postscript - bjo.bmj.com · bicarbonate producing a metabolic acidosis— ... physiological stress...

Acetazolamide, alternatecarbonic anhydrase inhibitorsand hypoglycaemic agents:comparing enzymatic withdiuresis induced metabolicacidosis following intraocularsurgery in diabetesWe describe a case of acetazolamide inducedacidosis associated with the precipitation of ahyperosmolar state in a diabetic patient6 weeks after routine phacoemulsification.While renal tubular acidosis is well reportedwith acetazolamide, this case suggests that adirect diuresis induced acidosis can also havesignificant effects, producing serious compli-cations when acetazolamide is prescribedto a diabetic patient, and those with renalimpairment, with important implications forprescribing.

Case reportA 47 year old female patient underwenttechnically uncomplicated left phacoemulsi-fication with intraocular lens implant in2002. Medical history included insulindependent diabetes since 1971. She hadtreated, stable proliferative diabetic retino-pathy, relatively mild diabetic nephropathy(proteinuria with a stable creatinine in theregion of 140 mmol/l for several months), andmild diabetic autonomic neuropathy. Serumurea had been slightly raised in the past,though had normalised. Serum electrolyteswere also within normal limits. The patientwas compliant with instructions and bloodglucose had been well controlled over manyyears with regular subcutaneous insulin, noepisodes of ketoacidosis or a non-ketotichyperosmolar state.

Six weeks after cataract surgery she devel-oped left cystoid macular oedema. Confirmedby fundus fluorescein angiography, treat-ment was started with topical ketorolac andfrequency of postoperative topical steroidincreased. Treatment was later started withacetazolamide 250 mg orally twice a day,with instructions to drink lots of sugar freefluids to compensate for the diuretic effect.Arrangements were made for regular mon-itoring of her electrolyte status.

The patient started to progressively dete-riorate over the next few days, reporting amassive diuresis. She required emergencyadmission 6 days after starting treatment.Biochemical results are shown in table 1.Subcutaneous insulin was administered andacetazolamide discontinued. A sliding scaleof insulin and intravenous saline drip werecommenced to resuscitate her. Full bloodcount was normal, with no evidence ofneutrophilia. Arterial blood gas analysis isshown in table 2. This shows that she had ametabolic acidosis. Arterial blood pH 7.3 afterinitial resuscitation implies that she was evenmore acidotic before fluid resuscitation. Thehyperglycaemia, absence of ketones, andraised osmolality led to the diagnosis ofhyperglycaemic hyperosmolar non-ketoticsyndrome (HONK)

The patient stabilised rapidly overnight,with normal blood gases, blood glucose, andan improving serum creatinine of 141 mmol/lby the next day. A sliding scale was discon-tinued 2 days following admission, when shewas recommenced on a subcutaneous insulinregimen and discharged as an inpatient.

CommentThis case suggests that the diuretic inducedmechanism for acetazolamide acidosis can bea cause of severe metabolic acidosis insusceptible patients, and that the diuresiscan be severe enough to precipitate a lifethreatening diabetic crisis. Carbonic anhy-drase inhibitors such as acetazolamide affectthe metabolism of carbonic acid, bicarbonate,and carbon dioxide within the proximaltubule cell, inducing a slight diuresis.1 It israre for severe metabolic acidosis to developoutside advanced renal failure, chronic dia-lysis, in the elderly and those on nephrotoxicdrugs.2–6 While the patient’s renal impair-ment was only moderate with serum creati-nine at 140 mmol/l, when acutely unwell itapproached 150 mmol/l, a level which wouldhave necessitated referral to a renal specialistto plan end stage renal replacement ther-apy.2 4 5 This is because patients with diabeticnephropathy tend to do less well than thosewith other causes of renal impairment and, infact, renal dialysis may in any case be

required at relatively low levels of creatininesuch as less than 200 mmol/l.5

Most reports in the literature do not specifythe underlying pathophysiological mechan-ism causing metabolic acidosis with acetazol-amide. Some cases have been suspected to bethe result of a biochemical effect operating atan enzymatic level to increase urinary loss ofbicarbonate producing a metabolic acidosis—for example, renal tubular acidosis, andpotentially also lactic acidosis, damage tothe tricarboxylic acid cycle, ketosis andinhibition of pyruvate carboxylase.2 7

However, the biochemical results in thispatient, together with the rapidity of acidosis,do not suggest a tubular origin for theacidosis.2 Instead the patient displayed analternative mechanism that accounts for themetabolic acidosis. This was causing thephysiological effect of diuresis causing lossof excess body water in a diabetic patient.Further, there was no history of biguanideuse; metformin is an oral hypoglycaemicagent that can cause lactic acidosis to theextent that it is contraindicated with acreatinine level of 150 mmol/l or more.

Basic physiological work suggests that adiuresis induced acidosis can be a significantfactor with acetazolamide.8 Biochemicalresults in this patient directly correspond tothose obtained when healthy subjects havebeen given three 250 mg doses of acetazol-amide.8 Acute clinical doses of the drug causea change in body fluid compartments leadingto a moderate isosmotic hypovolaemia withan intracellular volume expansion as well asmetabolic acidosis.8 Three 250 mg doses ofacetazolamide in healthy men are associatedwith a significant 1.7 litres reduction in bodywater, compartmentalised as a significantreduction in extracellular water and increasein intracellular water.8 In this patient such adiuresis would have been significant enoughwhen occurring over a few days to produceenough loss of body water to precipitatedehydration and lactic acidosis despite herdrinking large volumes of fluids.Physiological stress of this nature is a wellknown stimulus that can precipitate a dia-betic crisis in a susceptible patient, themassive rise in blood glucose largely account-ing for the high osmolality in the patient.Hyperglycaemic hyperosmolar non-ketoticsyndrome (HONK) does occur, although lesscommonly than ketoacidosis, in insulindependent diabetics.9 This makes plausiblethe postulate that acetazolamide was theculprit. Theoretically, a diabetic ketoacidosisis also possible, though we are unaware ofspecific reports to date in this context. HONKis arbitrarily defined as serum osmolality.320 mOsm/kg and a blood glucose level.33 mmol/l, without excessive ketones, and

Table 1 Biochemistry on admission

Serum glucose Serum Na+ Serum K+ Serum urea Serum creatinineUrineketones* Serum osmolality�

Patient 55 mmol/l 135 mmol/l 4.0 mmol/l 14.1 mmol/l 149 mmol/l 0 357 mOsm/kgNormal referencerange

3–6 mmol/l 135–145 mmol/l 3.5–5.0 mmol/l 2–8 mmol/l 60–120 mmol/l

*Normally no ketones are detected on stick testing of urine.�(2[Na++K+])+urea+glucose, using serum concentrations; dangerous if outside 240–330 mOsm/kg.

PostScript . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

LETTERS

If you have a burning desire to respond to apaper published in BJO, why not make useof our ‘‘rapid response’’ option?

Log on to our web site (www.bjophthalmol.com), find the paper that interests you, andsend your response via email by clicking onthe ‘‘eLetters’’ option in the box at the top righthand corner.

Providing it isn’t libellous or obscene, itwill be posted within seven days. You canretrieve it by clicking on ‘‘read eLetters’’ onour homepage.

The editors will decide as before whether toalso publish it in a future paper issue.

714 Br J Ophthalmol 2004;88:714–725

www.bjophthalmol.com

on August 13, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.2003.029124 on 16 A

pril 2004. Dow

nloaded from

http://bjo.bmj.com/

was clearly induced by the stress of diuresisin this patient, with which it is associated. Itwould also have compounded the patient’sexisting dehydration. Mortality from HONKcan be as high as 40% despite hospitaladmission.9

It is possible that the precise mechanismof metabolic acidosis seems not to have beenconsidered in most case reports as treatmentwas, in many ways, unaffected. Alternately,it may be that the effect reported in this caseis extremely rare. However, the clinicalfindings in this case are supported directlyby correlation with the findings of basicphysiological work on the pharmacody-namics of acetazolamide, together with workon the pathophysiology of HONK.8 9 Thissuggests that the observations made on thiscase are certainly of much broader signifi-cance and raise an issue of concern about thedrug’s prescription in both diabetes and renalfailure. While manufacturer’s recommenda-tions for acetazolamide in Britain includecontraindications to its use in suprarenaldysfunction, they do not issue cautions for itsuse in diabetics. Thus this case’s principalvalue lies in evaluating current prescribingpractice, particularly as diabetics are a verycommon group of patients in ophthalmicpractice, and acetazolamide is not uncom-monly prescribed in many different areas ofclinical ophthalmology, as well as by otherclinicians. Until further data are forthcoming,including data on newer slow release for-mulations, good practice should be to pre-scribe the drug with especial caution indiabetics, particularly for those conditions,including this case, where its prescription isnot routine. In the context of its use indiabetes it is also certainly worth comparingacetazolamide with other carbonic anhydraseinhibitors. One of the other carbonic anhy-drase inhibitors that have been used inclinical ophthalmology is methazolamide.The latter is associated with a less profoundreduction in intraocular pressure, but alsoless acidosis.6 10

This case should also serve as a reminderthat patients with any level of renal impair-ment are a group that are vulnerable toacetazolamide toxicity. The data sheet andelectronic medicines compendiums state thatacetazolamide is contraindicated in markedkidney and liver dysfunction, suprarenalgland failure, and hyperchloraemic acidosis.The British National Formulary is less specificand states that it is contraindicated in renalimpairment. We would suggest that diabeticpatients with a creatinine level of 140 mmol/lare at quite high risk of nephrotoxic drugreactions, though caution should be exercisedin even mild renal impairment.

