative of asubstantial delayed remission.150 However studies directly comparing such prolongedtreatment versus switching have not yet been performed in MDD.

- The controversial evidence advocating initial, high doses of SSRI in panic disorder,126;128 socialfobia,151 posttraumatic stress disorder,152 OCD130;131;134 and other anxiety disorders127 necessitatesa well-performed meta-analysis of fixed-dose, dose-finding studies in anxiety disorders,stratified by disorder and drug. In addition, dose-escalation studies in these disorders would beinteresting, especially when combined with the investigation of neurobiological mechanismsby neuroimaging. For example, the effects of a randomized, placebo-controlled dose-escalation on dopaminergic D2-like receptors, or dopamine transporters in OCD would beinteresting.

Future neurobiological studiesIn order to develop more successful treatment approaches, we need to understand better whicheffects of treatment are required for recovery, and which biomarkers are associated withtreatment response. Therefore, more fundamental, neurobiological studies should aim at severalgaps in the evidence.- In order to elucidate SSRI working-mechanisms, the increased coupling between amygadala

and frontal and cingulate cortices61 must be investigated with other SSRIs. Secondly, it shouldbe investigated whether 5-HTTLPR polymorphisms mediate this coupling sec,48 or the increaseafter SSRIs, or both. If the magnitude of coupling would be mediated by 5-HTTLPRpolymorphisms, this would also corroborate our findings for the association between SERToccupancy with clinical response in LA/LA carriers.45

Chap

ter

Chapter 12

12

246

- In order to study secondary effects of SSRI treatment, neuroimaging approaches to quantifythe effects on 5-HT1A receptors during treatment and after dose-escalation are warranted inpatients. Although a desensitization of 5-HT1A was demonstrated for chronic SSRI treatment inanimals,118;153 the question remains whether this desensitization is also associated with 5-HT1Adownregulation and/or whether dose-response effects occur. This study should ideally beperformed in combination with gepirone-challenges at different moments during treatment toassess 5-HT1A receptor sensitivity.125

- In order to investigate secondary effects of serotonin, e.g. on endogenous dopamine releasein the nucleus accumbens and other brain regions, microdialysis experiments in animals,chronically treated with SSRIs, will further investigate dose-escalation and time dependentchanges in SSRI treatment.

- The hypothesis of the involvement of dopamine in mood and MDD must be investigated.107

Serotonergic antidepressants indirectly influence the dopaminergic system. Responders toamitriptyline showed decreased [123I]IBZM SPECT binding to striatal dopamine D2 receptors,relative to pretreatment.154 This decrease correlated with a decrease in HDRS-score. This mightrepresent an increased tonic dopamine release after amitriptyline, but may also be attributedto improved psychomotor activity in responders. Nevertheless, a failure to achieve alterationsin the dopaminergic system by antidepressants, might explain non-response or residualsymptoms (sleep disturbances, diminished pleasure, loss of interest, fatigue, and decreasedmotivation), which are often observed in clinical practice.107 Therefore, investigation of thedopaminergic effects of SSRIs, in combination with or versus behavioural approaches (re-activation) in combined fMRI and SPECT/PET studies in (retarded) MDD-patients are necessary.

- The effects of SSRIs on neurogenesis also merit further research. After prolongedantidepressant administration in rodents, increased hippocampal neurogenesis was found.121;123

The second-messenger effects of SSRIs upregulate cAMP responsive element binding protein(CREB) in the neuron’s nucleus, which consequently regulates CREB-directed genetranscription.117;155 To date, it is unclear which specific genes transcription are affected by CREB.Therefore the exact role of second messenger systems (and CREB in particular) in response toSSRIs remain to be elucidated.117 One of the genes that is (positively) influenced by CREB is thebrain derived neurotrophic factor (BDNF) gene. BDNF is a plethoric factor, which regulatesneuronal survival, migration, phenotypic differentiation, axonal and dendritic growth andsynapse formation. As such BDNF has a role in cognitive functions (e.g. in memory andhippocampal plasticity).100 BDNF and serotonin are closely interlinked: BDNF promotesproduction of serotonin, upregulates serotonin uptake/release, and modifies firing rates ofneurons in the raphe nuclei, while serotonin upregulates BDNF. This stimulating feedbackmechanism may be involved in the selection of useful synapses, which could be hypothesizedto be a requisite for clinical response.119 Few studies investigated the long-term changes inBDNF measured in venous blood of MDD patients treated with antidepressants.156-158 Althoughdifferent effects for different drugs were seen, generally an increase in BDNF after 6 monthswas found, significantly associated with a decrease in HDRS. Nevertheless, the role of BDNF asbiomarker for antidepressant response is equivocal. Furthermore, due to dilution, peripheralmeasures may not be adequate, so sampling from an internal jugular vein might be better.159

Therefore, this line of research might need development of the methods to measure BDNF(and CREB) more locally in the brain of humans.


12

247

Conclusion

This thesis addresses the pharmacological treatment of Major Depressive Disorder (MDD), andfocuses on what to do when patients do not respond to a standard dose of an antidepressant. Intwo systematic reviews, we summarized the evidence for various switch strategies and dose-escalation, and concluded that for both strategies insufficient evidence existed to unambiguouslyguide clinicians. We thereafter performed a randomized controlled trial in which we studied dose-escalation in depressed patients who had not responded to 6 weeks of paroxetine (20 mg/day),and combined clinical measurements with SPECT and fMRI neuroimaging in subgroups of theparticipants (DELPHI-study). With the DELPHI-study, we show that dose-escalation of paroxetinehas no clinical benefit over placebo dose-escalation, and furthermore quantify the absence of apharmacological effect on the target of paroxetine: the SERT.

Another systematic review and four studies in this thesis, investigate the role of monoaminesand SERT in the etiology of MDD and the neurobiological effects of pharmacological treatmentwith paroxetine. From these studies we conclude that MDD is not necessarily caused by amalfunction of the serotonergic system perse. MDD may better be considered as a disturbance ofthe emotional and cognitive functioning in the limbic-subcortical-cortical network. The increase ofserotonergic neurotransmission by paroxetine (and likely other antidepressants), appears toimprove the constraint of neuronal information processing, with less emotional activation andmore cognitive control in treatment responders relative to non-responders. The findings of thisthesis can be the starting point for further investigation of biomarkers for treatment response, inorder to develop better understanding of MDD and a more effective treatment approaches.

References



3. Entsuah R, Shaffer M, Zhang J. A critical examination of the sensitivity of unidimensional subscales derived from theHamilton Depression Rating Scale to antidepressant drug effects. J Psychiatr Res. 2002; 36: 437-448.

4. O'Sullivan RL, Fava M, Agustin C, Baer L, Rosenbaum JF. Sensitivity of the six-item Hamilton Depression Rating Scale. ActaPsychiatr Scand. 1997; 95: 379-384.

5. Hooper CL, Bakish D. An examination of the sensitivity of the six-item Hamilton Rating Scale for Depression in a sample ofpatients suffering from major depressive disorder. J Psychiatry Neurosci. 2000; 25: 178-184.







12. Dornseif BE, Dunlop SR, Potvin JH, Wernicke JF. Effect of dose escalation after low-dose fluoxetine therapy.Psychopharmacol Bull. 1989; 25: 71-79.

13. Schweizer E, Rickels K, Amsterdam JD, Fox I, Puzzuoli G, Weise C. What constitutes an adequate antidepressant trial forfluoxetine? J Clin Psychiatry. 1990; 51: 8-11.

14. Fava M, Cohen L, Rosenbaum JF, Reiter S, et a. High-dose fluoxetine in the treatment of depressed patients notresponsive to a standard dose of fluoxetine. J Affect Disord. 1992; 25: 229-234.

15. Fava M, Rosenbaum JF, McGrath PJ, Stewart JW, Amsterdam JD, Quitkin et al. Lithium and tricyclic augmentation offluoxetine treatment for resistant major depression: a double-blind, controlled study. Am J Psychiatry. 1994; 151: 1372-1374.

Chap

ter

Chapter 12

12

248



18. Fava M, Alpert J, Nierenberg A, Lagomasino I, Sonawalla S, Tedlow J et al. Double-blind study of high-dose fluoxetineversus lithium or desipramine augmentation of fluoxetine in partial responders and nonresponders to fluoxetine. JClinPsychopharmacol. 2002; 22: 379-387.





23. Ruhe HG, Huyser J, Swinkels JA, Schene AH. Switching antidepressants after a first selective serotonin reuptake inhibitorin major depressive disorder: a systematic review. J Clin Psychiatry. 2006; 67: 1836-1855.



26. Fava M, Rush AJ, Wisniewski SR, Nierenberg AA, Alpert JE, McGrath PJ et al. A comparison of mirtazapine andnortriptyline following two consecutive failed medication treatments for depressed outpatients: a STAR*D report. Am JPsychiatry. 2006; 163: 1161-1172.

27. McGrath PJ, Stewart JW, Fava M, Trivedi MH, Wisniewski SR, Nierenberg AA et al. Tranylcypromine versus venlafaxineplus mirtazapine following three failed antidepressant medication trials for depression: a STAR*D report. Am J Psychiatry.2006; 163: 1531-1541.



30. Nolen WA, van de Putte JJ, Dijken WA, Kamp JS. L-5HTP in depression resistant to re-uptake inhibitors. An opencomparative study with tranylcypromine. Br J Psychiatry. 1985; 147: 16-22.

31. Nolen WA, van de Putte JJ, Dijken WA, Kamp JS, et a. Treatment strategy in depression: II. MAO inhibitors in depressionresistant to cyclic antidepressants: Two controlled crossover studies with tranylcypromine versus L-5-hydroxytryptophanand nomifensine. Acta Psychiatr Scand. 1988; 78: 676-683.


33. Bell C, Abrams J, Nutt D. Tryptophan depletion and its implications for psychiatry. Br J Psychiatry. 2001; 178: 399-405.34. Booij L, Van der Does AJ, Riedel WJ. Monoamine depletion in psychiatric and healthy populations: review. Mol Psychiatry.

2003; 8: 951-973.35. Van der Does AJ. The effects of tryptophan depletion on mood and psychiatric symptoms. J Affect Disord. 2001; 64: 107-

119.36. Moore P, Landolt HP, Seifritz E, Clark C, Bhatti T, Kelsoe J et al. Clinical and physiological consequences of rapid



Gen Psychiatry. 2001; 58: 200-202.39. Ruhe HG, Booij J, Reitsma JB, Schene AH. Serotonin transporter binding with [123I]ß-CIT SPECT in major depressive

disorder versus controls: effect of season and gender. [Submitted]. 40. Staley JK, Krishnan-Sarin S, Zoghbi S, Tamagnan G, Fujita M, Seibyl JP et al. Sex differences in [123I]β-CIT SPECT measures

of dopamine and serotonin transporter availability in healthy smokers and nonsmokers. Synapse. 2001; 41: 275-284.41. Pirker W, Asenbaum S, Hauk M, Kandlhofer S, Tauscher J, Willeit M et al. Imaging serotonin and dopamine transporters

with [123I]β-CIT SPECT: binding kinetics and effects of normal aging. J Nucl Med. 2000; 41: 36-44.42. van Dyck CH, Malison RT, Seibyl JP, Laruelle M, Klumpp H, Zoghbi SS et al. Age-related decline in central serotonin

transporter availability with [123I]β-CIT SPECT. Neurobiol Aging. 2000; 21: 497-501.43. Buchert R, Schulze O, Wilke F, Berding G, Thomasius R, Petersen K et al. Is correction for age necessary in SPECT or PET of

the central serotonin transporter in young, healthy adults? J Nucl Med. 2006; 47: 38-42.44. Staley JK, Sanacora G, Tamagnan G, Maciejewski PK, Malison RT, Berman RM et al. Sex differences in diencephalon

serotonin transporter availability in major depression. Biol Psychiatry. 2006; 59: 40-47.45. Ruhé HG, Ooteman W, Booij J, Michel MC, Moeton M, Baas F et al. Serotonin transporter gene promoter polymorphisms

modify the association between paroxetine serotonin transporter occupancy and clinical response in major depressivedisorder. Pharmacogenetics and Genomics. 2008; [In press].


12

249

46. Smits KM, Smits LJ, Schouten JS, Stelma FF, Nelemans P, Prins MH. Influence of SERTPR and STin2 in the serotonintransporter gene on the effect of selective serotonin reuptake inhibitors in depression: a systematic review. MolPsychiatry. 2004; 9: 433-441.





51. Ruhe HG, Booij J, Weert HC van, Reitsma JB, Fransen EJF, Michel MC et al. Evidence why dose-escalation of paroxetine inmajor depressive disorder is not effective: A 6-week, randomized-controlled trial with assessment of serotonintransporter occupancy. Neuropsychpharmacology.[In Press].



54. Ruhe HG, Booij J, Veltman DJ, Michel MC, Schene AH. Successful treatment of major depressive disorder by paroxetineattenuates amygdala activation to negative facial expressions. An fMRI study. [Submitted].






60. Anand A, Li Y, Wang Y, Wu J, Gao S, Bukhari L et al. Antidepressant effect on connectivity of the mood-regulating circuit:an FMRI study. Neuropsychopharmacology. 2005; 30: 1334-1344.


62. Abdelmalik N, Ruhé HG, Barwari K, Dool E-Jvd, Meijers JCM, Middeldorp S et al. The effect of the selective serotoninreuptake inhibitor paroxetine on platelet function is modified by a SLC6A4 serotonin transporter polymorphism. JThromb Haemost. 2008; [In Press].

63. Skop BP, Brown TM. Potential vascular and bleeding complications of treatment with selective serotonin reuptakeinhibitors. Psychosomatics. 1996; 37: 12-16.



66. Lenox-Smith AJ, Jiang Q. Venlafaxine extended release versus citalopram in patients with depression unresponsive to aselective serotonin reuptake inhibitor. Int Clin Psychopharmacol. 2008; 23: 113-119.

67. Thase ME, Lu Y, Joliat MJ, Detke MJ. Remission in placebocontrolled trials of duloxetine with an SSRI comparator.Presented atthe annual meeting of the American Psychiatric Association, May 2003, San Francisco, CA . 2003.

68. Malhi GS, Ng F, Berk M. Dual-dual action? Combining venlafaxine and mirtazapine in the treatment of depression. Aust N ZJ Psychiatry. 2008; 42: 346-349.

69. Perahia DG, Quail D, Desaiah D, Corruble E, Fava M. Switching to duloxetine from selective serotonin reuptake inhibitorantidepressants: a multicenter trial comparing 2 switching techniques. J Clin Psychiatry. 2008; 69: 95-105.

70. Papakostas GI, Fava M, Thase ME. Treatment of SSRI-resistant depression: a meta-analysis comparing within- versusacross-class switches. Biol Psychiatry. 2008; 63: 699-704.

71. Lenox-Smith AJ, Schaeffer P, Reynolds A, Williard L. A double blind trial of venlafaxine XR versus citalopram in patientswith severe depression. J Psychopharmacol. 2001; 15 Suppl 3: A11.

72. Thase ME, Kremer C, Rodrigues H. Mirtazapine versus sertraline after SSRI nonresponse [Abstract]. American PsychiatricAssociation.2001 annual meeting: New Orleans LA. 2001.