F H Zaidi, P E KinnearDepartment of Ophthalmology, Charing Cross,Chelsea and Westminster Hospital, London, UK

Correspondence to: Farhan H Zaidi, Department ofOphthalmology, Imperial College London, Charing

Cross and Hammersmith Hospitals, London, UK;[email protected]

doi: 10.1136/bjo.2003.027490

Accepted for publication 3 July 2003

References

1 Greger R, Lohrmann E, Schlatter E. Action ofdiuretics at the cellular level. Clin Nephrol1992;38(Suppl 1):S64–8.

2 Elinav E, Ackerman Z, Gottehrer NP, et al.Recurrent lifethreatening acidosis induced byacetazolamide in a patient with diabetic type IVrenal tubular acidosis. Ann Emerg Med2002;40:259–60.

3 Gabay EL. Metabolic acidosis fromacetazolamide therapy. Arch Ophthalmol1983;101:303–4.

4 Chapron DJ, Gomolin IH, Sweeney KR.Acetazolamide blood concentrations areexcessive in the elderly: propensity for acidosisand relationship to renal function. J ClinPharmacol 1989;29:348–53.

5 De Marchi S, Cecchin E, Tesio F. Intraocularpressure changes during hemodialysis: preventionof excessive dialytic rise and development ofsevere metabolic acidosis followingacetazolamide therapy. Ren Fail1989;11:117–24.

6 Sporn A, Scothorn DM, Terry JE. Metabolicacidosis induced by acetazolamide. J Am OptomAssoc 1991;62:934–7.

7 Filippi L, Bagnoli F, Margollicci M, et al.Pathogenic mechanism, prophylaxis, and therapyof symptomatic acidosis induced byacetazolamide. J Invest Med 2002;50:125–32.

8 Brechue WF, Stager JM, Lukaski HC. Body waterand electrolyte responses to acetazolamide inhumans. J Appl Physiol 1990;69:1397–401.

9 Levine SN, Sanson TH. Treatment ofhyperglycaemic hyperosmolar non-ketoticsyndrome. Drugs 1989;38:462–72.

10 Stone RA, Zimmerman TJ, Shin DH, et al. Low-dose methazolamide and intraocular pressure.Am J Ophthalmol 1977;83:674–9.

The art of retinal detachmentsurgery: a photoessaySubjective visual experience has beendescribed previously in patients undergoingintraocular surgery, and may occur duringeither topical anaesthesia or regional anaes-thesia (peribulbar, retrobulbar, subtenons).1–7

Published reports suggest most or all patientsundergoing cataract extraction under localanaesthesia will report some visual symp-toms when questioned immediately aftertheir procedures. These symptoms are com-mon therefore and range from perception oflight, photopsia, colours, and movement,through to more formed visual sensationssuch as patterns, instruments, and surgeon’sfingers/hands/detail. It is not surprising thatpatients undergoing vitreoretinal surgery

under local anaesthesia might also experi-ence visual symptoms.

We present illustrations and comments(figs 1–4) made by an artist who underwentretinal detachment surgery. He presentedwith macula-on retinal detachment success-fully repaired by vitrectomy, cryotherapy, and20% SF6 gas performed by peribulbar anaes-thesia. They provide an interesting insightinto previously unreported visual experienceduring vitreoretinal surgery. As visual symp-toms are both common, and may be per-ceived to be frightening in a small percentageof patients,2 we reinforce the view thatinformed patient consent procedures shouldinclude the possibility of visual experienceduring vitreoretinal surgery under localanaesthesia.

B J Vote, M B EcksteinDepartment of Ophthalmology, Sussex Eye Hospital,

Brighton, UK

Correspondence to: Mr B J Vote, Department ofOphthalmology, Sussex Eye Hospital, Brighton BN2

5BF, UK; [email protected]

doi: 10.1136/bjo.2003.029504

Accepted for publication 18 August 2003

Figure 1 (A) ‘‘Now we are going to shine avery bright light light into your eye’’ … andindeed it was bright. The light pulsated gentlyand sparkled at the edges where the heavenlyblue began. (B) It was outshone by an evenbrighter light … more pulsations andshimmerings of an intense whiteness followedand two glowing red circles appeared, thelower one seeming to be a secondary image orreflection of the upper one.

Table 2 Arterial blood gases on admission

pH pCO2 pO2 Base deficit (excess)�

Patient 7.3 3.4 kPa 15 kPa 210.3 mmol/lNormal referencerange

7.35–7.45* 4.5–6.0 kPa 12–15 kPa

*Life threatening at and beyond 7.2 and 7.6.�Normal .23 mmol/l: ,23 mmol/l = metabolic acidosis, .+3 mmol/l = metabolic alkalosis,23 mmol/l to +3 mmol/l = mild metabolic acidosis, severe metabolic alkalosis, or mixed metabolicdisturbance.

Disclaimer: the authors have no financial interest inthis manuscript.

PostScript 715




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

References

1 Prasad N, Kumar CM, Patil BB, et al. Subjectivevisual experience during phacoemulsificationcataract surgery under sub-Tenon’s block. Eye2003;17:407–9.

2 Au Eong KG, Low CH, Heng WJ, et al. Subjectivevisual experience during phacoemulsification and

intraocular lens implantation under topicalanesthesia. Ophthalmology 2000;107:248–50.

3 Au Eong KG, Lee HM, Lim AT, et al. Subjectivevisual experience during extracapsular cataractextraction and intraocular lens implantation underretrobulbar anaesthesia. Eye 1999;13:325–8.

4 Au Eong KG, Lim TH, Lee HM, et al. Subjectivevisual experience during phacoemulsification and

intraocular lens implantation using retrobulbaranesthesia. J Cataract Refract Surg2000;26:842–6.

5 Newman DK. Visual experience duringphacoemulsification cataract surgery undertopical anaesthesia. Br J Ophthalmol2000;84:13–15.

6 Tranos PG, Wickremasinghe SS, Sinclair N, et al.Visual perception during phacoemulsificationcataract surgery under topical and regionalanaesthesia. Acta Ophthalmol Scand2003;81:118–22.

7 Murdoch IE, Sze P. Visual experience duringcataract surgery. Eye 1994;8:666–7.

Apolipoprotein E polymorphismin patients with cataractBased on similarities in epidemiology andbiochemistry, it has been suggested thatcataract and Alzheimer’s disease (AD) sharethe same aetiological mechanisms.Comorbidity of cataract and AD in trisomy21 (Down’s syndrome) is well known1 2 andboth diseases are characterised by aggregatedproteins exhibiting excessive glycation andracemisation of aspartyl residues.3 Several AD

Figure 2 (A) From time to time liquid was splashed on to the eye, producing quite remarkablestarbursts of bubbles and comet trails of light. (B) The brilliant light faded and was replaced bydeliquescent magenta coloured shapes, glowing like neon, constantly forming and reforming andwith sparks coruscating round the edges … a real firework display. (C) Quite suddenly thefireworks gave way to this darker mood … the shapes still forming and reforming but more slowly,no longer glowing, but in a complementary colour to that of the previous image … a muted limeyellow. (D) Next came perhaps the most extraordinary and delightful of all the images … this nearperfect facsimile of a Chinese silk painting of bamboos gently swaying, as if in a slight breeze, infront of a full moon. (E) Alas, the ‘‘Chinese painting’’ gave way to the appearance of ‘‘longwhiskers’’—rather like huge stray eyelashes—moving slowly but sometimes jerkily, at the top rightof the field of vision. (F) Darker than this, richly beautiful and gently blurred, this next imagesomehow resembled a shallow stream. Clear water rippled over sunlit pebbles and water plants,long streamers of which gently swayed in the movement of the water. Sunlight flickered gently overeverything—in many ways, a surprisingly realistic image. (G) This very powerful yet momentaryimage occurred towards the end of the operation. The incandescent green cross burned powerfullylike an electric filament … an inspection light perhaps?

Figure 3 (A) After an initial period ofimpenetrable ‘‘fog’’, the gas bubble graduallyreceded so as to allow me to see over the top ofit. In these illustrations the bubble has recededto roughly halfway down the field of vision and Iam lying on my left hand side with both eyesclosed. The room is only semi-darkened. Thepink light is very beautiful and emanates from atiny convexity in the upper edge of the gasbubble. (B) The room is in complete darkness.The display is even more beautiful. Now it is atiny concavity in the upper surface of the bubblethat allows the violet ‘‘flare’’ of light to emerge.

716 PostScript




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

related proteins—amyloid precursor protein(APP), b amyloid (Ab), and presenilin (PS)—are expressed in the lens4 5 and Ab isaccumulated in the cytosol of lens fibres incataractous lenses of people with AD.6

Human apolipoprotein E (apoE) exists inthree major isoforms encoded by distinctalleles (APOE e2, e3, and e4). The differentAPOE alleles have been studied in relation toseveral human age related diseases: inheri-tance of the e4 allele is a strong risk factor forAD and influences Ab metabolism.7 8 Thepurpose of this study was to investigate theAPOE e2/e3/e4 polymorphism in patientswith cataract.

After informed consent, patients withsenile cataract and control individuals wererecruited from two ophthalmic clinics inTartu and the south Estonian area. The studywas approved by the ethics committee at theUniversity of Tartu, Estonia. Before surgery,the type of cataract was determined usingbiomicroscopy and ophthalmoscopy.Secondary cataracts were excluded. The casegroup included 502 patients; 77 with nuclear,155 with cortical, 119 with posterior sub-capsular, and 151 with mixed opacities. Meanage was 72.0 (SD 8.7) years (range 47–93years) and 348 (69.3%) were women. Thecontrol group consisted of 187 individualswithout cataract, uveitis, or glaucoma. Meanage was 65.8 (SD 6.9) years (range 43–90years) and 136 (72.7%) were women. Thepower of the study was .99% as calculatedaccording to Altman9 on the basis of APOE e4allele frequencies in a recent study on AD.10

The APOE alleles and genotypes weredetermined as previously described.11 Theallele and genotype frequencies of cataractcases and controls were compared using atwo tailed Fisher’s exact test, and odds ratios(ORs) and 95% confidence intervals (CIs)were calculated.9 All statistical analyses wereperformed using SYSTAT as software (SPSSInc, Chicago, IL, USA). Statistical significancewas defined as p,0.05.

APOE allele and genotype frequenciesfound in this study are well in accordancewith those reported in other NorthernEuropean populations.12 No significant differ-ences were seen between the control andcataract groups for any of the APOE alleles(table 1) or APOE genotypes (table 2).Neither were there any differences betweenthe control group and the specific cataractsubgroups. In order to prevent the data frombeing influenced by age differences between

the groups studied, age matched controlindividuals were selected and compared withthe cataract group and vice versa, withoutresulting in any significant changes in APOEallele or genotype frequencies.

Alzheimer’s disease and cataract bothexhibit large aggregates of aberrant proteins,senile plaques composed of Ab and neurofi-brillary tangles containing the cytoskeletalprotein tau in the former case, and lightscattering high molecular weight aggregatesof crystallins in the latter. Together withseveral other diseases characterised by pro-tein aggregates, such as amyloidosis andprion diseases, the term ‘‘conformationaldisease’’ has been created, suggesting acommon aetiology.3 13

The APOE e4 allele is a strong risk factorfor AD, and it is believed that in neuronaltissue, apoE is important for mobilisation andredistribution of lipids, and for maintenanceand repair of neuronal cell membranes.14

However, in age related macular degenera-tion (AMD)—a condition characterised byaccumulation of extracellular depositstermed drusen, containing among otherthings neutral lipids, cholesterol, andapoE—the e4 allele appears to confer protec-tion, whereas the e2 allele is associated with amoderately increased risk of AMD.15 16 TheAPOE e4 allele also seems to play a protectiverole during embryogenesis,17 suggesting dif-ferent effects of the gene early and late in life.

To our knowledge, this is the first study toinvestigate the APOE polymorphism in catar-act patients. No differences in the distribu-tion of APOE alleles and genotypes could beseen between controls and cataract patientsin spite of a large number of participants anda very high power. This indicates that if thereis a common pathogenic mechanism forcataract and AD, it does not involve the

Figure 4 (A) Considerable time has nowelapsed and the gas bubble has diminished agreat deal. This is my view of it with the eyeclosed and in strong sunlight. (B) This illustratesstill further diminution of the bubble and itsdetachment from the edge of the field of vision.(C) Eventually the gas bubble becameminiscule, and reversed its position in the visualfield of vision before disappearing altogether.

Table 1 APOE allele frequencies for control and cataract groups

APOEallele*

Observed frequency

Cataracts

Controls(n = 374)

Nuclear(n = 154)

Cortical(n = 310)

Posteriorsubcapsular(n = 238)

Mixed(n = 302)

All cases(n = 1004)

e2 0.112 0.110 0.113 0.071 0.126 0.107e3 0.773 0.792 0.774 0.832 0.768 0.789e4 0.115 0.097 0.113 0.097 0.106 0.105

n, number of alleles.*p.0.05 for all alleles when comparing controls and cataracts (all cases) or cataract subgroups. 95%confidence intervals of all odds ratios included 1.0 (no difference).

Table 2 APOE genotype distributions for control and cataract groups

APOEgenotype

Observed frequency

Cataracts

Controls(n = 187)

Nuclear(n = 77)

Cortical(n = 155)

Posteriorsubcapsular(n = 119)

Mixed(n = 151)

All Cases(n = 502)

e2/e2 0.011 0.013 0.013 0.017 0.020 0.016e2/e3 0.182 0.156 0.181 0.101 0.179 0.157e2/e4 0.021 0.039 0.019 0.008 0.033 0.024e3/e3 0.588 0.636 0.587 0.697 0.596 0.624e3/e4 0.187 0.156 0.194 0.168 0.166 0.173e4/e4 0.011 NO 0.006 0.008 0.007 0.006

p.0.05 for all genotypes when comparing controls and cataracts (all cases) or cataract subgroups.95% confidence intervals of all odds ratios included 1.0 (no difference).NO, not observed.

PostScript 717




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

APOE polymorphism. Of course the resultsneed to be confirmed by other groups beforethe APOE polymorphism can be regarded asinsignificant in cataractogenesis. Bearing inmind the similarities between cataract andAD is very important, however, as progress inaetiological research of one disease maycontribute to elucidating the pathogenesis ofthe other.

AcknowledgementsThis work was supported by grants from theSwedish Medical Research Council (projects#02226, #5932, and #12103), the SahlgrenskaUniversity Hospital, the Goteborg Medical Society,Stiftelsen Kronprinsessan Margaretas Arbetsnamndfor Synskadade, De Blindas Vanner, StiftelsenHjalmar Svenssons forskningsfond, and ToreNilsons Stiftelse for Medicinsk Forskning.

M ZetterbergInstitute of Clinical Neuroscience, Section of

Ophthalmology, Sahlgrenska University Hospital,Goteborg University, Molndal, Sweden

H Zetterberg, M Palmer, L Rymo, K BlennowDepartment of Clinical Chemistry and Transfusion

Medicine, Sahlgrenska University Hospital, GoteborgUniversity, Goteborg, Sweden

G Tasa, E Juronen, S VeromannDepartment of Human Biology and Genetics,

University of Tartu, Tartu, Estonia

P TeesaluDepartment of Ophthalmology, University of Tartu,

Tartu, Estonia

M Zetterberg, J-O KarlssonDepartment of Anatomy and Cell Biology, Medical

Faculty, Goteborg University, Goteborg, Sweden

K Hoglund, K BlennowInstitute of Clinical Neuroscience, Department of

Experimental Neuroscience, Sahlgrenska UniversityHospital, Goteborg University, Goteborg, Sweden

Correspondence to: Dr Madeleine Zetterberg, Instituteof Clinical Neuroscience, Section of Ophthalmology,

Sahlgrenska University Hospital, SE-431 80 Molndal,Sweden; [email protected]

doi: 10.1136/bjo.2003.032698

Accepted for publication 1 September 2003

References

1 da Cunha RP, Moreira JB. Ocular findings inDown’s syndrome. Am J Ophthalmol1996;122:236–44.