73. Cipriani A, Barbui C, Brambilla P, Furukawa TA, Hotopf M, Geddes JR. Are all antidepressants really the same? The case offluoxetine: a systematic review. J Clin Psychiatry. 2006; 67: 850-864.


Chap

ter

Chapter 12

12

250

75. Papakostas GI, Charles D, Fava M. Are typical starting doses of the selective serotonin reuptake inhibitors sub-optimal? Ameta-analysis of randomized, double-blind, placebo-controlled, dose-finding studies in major depressive disorder. WorldJ Biol Psychiatry. 2007; 1-8.


77. de Abajo FJ, Montero D, Rodriguez LA, Madurga M. Antidepressants and risk of upper gastrointestinal bleeding. Basic ClinPharmacol Toxicol. 2006; 98: 304-310.

78. Wessinger S, Kaplan M, Choi L, Williams M, Lau C, Sharp L et al. Increased use of selective serotonin reuptake inhibitors inpatients admitted with gastrointestinal haemorrhage: a multicentre retrospective analysis. Aliment Pharmacol Ther.2006; 23: 937-944.



81. van Walraven C, Mamdani MM, Wells PS, Williams JI. Inhibition of serotonin reuptake by antidepressants and uppergastrointestinal bleeding in elderly patients: retrospective cohort study. BMJ. 2001; 323: 655-658.


83. van Marwijk HW, Grundmeijer HG, Bijl D, van Gelderen MG, de Haan M, van Weel-Baumgarten EM et al. NHG-StandaardDepressieve stoornis (depressie) (eerste herziening). Huisarts Wet. 2003; 46: 614-623.



86. Meyer JH, Wilson AA, Sagrati S, Hussey D, Carella A, Potter WZ et al. Serotonin Transporter Occupancy of Five SelectiveSerotonin Reuptake Inhibitors at Different Doses: An [11C]DASB Positron Emission Tomography Study. AmJPsychiatry.2004; 161: 826-835.




90. Hiltunen J, Akerman KK, Kuikka JT, Bergstrom KA, Halldin C, Nikula T et al. Iodine-123 labeled nor-beta-CIT as a potentialtracer for serotonin transporter imaging in the human brain with single-photon emission tomography. Eur J Nucl Med.1998; 25: 19-23.








98. Jans LA, Riedel WJ, Markus CR, Blokland A. Serotonergic vulnerability and depression: assumptions, experimentalevidence and implications. Mol Psychiatry. 2007; 12: 522-543.

99. Fox MA, Andrews AM, Wendland JR, Lesch KP, Holmes A, Murphy DL. A pharmacological analysis of mice with a targeteddisruption of the serotonin transporter. Psychopharmacology (Berl). 2007; 195: 147-166.


101. Murphy DL, Lesch KP. Targeting the murine serotonin transporter: insights into human neurobiology. Nat Rev Neurosci.2008; 9: 85-96.

102. Holmes A, Yang RJ, Murphy DL, Crawley JN. Evaluation of antidepressant-related behavioral responses in mice lackingthe serotonin transporter. Neuropsychopharmacology. 2002; 27: 914-923.


12

251

103. Holmes A, Murphy DL, Crawley JN. Abnormal behavioral phenotypes of serotonin transporter knockout mice: parallelswith human anxiety and depression. Biol Psychiatry. 2003; 54: 953-959.

104. Spoont MR. Modulatory role of serotonin in neural information processing: implications for human psychopathology.Psychol Bull. 1992; 112: 330-350.

105. Mayberg HS. Modulating dysfunctional limbic-cortical circuits in depression: towards development of brain-basedalgorithms for diagnosis and optimised treatment. Br Med Bull. 2003; 65: 193-207.

106. Phillips ML, Drevets WC, Rauch SL, Lane R. Neurobiology of emotion perception II: Implications for major psychiatricdisorders. Biol Psychiatry. 2003; 54: 515-528.

107. Dunlop BW, Nemeroff CB. The role of dopamine in the pathophysiology of depression. Arch Gen Psychiatry. 2007; 64: 327-337.

108. D'Aquila PS, Collu M, Gessa GL, Serra G. The role of dopamine in the mechanism of action of antidepressant drugs. Eur JPharmacol. 2000; 405: 365-373.

109. Knutson B, Adams CM, Fong GW, Hommer D. Anticipation of increasing monetary reward selectively recruits nucleusaccumbens. J Neurosci. 2001; 21: RC159.



112. Neumeyer JL, Tamagnan G, Wang S, Gao Y, Milius RA, Kula NS et al. N-substituted analogs of 2 beta-carbomethoxy-3 beta-(4'-iodophenyl)tropane (β-CIT) with selective affinity to dopamine or serotonin transporters in rat forebrain. J Med Chem.1996; 39: 543-548.


114. de Win MM, Habraken JB, Reneman L, Brink Wvan den, den Heeten GJ, Booij J. Validation of [123I]β-CIT SPECT to assessserotonin transporters in vivo in humans: a double-blind, placebo-controlled, crossover study with the selective serotoninreuptake inhibitor citalopram. Neuropsychopharmacology. 2005; 30: 996-1005.



117. Blendy JA. The role of CREB in depression and antidepressant treatment. Biol Psychiatry. 2006; 59: 1144-1150.118. Blier P, De Montigny C. Current advances and trends in the treatment of depression. Trends Pharmacol Sci. 1994; 15: 220-

226.119. Castren E. Is mood chemistry? Nat Rev Neurosci. 2005; 6: 241-246.120. Davidson C, Stamford JA. Effect of chronic paroxetine treatment on 5-HT1B and 5-HT1D autoreceptors in rat dorsal raphe

nucleus. Neurochem Int. 2000; 36: 91-96.121. Djavadian RL. Serotonin and neurogenesis in the hippocampal dentate gyrus of adult mammals. Acta Neurobiol Exp

(Wars). 2004; 64: 189-200.122. Kim SW, Park SY, Hwang O. Up-regulation of tryptophan hydroxylase expression and serotonin synthesis by sertraline.

Mol Pharmacol. 2002; 61: 778-785.123. Malberg JE, Eisch AJ, Nestler EJ, Duman RS. Chronic antidepressant treatment increases neurogenesis in adult rat

hippocampus. J Neurosci. 2000; 20: 9104-9110.124. Meyer JH, Kapur S, Eisfeld B, Brown GM, Houle S, DaSilva J et al. The effect of paroxetine on 5-HT2A receptors in

depression: an [18F]setoperone PET imaging study. Am J Psychiatry. 2001; 158: 78-85.125. Sargent P, Williamson DJ, Pearson G, Odontiadis J, Cowen PJ. Effect of paroxetine and nefazodone on 5-HT1A receptor

sensitivity. Psychopharmacology (Berl). 1997; 132: 296-302.126. Mavissakalian MR, Ryan MT. Rational treatment of panic disorder with antidepressants. Ann Clin Psychiatry. 1998; 10: 185-

195.127. Rickels K, Zaninelli R, McCafferty J, Bellew K, Iyengar M, Sheehan D. Paroxetine Treatment of Generalized Anxiety

Disorder: A Double-Blind, Placebo-Controlled Study. Am J Psychiatry. 2003; 160: 749-756.128. Practice guideline for the treatment of patients with panic disorder. Work Group on Panic Disorder. American Psychiatric

Association. Am J Psychiatry. 1998; 155: 1-34.129. Richtlijn Angststoornissen. Multidisciplinaire richtlijn Angststoornissen. Richtlijn voor de diagnostiek en behandeling van

volwassen cliënten met een angststoornis. Velde Vvd.2003. Utrecht, Trimbos-instituut, in opdracht van de LandelijkeStuurgroep Multidisciplinaire Richtlijnontwikkeling in de GGZ.

130. Denys D. Pharmacotherapy of obsessive-compulsive disorder and obsessive-compulsive spectrum disorders. PsychiatrClin North Am. 2006; 29: 553-584.

131. Figee M, Denys D. New pharmacotherapeutical approaches to obsessive compulsive disorder. CNS Spectr. 2008; [Inpress].

132. Ninan PT, Koran LM, Kiev A, Davidson JR, Rasmussen SA, Zajecka JM et al. High-dose sertraline strategy fornonresponders to acute treatment for obsessive-compulsive disorder: a multicenter double-blind trial. J Clin Psychiatry.2006; 67: 15-22.

133. Dougherty DD, Rauch SL, Jenike MA. Pharmacotherapy for obsessive-compulsive disorder. J Clin Psychol. 2004; 60: 1195-1202.

Chap

ter

Chapter 12

12

252

134. American Psychiatric Association. Practice guideline for the treatment of patients with obsessive-compulsive disorder.Available online at http//www.psych.org/psych_pract/treatg/pg/prac_guide.cfmArlington, VA: American PsychiatricAssociation; 2007:1-96.

135. Wittkampf KA, Ruhé HG, Huyser J, Schene AH. De Creatieve Psychiater? Een onderzoek onder Nederlandse psychiatersnaar de behandeling van depressieve stoornissen bij het falen van de eerste SSRI.2001. Amsterdam, ZorgprogrammaStemmingsstoornissen MFO Psychiatrie AMC/ De Meren.





140. Cochrane Collaboration. Preparing, maintaining and promoting the accessibility of systematic reviews of the effects ofhealth care interventions. Internet: http://www.cochrane.org/. 2001. Cochrane Collaboration.

141. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986; 7: 177-188.142. Rohleder N, Nater UM, Wolf JM, Ehlert U, Kirschbaum C. Psychosocial stress-induced activation of salivary alpha-amylase:

an indicator of sympathetic activity? Ann N Y Acad Sci. 2004; 1032: 258-263.143. Schmidt-Reinwald A, Pruessner JC, Hellhammer DH, Federenko I, Rohleder N, Schurmeyer TH et al. The cortisol response

to awakening in relation to different challenge tests and a 12-hour cortisol rhythm. Life Sci. 1999; 64: 1653-1660.144. Owens MJ, Krulewicz S, Simon JS, Sheehan DV, Thase ME, Carpenter DJ et al. Estimates of Serotonin and Norepinephrine

Transporter Inhibition in Depressed Patients Treated with Paroxetine or Venlafaxine. Neuropsychopharmacology. 2008.145. Gilmor ML, Owens MJ, Nemeroff CB. Inhibition of norepinephrine uptake in patients with major depression treated with

paroxetine. Am J Psychiatry. 2002; 159: 1702-1710.146. Parker G, Gibson NA, Brotchie H, Heruc G, Rees AM, Hadzi-Pavlovic D. Omega-3 fatty acids and mood disorders. Am J

Psychiatry. 2006; 163: 969-978.147. du Bois TM, Deng C, Bell W, Huang XF. Fatty acids differentially affect serotonin receptor and transporter binding in the

rat brain. Neuroscience. 2006; 139: 1397-1403.148. Rush AJ, Fava M, Wisniewski SR, Lavori PW, Trivedi MH, Sackeim HA et al. Sequenced treatment alternatives to relieve

depression (STAR*D): rationale and design. Control Clin Trials. 2004; 25: 119-142.149. Rush AJ, Trivedi MH, Wisniewski SR, Nierenberg AA, Stewart JW, Warden D et al. Acute and longer-term outcomes in

depressed outpatients requiring one or several treatment steps: a STAR*D report. Am J Psychiatry. 2006; 163: 1905-1917.150. Quitkin FM, Petkova E, McGrath PJ, Taylor B, Beasley C, Stewart J et al. When should a trial of fluoxetine for major

depression be declared failed? Am J Psychiatry. 2003; 160: 734-740.151. Stein DJ, Ipser JC, Balkom AJv. Pharmacotherapy for social anxiety disorder. Cochrane Database of Systematic Reviews.

2000; Issue 4.: Art. No.: CD001206. DOI: 10.1002/14651858.CD001206.pub2.152. Stein DJ, Ipser JC, Seedat S. Pharmacotherapy for post traumatic stress disorder (PTSD). Cochrane Database of Systematic

Reviews. 2006; Issue 1: Art. No.: CD002795. DOI: 10.1002/14651858.CD002795.pub2.153. Blier P, De Montigny C. Serotonin and drug-induced therapeutic responses in major depression, obsessive-compulsive

and panic disorders. Neuropsychopharmacology. 1999; 21: 91S-98S.154. Ebert D, Feistel H, Loew T, Pirner A. Dopamine and depression--striatal dopamine D2 receptor SPECT before and after

antidepressant therapy. Psychopharmacology (Berl). 1996; 126: 91-94.155. Belmaker RH, Agam G. Major depressive disorder. N Engl J Med. 2008; 358: 55-68.156. Aydemir O, Deveci A, Taneli F. The effect of chronic antidepressant treatment on serum brain-derived neurotrophic factor

levels in depressed patients: a preliminary study. Prog Neuropsychopharmacol Biol Psychiatry. 2005; 29: 261-265.157. Matrisciano F, Bonaccorso S, Ricciardi A, Scaccianoce S, Panaccione I, Wang L et al. Changes in BDNF serum levels in

patients with major depression disorder (MDD) after 6 months treatment with sertraline, escitalopram, or venlafaxine. JPsychiatr Res. 2008.

158. Piccinni A, Marazziti D, Catena M, Domenici L, Del Debbio A, Bianchi C et al. Plasma and serum brain-derived neurotrophicfactor (BDNF) in depressed patients during 1 year of antidepressant treatments. J Affect Disord. 2008; 105: 279-283.

159. Barton DA, Esler MD, Dawood T, Lambert EA, Haikerwal D, Brenchley C et al. Elevated brain serotonin turnover in patientswith depression: effect of genotype and therapy. Arch Gen Psychiatry. 2008; 65: 38-46.

CHA

PTER

SAMENVATTING EN CONCLUSIES

ChapterSamenvatting en conclusies

13

255

Samenvatting en conclusies

In deze Nederlandse samenvatting wordt een overzicht gegeven van een aantal basiskenmerkenen een aantal problemen in de farmacologische behandeling van depressie. Deze problemenstonden aan de basis van de studies beschreven in dit proefschrift. Hiervoor is eenonderzoeksproject opgezet met verschillende substudies (beschreven in hoofdstuk 2). Ditonderzoek had de overkoepelende naam DELPHI-studie, wat staat voor ‘Dose-escalationLegitimate? Pharmacology and Imaging studies in depression’. In het Nederlands: ‘Isdosisverhoging legitiem? Farmacologische en neuroimaging studies bij depressie.

Na de achtergrond wordt een samenvatting gegeven van de belangrijkste resultaten van destudies, de klinische relevantie ervan, gevolgd door een afsluitende conclusie.

Achtergrond en problemen in de behandeling van depressie

Major Depressive Disorder (MDD), in het Nederlands depressieve stoornis, verder aangeduid metdepressie, is een veelvoorkomende en invaliderende ziekte die in potentie recidiveert of verwordttot een chronische aandoening. Depressie staat momenteel wereldwijd op de tweede plaats vaninvaliderende ziekten en komt per jaar bij ±5.5% van de volwassenen voor. Ongeveer 12-14% vanalle mannen en 22-24% van alle vrouwen wordt gedurende hun leven getroffen door depressie.1;2

Depressie gaat gepaard met hoge directe kosten betreffende de behandeling en indirecte kostenals gevolg van verminderde arbeidsproductiviteit en het verlies van kwaliteit van leven. Naastverschillende vormen van psychotherapie wordt farmacotherapie met antidepressiva vaakaangewend als behandeling.3-8

Antidepressiva kunnen worden ingedeeld in vier groepen of ‘klassen’: Selectieve serotonineheropname remmers (SSRIs), tricyclische antidepressiva (TCAs), monoamine oxidase remmers(MAOIs) en een restgroep, waaronder de zogenaamde ‘dual action’ antidepressiva. SSRIs, 'dual-action' antidepressiva en TCAs worden het meest gebruikt.9 De meeste antidepressivabeïnvloeden de balans van verschillende boodschapperstoffen in het brein, de‘neurotransmitters’ serotonine (5-HT) en noradrenaline (zogenaamde monoaminen). Deboodschapperstoffen zijn noodzakelijk voor de communicatie tussen zenuwcellen. Communicatievindt plaats door het uitscheiden van de boodschapperstoffen in de ruimte tussen dezenuwcellen: de synaps. De meeste antidepressiva blokkeren een belangrijk onderdeeltje van dezenuwcel: de serotonine (of noradrenaline) heropname transporter (respectievelijk SERT of NAT),waardoor de hoeveelheid serotonine en/of noradrenaline in de synaps wordt verhoogd. Op dezemanier verhogen de meeste antidepressiva de serotonerge en/of noradrenerge signaaloverdrachtof neurotransmissie. Als dit echter nader wordt bekeken, hebben verschillende groepenantidepressiva ook specifieke, aanvullende effecten op de zenuwcellen; zogenaamde pre- enpost-synaptische effecten (zie hoofdstuk 1, pagina 19).10;11 In reactie op de verhoogde serotonergeen/of noradrenerge signaaloverdracht worden ‘second messenger’ systemen geactiveerd, diemen op dit moment slechts gedeeltelijk begrijpt.12

Het brein is ook een systeem van met elkaar verbonden kernen of schakelcentra, ook welneuronale netwerken genoemd. Behandeling met antidepressiva blijkt aanzienlijke veranderingenin de functie van dergelijke neuronale netwerken te veroorzaken (zie hoofdstuk 1, pagina 21).13-15

Tot op heden is echter onvoldoende duidelijk welke veranderingen noodzakelijk en specifiek zijnom een depressie te genezen. Het lijkt erop dat de effecten van een toegenomensignaaloverdracht gemedieerd worden door de genetische eigenschappen van patiënten(bijvoorbeeld het 5-HTTLPR polymorphisme16) (zie hoofdstuk 1, pagina 22). Er is namelijk eenverband tussen genetische eigenschappen van patiënten en de effecten van behandeling metantidepressiva.17 Ook blijken de ontwikkeling van het brein en verbindingen in de neuronalenetwerken genetisch bepaald te zijn.18

Chap

ter

Chapter 13

13

256

Antidepressiva hebben een vergelijkbare effectiviteit; ongeveer 50% van de depressievepatiënten reageert op het eerste voorgeschreven antidepressivum. Bij slechts een kleinerpercentage van de patiënten wordt volledige remissie van de depressieve klachten bereikt.19;20

Daarom is onvoldoende verbetering op een eerste antidepressivum een groot klinisch probleem.Ten eerste wordt een behandelaar geconfronteerd met de vraag hoe de verandering van

symptomen goed en efficiënt te meten. Om op juiste gronden klinische beslissingen te nemen, isde beschikbaarheid en implementatie van valide, korte en eenvoudige meetinstrumentenbelangrijk. Ten tweede is het de vraag wat het optimale moment is om de gestarteantidepressieve behandeling te evalueren en zonodig te wijzigen: aangezien antidepressiva enigeweken tijd nodig hebben om een verbetering te bewerkstelligen, moet dit niet te vroeg, maar ookniet te laat plaatsvinden.21

Als een patiënt geen verbetering (veelal gedefinieerd als 50% verbetering van de klachten;‘respons’) ervaart op de gestarte SSRI, zijn er vijf farmacologische strategieën: 1) nog enkeleweken ongewijzigd voortzetten van de aanvankelijke behandeling, 2) dosisverhoging, 3) hetveranderen of ‘switchen’ van het antidepressivum, 4) het toevoegen van een middel dat dewerking van het antidepressivum versterkt (augmentatie) en 5) het combineren van verschillendeantidepressiva. Over al deze strategieën is de wetenschappelijke kennis beperkt. Om goede‘evidence-based’ behandelalgoritmen te ontwikkelen zijn goede systematischeliteratuuroverzichten noodzakelijk. Naast het overzicht over de effectiviteit van behandelingen,verduidelijken deze overzichten ook waar zich lacunes in de kennis over de behandeling metantidepressiva bevinden. Deze lacunes kunnen in nieuwe onderzoeksprojecten wordenbestudeerd.

De verlate start van de werking van antidepressiva, de beperkte werkzaamheid vanantidepressiva en de verscheidenheid aan farmacologische effecten in het brein, roepen de meerfundamentele vraag op wat het pathofysiologische model voor depressie is. Meer specifiek kanmen zich afvragen: wat is precies de onderbouwing van het uitgangspunt van de werking van demeeste antidepressiva: de monoamine hypothese en de betrokkenheid van het serotonergesysteem als een oorzakelijke verklaring van het ontstaan van depressie?

Samenvatting en conclusies van de studies in dit proefschrift

Dit proefschrift behandelt verschillende vragen:

1 Is een korte, door de behandelaar in te vullen vragenlijst even effectief en precies als degebruikelijke Hamilton depressieschaal (HDRS)?In hoofdstuk 3 heranalyseren wij de behandeluitkomsten van twee antidepressiva/psychotherapie onderzoeken met als doel de evaluatie van de uitkomsten bepaald doorverschillende subschalen van de oorspronkelijke HDRS.22 Conform andere studies,23-26 vondenwij dat de twee 6-items bevattende subschalen van Maier27 en Bech28 overeenkomen met deoriginele 17-items bevattende HDRS. Dit gold voor het meten van de ernst van depressie en degevoeligheid om veranderingen in de tijd te meten. Wij berekenden een conversietabel vanoorspronkelijke HDRS-scores naar de Maier en Bech scores (en vice versa). Ook bepaalden weafkappunten voor remissie bij deze subschalen.29-31 Wij concluderen dat met deze subschalenclinici de ernst en respons van patiënten efficiënter kunnen meten dan met de oorspronkelijkeHDRS.

2 Wat is het wetenschappelijke bewijs voor dosisverhoging als strategie voor patiënten die nietverbeterden op een eerste SSRI?In hoofdstuk 4 beschrijven wij een systematisch literatuuroverzicht van het wetenschappelijkebewijs voor een dosis-respons relatie voor SSRIs bij de behandeling van depressie.32 Wijconcludeerden dat er géén wetenschappelijk bewijs bestaat dat dosisverhoging van SSRIseffectief is in de eerste 4 weken van de behandeling met een standaarddosering. Verder lijktdosisverhoging na 6 weken behandeling met een standaarddosering zelfs minder effectief


13

257

dan het ongewijzigd doorgebruiken van de oorspronkelijke dosering. Dosisverhoging gaatgepaard met een toename van de bijwerkingen zonder dat patiënten de behandeling stakenvanwege deze bijwerkingen. Vanwege methodologische tekortkomingen concludeerden wijdesondanks dat het wetenschappelijke bewijs voor dosisverhoging noch een duidelijktoegenomen werkzaamheid, noch een duidelijke onwerkzaamheid aantoont. Deze conclusievormde het startpunt voor de DELPHI-studie, beschreven in de hoofdstukken 2 en 7 tot enmet 10.

3 Wat is het wetenschappelijke bewijs voor het ‘switchen’ van antidepressiva als strategie voorpatiënten die niet verbeterden op een eerste SSRI?In hoofdstuk 5 beschrijven wij een systematisch literatuuroverzicht waarin hetwetenschappelijke bewijs voor het wisselen (‘switchen’) van antidepressiva in het geval dat eronvoldoende verbetering van de depressieve klachten optreedt bij het eersteantidepressivum.33 Voor drie gerandomiseerde switch-studies die het switchen naar een 2e

SSRI vergeleken met een switch naar een SNRI (venlafaxine) konden wij een meta-analyseuitvoeren.34-36 Deze meta-analyse liet alleen in de meest rooskleurige analyse een beperktemeerwaarde van het switchen naar venlafaxine als tweede antidepressivum zien. Wijconcluderen dat bij onvoldoende respons op een eerste SSRI, het ‘switchen’ naar elk andermiddel kan plaatsvinden, zonder duidelijke verschillen in werkzaamheid en dus vergelijkbaar ismet de keuze van het eerste middel.

4 Verlaagt de depletie van monoaminerge systemen de stemming van mensen en is dezestemmingsdaling verschillend voor verschillende subgroepen van mensen?In hoofdstuk 6 beschrijven wij een systematisch literatuuroverzicht van hetwetenschappelijke bewijs dat zogenaamde monoamine depletie studies de stemming vanonderzochte proefpersonen verminderen.37 Hierbij voerden wij een meta-analyse uit, zodat deresultaten van 53 kleine depletie studies konden worden samengevat (‘gepooled’). Wijvonden dat serotonerge, noch noradrenerge/dopaminerge depletie een stemmingsdalinggeeft in gezonde controlepersonen, terwijl depletie de stemming van gezondecontrolepersonen met een eerstegraads familielid dat depressief is geweest enigszinsverminderd. Bij medicatievrije patiënten die na een depressie een volledige remissie van hunklachten hebben bereikt vonden we een matige stemmingsdaling. Monoamine depletieveroorzaakt een terugval bij patiënten die na een depressie een volledige remissie van hunklachten hebben bereikt, echter alleen als deze patiënten een serotonerg werkendantidepressivum gebruiken. Bij depressieve patiënten vindt geen verdere stemmingsdalingplaats na depletie.

Wij concluderen dat stemming niet direct is gecorreleerd aan de hoeveelheid serotonine,noradrenaline of dopamine in het brein. Wel lijken monoaminerge systemen een belangrijkerol te spelen in de kwetsbaarheid om depressief te worden.

5 Verschillen depressieve patiënten en gezonde controlepersonen in het aantal serotoninetransporters en is het aantal serotonine transporters gecorreleerd met de ernst van dedepressie?In hoofdstuk 7 onderzoeken we het aantal serotonine transporters (SERTs) in het brein vandepressieve patiënten, zoals gemeten in de eerste (baseline) SPECT scans van de deelnemersaan de DELPHI-SPECT studie. Wij vonden diverse interactie-effecten tussen geslacht enziektestatus met betrekking tot het aantal SERTs in zowel het diencephalon ((hypo)-thalamusregio) als het midbrain (de hersenstam), met daarnaast interactie-effecten tussenroken en ziektestatus (alleen in het diencephalon). Deze interactie-effecten resulteerden ineen lager aantal SERTs bij depressieve mannen maar juist meer SERTs bij depressieve vrouwen(ten opzichte van gezonde controlepersonen). Daarnaast bleek het seizoen waarin de scangemaakt wordt een duidelijke verandering van het aantal SERTs in het midbrain te geven.Deze bevindingen wijzen op complexe effecten van geslacht, roken en het seizoen waarin descan gemaakt wordt op het aantal SERTs bij depressieve patiënten en gezonde controles.Deze effecten moeten in toekomstig onderzoek van het aantal SERTs bij depressievepatiënten versus gezonde controlepersonen worden meegewogen in de analyses.

Chap

ter

Chapter 13

13

258

6 Verandert een veelvoorkomend genetisch polymorfisme van de promotor regio van het genvoor de serotoninereceptor (SLC6A4) de associatie tussen de SERT bezettingsgraad doorparoxetine en de klinische respons?In hoofdstuk 8 onderzoeken wij de relatie tussen klinische respons na 6 weken paroxetinegebruik en de mate van bezetting van het farmacologische doelwit van paroxetine: deserotonine transporter. De mate van bezetting, ofwel bezettingsgraad van de SERT is hierinde verandering van de hoeveelheid gemeten SERTs na 6 weken behandeling ten opzichte vande hoeveelheid gemeten SERTs vóór behandeling (baseline).38 Daarbij onderzoeken we of derelatie tussen klinische respons en bezettingsgraad wordt beïnvloedt door genetischevarianten in het SERT-gen, de 5-HTTLPR polymorfismen.

Wij vonden dat voor alle patiënten samen, de SERT bezettingsgraad niet geassocieerd ismet de klinische respons. Echter, nadat we de patiënten onderverdeelden op basis van hun 5-HTTLPR polymorfisme, vonden we alleen bij patiënten met het (gunstige)17;39 LA/LA genotypeeen significante relatie tussen SERT bezettingsgraad en klinische respons. Wij verklaren onzebevindingen met de hypothese dat patiënten met het 5-HTTLPR LA/LA genotype mogelijk eenmeer flexibel serotonerg systeem bezitten,18 dat beter beïnvloed kan worden doorserotonerge antidepressiva.

7 Is dosisverhoging van paroxetine een effectieve klinische strategie bij onvoldoende responsbij de behandeling van depressie?In hoofdstuk 9 onderzoeken wij dosisverhoging als strategie voor patiënten die onvoldoendeverbeteren op een standaarddosering paroxetine gedurende 6 weken in eengerandomiseerde klinische studie (RCT). Dit onderzoek was een gevolg van de twijfels over deeffectiviteit van dosisverhoging (zie ook hoofdstuk 4).32 Als meest belangrijke aanvullingonderzoeken we in een subgroep van de patiënten of een echte dosisverhoging vanparoxetine de SERT bezettingsgraad meer doet toenemen dan een placebo-verhoging.40

In een interim analyse analyseerden wij 57 gerandomiseerde patiënten en vonden dat ergeen verschil in werkzaamheid bestaat tussen een echte en een placebo dosisverhoging vanparoxetine. De studie werd hierop afgebroken in verband met futiliteit. Als belangrijkstebevinding laat ons onderzoek zien dat een echte dosisverhoging van paroxetine geen groteretoename van de SERT bezettingsgraad geeft dan een placebo dosisverhoging. Dit is teverklaren doordat waarschijnlijk al bij een standaarddosering een plateau van SERT bezettingwordt bereikt. We concluderen, in overeenstemming met eerdere, echter meer twijfelachtigegegevens,41;42 dat dosisverhoging geen meerwaarde heeft in de behandeling van depressie.