2 Lott IT. Down’s syndrome, aging, and Alzheimer’sdisease: a clinical review. Ann N Y Acad Sci1982;396:15–27.

3 Harding J. Cataract, Alzheimer’s disease, andother conformational diseases. Curr OpinOphthalmol 1998;9:10–13.

4 Frederikse PH, Zigler Jr JS. Presenilin expresssionin the ocular lens. Curr Eye Res 1998;17:947–52.

5 Frederikse PH, Garland D, Zigler J Jr, et al.Oxidative stress increases production of b-amyloid (Ab) in mammalian lenses, and Ab hastoxic effects on lens epithelial cells. J Biol Chem1996;271:10169–74.

6 Goldstein LE, Muffat JA, Cherny RA, et al.Cytosolic b-amyloid deposition and supranuclearcataracts in lenses from people with Alzheimer’sdisease. Lancet 2003;361:1258–64.

7 Holtzman DM, Bales KR, Tenkova T, et al.Apolipoprotein E isoform-dependent amyloiddeposition and neuritic degeneration in a mousemodel of Alzheimer’s disease. Proc Natl AcadSci U S A 2000;97:2892–7.

8 Corder EH, Saunders AM, Strittmatter WJ, et al.Gene dose of apolipoprotein E type 4 allele andthe risk of Alzheimer’s disease in late onsetfamilies. Science 1993;261:921–3.

9 Altman DG. Practical Statistics for MedicalResearch. First edition. London: Chapman andHall/CRC, 1991.

10 Johansson A, Hampel H, Faltraco F, et al.Increased frequency of a new polymorphism inthe cell division cycle 2 (cdc2) gene in patientswith Alzheimer’s disease and frontotemporaldementia. Neurosci Lett 2003;340:69–73.

11 Blennow K, Ricksten A, Prince JA, et al. Noassociation between the a2-macroglobulin (a2M)deletion and Alzheimer’s disease, and no changein a2M mRNA, protein, or protein expression.J Neural Transm 2000;107:1065–79.

12 Eichner JE, Dunn ST, Perveen G, et al.Apolipoprotein E polymorphism andcardiovascular disease: a HuGE review.Am J Epidemiol 2002;155:487–95.

13 Carrell RW, Lomas DA. Conformational disease.Lancet 1997;350:134–8.

14 Mahley RW. Apolipoprotein E: cholesteroltransport protein with expanding role in cellbiology. Science 1988;240:622–30.

15 Souied EH, Benlian P, Amouyel P, et al. Theepsilon4 allele of the apolipoprotein E gene as apotential protective factor for exudative age-related macular degeneration. Am J Ophthalmol1998;125:353–9.

16 Klaver CCW, Kliffen M, van Duijn CM, et al.Genetic association of apolipoprotein E with age-related macular degeneration. Am J Hum Genet1998;63:200–6.

17 Zetterberg H, Palmer M, Ricksten A, et al.Influence of the apolipoprotein E e4 allele onhuman embryonic development. Neurosci Lett2002;324:189–92.

Mitomycin C in sebaceous glandcarcinoma with pagetoid spreadSebaceous gland carcinoma is a rare eyelidtumour comprising less than 1% of all eyelidmalignancies.1 It commonly arises from themeibomian glands of the tarsus, but may alsoarise from the glands of Zeis or from thesebaceous glands of caruncle.2 It can presentin a nodular or diffuse infiltrative form. Thelatter form with intraepithelial (pagetoid)invasion has poor prognosis as a result ofdelay in diagnosis as well as more extensiveinvolvement of ocular tissues. Topical appli-cation of mitomycin C, a non-cell cyclespecific alkylating agent, has been advocatedfor pagetoid spread of sebaceous glandcarcinoma.3 We report the use of mitomycinC as adjuvant therapy in a patient withcompletely excised sebaceous gland carci-noma and pagetoid spread.

Case reportA 78 year old man was referred to theoculoplastic clinic with epiphora and irrita-tion of right eye for 2 years. There was noprevious ocular or medical history. Clinicallyhe had a unilateral right upper lid entropionwith tarso-conjunctival cicatrisation (fig 1)and bilateral dermatochalasis. The patient

underwent bilateral blepharoplasty andbiopsy of right upper lid tarsal plate andconjunctiva. The biopsy confirmed sebaceousgland carcinoma with pagetoid invasion ofthe conjunctival epithelium (fig 2).

He had a full thickness wedge excision ofthe right upper lid with tarsoconjunctivalbiopsies. These showed sebaceous glandcarcinoma to the margin of the excision withpagetoid invasion of the conjunctiva andepidermis of the lid margin. A wider excisionof the lid and further conjunctival biopsieswere performed with frozen section revealingcomplete excision of the tumour.Reconstruction of the posterior lamellae wasachieved using a hard palate graft and theanterior lamellae was repaired by a myocu-taneous flap with post auricular skin graftand a bilobed flap medially.

Conjunctival map biopsies were clear oftumour 1 and 6 months post excision. Inview of pagetoid spread, the patient wascommenced on three cycles of topical mito-mycin C 0.02% four times a day. Each cycleconsisted of 2 weeks of mitomycin C and2 weeks off therapy. Corneal epithelial toxi-city and ulceration was noted with mitomy-cin C therapy, requiring preservative freelubricants and lateral tarsorrhaphy. Twoyears after excision of tumour, the patientremains disease free.

CommentIntraepithelial invasion in sebaceous glandcarcinoma is noted to occur in 41–80% ofcases.1 4 Diagnosis may be delayed as thepresenting symptoms are often benign andnon-specific such as blepharoconjunctivitis.Diagnosis requires biopsy of the abnormalarea and conjunctival map biopsies in thepresence of intraepithelial invasion.3 5

Various treatments have been used forpagetoid invasion including surgical excisionwith cryotherapy, external beam radiother-apy, and orbital exenteration.6 Eyes withpagetoid invasion are more likely to undergoexenteration.7

In our case the suspicion of malignancywas raised because of the unilaterality of theclinical features. Our patient underwentextensive excision of the tumour withtumour free conjunctival biopsies.Mitomycin C as adjuvant treatment wascommenced as a result of the difficulty inclinically assessing for recurrence with page-toid invasion. Mitomycin C was associatedwith moderate epithelial toxicity which wasself limiting.

Mitomycin C is a non-cell cycle specificalkylating agent which acts to inhibit cellproliferation, and is used successfully in thetreatment of corneal intraepithelial neopla-sia.8 This is only the second reported articlewhere mitomycin C has been used in the

Figure 1 Upper lid tarso-conjunctivalscarring.

Figure 2 Conjunctival biopsy demonstratingintraepithelial invasion of malignant cells.

718 PostScript




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

treatment of sebaceous gland carcinoma. Inthe pilot study by Shields et al3 there wascomplete resolution of tumour with norecurrences over 12 month follow up.Mitomycin C as adjuvant treatment inpagetoid spread of sebaceous gland carci-noma may reduce the need for more invasivetreatment options.

K Tumuluri, G Kourt, P MartinDepartment of Oculoplastic and Orbital Surgery,

Sydney Eye Hospital, Sydney, Australia

Correspondence to: Dr Georgina Kourt, EyeAssociates, 149 Macquarie Street, Sydney 2000,

Australia; [email protected]

doi: 10.1136/bjo.2003.034215


References

1 Doxanas MT, Green WR. Sebaceous glandcarcinoma: review of 40 cases. Arch Ophthalmol1984;102:245–9.

2 Cook B, Bartley G. Treatment options and futureprospects for the management of eyelidmalignancies. Ophthalmology2001;108:2088–100.

3 Shields CL, Naseripour M, Shields J, et al. Topicalmitomycin C for pagetoid invasion of theconjunctiva by eyelid sebaceous glandcarcinoma. Ophthalmology 2002;109:2129–33.

4 Rao NA, Hidayat AA, McLean JW, et al.Sebaceous carcinoma of the ocular adnexa: aclinicopathologic study of 104 cases with five yearfollow-up data. Human Pathol 1982;13:113–22.

5 Putterman AM. Conjunctival map biopsy todetermine pagetoid spread. Am J Ophthalmol1986;102:87–90.

6 Kass LG, Hornblass A. Sebaceous carcinoma ofthe ocular adnexa. Surv Ophthalmol1989;33:477–90.

7 Chao AN, Shields CL, Krema H, et al. Outcome ofpatients with periocular sebaceous glandcarcinoma with and without conjunctivalintraepithelial invasion. Ophthalmology2001;108:1877–83.

8 Frucht-Pery J, Sugar J, Baum J, et al. Mitomycin Ctreatment for conjunctival-corneal intraepithelialneoplasia: a multicenter experience.Ophthalmology 1997;104:2085–93.

A questionnaire survey of patientacceptability of optic discimaging by HRT II and GDxGlaucoma is an insidious condition whichremains asymptomatic until very advancedwith nerve damage occurring before detect-able visual field loss.1 Early detection andtreatment result in a better prognosis withretardation of progression.2

The Heidelberg retinal tomograph (HRT) II(Heidelberg Engineering, Germany) and theGDx Nerve fibre analyser (Laser DiagnosticTechnologies Inc, San Diego, CA, USA) areinstruments which use scanning laser tech-nology to diagnose and monitor the progres-sion of glaucoma.

We conducted a questionnaire survey ofsubjects undergoing imaging by these meth-ods in a primary care setting to comparepatient acceptability of the two tests.

MethodsSeventy new patients referred with a possiblediagnosis of glaucoma were asked to com-plete a questionnaire about their experienceof optic disc imaging. Informed consent wasobtained and the study had approval fromthe Moorfields Eye Hospital research andethics committee. None of the subjects had

undergone disc imaging previously. Subjectsunderwent sequential disc imaging by experi-enced technicians using HRT II then GDx orvice versa in approximately equal numbers.Only subjects who had vision of at least 6/12and who had successful imaging by bothmethods were included.

The questionnaires consisted of two iden-tical sets of six direct questions using a size14 font (Appendix 1). Questionnaires werecompleted immediately after imaging toreduce the potential for recall bias.Statistical significance was determined usingBinomial and McNemar’s tests.

ResultsSixty seven questionnaires were completed.Demographic and diagnostic data are shownin table 1 and patient responses in table 2.The majority of patients found both testsagreeable with regards to each characteristicunder study other than chin rest comfort.Sixteen patients found the HRT II comfor-table but not the GDx, compared with justone patient finding the GDx comfortable butnot the HRT II (p = 0.0003). Despite this,

there were slightly more patients reportingthat the HRT II chin rest only was uncomfor-table (p = 0.052). Nine patients found thatimaging with GDx but not HRT II took toolong compared with no patients findingthe HRT II but not GDx too long(p = 0.0039). Similar numbers of patientsreported that only one of the imagingtechniques was too bright. Nineteen patientshad trouble fixating with the GDx but not theHRT II compared with just three patientshaving trouble fixating with HRT II but notGDx (p = 0.009).

Despite these differences, 28 subjects(42%) stated no preference for either imagingtechnique. Of the 39 subjects who did state apreference, 31 (79%) preferred HRT II com-pared with eight (21%) who preferred GDx(p = 0.0003, Binomial test). Twenty of the 31subjects (65%) who chose HRT II did sobecause it was of a shorter duration, five ofthe 31 (16%) said that it was more comfor-table, and six of the 31 (19%) said that it waseasier to perform the test. Four of the eightsubjects (50%) who chose GDx did so becauseit was easier, two (25%) said that it was morecomfortable and two (25%) did not give areason.

CommentDiagnostic and screening tests should besafe, specific, sensitive, and acceptable topatients. The HRT and GDx have citedsensitivities of 0.42–0.88 and 0.64–0.96 andspecificities of 0.84–0.90 and 0.74–0.96respectively.3–9

Most patients found both tests to be fairlyacceptable. Twenty eight (42%) subjectsstated no preference but of those who did, asignificant proportion of patients preferredHRT II over GDx. The most common reasongiven was a shorter test duration implyingthat acquisition time may have an impact onacceptability. Examination with the GDx maybe longer because of the external fixationtarget, which a greater proportion of subjectsfound difficult to focus on. In contrast, theHRT II has an internal fixation target.

Table 1 Demographic anddiagnostic data

Demographic features (n = 67)

Male 34 (50.7%)Female 33 (49.3%)Age (years) Mean 57.1,

SD 14.3,range 18–85

DiagnosisPrimary open angleglaucoma

12 (18%)

Glaucoma suspect 18 (27%)Ocular hypertension 16 (24%)Non-glaucomatous opticneuropathy

3 (4.5%)

Normal 15 (22%)No diagnosis given 3 (4.5%)

Table 2 Patient responses

Question p Value (McNemar’s)

1. Was the test comfortable? (n = 65)HRT 0.0003Yes No

GDX Yes 46 1No 16 2

2. Was the light too bright? (n = 58)HRT 0.3877Yes No

GDX Yes 2 4No 8 44

3. Was the chin rest uncomfortable? (n = 65)HRT 0.0522Yes No

GDX Yes 26 8No 19 12

4. Was the test too long? (n = 65)HRT 0.0039Yes No

GDX Yes 1 9No 0 55

5. Did you have trouble keeping your eye still?(n = 65)

HRT 0.0009Yes No

GDX Yes 14 19No 3 29

PostScript 719




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

Although the HRT II was found to be a morecomfortable test, a higher proportion foundthe HRT II chinrest to be uncomfortablepossibly because of the forward slopingangulation.

We did not correlate patient preferencewith image acquisition time. More openquestions may have helped to find thereasons for certain preferences. We wereunable to determine the strength of prefer-ence from the collected data. Our study wasalso not randomised but roughly equalnumbers had either HRT II or GDx first. Wedo not feel that there was a significant ordereffect. As our patients were new referrals,their responses were not biased by familiaritywith previous tests. Patients underwent bothtests sequentially on the same day by trainedtechnicians reducing the likelihood of pro-longed acquisition time due to inexperiencedoperators. All patients had good vision solocating the target was not an issue.Additional work examining the factors whichaffect acquisition time (for example, refrac-tive error, presence of media opacity, pupilsize) is needed to further understand patientpreference.

It is uncertain if the differences in pre-ference between the two tests will have asignificant impact on patient satisfaction andcompliance with clinic visits as a whole.Other factors, such as waiting time andcomfort of waiting room, will have to beexamined as well.

In conclusion, our study highlights theimportance of both test characteristics andcomfort in instrument design. It is hoped thatmanufacturers take into account these fac-tors in the design of the next generation ofglaucoma imaging devices.

Appendix 1The questions were:

1. Was the test comfortable? Yes or No?

2. Was the light too bright? Yes or No?

3. Was the chin rest uncomfortable? Yesor No?

4. Was the test too long? Yes or No?

5. Did you have trouble keeping your eyestill? Yes or No?

6. (a) Which test did you prefer? HRT orGDx? (b) Why?

E Tay, P Andreou, W Xing, C Bunce, T Aung,W A Franks

Glaucoma Service, Moorfields Eye Hospital, London,UK

Correspondence to: E Tay, Moorfields Eye Hospital,City Road, London EC1V 2PD, UK; dr_eugenetay@

yahoo.com

doi: 10.1136/bjo.2003.034975


References

1 American Academy of Ophthalmology PreferredPractice Pattern Committee Glaucoma Panel.Primary open angle glaucoma. San Francisco,CA: American Academy of Ophthalmology,1996.

2 Jay J, Allen D. The benefit of early trabeculectomyversus conventional management in primary openangle glaucoma relative to severity of disease.Eye 1989;3:528–35.

3 Mikelberg FS, Parfitt CM, Swindale NV, et al.Ability of the Heidelberg retina tomograph todetect early glaucomatous visual field loss.J Glaucoma 1995;4:242–7.

4 Lester M, Mikelberg FS, Drance SM. The effect ofoptic disc size on diagnostic precision with theHeidelberg Retinal Tomograph. Ophthalmology1997;104:545–8.

5 Poinoosawmy D, Tan JC, Bunce C, et al. Theability of the GDx nerve fibre analyser neuralnetwork to diagnose glaucoma. Graefe’s ArchClin Exp Ophthalmol 2001;239:112–27.

6 Kamal DS, Bunce C, Hitchings RA. The use of theGDx to detect retinal nerve fibre thicknessbetween normal, ocular hypertension and earlyglaucoma. Eye 2000;14:367–70.

7 Weinreb RN, Zangwill L, Berry CC, et al.Detection of glaucoma with scanning laserpolarimetry. Arch Ophthalmol1998;116:1583–90.

8 Choplin NT, Lundy DC. The sensitivity andspecificity of scanning laser polarimetry in thedetection of glaucoma in a clinical setting.Ophthalmology 2001;108:899–904.

9 Tjon-Fo-Sang M, Lemij HG. The sensitivity andspecificity of nerve fibre layer measurements inglaucoma as determined with scanning laserpolarimetry. Am J Ophthalmol 1997;123:62–9.