8 Wordt de hyperactivatie van de amygdala bij depressie door een behandeling met paroxetinegenormaliseerd?In hoofdstuk 10 onderzoeken we de emotionele respons van het brein op een (negatieve)gezichtentaak in de DELPHI-fMRI substudie.43 We onderzoeken hierin de effecten vanbehandeling met paroxetine op de vaak vastgestelde hyperactivatie van de amygdala* tijdenseen depressie. Ten opzichte van gezonde controlepersonen vonden we bij depressievepatiënten een toegenomen activiteit in beide amygdala’s en de daaraan grenzende regio. Wijvonden aanvankelijk een toename van de activatie van de rechter amygdala na 6 wekenbehandeling, die kon worden toegeschreven aan de patiënten (12/20) die onvoldoendeverbeterden (non-responders). Deze toegenomen amygdala-activatie nam af in de periode totweek 12. Vooral bij de patiënten die in week 6 en/of week 12 ≥50% verbetering van hunklachten ervaren, nam de amygdala-activatie beiderzijds af, terwijl de amygdala-activatie inweek 6 en 12 geassocieerd was met de HDRS-score. Patiënten met onvoldoende responshadden meer activatie in andere emotie ‘genererende’ (ventrale; zie hoofdstuk 1, pagina 18)gebieden, terwijl responderende patiënten meer activatie toonde in emotie ‘regulerende’(dorsale; zie hoofdstuk 1, pagina 18) gebieden. We concluderen dat, in overeenstemming meteerdere studies,44-47 12 weken behandeling met paroxetine de amygdala en andere emotie

* De amygdala is een amandelvormig onderdeel van de hersenen, dat zich zowel in de linker als rechter hersenhelft bevindt,en onder andere betrokken is bij automatische reacties op angstige situaties.


13

259

genererende gebieden minder actief doet zijn en dat de corticale, regulerende functieverbeterd. Dit gebeurt vooral bij patiënten die een klinische respons ervaren. Dezebevindingen wijzen op een toegenomen frontaal-limbische controle als werkingsmechanismevoor paroxetine geïnduceerde behandeleffecten (zie hoofdstuk 1, pagina 21).

9 Hoe veranderen de haemostase en stollingsparameters in bloedplaatjes gedurende eenbehandeling met paroxetine en worden deze veranderingen gemedieerd doordosisverhoging of een genetisch polymorfisme van de promotor regio van het serotoninetransporter gen?In hoofdstuk 11 beschrijven we de resultaten van een weinig voorkomende, maar potentieelgevaarlijke bijwerking van SSRIs: toegenomen bloedingsneiging.48 We vonden dat eenstandaard dosis paroxetine (20 mg/dag) al een significante daling van serotonine en β-thromboglobuline (β-TG) in de plaatjes veroorzaakt, zonder dat dit verder afneemt bijdosisverhoging. Bovendien vonden we dat het 5-HTTLPR polymorfisme deze effectenmedieert: ten opzichte van LA/LA-dragers, neemt de bloedingstijd significant toe in patiëntenmet <2LA-allelen en hadden patiënten zonder een LA-allel, sterkere veranderingen in plaatjesaggregatie gerelateerde parameters. Deze bevindingen zijn van belang voor patiënten meteen eerdere ernstige bloeding; bij hen kan genotypering overwogen worden, waarbij bij eenS’/S’ genotype het starten van, of het switchen naar een niet-serotonerg antidepressivummoet worden overwogen.

Klinische relevantie

Switchen Ondanks een klein voordeel van een eventuele switch naar venlafaxine in plaats van een tweedeSSRI (slechts in de meest gunstige analyse), concluderen wij dat de switch-opties na een eersteSSRI identiek zijn aan de keuze voor het eerste middel. Op basis van de literatuur kunnen geenduidelijke aanbevelingen worden gegeven voor een keuze na een eerste SSRI. Sinds onssystematische literatuuroverzicht verschenen er nog één nieuwe RCT die switchen naarvenlafaxine vergeleek met switchen naar citalopram,49 en drie open-label studies.50-52

Recent werd er een andere meta-analyse gepubliceerd, waarin een niet-SSRI switch(venlafaxine, mirtazapine of buproprion) werd vergeleken met switchen naar een tweede SSRI.53

In deze meta-analyse includeerdende auteurs 2 extra studies (abstracts van symposia)54;55 enexcludeerdenzij (net als wij in onze meest rooskleurige analyse34) een methodologischtwijfelachtige studie. Deze meta-analyse concludeert dat er bij switch naar venlafaxine eenstatistisch significant, doch klinisch niet relevant voordeel bestaat in het aantal patiënten dat eenremissie van de klachten bereikt.5;56 Voor het uitkomstcriterium respons werd geen significantverschil vastgesteld.53 Onze conclusie blijft met deze aanvullende bevindingen ongewijzigd; er lijktvoor individuele patiënten geen duidelijk verschil tussen de verschillende switch-opties tebestaan.

Dosisverhoging Na ons systematische literatuuroverzicht concluderen wij dat er geen overtuigend bewijs tenfaveure of tegen dosisverhoging bestaat. Een onafhankelijke onderzoeksgroep concludeerdehetzelfde op basis van een ander overzichtsartikel.41 De resultaten van onzedosisverhogingsstudie (hoofdstuk 9) verdisconteerde de methodologische beperkingen vaneerdere studies en werd vanwege futiliteit afgebroken na een interim-analyse.40 Hoewel hetaantal gerandomiseerde patiënten beperkt was, waren de verschillen tussen echte en placebodosisverhoging zodanig klein (en niet significant) dat de studie voortijdig moest wordenbeëindigd vanwege futiliteit. Dit houdt in dat het gevonden verschil in HDRS-scores kleiner wasdan een vooraf berekend noodzakelijk verschil om nog enige kans op een daadwerkelijk verschilte verwachten, indien het onderzoek voltooid zou worden conform de oorspronkelijke planning.

Chap

ter

Chapter 13

13

260

Zodoende beschermde deze interim-analysepatiënten om blootgesteld te worden aan eeninterventie waarvan op dat moment bijna 100% zeker was dat er uiteindelijk geen meerwaarde vanzou worden vastgesteld. Deze uitkomst is in overeenstemming met het eerdere onderzoek vanLicht et al.57 Helaas maakt het voortijdig staken van onze studie het onmogelijk om eventuelesubgroepen te identificeren waarin dosisverhoging wellicht toch effectief is (bijvoorbeeld depatiënten die gedeeltelijk (partieel) verbeterden).58;59

De substudie waarin wij de verandering in de SERT bezettingsgraad bepaalden bij patiëntendie een dosisverhoging kregen, versterkt de hierboven beschreven klinische bevindingen. Wijconcluderen op basis van ons systematisch literatuuroverzicht, onze RCT en onze imaging studie,dat dosisverhoging van SSRIs geen effectieve strategie is voor depressieve patiënten die na 6weken onvoldoende verbetering ervaren op een standaarddosering.

Een recente meta-analyse van 9 placebo-gecontroleerde, ‘fixed-dose’ studies, gericht op hetvinden van de optimale dosering (‘dose-finding’) van SSRIs,60 concludeerde dat er een gering,maar statistisch significant voordeel was van hogere doseringen van citalopram, escitalopram,fluoxetine, fluvoxamine, paroxetine en/of sertraline, gegeven vanaf het begin van de behandeling(4% meer patiënten met een respons; p= 0.04). Het kleine verschil wordt echter als klinisch nietrelevant beoordeeld. De auteurs konden niet aangeven dat de significantie van de resultaten aanéén van de SSRIs kon worden toegeschreven. Een nadere inspectie van hun gegevens laat in iedergeval geen enkel voordeel van hogere doseringen paroxetine zien, maar wel een duidelijk grotereuitval vanwege bijwerkingen (16.5% ten opzichte van 9.8% bij standaarddoseringen). Daarom blijfthet volgens ons dubieus of hogere doseringen SSRIs vanaf het begin van de behandeling zinvolzijn bij depressieve patiënten. Het zou zelfs zo kunnen zijn dat hogere doseringen mensen eerderdoen stoppen vanwege bijwerkingen, zodat een adequate, langerdurende blootstelling aan (quabijwerkingen acceptabele) standaarddoseringen voorkomen wordt.

Andere klinisch relevante resultaten Twee andere relevante bevindingen komen in dit proefschrift naar voren. Ten eerste blijkt dat deMaier en Bech subschalen resultaten geven die overeenkomen met de volledige HDRS17.22 Andereauteurs suggereren dat deze subschalen beter de kernsymptomen van depressie meten en dathet gebruik van subschalen zelfs efficiëntere metingen geeft23;24 vanwege de afname invariantie.61 De Maier of Bech subschalen kunnen dus gebruikt worden in de klinische beoordelingvan (de behandeling van) depressieve patiënten, naast zelfinvullijsten als de ‘Inventory fordepressive symptomatolgy self-rated’ (IDS-SR) of de ‘Beck depression inventory’ (BDI). Dit isvooral belangrijk bij mensen die zelfinvullijsten niet betrouwbaar kunnen invullen (bijvoorbeeld bijpsychotische kenmerken of bij ernstige cognitieve beperkingen).

Ten tweede verdient de verhoogde bloedingsneiging bij SSRI-gebruik meer aandacht.48 Ditwas al bekend vooral in combinatie met NSAIDs.62-67 De standaard van het Nederlands HuisartsGenootschap adviseert een TCA als de patiënt tevens NSAIDs gebruikt,4 hetgeen volgens ons nietgeheel correct is, aangezien een serotonerg werkend TCA als clomipramine vermoedelijkvergelijkbare problemen geeft. Vanwege de relatief lage incidentie van potentieellevensbedreigende gastro-intestinale bloedingen (4.3/1000 SSRI behandeljaren; 14.5/1000SSRI+NSAID/aspirine behandeljaren; ten opzichte van 1.2/1000/jaar in onbehandeldevolwassenen),65 lijkt de toegenomen bloedingsneiging bij SSRIs irrelevant bij patiënten die geeneerdere ernstige bloedingsincidenten hebben doorgemaakt, of een trauma of operatiemeemaken. Echter, bij patiënten met anamnestisch een eerder ernstig bloedingsincident, of bijpatiënten die een electieve operatie zullen ondergaan en wellicht ook bij patiënten die regelmatigNSAIDs gebruiken, is genotypering van het 5-HTTLPR een overweging. Bij het vaststellen van hetS’/S’ genotype zou dan een switch naar een niet-serotonerg werkend antidepressivum kunnenworden geadviseerd. De (kosten-) effectiviteit van deze benaderingswijze dient nog wel teworden onderzocht.


13

261

Neurobiologie van de behandeling van depressie

Met ons onderzoek onderzochten we ook de neurobiologie van depressiebehandeling metparoxetine in vivo. De algemene beschouwing omtrent dit deel van het proefschrift vraagtaanzienlijke uitleg en enige voorkennis, welke gedetailleerder uiteen wordt gezet in deEngelstalige ‘general discussion’.

Met de SPECT-scans gedurende de gerandomiseerde, dubbelblinde dosisverhogings fase40

laten wij als eersten ondubbelzinnig zien dat er bij dosisverhoging geen sprake is van een stijgingvan de SERT-bezettingsgraad. Dit kon op basis van eventuele vertekening in eerder onderzoekniet zo eenduidig worden geconcludeerd.68-80 Bovendien wijzen onze gegevens op een matigerelatie tussen bezettingsgraad en klinische respons, in tegenstelling tot een eerder gepostuleerde80% bezettingsgraad die hiervoor vereist zou zijn.38;81 De feiten dat de relatie tussen bezetting enklinische respons alleen lijkt te bestaan voor dragers van het LA/LA genotype van het 5-HTTLPRgen en dat datzelfde genotype de anatomie en connectiviteit van emotiegerelateerde netwerkenbeïnvloedt,18 wijzen erop dat de (heilzame) effecten van een verhoogde serotonergesignaaloverdracht waarschijnlijk afhankelijk zijn van de ontwikkeling van het serotonergesysteem.

Een toename van serotonine geeft in het algemeen een beter controle of beteugeling van deinformatieverwerking in het brein,82 met 1) een vermindering van de kans op doorschieten in eenreactie van andere systemen (bijvoorbeeld het dopaminesysteem) en 2) een verminderdegevoeligheid voor verstorende in- en externe invloeden op het brein. Deze effecten worden metdepletie van serotonine omgekeerd beïnvloedt. Daarnaast beïnvloedt serotonine de aanmaak van‘brain derived neurotrophic factor’ (BDNF),83 een eiwit dat betrokken is bij neuroplasticiteit. Onzeen andere fMRI-studies kunnen in dat licht worden geïnterpreteerd. Bij responders verbeterde decognitieve controle en nam de emotionele hyperactivatie af.43 Uit een eerder onderzoek bleek datbehandeling met sertraline een toename veroorzaakte in de functionele koppeling tussen deamygdala en de frontale en (pregenuale) cingulaire cortex,84 respectievelijk emotionele encontrolerende, cognitieve onderdelen van het limbische-subcortico-corticale netwerk. Mogelelijkis deze toegenomen functionele koppeling een gevolg van de verhoogde serotonergeneurotransmissie, mogelijk het gevolg van een toename van BDNF, of beide. Dit dient nog verderonderzocht te worden.

Beperkingen van het onderzoek

In ieder hoofdstuk en de Engelstalige ‘general discussion’ zijn de mogelijke beperkingen van destudies van dit proefschrift afzonderlijk beschreven. De belangrijkste hiervan worden in dezeNederlandse samenvatting aangestipt.

Voor de SPECT-studies gebruikten wijeen niet-selectieve radioligand ([123I]β-CIT), die ook bindtaan dopamine transporters (bijvoorbeeld in de substantia nigra) en noradrenaline transporters(bijvoorbeeld in de locus coeruleus).85-88 Inmiddels zijn er selectieve SERT liganden zoals [11C]DASBvoor PET of [123I]ADAM voor SPECT beschikbaar. Eerdere beeldvormende onderzoeken in apenlieten zien dat de opname van [123I]β-CIT in het midbrain en diencephalon voornamelijk bindingaan SERTs vertegenwoordigt.89 Wij menen daarom dat onze bevindingen in het diencephalon enmidbrain vooral de SERT-bezettingsgraad weergeven. Tevens reduceert het gebruik van deSPECT-scans ten tijde van randomisatie (als een uitgangspunt voor de veranderingen doordosisverhoging) een mogelijke vertekening door de niet-selectieve ligand. Uiteraard zou intoekomstig onderzoek het gebruik van selectieve liganden ([11C]DASB of [123I]ADAM) overwogenmoeten worden en is het in meerdere opzichten een uitdaging om ons onderzoek met dergelijkeliganden te herhalen.

Chap

ter

Chapter 13

13

262

In het case-control onderzoek naar de beschikbaarheid van SERTs bij depressieve patiëntenversus gezonde proefpersonen bleven kleine subgroepen over bij het onderzoeken van effect-mediërende variabelen. Daarom hebben wij deze verschillen als exploratieve analysesgepresenteerd. Ten aanzien van de associatie tussen SERT bezettingsgraad en klinische responsvergeleken wij ook kleine groepen, maar vonden wij statistisch sterk significante verbanden diewij daarom als valide beschouwen. Om toevallige bevindingen uit te sluiten is replicatie van dezebevindingen wenselijk. Na de interim-analyse staakten wij onze vergelijkende studie vanwegefutiliteit, wat verdere analyse van subgroepen niet toeliet. De fMRI-studie was een voorneuroimaging vrij grote studie, waarin wij robuuste effecten vonden. Echter ook hierbij was hetaantal patiënten in de dosisverhogingsgroep beperkt voor het onderzoeken van drie-weginteracties tussen dosisverhoging, tijd en respons.