A novel mutation in thealternative splice region of thePAX6 gene in a patient withPeters’ anomalyThe PAX6 gene is involved in ocular embryo-genesis. This gene seems to be the mastercontrol gene for morphogenesis of the eye.Mutations in the PAX6 gene have beendetected in various ocular anomalies sus-pected to have bilateral genetic backgroundsduring development, including aniridia,Peters’ anomaly,1 and foveal hypoplasia.2

In 1994, a sporadic case of Peters’ anomalyand a small family with a range of anteriorsegment malformations, including Peters’anomaly, were shown to have a mutation ofthe PAX6 gene.1 More recently, Azuma et alreported a subject with Peters’ anomalyhaving a missense mutation in the alternativesplice region of the PAX6 gene in 1999.3 Herewe report a novel PAX6 gene mutation in apatient with Peters’ anomaly.

Case reportThe present study had the approval of KyotoPrefectural University of Medicine ethicscommittee and was conducted in accordancewith the World Medical AssociationDeclaration of Helsinki. Genomic DNA sam-ples were isolated from the whole blood ofpatients and their relatives after informedconsent. Each exon of the PAX6 gene and itsimmediate flanking sequence were amplifiedby polymerase chain reaction (PCR). Purifiedamplified fragments were subjected tosequenced using an ABI Prism 3100 geneticanalyser (Applied Biosystems, Foster City,CA, USA). To confirm the sequence ofmutations, the SNaPshot method4 was per-formed.

Of the four patients studied, we detected anovel missense mutation in one patient. Thepatient, a 20 year old girl, had bilateralPeters’ anomaly showing corneal opacitywith iridocorneal adhesion and nystagmus(fig 1). The fundus of both eyes could not beseen because of corneal opacity. No systemicassociations with Peters’ anomaly were iden-tified. Sequence analyses revealed a hetero-zygous mutation as A/G at the 38th positionwhich resulted in Q13R substitution in thePAX6 gene. No mutation was found in herparents and elder brother, which is consistentwith the fact that they show no abnormalfindings on clinical examination (fig 2).

CommentWe have identified one missense mutation inthe alternative splice region (exon5a) of thePAX6 gene in a subject with Peters’ anomaly.This missense mutation was the substitutionfrom glutamine to arginine at the 13th codon,which is the second reported position in theexon5a.

The Pax family of developmentally regu-lated transcription factors share an aminoterminal DNA binding motif known as thepaired domain. The paired domain has abipartite structure with a highly conserved N-terminal subdomain (NTS) and C-terminalsubdomain, which bind distinctive consensussequences, and the insertion of exon5a into

Figure 1 Photographs show the anterior segment region. (A) The anterior segment of the patient.She had bilateral Peters’ anomaly and showed corneal opacity with iridocorneal adhesion andnystagmus. The best corrected visual acuity was 20/100 (right) and 20/200 (left). The fundus ofboth eyes could not be seen because of corneal opacity. (B) The anterior segment of patient’s father.(C) The anterior segment of patient’s mother. (D) The anterior segment of patient’s elder brother. Nocongenital ocular abnormalities of anterior segment region were found in her parents or elderbrother.

No financial interest or support.

720 PostScript




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

the N-terminal subdomain abolishes theDNA binding activity of the NTS.5

Interestingly, in 1999, Azuma et al provedthat the mutation in the NTS of the paireddomain partially restored the DNA bindingactivity of the NTS, using functional ana-lyses.3 In addition, because the amino acidsglutamine and arginine belong to hydrophilicand basic amino acid respectively, they havedifferent electronic charges. Therefore, thisamino acid substitution may affect thestructure of the PAX6 protein and thenchange DNA binding activity of the paireddomain. In other words, such a single aminoacid substitution is an important motif as thepaired domain such as we have found mayseverely damage the normal protein function.In fact, except for aniridia, other missensemutations that result in a serious congenitaleye disease such as Peters’ anomaly or fovealhypoplasia, also exist in the paired domain.The typical clinical presentation with amissense mutation in a highly conservedand functionally important region suggeststhere is a reasonable likelihood that thissequence variant is caused by the patient’sphenotype. Our report adds a novel andpotentially structurally significant mutationin the PAX6 gene to the present spectrum ofmutations.

Y Nanjo, S Kawasaki, K Mori, C Sotozono,T Inatomi, S Kinoshita

Kyoto Prefectural University of Medicine, 465 KajiichoKawaramachi Kamigyo-ku Kyoto, 602-0841, Japan

Correspondence to: Ms Yukako Nanjo, KyotoPrefectural University of Medicine, 465 Kajiicho

Kawaramachi Kamigyo-ku Kyoto, 602-0841, Japan;[email protected]

Accepted for publication 2 October 2003

References

1 Hanson IM, Fletcher JM, Jordan T, et al.Mutations at the PAX6 locus are found inheterogeneous anterior segment malformationsincluding Peters’ anomaly. Nature Genet1994;6:168–73.

2 Azuma N, Nishina S, Yanagisawa H, et al. PAX6missense mutation in isolated foveal hypoplasia.Nature Genet 1996;13:141–2.

3 Azuma N, Yamaguchi Y, Handa H, et al.Missense mutation in the alternative splice regionof the PAX6 gene in the eye anomalies. Am J HumGenet 1999;65:656–63.

4 Saiki RK, Gelfand DH, Stoffel S, et al. Primer-directed enzymatic amplification of DNA with athermostable DNA polymerase. Science1998;239:487–91.

5 Epstein JA, Glaser T, Cai J, et al. Twoindependent and interactive DNA-bindingsubdomains of the Pax6 paired domain areregulated by alternative splicing. Genes Dev1994;8:2022–34.

Tuberculous intraocular infectionpresenting with pigmentedhypopyon: a clinicopathologicalcase reportTuberculosis still remains a major cause ofmorbidity and mortality today. Globally theincidence of this disease is increasing by eightmillion new cases annually and is a cause ofdeath for two to three million patients everyyear.1 The ocular manifestations of tubercu-losis are diverse, and depend on the immu-nological, bacteriological, andepidemiological variables.2 Individuals withcompromised immune status usually present

with atypical presentations.3 This clinico-pathological report of a patient treated withimmunosuppressive agents shows intraoculartuberculosis presenting with pigmentedhypopyon.

A 38 year old female patient with a historyof polyarthralgia, anaemia, hypertension, andan impaired renal function with a possibleclinical diagnosis of systemic lupus nephro-pathy underwent renal biopsy, which dis-closed membranous glomerulonephropathywith peripheral granular deposits of IgG,Clq, and IgM on immunofluorescence. Hererythrocyte sedimentation rate was elevated(74 mm in the first hour) and she hadpositive antinuclear antibody; negative rheu-matoid factor, VDRL, HIV, and tuberculinskin test (PPD). She was treated withintravenous cyclophosphamide 1 g per dayonce every month for 3 months and cortico-steroids 30 mg/day. At the time of the thirdintravenous injection of cyclophosphamide,she noticed deterioration of vision in the righteye. On examination, right eye visual acuitywas hand movements close to face. Theconjunctiva was congested and the corneawas oedematous. The anterior chamber wasshallow, and a 3 mm pigmented hypopyonwas noted (fig 1). The left eye was unremark-able and the vision was 6/6. Blood and urinecultures showed no growth, and smears ofanterior chamber fluid were negative forbacteria and fungi. Oral ciprofloxcin(500 mg twice per day) was started inaddition to topical corticosteroids andmydriatrics. A week later, the pigmentedhypopyon had increased to 5 mm; it wasaspirated and submitted for cultures andstaining. Ziehl-Nielsen’s stain revealed sev-eral acid fast bacilli (AFB). The cultures werepositive for AFB and the TuberculosisResearch Centre in Chennai, India, identifiedthe organisms as Mycobacterium tuberculosisbased on pigment production, positive niacin,and catalase test. The patient was re-exam-ined for evidence of systemic tuberculosis.Her PPD was negative and there were noradiological or clinical evidence of extraoculartuberculosis. Despite treatment with fourantituberculous drugs (rifamycin 450 mg,isoniazid 300 mg, ethambutol 800 mg, andpyrazinamide 1500 mg), and oral steroids(20 mg) for her polyarthralgia, the patientdeveloped multiple scleral abscesses and lostthe remaining vision. She underwent enu-cleation of the right eye and was continuedon antituberculous agents for 6 months. Shewas continued on tapering dose of systemiccorticosteroids for 3 months following afourth intravenous cyclophosphamide injec-tion. She was followed for two more yearsand there were no signs of disseminatedtuberculosis during that time.

Figure 2 (A) Pedigree of the patient (arrow). All of the members represented here were examined.Solid symbol indicates Peters’ anomaly and open symbols indicate normal phenotype. Sequenceanalysis revealed a heterozygous mutation in the alternative splicing region in the patient.However, no mutation was found in her parents or elder brother. (B) Direct sequencing of PAX6PCR product from the patient with Peters’ anomaly showed an A/G heterozygosity at the 38thposition which resulted in Q13R in exon 5a (arrow). (B-1) Electropherogram of forward sequenceof PAX6 exon 5a. (B-2) Electropherogram of reverse sequence of PAX6 exon 5a. (C) The result ofSNaPshot method. (C-1) The patient represents double peaks with green corresponding to A andblue corresponding to G, indicating heterozygous mutation at the base position of 38 which isconsistent with sequencing results. (C-2) Normal subjects represent a single peak corresponding toA indicating no mutation at the position. (D) Direct sequencing of PAX6 PCR product from thepatient’s relatives indicates that they do not have the same mutation.

Figure 1 The right eye shows oedematouscornea with presence of pigmented hypopyon.

PostScript 721




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

Histopathological examination of the enu-cleated eye showed infiltration of acuteinflammatory cells and macrophages in theposterior half of the corneal stroma (fig 2).The anterior chamber was filled with pigmentcontaining necrotic cells, macrophages, andproteinaceous exudate. The iris and ciliarybody were necrotic and were infiltrated bypigment laden histiocytes. The sclera revealednecrosis with infiltration of acute inflamma-tory cells. The vitreous cavity containedproteinaceous exudate without significantinflammatory cell infiltration. Acid fast stainsdisclosed an abundance of AFB deep in thecorneal stroma, in the anterior chamberexudates, and in the necrotic iris (fig 2).Histopathological diagnosis was tuberculousnecrotising keratouveitis.

CommentIn this case, the pigmented hypopyon wasmade up of melanophages. Darkly pigmentedhypopyon may appear in eyes harbouringnecrotic uveal melanomas in endogenousendophthalmitis caused by Listeria monocyto-genes and Serratia marcescens.4 5 The cause ofdark hypopyon in the endophthalmitis caseswas assumed to be a dispersion of melaninfrom the necrotic iris.5 The present case alsoshowed necrotic iris and dispensed melaningranules in the anterior chamber, suggestinga common underlying pathology for theformation of pigmented hypopyon. To thebest of our knowledge this is the first knowncase of pigmented hypopyon in a biopsy andculture proved intraocular tuberculosis, andhighlights the need for anterior chamberfluid analysis in arriving at the diagnosis.

The clinical spectrum of ocular tuberculousinfection includes chronic uveitis, interstitialkeratitis, scleritis, sclerouveitis, optic neuritis,choroiditis, retinitis, chorioretinitis, andpanophthalmitis.2 3 6 Hypopyon is rarelynoted in tuberculosis. Ni et al presented casesof intraocular tuberculous with turbid,haemorrhagic, greyish yellow exudate in theanterior chamber in one case, and fibrinoushypopyon in three other cases.7 Hypopyonmay appear in rifabutin treated patients whohad Mycobacterium avium complex infection.8

In all instances, the hypopyon was not darklypigmented. The clinical and histopathologicalfeatures suggest that the ocular infectioncould be endogenous; however, systemicevaluation did not disclose extraocular focus.The presence of large numbers of acid fastorganisms in the histological sections sug-gests that the organisms could be atypicalmycobacteria. However, the cultures showedthat the organisms were Mycobacterium tuber-culosis. Presence of such large numbers of the

organisms in the ocular tissue could be fromtreatment induced immunosuppression.9

AcknowledgementsWe thank the Tubercular Research Center (TRC)

Chennai, Tamilnadu for conducting biochemical

tests for Mycobacterium tuberculosis.

S R RathinamAravind Eye Hospital, Madurai, India

N A RaoA Ray Irvine Ocular Pathology Laboratory, Doheny

Eye Institute, Los Angeles, CA, USA

Correspondence to: S R Rathinam, Aravind EyeHospital and Postgraduate Institute of Ophthalmology,1 Anna Nagar Madurai 625 020, Tamil Nadu, India;

[email protected]

doi: 10.1136/bjo.2003.029124


References

1 Raviglione MC, Snider DE Jr, Kochi A. Globalepidemiology of tuberculosis. Morbidity andmortality of a worldwide epidemic. JAMA1995;273:220–6.

2 Dunn JP, Helm CJ, Davidson PT. Tuberculosis. In:Pepose JS, Holland GN, Willhelmus KR, eds.Ocular infection and immunity. St. Louis: Mosby,1996:1421–9.

3 Helm CJ, Holland GN. Ocular tuberculosis (majorreview). Surv Ophthalmol 1993;38:229–52.

4 Reese AB, Archila EA, Jones IS, et al. Necrosis ofmalignant melanoma of the choroid.Am J Ophthalmol 1970;69:91–104.

5 Eliott D, O’Brien TP, Green WR, et al. Elevatedintraocular pressure, pigment dispersion and darkhypopyon in endogenous endopthalmitis fromListeria monocytogenes. Surv Ophthalmol1992;37:117–24.

6 Biswas J, Madavan HN, Gopal L, et al.Intraocular tuberculosis. Clinicopathologic studyof five cases. Retina 1995;15:461–8.

7 Ni C, Palpale JJ, Robinson JL, et al. Ocular tumorsand other ocular pathology: a Chinese-Americancollaborative Study. Int Ophthalmol Clin1982;22:103–24.

8 Fineman MS, Vander J, Regillo CD, et al.Hypopyon uveitis in immunocompetent patientstreated for Mycobacterium avium complexpulmonary infection with rifabutin. Retina2001;21:531–3.

9 Zamir E, Hudson H, Ober R, et al. Massivemycobacterial choroiditis during HAART: anotherimmune-recovery uveitis. Ophthalmology2002;109:2144–8.

Spontaneous stabilisation ofsymptomatic schisis detachmentsAcquired retinoschisis affects 7% of peopleaged .40 years and is bilateral in 85%.1 2

Although retinoschisis is generally asympto-matic and stable, retinal detachment cansupervene in one of three ways. Two of theseare rare, are associated with posterior vitr-eous detachment (PVD), and justify surgicalcorrection; thus, a retinal tear and detach-ment may originate within non-schitic retinaor open breaks in the inner leaf of a schisismay allow fluid vitreous to be recruited intothe cyst and thence to pass through breaks inthe outer leaf, causing it to separate from theretinal pigment epithelium (RPE) over a widearea. The third mechanism—‘‘schisis detach-ment’’—is quite common (for example, 9%prevalence in Byer’s series2) and reflects theopening of outer leaf breaks alone. No fluidvitreous gains access to the cyst or subretinalspace whether or not inner leaf breaks arepresent. Schisis detachments remain rela-tively localised owing to the viscous natureand limited volume of cyst fluid,3 and theaction of the RPE pump can eventually leadto reattachment of the retina and collapse ofthe cyst. Byer’s natural history studiesresulted in his recommending that surgeryfor schisis detachments be limited to eyeswith symptomatic progression.2 3 We present apatient with sequential bilateral schisisdetachments, each of which was sympto-matic at onset but which settled sponta-neously.

Case reportIn March 2000 a 46 year old man wasreferred with a 2 day history of photopsiaand a disturbance in the upper part of thevisual field in his right eye. Visual acuity was6/5 unaided in each eye, and neither eye hada PVD. Funduscopy revealed a schisis detach-ment inferiorly in the right eye with acurvilinear outer leaf break at the posteriorlimit of the cyst (fig 1A). The detachmentextended midway between the main infer-otemporal branch retinal artery and thefovea. The full thickness of retina delineatingthe superotemporal edge of the outer leafbreak had a sawtooth pattern of outer retinalshagreen, while the free inferonasal edge ofthe break was rolled over. The oedemaresolved within a week and no surgicalintervention was recommended, merelyobservation. A bullous inferotemporal reti-noschisis was also noted in the left eye,extending almost to the major vasculararcade (fig 1B). During 3 years of follow up,a clinically obvious decrease in the height ofthe cyst and spontaneous closure of the outerleaf break were observed in the right eye, RPEatrophy indicating the extent of previousretinal detachment (fig 1C and E). Rightvision has remained 6/5 albeit with visualfield loss superiorly.

In November 2001, photopsia and visualfield disturbances were experienced in theleft eye. Examination revealed a curvilinearouter leaf break at the posterior limit of theinferotemporal cyst (fig 1D). The associatedschisis detachment encroached upon, but didnot involve, the left fovea. Given the patient’shistory, surgery was considered unnecessary.Again the retinoschisis cavity deflated andsubretinal fluid slowly absorbed leaving RPEatrophy in its wake (fig 1F). Left vision hasremained 6/5 but with an absolute super-onasal scotoma.