Ons neurobiologische onderzoek onderzocht geen van de specifieke secundaire effecten vanSSRIs. Er zijn vele adaptieve pre- en post-synaptische effecten van SSRIs bekend,10;83;90-97 dezewerden meestal bij knaagdieren onderzocht.98 De maximale stralingsbelasting voorwetenschappelijk onderzoek liet geen ander scanonderzoek met radioliganden toe, evenminonderzochten wij de effecten van dosisverhoging in diermodellen. Vooropgesteld moet echterblijven, dat de klinische effecten zowel in ons, als in eerder onderzoek geen effectiviteit vandosisverhoging laten zien. Aanvullend dierexperimenteel onderzoek zou daarom vooral meerfundamentele kennis over het werkingsmechanisme van SSRIs en de neurobiologie vandosisverhoging geven.

Tenslotte kunnen clinici zich bij het lezen van het onderzoek in dit proefschrift af gaan vragenwaarom initieel hogere doseringen van SSRIs effectief zijn gebleken in (sommige) ‘fixed-dose,dose-finding’ studies bij patiënten met een angststoornis,99-102 en vooral bij de obsessiefcompulsieve stoornis (OCD).103-107 Dit vormt een intrigerend vraagstuk dat buiten het kader van ditproefschrift valt, maar een vergelijkbare aanpak als in ons onderzoek rechtvaardigt.

Methodologisch sterke kenmerken van het onderzoek

Dit proefschrift is gebaseerd op een relevant klinisch vraagstuk. Alle behandelaren kennen hetprobleem dat patiënten onvoldoende verbeteren op een eerste SSRI en dosisverhoging is bijnaaltijd de eerste stap, gevolgd door een switch van antidepressivum.108-112 Dit proefschrift kandaarom relevant zijn voor toekomstige richtlijnen.

Onze systematische literatuuroverzichten werden uitgevoerd conform de strenge richtlijnenvan de Cochrane Collaboration.113 Wij voerden ‘sensitieve’ zoekacties van de literatuur, zodat dekans op het missen van waardevolle studies geminimaliseerd werd. Bovendien werden allegevonden studies kritisch beoordeeld en samengevat, waarbij wij alleen een meta-analyseuitvoerden als de klinische homogeniteit dit toeliet. In onze meta-analyse van de monoaminedepletiestudies hanteerden wij bovendien het statistische probleem van het ‘poolen’ van ‘cross-over’ studies.37

De empirische studies uit dit proefschrift werden zorgvuldig opgezet, waarbij demethodologische problemen van eerdere dosisverhogingsstudies werden voorkomen.32;114

Hiertoe randomiseerden wij patiënten pas na 6 weken behandeling, zodat de kans op een verlaterespons werd verminderd. Ook leidde ons dosisverhogingsschema niet tot een toename van deuitval (door bijwerkingen) in de dosisverhogingsgroep. Tenslotte was onze SPECT-studie die heteffect op de SERT bezettingsgraad kwantificeerde uniek en gaf dit een duidelijke rationalewaarom dosisverhoging niet werkzaam is.

Met het uitvoeren van de fMRI-studieonderzochten we de effecten op de emotieregulatie alsgevolg was van de behandeling met paroxetine. Dit was de zesde fMRI studie die debehandeleffecten van SSRIs bij depressie onderzocht. Het feit dat wij 3 scans maakten en dat veelpatiënten initieel niet voldoende verbeterden gaf ons de mogelijkheid om het verschil inhersenactivatie tussen responders en non-responders bij paroxetinebehandeling te onderzoeken.Dit is een belangrijke aanvulling op de bestaande literatuur.


13

263

Verder keken wij ook naar de perifere effecten van paroxetine, buiten het brein, maar methetzelfde mechanisme: de blokkade van de opname van serotonine in de plaatjes en de effectendaarvan op de bloedstolling. Toegenomen bloedingsneiging is een infrequente, maar potentieelgevaarlijk complicatie. Wij lieten zien, dat ook de standaarddosering van paroxetine invloed kanhebben op de haemostase, vooral in de subgroep van patiënten die drager zijn van het S’/S’polymorfisme van het 5-HTTLPR gen. Deze groep is mogelijk ook gebaat bij secundaire preventievan toekomstige bloedingsproblemen bij het gaan of blijven gebruiken van SSRIs.

Toekomstig onderzoek

De studies in dit proefschrift stimuleren om nieuwe studies en systematischeliteratuuroverzichten te starten. Een aantal onderdelen van de DELPHI-studie dienen noguitgewerkt te worden, terwijl andere vraagstukken met deze studies onbeantwoord zullenblijven. In de Engelstalige ‘general discussion’ wordt een aanzet gegeven voor toekomstigestudies, die zich met name zullen moeten richten op de vergelijking van verschillende strategiëenvoor onvoldoende respons, en het meer fundamentele begrip van de dwarsverbanden tussenverschillende neurobiologische systemen. Hierbij valt te denken aan een RCT waarin switchen na6 weken wordt vergeleken met 6 weken doorbehandelen, onderzoek bij depressieve patiëntennaar de beïnvloeding van het dopaminerge systeem door SSRIs, en de effecten van antidepressivaop de neurogenese.

Conclusie

Dit proefschrift heeft betrekking op de behandeling van depressie met antidepressiva en vooralop het vraagstuk wat te doen als patiënten niet verbeteren op een standaarddosering. Op basisvan twee systematische literatuuroverzichten concluderen wij dat er onvoldoendewetenschappelijk bewijs bestaat voor een duidelijke aanbeveling over het switchen naar eenander antidepressivum en het verhogen van de dosering. In de hierop volgende gerandomiseerdeplacebo-gecontroleerde studie (de DELPHI-studie) onderzochten wij de klinische werkzaamheidvan dosisverhogingen bij depressieve patiënten die onvoldoende waren verbeterd na 6 wekenparoxetine gebruik (20 mg/dag). Daarnaast onderzochten we in een subgroep deneurobiologische effecten van dosisverhoging in SPECT en fMRI neuroimaging studies. Met deDELPHI-studie tonen we aan dat dosisverhoging bij paroxetine geen meerwaarde heeft tenopzichte van een placebo-verhoging. Bovendien kwantificeren we dat dosisverhoging deblokkade van het farmacologische doel van paroxetine, de SERT, niet doet toenemen.

Eveneens onderzochten wij in dit proefschrift de rol van monoaminen en de SERT in deetiologie van depressie en de neurobiologische effecten van behandeling met paroxetine. Opbasis van deze studies concluderen wij dat depressie niet noodzakelijkerwijs veroorzaakt wordtdoor een slecht werkend serotonerg systeem. Depressie kan beter worden gezien als eenverstoring van de balans tussen het emotionele en cognitieve functioneren van het limbisch-subcortico-corticale netwerk. De toename van de serotonergeneurotransmissie als gevolg vanbehandeling met paroxetine (en waarschijnlijk andere SSRIs) lijkt te liggen in een beterebeteugeling van de neuronale verwerking van informatie. Patiënten die verbeteren na debehandeling laten een afname van de emotionele activiteit (bijv. in de amygdala) en een toenamevan de cognitieve controle zien. De bevindingen van dit onderzoek kunnen dienen alsuitgangspunt voor het verder ontwikkelen van biomarkers voor behandel-uitkomsten, met alsdoel het ontwikkelen van een beter begrip van het ziektebeeld depressie en effectieverebehandelingen.

Chap

ter

Chapter 13

13

264

Referenties

1. Bijl RV, Zessen Gv, Ravelli A. Psychiatrische morbiditeit onder volwassenen in Nederland: het NEMESIS-onderzoek. II.Prevalentie van psychiatrische stoornissen. NedTijdschr Geneesk. 1997; 141: 2453-2460.

2. Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med. 2006; 3: e442.3. Richtlijn Depressie. Multidisciplinaire Richtlijn Depressie. Richtlijn voor de diagnostiek en behandeling van volwassen

cliënten met een depressie. Brandt-Dominicus JC.2005. Utrecht, Trimbos-instituut, in opdracht van de LandelijkeStuurgroep Multidisciplinaire Richtlijnontwikkeling in de GGZ.

4. van Marwijk HW, Grundmeijer HG, Bijl D, van Gelderen MG, de Haan M, van Weel-Baumgarten EM et al. NHG-StandaardDepressieve stoornis (depressie) (eerste herziening). Huisarts Wet. 2003; 46: 614-623.





9. Ruhe HG, Rijswijk Ev, Verkes RJ. Biologische interventies bij depressie. In: Huyser J, Schene AH, Sabbe B, Spinhoven Ph,eds. Handboek depressieve stoornissen. Utrecht: De Tijdstroom; 2008:201-218.

10. Blier P, De Montigny C. Current advances and trends in the treatment of depression. Trends Pharmacol Sci. 1994; 15: 220-226.

11. Nemeroff CB, Owens MJ. Neuropharmacology of paroxetine. Psychopharmacol Bull. 2003; 37 Suppl 1: 8-18.12. Belmaker RH, Agam G. Major depressive disorder. N Engl J Med. 2008; 358: 55-68.13. Malhi GS, Lagopoulos J. Making sense of neuroimaging in psychiatry. Acta Psychiatr Scand. 2008; 117: 100-117.14. Mayberg HS. Modulating dysfunctional limbic-cortical circuits in depression: towards development of brain-based

algorithms for diagnosis and optimised treatment. Br Med Bull. 2003; 65: 193-207.15. Phillips ML, Drevets WC, Rauch SL, Lane R. Neurobiology of emotion perception II: Implications for major psychiatric

disorders. Biol Psychiatry. 2003; 54: 515-528.16. Hu XZ, Lipsky RH, Zhu G, Akhtar LA, Taubman J, Greenberg BD et al. Serotonin transporter promoter gain-of-function

genotypes are linked to obsessive-compulsive disorder. Am J Hum Genet. 2006; 78: 815-826.17. Serretti A, Kato M, De Ronchi D, Kinoshita T. Meta-analysis of serotonin transporter gene promoter polymorphism (5-

HTTLPR) association with selective serotonin reuptake inhibitor efficacy in depressed patients. Mol Psychiatry. 2007; 12:247-257.







24. Entsuah R, Shaffer M, Zhang J. A critical examination of the sensitivity of unidimensional subscales derived from theHamilton Depression Rating Scale to antidepressant drug effects. J Psychiatr Res. 2002; 36: 437-448.

25. O'Sullivan RL, Fava M, Agustin C, Baer L, Rosenbaum JF. Sensitivity of the six-item Hamilton Depression Rating Scale. ActaPsychiatr Scand. 1997; 95: 379-384.

26. Hooper CL, Bakish D. An examination of the sensitivity of the six-item Hamilton Rating Scale for Depression in a sample ofpatients suffering from major depressive disorder. J Psychiatry Neurosci. 2000; 25: 178-184.







13

265


33. Ruhe HG, Huyser J, Swinkels JA, Schene AH. Switching antidepressants after a first selective serotonin reuptake inhibitorin major depressive disorder: a systematic review. J Clin Psychiatry. 2006; 67: 1836-1855.





38. Ruhé HG, Ooteman W, Booij J, Michel MC, Moeton M, Baas F et al. Serotonin transporter gene promoter polymorphismsmodify the association between paroxetine serotonin transporter occupancy and clinical response in major depressivedisorder. Pharmacogenetics and Genomics. 2008; [In press].

39. Smits KM, Smits LJ, Schouten JS, Stelma FF, Nelemans P, Prins MH. Influence of SERTPR and STin2 in the serotonintransporter gene on the effect of selective serotonin reuptake inhibitors in depression: a systematic review. MolPsychiatry. 2004; 9: 433-441.

40. Ruhe HG, Booij J, Weert HCv, Reitsma JB, Fransen EJF, Michel MC et al. Evidence why dose-escalation of paroxetine inmajor depressive disorder is not effective: A 6-week, randomized-controlled trial with assessment of serotonintransporter occupancy. Neuropsychopharmacology. 2008; [In Press].



43. Ruhe HG, Booij J, Veltman DJ, Michel MC, Schene AH. Successful treatment of major depressive disorder by paroxetineattenuates amygdala activation to negative facial expressions. An fMRI study. 2008; [Submitted].





48. Abdelmalik N, Ruhé HG, Barwari K, Dool E-Jvd, Meijers JCM, Middeldorp S et al. The effect of the selective serotoninreuptake inhibitor paroxetine on platelet function is modified by a SLC6A4 serotonin transporter polymorphism. JThromb Haemost. 2008; [In Press].

49. Lenox-Smith AJ, Jiang Q. Venlafaxine extended release versus citalopram in patients with depression unresponsive to aselective serotonin reuptake inhibitor. Int Clin Psychopharmacol. 2008; 23: 113-119.

50. Thase ME, Lu Y, Joliat MJ, Detke MJ. Remission in placebocontrolled trials of duloxetine with an SSRI comparator.Presented atthe annual meeting of the American Psychiatric Association, May 2003, San Francisco, CA . 2003.

51. Malhi GS, Ng F, Berk M. Dual-dual action? Combining venlafaxine and mirtazapine in the treatment of depression. Aust N ZJ Psychiatry. 2008; 42: 346-349.

52. Perahia DG, Quail D, Desaiah D, Corruble E, Fava M. Switching to duloxetine from selective serotonin reuptake inhibitorantidepressants: a multicenter trial comparing 2 switching techniques. J Clin Psychiatry. 2008; 69: 95-105.

53. Papakostas GI, Fava M, Thase ME. Treatment of SSRI-resistant depression: a meta-analysis comparing within- versusacross-class switches. Biol Psychiatry. 2008; 63: 699-704.

54. Lenox-Smith AJ, Schaeffer P, Reynolds A, Williard L. A double blind trial of venlafaxine XR versus citalopram in patientswith severe depression. J Psychopharmacol. 2001; 15 Suppl 3: A11.

55. Thase ME, Kremer C, Rodrigues H. Mirtazapine versus sertraline after SSRI nonresponse [Abstract]. American PsychiatricAssociation.2001 annual meeting: New Orleans LA . 2001.

56. Cipriani A, Barbui C, Brambilla P, Furukawa TA, Hotopf M, Geddes JR. Are all antidepressants really the same? The case offluoxetine: a systematic review. J Clin Psychiatry. 2006; 67: 850-864.


58. Fava M, Rosenbaum JF, McGrath PJ, Stewart JW, Amsterdam JD, Quitkin et al. Lithium and tricyclic augmentation offluoxetine treatment for resistant major depression: a double-blind, controlled study. Am J Psychiatry. 1994; 151: 1372-1374.

59. Fava M, Alpert J, Nierenberg A, Lagomasino I, Sonawalla S, Tedlow J et al. Double-blind study of high-dose fluoxetineversus lithium or desipramine augmentation of fluoxetine in partial responders and nonresponders to fluoxetine. J ClinPsychopharmacol. 2002; 22: 379-387.

Chap

ter

Chapter 13

13

266

60. Papakostas GI, Charles D, Fava M. Are typical starting doses of the selective serotonin reuptake inhibitors sub-optimal? Ameta-analysis of randomized, double-blind, placebo-controlled, dose-finding studies in major depressive disorder. WorldJ Biol Psychiatry. 2007; 1-8.


62. de Abajo FJ, Montero D, Rodriguez LA, Madurga M. Antidepressants and risk of upper gastrointestinal bleeding. Basic ClinPharmacol Toxicol. 2006; 98: 304-310.

63. Wessinger S, Kaplan M, Choi L, Williams M, Lau C, Sharp L et al. Increased use of selective serotonin reuptake inhibitors inpatients admitted with gastrointestinal haemorrhage: a multicentre retrospective analysis. Aliment Pharmacol Ther.2006; 23: 937-944.



66. van Walraven C, Mamdani MM, Wells PS, Williams JI. Inhibition of serotonin reuptake by antidepressants and uppergastrointestinal bleeding in elderly patients: retrospective cohort study. BMJ. 2001; 323: 655-658.








74. Hiltunen J, Akerman KK, Kuikka JT, Bergstrom KA, Halldin C, Nikula T et al. Iodine-123 labeled nor-β-CIT as a potentialtracer for serotonin transporter imaging in the human brain with single-photon emission tomography. Eur J Nucl Med.1998; 25: 19-23.








82. Spoont MR. Modulatory role of serotonin in neural information processing: implications for human psychopathology.Psychol Bull. 1992; 112: 330-350.






13

267


88. Neumeyer JL, Tamagnan G, Wang S, Gao Y, Milius RA, Kula NS et al. N-substituted analogs of 2 beta-carbomethoxy-3 beta-(4'-iodophenyl)tropane (β-CIT) with selective affinity to dopamine or serotonin transporters in rat forebrain. J Med Chem.1996; 39: 543-548.



91. Benmansour S, Owens WA, Cecchi M, Morilak DA, Frazer A. Serotonin clearance in vivo is altered to a greater extent byantidepressant-induced downregulation of the serotonin transporter than by acute blockade of this transporter. JNeurosci. 2002; 22: 6766-6772.

92. Blendy JA. The role of CREB in depression and antidepressant treatment. Biol Psychiatry. 2006; 59: 1144-1150.93. Castren E. Is mood chemistry? Nat Rev Neurosci. 2005; 6: 241-246.94. Davidson C, Stamford JA. Effect of chronic paroxetine treatment on 5-HT1B and 5-HT1D autoreceptors in rat dorsal raphe

nucleus. Neurochem Int. 2000; 36: 91-96.95. Djavadian RL. Serotonin and neurogenesis in the hippocampal dentate gyrus of adult mammals. Acta Neurobiol Exp

(Wars). 2004; 64: 189-200.96. Kim SW, Park SY, Hwang O. Up-regulation of tryptophan hydroxylase expression and serotonin synthesis by sertraline.

Mol Pharmacol. 2002; 61: 778-785.97. Malberg JE, Eisch AJ, Nestler EJ, Duman RS. Chronic antidepressant treatment increases neurogenesis in adult rat

hippocampus. J Neurosci. 2000; 20: 9104-9110.98. Meyer JH, Kapur S, Eisfeld B, Brown GM, Houle S, DaSilva J et al. The effect of paroxetine on 5-HT(2A) receptors in

depression: an [18F]setoperone PET imaging study. Am J Psychiatry. 2001; 158: 78-85.99. Mavissakalian MR, Ryan MT. Rational treatment of panic disorder with antidepressants. Ann Clin Psychiatry. 1998; 10: 185-

195.100. Rickels K, Zaninelli R, McCafferty J, Bellew K, Iyengar M, Sheehan D. Paroxetine Treatment of Generalized Anxiety

Disorder: A Double-Blind, Placebo-Controlled Study. Am J Psychiatry. 2003; 160: 749-756.101. Practice guideline for the treatment of patients with panic disorder. Work Group on Panic Disorder. American Psychiatric

Association. Am J Psychiatry. 1998; 155: 1-34.102. Richtlijn Angststoornissen. Multidisciplinaire richtlijn Angststoornissen. Richtlijn voor de diagnostiek en behandeling van

volwassen cliënten met een angststoornis. Velde V vd. 2003. Utrecht, Trimbos-instituut, in opdracht van de LandelijkeStuurgroep Multidisciplinaire Richtlijnontwikkeling in de GGZ.

103. Denys D. Pharmacotherapy of obsessive-compulsive disorder and obsessive-compulsive spectrum disorders. PsychiatrClin North Am. 2006; 29: 553-84.

104. Figee M, Denys D. New pharmacotherapeutical approaches to obsessive compulsive disorder. CNS Spectr. 2008; [Inpress].

105. Ninan PT, Koran LM, Kiev A, Davidson JR, Rasmussen SA, Zajecka JM et al. High-dose sertraline strategy fornonresponders to acute treatment for obsessive-compulsive disorder: a multicenter double-blind trial. J Clin Psychiatry.2006; 67: 15-22.

106. Dougherty DD, Rauch SL, Jenike MA. Pharmacotherapy for obsessive-compulsive disorder. J Clin Psychol. 2004; 60: 1195-1202.

107. American Psychiatric Association. Practice guideline for the treatment of patients with obsessive-compulsive disorder.Available online at http//www.psych.org/psych_pract/treatg/pg/prac_ guide.cfm ed. Arlington, VA: American PsychiatricAssociation; 2007:1-96.

108. Wittkampf KA, Ruhé HG, Huyser J, Schene AH. De Creatieve Psychiater? Een onderzoek onder Nederlandse psychiatersnaar de behandeling van depressieve stoornissen bij het falen van de eerste SSRI.2001. Amsterdam, ZorgprogrammaStemmingsstoornissen MFO Psychiatrie AMC/ De Meren.





113. Cochrane Collaboration. Preparing, maintaining and promoting the accessibility of systematic reviews of the effects ofhealth care interventions. Internet: http://www.cochrane.org/. 2001. Cochrane Collaboration.


ABBREVIATIONS, COLOR FIGURES, PUBLICATIONS, ACKNOWLEDGEMENT AND CURRICULUM VITAE

PART

Abbreviations, color figures, publications, acknowledgement and curriculum vitae

ABBREVIATIONS

List of abbreviations

273


[123I]ß-CIT [123I]methyl 3ß-(4-iodophenyl) tropane-2 ß –carboxylate 5-HT serotonin 5-HTTLPR Serotonin transporter promoter polymorphism95% CI 95% confidence interval Β-TG β-thromboglobulinAHCPR Agency of healthcare policy and researchAIC Akaike’s information criterionAMC Academic Medical Center AMPT Alpha-methyl-para-tyrosine ANOVA Analysis of variance model ATD Acute tryptophan depletion APTD Acute phenylalanine/tyrosine depletion AUC Area under the curve BDNF Brain derived neurotrophic factor BDI Beck depression inventoryBOLD Blood oxygen level dependentBPND Binding potential (non-displacable) cAMP Cyclic adenosine monophosphateCCDAN Cochrane collaboration depression anxiety and neurosis groupCER CerebellumCGI-I/S Clinical global impression scale (improvement/severity) COMT Catechol-O-methyltransferase CI95 95% Confidence interval CREB cAMP response element binding protein CYP P450 Cytochrome P450DA Dopamine DAT Dopamine transporter DE Dose-escalationDELPHI Dose-escalation legitimate? Pharmacology and imaging studies in depressionDIENC DiencephalonDLPFC Dorsolateral prefrontal cortexDMPFC Dorsomedial prefrontal cortexDNA Deoxyribonucleic acidDLPFC Dorsolateral prefrontal cortex DSM-IV Diagnostic and statistical manual, 4th edition E-S Effect sizeEEG Electric encephalogramEMEA European agency for the evaluation of medicinal productsEPI Echoplanar imageFDA Food and drug administrationfMRI functional Magnetic Resonance Imaging HC Healthy controlsHDRS Hamilton depression rating scale IDS-C/SR Inventory for depressive symptoms – clinician / self-rated IQR Inter Quartile RangeIRT Item response theoryITT Intention to treatIU International unitsLOCF Last observation was carried forward LoE Level of evidenceMAACL Multiple affect adjective checklistMADRS Montgomery Åsberg depression rating scale MAO(-I) Monoamine oxidase (inhibitor)MHRA (British) Medicines and healthcare products regulatory agencyMID Midbrain MDD Major depressive disorder


274

MDMA 3,4-methylenedioxy methamphe tamine (XTC)MNI Montreal Neurological Institute MOS-SF36 Medical outcome study short form (measuring health related quality of life)MRI, sMRI/fMRI Magnetic resonance imaging, structural/functional N/A Not appropriateNA/NE Noradrenaline (=norepinephrine)NAT/NET Nordrenaline transporter NERI Norepineprine reuptake inhbitorsNNT/NNH Number needed to treat/ harm NSAID Non-steroidal anti-inflammatory drugOCC Occupancy OCD Obsessive compulsive disorderOFC Orbitofrontal cortex PCPA Para-chlorophenylalaninePCR Polymerase chain reactionPET Positron emission tomography PFA Platelet function analyzerPF4 Platelet factor 4PFC Prefrontal cortexPOMS Profile of Mood StatesPRP Platelet rich plasmaPSC Paroxetine serum concentrationsPUFAs Poly unsaturated fatty acidsQIDS-SR Quick-inventory for depressive symptoms RCT Randomized controlled trial RIMA Reversible inhibitor of monoamine oxidase A ROC Receiver operator characteristic (curve)RoI Regions of interestRR Relative risk RT Reaction timeSCID Structured clinical interview for DSM-IVSCL Symptoms check list (90 items)SCN Suprachiasmatic nucleusSD Standard deviation SE(M) Standard error (of the mean)SERT Serotonin transporter SLC6A4 Serotonin transporter geneSNP Single nucleotide polymorphismSNRI Serotonin and norepinephrine reuptake inhibitor SPECT Single photon emission computed tomography SPM5 Statistical Parametric Mapping version 5SPSP Short psychodynamic supportive psychotherapySSRI Selective serotonin reuptake inhibitor STAR*D Sequenced treatment alternatives to relieve depressionTCA Tricyclic antidepressant TE Echo time TR Repetition time TRD Treatment-resistant depressionVA(M)S Visual analogue (mood)scalesVLPFC Ventrolateral prefrontal cortex VMPFC Ventromedial prefrontal cortex

COLOR FIGURES

Color Figures

277

Figure 1.2. Transporters, receptors and second messenger systems involved in the effects of antidepressants.

Color Figures

278

Figure adapted from Belmaker and Agam.79 The left half of the presynaptic neuron representsa serotonergic neuron, the right half a norepinephrinergic neuron.

In the presynaptic neuron, serotonin is synthesized from tryptophan by tryptophanhydroxylase and stored in vesicles. Likewise, norepinephrine is synthesized from tyrosine bytyrosine hydroxylase. These vesicles merge with the cell membrane when the neuron isdepolarized, thereby releasing their contents into the synaptic cleft.

After release, serotonin and norepinephrine are transported back into the presynaptic neuronby serotonine and norerepinephrine transporters. Furthermore, serotonin and norepinephrineare catabolized by the monoamine-oxidase A (MAO-A) enzyme.

In the synaptic cleft, serotonin and norepinephrine affect both the pre-and post-synapticneuron. The pre-synaptic 5-HT1A and 5-HT1B auto-receptors decrease serotonin release byinhibitory feedback; the α2-adrenergic receptor does the same for the release of norepinephrine.

Post-synaptically, serotonin and norepinephrine bind to G-protein-coupled monoaminereceptors (MARs): the cyclicAMP (cAMP)-coupled receptor, which activates protein kinase A(PKA), and the Phosphatidylinositol (PI)-coupled receptor, which activates phospholipase C (PLC)which thereafter form inositol triphosphate (IP3) and diacylglycerol (DAG). IP3 and DAG activateprotein kinase C (PKC). Both PKA and PKC finally activate cAMP responsive element binding(CREB) protein, which stimulates DNA transcription. For example, this might result in theproduction of brain derived neurotrophic factor (BDNF).

Color Figures

279

DELPHI-SPECT and DELPHI-fMRI represent sub-studies nested within the total study. For these2 sub-studies drug-free patients were recruited and treated in the Academic Medical Center.

Color Figures

280

Figure 2.1. Design of the DELPHI-study.

��

��

��

�� !" ��

��

� �#�$��% #& #� ' &�� '��(")

��

��

�� *��(�+�,��(� ��"��"��(�

�� -.�� -.�� !" -.��

Color Figures

281

Figure 2.3. Regions of Interest (RoI) for Midbrain, Cerebellum and Diencephalon.

Color Figures

282

Example of SPECT images after 3D reconstruction. Templates with fixed RoIs are shown ingreen.

A. Midbrain (circle) and cerebellum. B. Striatum (for demarcation midbrain-diencephalon) anddiencephalon (circle). RoIs were positioned by hand and situated based on anatomy andmaximum concentration of activity/ml in the RoI.

Color Figures

283

Negative faces contrast. Week 6 and 12 scans and response status were combined in a fullfactorial model.

A. SPM-results showing design matrix, contrast and activation in right amygdala (MNI 18,-2,-18).

B. Activations of left (MNI -22,2,-18; blue) and right (MNI 18,-2,-18; red) amygdala stratified fornon-response and response at 6 and 12 weeks. Significant difference between non-respondersand responders (F1,32= 4.591; p<0.0001); no significant lateralization (F1,32= 0.1517; p= 0.70).

C. Plot of contrast estimate by HDRS17-score (centralized to mean). Significant positivecorrelations for left (0.33 ±0.12 [SE]; F1,31= 7.076; p= 0.0123; r2= 0.19; shown in blue) and right (0.38±0.14 [SE]; F1,31= 6.929; p= 0.0131; r2= 0.18; shown in red) amygdala. Regression lines for left andright amygdala are not significantly different (F1,62= 0.0683; p= 0.79).

Color Figures

284

Figure 10.1. Amygdala response relative to clinical response (HDRS17 ≥50% decrease) after 6 and 12 weeks of treatment with paroxetine.

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

PUBLICATIONS

Publications

287

Papers and chapters

Ruhe HG, Booij J, Weert HCv, Reitsma JB, Fransen EJF, Michel MC, Schene AH. Evidence why dose-escalation of paroxetine in major depressive disorder is not effective: A 6-week, randomized-controlled trial with assessment of serotonin transporter occupancy. Neuropsychopharmacology.2008; [In Press].

Ruhe HG, Booij J, Reitsma JB, Schene AH. Serotonin transporter binding with [123I]ß-CIT SPECT inmajor depressive disorder versus controls: effect of season and gender. 2008; [Submitted].

Abdelmalik N, Ruhé HG, Barwari K, Dool E-Jvd, Meijers JCM, Middeldorp S, Büller HR, Schene AH,Kamphuisen PW. The effect of the selective serotonin reuptake inhibitor paroxetine on plateletfunction is modified by a SLC6A4 serotonin transporter polymorphism. J Thromb Haemost. 2008;[In Press].

Visser I, Wekking EM, de Boer AGEM, van Hout MSE, van Dorsselaar S, Ruhé HG, Huyser J, van derLaan G, van Dijk FJH, Schene AH. Mental disorders in chronic solvent induced encephalopathy:prevalence and symptom profiles. 2008; [Submitted].

Baas K, Wittkampf KA, Weert HC van, Lucassen P, Huyser J, Hoogen H van den, Lisdonk, E van de,Bindels P, Bockting CL, Ruhé HG, Schene AH. Screening on depression in high risk groups ingeneral practice is not useful: results from a prospective cohort study in general practice. Br JPsychiatry. 2008; [In press].

Ruhe HG, Rijswijk Ev, Verkes RJ. Biologische interventies bij depressie. In: Huyser J, Schene AH,Sabbe B, Spinhoven Ph, eds. Handboek depressieve stoornissen. Utrecht: De Tijdstroom; 2008:201-218.

Ruhe HG, Booij J, Veltman DJ, Michel MC, Schene AH. Successful treatment of major depressivedisorder by paroxetine attenuates amygdala activation to negative facial expressions. An fMRIstudy. 2008; [Submitted].

Ruhé HG, Ooteman W, Booij J, Michel MC, Moeton M, Baas F, Schene AH. Serotonin transportergene promoter polymorphisms modify the association between paroxetine serotonin transporteroccupancy and clinical response in major depressive disorder. Pharmacogenetics and Genomics.2008; [In press].

Ruhe HG, Mason NS, Schene AH. Mood is indirectly related to serotonin, norepinephrine anddopamine levels in humans: a meta-analysis of monoamine depletion studies. Mol Psychiatry.2007; 12: 331-359.

Baas KD, Ruhé HG. Gedragsactivatie effectief in de behandeling van depressie. Tijdschr vPsychiatrie.2007; 49: 411.

Ruhé HG. Polymorfisme van het serotoninetransporteiwit niet gerelateerd aan respons, maar welaan bijwerkingen bij citalopramgebruik. Ned Tijdschr Geneeskd. 2007;151: 2641.

Ruhe HG, Huyser J, Swinkels JA, Schene AH. Dose escalation for insufficient response to standard-dose selective serotonin reuptake inhibitors in major depressive disorder: Systematic review. Br JPsychiatry. 2006; 189: 309-316.

Ruhe HG, Huyser J, Swinkels JA, Schene AH. Switching antidepressants after a first selectiveserotonin reuptake inhibitor in major depressive disorder: a systematic review. J Clin Psychiatry.2006; 67: 1836-1855.

Publications

288

Ruhé HG. Suïciderisico door antidepressiva opnieuw onder de loep genomen. Tijdschr vPsychiatrie.2006;48: 494-495.

Ruhe HG, Dekker JJ, Peen J, Holman R, de Jonghe F. Clinical use of the Hamilton DepressionRating Scale: is increased efficiency possible? A post hoc comparison of Hamilton DepressionRating Scale, Maier and Bech subscales, Clinical Global Impression, and Symptom Checklist-90scores. Compr Psychiatry. 2005; 46: 417-427.

Ruhe HG, Swinkels JA, Schene AH. Vervolgstrategieën bij onvoldoende respons op een eersteSSRI. Pharmaceutisch Weekblad. 2005; 44: 1382-1385.

Schene AH, Huyser J, Ruhe HG. Validiteit van de diagnose depressie. In: Koerselman F, VergouwenAC, eds. De gecompliceerde kliniek 2005: over de geldigheid van diagnoses. Amsterdam: Benecke;2005.

Ruhé HG. Farmacotherapie bij unipolaire depressie. Bijblijven. 2003;19:12-25.

Ruhe HG. Review: Low dose tricyclic antidepressants may be effective for adults with acutedepressive disorder. Evid Based Ment Health. 2003;6:46.

Ruhé HG, Huyser J, Swinkels JA, Schene AH. Wat te doen als de eerste pil niet werkt? EenEvidence-Based richtlijn voor de biologische behandeling van depressie bij onvoldoende responsop een standaarddosering van een eerste SSRI. 2002. Amsterdam, ZorgprogrammaStemmingsstoornissen MFO Psychiatrie AMC/ De Meren.

Ruhé HG, Becker HE, Jessurun P, Marees CH, Heeringa M, Vermeulen HD. Agranulocytosis andgranulocytopenia associated with quetiapine. Acta Psychiatr Scand. 2001;104:311-3.

Swinkels JA, Schene AH, Ruhé HG. Evidence Based Mental Health: inhoud en methode. In: ScheneAH, Boer F, Heeren TJ, Henselmans HWJ, Trijsburg RW, Vandereycken W, Velden K van der, eds.Jaarboek voor psychiatrie en psychotherapie 2001-2002. Houten/Diegem: Bohn Stafleu van Loghum.2001: 298-313.

de Vegt F, Dekker JM, Ruhe HG, Stehouwer CD, Nijpels G, Bouter LM, Heine RJ. Hyperglycaemia isassociated with all-cause and cardiovascular mortality in the Hoorn population: the Hoorn Study.Diabetologia. 1999;42:926-31.

Jager A, Kostense PJ, Ruhe HG, Heine RJ, Nijpels G, Dekker JM, Bouter LM, Stehouwer CD.Microalbuminuria and peripheral arterial disease are independent predictors of cardiovascularand all-cause mortality, especially among hypertensive subjects: five-year follow-up of the HoornStudy. Arterioscler Thromb Vasc Biol. 1999;19:617-24.

Dekker J, Ruhé E, Nijpels G, Heine R. Diabetes en cardiovasculaire complicaties. In: Broese vanGroenau M, Deeg DJH, Knipscheer CPM, Ligthart GJ, eds. VU-visies op verandering. Amsterdam:Thela-Thesis 1998: 21-24.

Publications

289

Published abstracts

Ruhé HG, Booij J, v Weert HC, Reitsma J, Michel MC, Schene AH. Randomized dose escalation ofparoxetine in depressed patients visualized by [123I]β-CIT SPECT. [Abstract P1.22]. J AffectDisord.2008;107 Suppl 1:S77.

RuhéHG, Booij J, ScheneAH. Season and gender modify in-vivo serotonin transporter binding indepressed patients and healthy controls. [Abstract P1.23]. J Affect Disord.2008;107 Suppl 1:S78.

Ruhé HG, Booij J, VeltmanDJ, Reitsma J, MichelMC, ScheneAH. The effect of paroxetine onamygdala reactivity after emotional faces measured with fMRI. [Abstract P1.24]. J AffectDisord.2008;107 Suppl 1:S78.

Schene AH, Ruhé HG, Huyser J, Speckens A, Nolen WA. Wat leert STAR*D ons over de behandelingvan depressie, nu en in de toekomst? [Abstract S-39].Tijdschr v Psychiatrie.2008;50 Suppl 1: S106-109.

Ruhé HG, Ooteman W, Visser I. Etiologie en neuroimaging van depressie. [Abstract PO-63].Tijdschrv Psychiatrie.2008;50 Suppl 1: S185-188.

Ruhé HG, Booij J, v Weert HC, Michel MC, Schene AH. Randomized dose escalation of paroxetinein depressed patients visualized by [123I]β-CIT SPECT [Abstract 344]. Eur J Nucl Med Mol Imag.2007;34 Suppl 2:S195.

Ruhé HG, Booij J, Schene AH. Effect of season and gender on in-vivo serotonin transporter bindingin depressed patients versus healthy controls. A [123I]β-CIT SPECT study [Abstract 345].Eur J NuclMed Mol Imag.2007;34 Suppl 2:S195.

Ruhé HG, Huyser J, Swinkels JA, Schene AH. Failure to respond to the first SSRI in majordepressive disorder: an evidence-based guideline for timing of treatment changes, increasing thedose and switching. [Abstract]. J Affect Disord.2004;78 Suppl 1:S89.

Publications

290

DANKWOORD

Dankwoord

293

Dankwoord

Het beste bewaar je tot het laatste. In de afgelopen 5 jaar heb ik met veel plezier met vele mensensamengewerkt zonder wie dit proefschrift onmogelijk zou zijn geweest. In mijn gedachten heb ikdit dankwoord al vele malen geschreven.

In de eerste plaats is deze studie alleen mogelijk geworden dankzij heel veel patiënten, dieondanks hun depressie bereid waren verwezen te worden naar het AMC. Ik ben u allenbuitengewoon erkentelijk. Zonder deze deelname was het onderzoek nooit geslaagd. Allepatiënten die naast de ‘gewone’ dosisverhogingsstudie ook nog meededen aan het SPECT en/offMRI onderzoek wil ik dubbel bedanken. U kwam op moeilijke tijden, soms meerdere keren perweek terug voor deze scans, soms ook als de somberheid nog niet zo was verbeterd.

Ook de gezonde proefpersonen die bereid waren SPECT en fMRI scans te ondergaan wil ikhartelijk bedanken. Ik hoop uw interesse in wetenschappelijk onderzoek te hebben gestimuleerdmet mijn uitleg over wat we onderzochten. Dankzij uw bereidheid is het mogelijk resultaten tevergelijken met een niet-zieke toestand.

Graag wil ik ook alle huisartsen, psychiaters en sociaal psychiatrisch verpleegkundigenbedanken die in de periode dat wij op zoek waren naar deelnemers voor ons onderzoek trouwbleven verwijzen. Wellicht dat wij in de toekomst opnieuw met jullie samen kunnen werken.

Prof.dr. Aart Schene, mijn promotor, jou ben ik veel dank verschuldigd. Om te beginnenrealiseerde jij mijn aanstelling als kersverse psychiater bij het zorgprogrammaStemmingsstoornissen, waarmee jij direct jouw vertrouwen aangaf. Vervolgens je keuze omondanks onvolledige financiering – tegen jouw gewoonte in – toch een aanvang te maken met deprojecten waar ik mijn zinnen op had gezet. Jouw no-nonsense houding bij het bespreken vanzakelijke, technische, en organisatorische aspecten van onderzoek en ondertussen ookmanagement zijn prijzenswaardig en leren mij veel. Bovendien kan ik het erg goed met je vinden.Het veel samenwerken aan projecten, op de gekste tijden bereikbaar zijn, snel reageren opgeschriften en jouw positieve benadering zijn kenmerkend en geven mij inspiratie om hier noglang mee door te gaan.

Dr. Jan Booij, mijn co-promotor, dank ook voor jouw denkwerk, enthousiasme engelijkwaardige begeleiding tijdens het uitvoeren van dit onderzoek. Jij regelde de infrastructuur inde samenwerking met de XTC-studie, de scan-faciliteiten, de uitleg over de analyses van de scansen mijn verdere scholing op het gebied van de nucleaire geneeskunde. Met jou is het altijd prettigsamenwerken geweest, laten we dit vooral voortzetten.

De leden van de promotiecommissie hebben als eersten het product van mijn werk mogenlezen en beoordelen: Prof.dr. Damiaan Denys, dank dat je in de commissie wilde plaatsnemen, endank voor de interessante wetenschappelijke discussies die we sinds jouw komst alskernhoogleraar al hebben gevoerd. Ik verheug me op de toekomst, waarin onze interessessamenkomen in nieuwe onderzoeksprojecten, hetgeen ongetwijfeld tot meer discours zal leiden.

Prof.dr. Rudi Dierckx, ook dank aan jou voor het beoordelen van het manuscript. Hopelijk gaatons andere project door, zodat we in de toekomst aan gezamenlijke studies kunnen werken.

Prof.dr. Berthe van Eck-Smit, dank voor de faciliteiten die door de afdeling Nucleairegeneeskunde werden geschapen om dit onderzoek te kunnen doen. Dank voor het beoordelenvan het manuscript, maar ook mijn dank voor de interesse in mijn onderzoek, als wij elkaar op F2ontmoetten.

Prof.dr. C.G. Kruse, mijn hartelijke dank voor uw beoordeling van mijn manuscript. Ik waardeerhet enorm dat u ondanks uw drukke reisschema aanwezig kan zijn bij de verdediging.

Prof.dr. Willem Nolen, biologische psychiater van het eerste uur en daarbij ervaren clinicus ophet gebied van de stemmingsstoornissen. Hartelijk dank voor jouw interesse in mijn activiteiten,waar je al sinds 2001 bij betrokken bent. Dat jij voor mijn commissie zou worden gevraagd wasvoor mij vanzelfsprekend, ik vind het een eer dat je mijn manuscript zo spoedig van een oordeelvoorzag. Het proefschrift is af, de studies zijn (bijna) gepubliceerd, en we moeten voort, op zoeknaar nieuwe strategieën in de behandeling van depressie.

Dankwoord

294

Dr. Robbert-Jan Verkes, ook jou wil ik hartelijk danken voor je bereidheid vragen te stellen bijmijn verdediging. Dank voor de plezierige samenwerking tijdens het schrijven van het hoofdstukin het Handboek Depressie, dat smaakte naar meer.

Jochanan Huyser, tot drie keer toe was jij op de jou toevertrouwde wijze - strategisch aan dezijlijn - een beïnvloeder van formaat. Je nam mij eind 1996 aan als AGNIO. Vervolgens bracht jij mede beginselen van depressiebehandeling bij. Je kwam, kort na een jaar zoeken naar een‘kristallisatiepunt’, met het aanbod om het richtlijnproject te gaan doen. Ik heb genoten van jestimulerende enthousiasme tijdens dit project, en denk met bewondering terug aan jouw kundein het voorzitten van de commissievergaderingen. Toen we collegae psychiaters waren, heb jijmijn keuze om de neuroimaging van de DELPHI-studie te gaan doen met een kleine suggestierichting gegeven. Dat ik, nu jij het AMC hebt verlaten, de dagelijkse leiding van ‘jouw’behandelcentrum Stemmingsstoornissen over heb mogen nemen beschouw ik als een grote eer.

Voor alle praktische en logistieke support was een telefoontje naar Marianne Haages meestalvoldoende. Marianne, fantastisch hoe jij al sinds het richtlijnproject enthousiast hebtmeegewerkt, en daarvoor ben ik je duizendmaal dank verschuldigd. Met jouw invulling krijgt deprofielschets voor een ervaren secretaris/secretaresse een onbereikbare dimensie.

Mijn collegae van de afdeling stemmingsstoornissen, Guido Nabarro, Prof.dr. Jan Swinkels,Mirjam van Zon, alle verpleegkundigen, assistenten in opleiding, ergotherapeuten,maatschappelijk werkers, psychomotore therapeuten en psychologen, jullie hebben allemaalbijgedragen aan het tot stand komen van dit boek. Ook de ex-collegae Nicole Eussen, ArjenRonner, Nicole de Boer en Pierre Delfgauw horen in dit rijtje thuis. Dank voor jullie interesse,waarneming, vragen, samenwerking op het gebied van onderzoek en ingewikkelde casuïstiek, hetverdragen van mijn afwezigheid, maar vooral jullie degelijke vakkennis en enthousiasme in debehandeling van onze depressieve patiënten. Dagelijks laten jullie zien waar het bij behandelenvan depressieve, zieke mensen echt om gaat: betrokkenheid met gepaste distantie, veelvuldigproberen en herhalen, gedoseerde activatie, soms toegeeflijk zijn maar zonder hierbij de eigengrenzen te overschrijden, gecombineerd met een professionele aandacht voor stagnatie enalarmsignalen. Ik vind het, nogmaals, een eer om met jullie te werken en hoop daar mee door tekunnen gaan.

Een speciaal woord van dank gaat uit naar Mascha ten Doesschate en Elizabeth Miedema, dieveel bijdroegen aan de opzet en de dataverzameling voor de DELPHI-studie. Mascha, veel succesmet de afronding van jouw promotie. Elizabeth, gezamenlijk werkten wij anderhalf jaar aan deinclusie. Dank voor je toewijding en bereidheid om logistieke zaken omtrent de patiëntenzorg vanmij over te nemen. Andere leden van onze onderzoeksgroep: Karin, Kim, Ieke, Anja, Wendy,Hanneke, Maarten, Monique, Judith en sinds kort Maaike en Mathilda, dank voor de prettigesamenwerking, jullie bijdragen aan het onderzoekersoverleg, de gezamenlijke projecten die zijnafgerond, nog lopen of gaan komen. Ik verheug me nu al op een volgend bezoek aan ISAD. Nada,Michiel, Koen en Max, dank voor jullie input bij mijn projecten. Maaike, sinds kort toegetreden alsAGIKO, jij startte jouw promotie in een tijd dat ik nog druk was/ben met de afronding van ditproefschrift. Ik bewonder je degelijkheid waarmee je complexe zaken uitdiept en kijk uit naar deverdere ontwikkeling van onze biomarker plannen.

Alle collegae van andere afdelingen ben ik dankbaar voor de prettige contacten, expertise entoegankelijkheid. Dr. Henk van Weert, met jou heb ik veel werk verzet om ons verwijzersnetwerkop te bouwen. Dank voor je betrokkenheid, je inzet en je hulp om DELPHI tot een succes temaken. Prof.dr. Dick Veltman, jij leerde mij in meerdere middagen één-op-één onderwijs hetwerken met SPM5. Het lijkt me goed hiermee door te gaan en vooral samen nieuwe fMRI studieste gaan doen. Prof.dr. Martin Michel, jij besteedde als co-auteur bijzonder veel aandacht aan mijnmanuscripten. Dank voor je kritische commentaren, het wordt er inderdaad alleen maar betervan. Dr. Hans Reitsma, “gelijk hebben en krijgen is niet altijd hetzelfde”, dank voor al jouw hulp bijook voor een gewone epidemioloog ingewikkelde statistiek; ik heb veel van je geleerd. Prof.dr.Koos Zwinderman, dat geldt ook voor jou. Dr. Maartje de Win, aanvankelijk kon ik voor de SPECT-scans meeliften op jouw XTC-studie. Dank hiervoor en voor jouw ordentelijke planning waar altijdop te bouwen was. De medewerkers van de afdeling Nucleaire geneeskunde, dank voor hetverdragen van al mijn ‘geregel’ om de scans op de juiste momenten te kunnen maken, en dank

Dankwoord

295

voor jullie hulp bij het scannen in de avonduren. Dr. Rob Scholten, ik ken je vanaf mijn eersteserieuze onderzoeksinspanningen op het EMGO. Ondanks dat ik door jou gedegen aan de tandwerd gevoeld tijdens mijn epidemiolgie-examen moet ik je nog steeds lastig vallen voor meta-analytische problemen, waarvan jij zelfs verzuchtte dat ze zo ingewikkeld zijn. Dank je voor jeprettige, relativerende en snelle hulp. Ook Dr. Astrid Goossens van de KEB wil ik bedanken voorhaar enthousiasme tijdens het richtlijnproject. Dr. Aart Nederveen, dank voor je rustige, gedegenhulp bij het instellen van de MRI-scanner en alle antwoorden op andere vragen die ik je hebgesteld. Raschel en Sandra, dank voor de eerste hulp bij MRI-problemen. Tessa, jij was nooit teberoerd om extra te komen scannen als MRI-piloot. Michiel de Ruiter, veel dank voor het delenvan jouw kennis van SPM en fMRI, ook als ik je op vakantie of in de supermarkt belde. Zsuzsika,dank voor je expertise bij het voorbewerken van de ToL, en succes met je eigen studies. EricFranssen, dank voor de snelle bepalingen van de paroxetine-spiegels en je commentaren opeerdere versies van het artikel. Remco Knol, dank voor je prachtige illustratie (figuur 1.2) en veelsucces de dag na mij. Peter Remijnse, we hadden een goede inclusie luchtbrug, moeten we vakerdoen. Ik ben zeer benieuwd naar jouw vorderingen. Prof.dr. Frank Baas, Marja Jakobs en MartinaMoeton, dank voor jullie hulp bij de genetische bepalingen, en de interpretatie ervan. Nadia,Pieter-Willem, Saskia, Joost, Eric-Jan, Kurdo en Prof.dr. Harry Büller, wat een mooi project is onzestollings-studie geworden, dank voor de samenwerking. Bep de Lange, sinds anderhalf jaar mijnbuurvrouw, dank voor je zorg, je interesse en natuurlijk de schutbladen op momenten dat hetecht nodig is. Prof.dr. Frans de Jonghe, Prof.dr. Jack Dekker en Jaap Peen van de MentrumDepressie groep, dank voor het gezamenlijk werken aan het HDRS-artikel, meten is weten, enefficiënt meten is sneller weten.

Eugenie Oosterhuis, u verdient een aparte vermelding in dit dankwoord. Het bereiken van eenplateau na een stevige beklimming wil niet zeggen dat je er niet alsnog af kan vallen; wel weet jebeter waar de gevaarlijke randen zitten. Dank dus opnieuw.

Leden van de intervisiegroep, Ethy, Agnes, Joop, Saskia en Hiske, gelukkig dat ik van jullie‘mag blijven’ als ik gepromoveerd ben. Dank voor jullie begripvolle, constructieve doch stevigeprobleemanalyses, waaruit nieuwe oplossingen opborrelden. Gelukkig is het ook erg gezellig ombij elkaar te komen.

De afgelopen jaren heeft het proefschrift een steeds groter deel van mijn tijd in beslaggenomen. Vele vrienden zal dat opgevallen zijn. Wouter en Jet, Mark, Cees en Dorte, Henk enMary, Mary en Bob, Miriam, Tamar en Ilan, Laetitia en Erik, Lilian en Frank, Ellen en Edwin, Stephanen Ellen, Natascha en Thijs, Michiel en Mathieu, Jef en Sylvia: het is af. Laten we dit vieren met eengroot feest. Graag nodig ik jullie deze winter uit om lekker bij de open haard te zitten, te eten ente kletsen.

Ook mijn (schoon-)familie wil ik bedanken voor de support van de laatste maanden ennatuurlijk de jaren ervoor. Pap en mam, van jullie heb ik heel veel meegekregen, maar absoluutook een zeker fanatisme, dat onontbeerlijk is voor een promotie. Dank daarvoor, en voor deinvesteringen die jullie in de afgelopen 40 jaar op alle terreinen in mij gedaan hebben. Dat de wegeigenlijk belangrijker is dan het doel, staat soms in schril contrast met ons fanatisme; laten we dusnog maar lang onderweg zijn, opdat we dat beter leren. Toine en Quinten, wat bof ik met jullie alsbroers, en dat gevoel wordt alleen maar sterker. Jullie hebben allebei een geweldige vrouw inonze familie geïntroduceerd. Toine en Andrea, geweldig dat jullie voor deze gelegenheid hierheenwillen komen. Quinten en Miriam, bij jullie is het altijd gezellig. Laten we elkaar vaak zien, ondanksalle opleiding, promotie en andere drukte. Miriam, heel veel succes ook met jouw laatste loodjes.Toon en Johanna, dank voor alle support bij dit proefschrift, praktisch, logistiek en geestelijk.Jullie voelden op de een of andere manier vaak goed aan wanneer een handje hulp gewenst was.Het is heerlijk dat ik jullie met Katelijne heb leren kennen.

Jan Willem, beste vriend sinds de Bhagwan-discotheek in 1986. Wat hebben wij al veel metelkaar meegemaakt en wat ben ik blij dat jij bij deze gelegenheid mijn paranimf bent. Quinten, jeblijft mijn broertje, ook jij bent mijn ideale paranimf. Ik wil jullie beiden speciaal bedanken voor hetovernemen van een aantal belangrijke klussen rondom de promotie, waarbij jullie ook nog evende redactie van de laatste hoofdstukken deden. Dat heeft mij veel ruimte gegeven om me op delaatste loodjes te concentreren.

Dankwoord

296

Het beste bewaar je tot het laatste. Laura en Merle, wat een feest dat jullie bij ons zijngeboren; we zouden nooit meer zonder jullie kunnen. Ik weet zeker dat ik het meest trots op julliezal zijn, als jullie op woensdag op de eerste rij zitten. Dank voor de tientallen tekeningen, dezaterdagochtenden dat we met de lego speelden (en nog gaan spelen), jullie verhalen enongelofelijke creativiteit. Hiermee hebben jullie mij geholpen om te blijven zien dat je somsgewoon even moet gaan spelen. Laura, jouw zelfvertrouwen is een lust om te zien, evenals je lachals je weer iets grappigs herkent. Merle, ik zou nog wel honderd jaar met jouw gezang wakkerwillen worden. En…. gelukkig zijn jullie veel sneller dan ik in het maken van boeken!

Katelijne, dank voor de mogelijkheid die jij biedt om mij op de wetenschap te richten en al hetandere dat je voor me betekent. Hoewel je mij ook op gezette tijden, op meer of minderdiplomatiek wijze, naar het nut van mijn exercities vraagt, heb ook jij de afgelopen jaren veel in ditproefschrift geïnvesteerd. Ik realiseer me maar al te goed dat jij mij op meerdere momentenbehoedt voor doordraven, wat vermoedelijk in mijn genen zit, maar niet minder mijn eigenverantwoordelijkheid is. Ik ga je niet beloven dat het allemaal anders gaat worden na 5 November,daarvoor ben ik te nieuwsgierig, heb ik teveel plezier in dit werk en hebben we niets aan lozebeloftes. Het vinden van een juiste, iets evenwichtiger balans moet wel weer kunnen. Gelukkig zalik maar eenmaal promoveren. Kus.

Eric Ruhe

Amstelveen, September 2008

CURRICULUM VITAE

Curriculum Vitae

299

Curriculum Vitae

Eric Ruhé werd geboren op 18 November 1968 in Amsterdam, als oudste van drie zoons. Toen hijvijf was verhuisde het gezin naar Overijssel, waar zij tien jaar op het platteland woonden. Hiernaverhuisden zij naar Den Haag, waar hij in 1986 het eindexamen VWO behaalde.

Vanaf 1986 studeerde hij geneeskunde aan de Universiteit van Amsterdam, waar hij in 1991 hetdoctoraal examen behaalde. Hierna vond een reis naar Kenia en Tanzania plaats waarbij hij zijneerste kennismaking met wetenschappelijk onderzoek had bij het uitvoeren van deresistentiebepaling van de malaria parasiet voor de middelen chloroquine en amodiaquine via deAfrikaanse flying docters (AMREF).

In 1992 startte hij zijn co-schappen en behaalde in Juli 1995 het artsexamen, na opnieuw 6maanden in Tanzania te hebben doorgebracht tijdens een tropen co-schap. Als een der laatste inNederland vervulde hij vanaf 1995 zijn vervangende militaire dienst bij het instituut voorExtramuraal Geneeskundig Onderzoek (EMGO) van de Vrije Universiteit, waar hij de toenmaligecross-sectionele gegevens van de Hoorn Studie voorzag van een mortaliteits- enmorbiditeitsregistratie. Deze mortaliteits- en morbiditeitsregistratie is tot op heden operationeel.In deze periode volgde hij de postacademische opleiding tot klinisch-epidemioloog.

Vanaf 1997 startte hij zijn carrière in de psychiatrie met een baan als AGNIO bij hetzorgprogramma Stemmingsstoornissen van het Academisch Medisch Centrum (AMC). Van 1998tot 2002 was hij in het AMC in opleiding tot psychiater (opleider Prof.dr. B.P.R. Gersons),aangevuld met twee keuzestages bij de stichting Mentrum. Tijdens de opleiding psychiatrie konzijn interesse in onderzoek bij stemmingsstoornissen verder worden ontwikkeld door eensubsidie vanuit het AMC. Met deze subsidie werd in 2001-2002 een richtlijn voor het klinischhandelen bij onvoldoende respons op het eerste antidepressivum opgesteld, waarvan enkelepublicaties in dit proefschrift zijn opgenomen. Deze AMC-richtlijn werd eveneens gebruikt bij desamenstelling van de multidisciplinaire richtlijn Depressie die in deze periode werd ontwikkeld.Deze richtlijn vormde tevens het uitgangspunt voor subsidieaanvragen en de andere studies die indit proefschrift staan beschreven. Sinds zijn inschrijving in het specialistenregister in Juni 2003,werkt hij als stafpsychiater bij het zorgprogramma Stemmingsstoornissen van het AMC, en heefthij zich verder gespecialiseerd in de behandeling van uni- en bipolaire depressies.

Eric Ruhé leeft sinds 1994 samen met Katelijne Schmeink, met wie hij in 1998 is getrouwd. Zijhebben samen twee prachtige dochters: Laura (2000) en Merle (2002).

300

Curriculum Vitae

Eric Ruhé was born in Amsterdam on November 18th, 1968. He is the eldest of three sons. At theage of five, the family moved to Overijssel, a rural part of the Netherlands, where they lived forten years. The family then moved to The Hague, where he completed his secondary school (VWO).

From 1986 until 1991, he studied medicine at the University of Amsterdam. His first encounterwith scientific research was in Kenya and Tanzania, where he participated in the determination ofresistance of plasmodium Falciparum against chloroquine and amodiaquine, a project conductedfor the African flying doctors service (AMREF).

He started his internships in 1992 and received his medical degree in 1995, after spendinganother 6 months in Tanzania, performing a special internship in tropical medicine. He thereafterserved his civil service at the Vrije Universiteit, at the institute for extramural medicine (EMGO).He developed a mortality and morbidity registration for the Hoorn Studie, which was until then across-sectional study. This mortality and morbidity registration is still operational until now.During this civil service, he received his masters in clinical epidemiology.

Since 1997, he has been working in psychiatry, first as a non-resident (AGNIO) at the programfor Mood Disorders of the Academic Medical Center (AMC). From 1998 until 2002, he specializedin psychiatry at the AMC and Mentrum. During his residency he combined his epidemiological andpsychiatric knowledge, and performed a clinical practice guideline project, funded by the AMC.This local guideline was also incorporated in the Dutch multidisciplinary guideline for depressionwhich was developed at the same time. Three publications of this thesis result form the guidelineproject. Thereafter two grant applications were written and awarded, which enabled the otherstudies of this thesis. Since his registration as a psychiatrist in 2003, he has been working as aspecialized clinical psychiatrist at theprogram for Mood Disorders at the AMC.

He shares his life with Katelijne Schmeink, they were married in 1998. Together they have twowonderful daughters: Laura (2000) and Merle (2002).

evidence why paroxetine dose escalation is not effective in major depressive disorder: a randomized...

Documents