Figure 2 (A) Haematoxylin and eosin stain reveals necrosis of iris and inflammatory exudates inthe anterior chamber. (B) The anterior chamber exudate displays several melanophages andnecrotic cells (haematoxylin and eosin). (C) Ziehl-Nielsen’s stain shows a myriad of acid fast bacilliin the deep corneal stroma and in the anterior chamber exudates.

Supported in part by Research to Prevent BlindnessInc, New York, NY, USA and core grant EYO3040from the National Institute of Health, Bethesda, MDand Aravind Medical Research Foundation, TamilNadu, India

722 PostScript




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

CommentMost outer leaf breaks develop well withinthe confines of a retinoschisis, and cyst fluidseparates the RPE and outer leaf only in theimmediate vicinity of the breaks.3 However,the giant outer leaf breaks responsible for theschisis detachments in our patient were eachlocated at the posterior limit of a large retinalcyst. It is unsurprising, therefore, that thedetachments progressed beyond the reti-noschisis and were symptomatic.

This is the first report of symptomaticschisis detachments that settled withoutsurgery. We agree with Byer2 3 that theappropriate management for non-progressiveschisis detachments is ‘‘to do nothing,’’ andbelieve this policy can be extended tosymptomatic, inferior schisis detachmentsthat do not involve the fovea. Surgicalintervention, including retinopexy aroundthe breaks, might well have induced sightthreatening complications in our patient2 4 5

while offering no real prospect of a betteroutcome or prognosis.

J Durnian, W PollockDepartment of Ophthalmology, Blackpool Victoria

Hospital, Blackpool, UK

D McLeodAcademic Department of Ophthalmology, Manchester

Royal Eye Hospital, Manchester, UK

Correspondence to: Professor D McLeod, AcademicDepartment of Ophthalmology, Manchester Royal EyeHospital, Oxford Road, Manchester M13 9WH, UK;

[email protected]

doi: 10.1136/bjo.2003.029876


References

1 Byer NE. Clinical study of senile retinoschisis.Arch Ophthalmol 1968;79:36–44.

2 Byer NE. Long term natural history study of senileretinoschisis with implications for management.Ophthalmology 1986;93:1127–37.

3 Byer NE. Perspectives on the management of thecomplications of senile retinoschisis. Eye2002;16:359–64.

4 Sulonen JM, Wells CG, Barricks ME, et al.Degenerative retinoschisis with giant outer layerbreaks and retinal detachment. Am J Ophthalmol1985;99:114–21.

5 Hoerauf H, Joachimmeyer E, Laqua H. Senileschisis detachment with posterior outer layerbreaks. Retina 2001;21:602–12.

Trypan blue: authors’ replyWe would like to thank Dr Rodrigues andcolleagues for bringing up this interestingpoint of what exactly trypan blue stains.1

In our study, immunohistochemistry wasperformed to determine the nature of cellsinvolved in the epiretinal membranes (ERM)—not to determine the presence or absence ofthe ERM. Presence or absence of ERM wasdetermined by examining routinely stainedsections (haematoxylin and eosin, periodicacid Schiff) for cytoplasm/nuclei of epiretinalcell elements. All four of the macular holeinternal limiting membrane (ILM) specimenswere examined in this way.2 Furthermoretrypan blue (in low concentrations) stainsthe anterior lens capsule.3 Since this capsulelacks glia, we do not believe that the evidencesupports the contention of the correspondentsthat the staining of our ILM specimens is dueto undetected ‘‘glial cell elements of the highlycellular ERM’’ rather than ILM.

Clinically two features are observed withthe use of trypan blue. Firstly, the wholeposterior pole that comes into contact withtrypan blue is stained a faint blue in all cases.The staining pattern is diffuse and not pat-chy, suggesting trypan blue staining is indis-criminate of ERM or ILM. Secondly, in casesof macular pucker, the trypan blue stainedERM can be removed separately, leavingintact ILM behind, which can be furtherstained and removed. In cases of macularhole where a clinical ERM is not present, itappears that only the ILM is stained andpeeled. We have harvested these membranesand confirmed that the membranes onlyconsist of ILM and without a secondary ERM.

There is no doubt that trypan blue stainsboth ERM and ILM. We, however, have noknowledge as to what the structural elementsof these membrane that the dye is attachedto. We concede that staining of ILM withtrypan blue can be variable and sometimesrather faint. Since our publication, Perrierand Sebag have also reported their experiencewith trypan blue in staining ILM and ERM.4 5

Although histological findings were not givenin these studies, clinically the authors foundthe dye to be useful in both types ofmembranes. Given the many concernsregarding the use of indocyanine green,6 webelieve it is a positive development that analternative clinically useful dye is available.

K K Li, P Hiscott, D WongThe Royal Liverpool University Hospital, Prescot Street,

Liverpool L7 8XP, UK

Correspondence to: Dr Kenneth K Li, Prince of WalesHospital, Shatin, NT, Hong Kong;

[email protected]


References

1 Rodrigues EB, Meyer CH, Schmidt JC, et al.Trypan blue stains the epiretinal membrane butnot the internal limiting membrane.Br J Ophthalmol 2003;87:1431–2.

2 Li K, Wong D, Hiscott, et al. Trypan blue stainingof internal limiting membrane and epiretinalmembrane during vitrectomy: visual results andhistopathological findings. Br J Ophthalmol2003;87:216–19.

3 Melles GR, de Waard PW, Pameyer JH, et al.Trypan blue capsule staining to visualize thecapsulorhexis in cataract surgery. J CataractRefract Surg 1999;25:7–9.

Figure 1 Retinal photographs; (A), (C), and (E) are from the right eye in March 2000, November2001, and October 2002 respectively; (B), (D), and (F) are from the left eye in March 2000,November 2001, and October 2002 respectively.

MAILBOX

PostScript 723




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

4 Perrier M, Sebag M. Trypan blue-assisted peelingof the internal limiting membrane during macularhole surgery. Am J Ophthalmol2003;135:903–5.

5 Perrier M, Sebag M. Epiretinal membranesurgery assisted by trypan blue. Am J Ophthalmol2003;135:909–11.

6 Gandorfer A, Haritoglou C, Gass CA, et al.Indocyanine green-assisted peeling of the internallimiting membrane may cause retinal damage.Am J Ophthalmol 2001;132:431–3.

Charles Bonnet syndrome andbrimonidine: commentsWe read, with interest the article published inthe BJO by Tomsak et al.1 Interest in theCharles-Bonnet syndrome (CBS) has esca-lated of late, highlighting the probable 15%incidence of the condition in patients withsignificant visual impairment coupled with aclear sensorium.2

The authors implied that CBS was inducedin four patients by brimonidine tartrate onthe basis of patient age and the instigation ofbrimonidine therapy, with discontinuationresulting in eventual resolution of the hallu-cinations. Firstly, the diagnostic criteriaproposed by Gold and Rabins3 and Podell etal,4 quite rightly made no reference to agebeing indicative of CBS, although incidencecertainly increases with age. Schwartz andVahgei5 found that CBS also occurred inchildren following profound visual loss. Thissuggests that the high incidence in the elderlypopulation is possibly attributable to theincreased incidence of acquired visual lossoccurring with age; therefore, age is not acriterion for diagnosis. Further, although theSnellen acuity of all four patients wasreasonably good in at least one eye of eachpatient, it may be surmised that severe visualimpairment may have been due to visual fieldloss secondary to glaucomatous damage.Although this is not clear from the article,the cause of visual impairment and bilater-ality are important in the diagnosis of CBS.Indeed, bilateral advanced visual field defectsinduced by glaucoma and homonymoushemianopia have resulted in CBS.6 7 A pre-vailing theory suggests sensory visual depri-vation as an integral causative factor in CBS.Interestingly, and supportive of this theory,musical pseudohallucinations have beendocumented in cases of acquired deafness.8

Sensory deprivation in the presence of a clearsensorium will be necessary bilaterally toinduce CBS, although no lower limit ofSnellen visual acuity has been defined as alevel for which CBS symptoms are stimu-lated. In the article case 4 seems to havesufficiently adequate visual function in theright eye to justify a definite misdiagnosis ofCBS.

Secondly, as mentioned by the authors, a-2agonists have been shown to cause systemicand neuropsychiatric phenomena.9 As withthe discontinuation of any medication, theexpectation would be resolution of inducedsymptoms, and as such we believe thehallucinations may easily be explained as aside effect of the medication. Brimonidine isa known lipophilic compound able to pene-trate the blood-brain barrier. Through theaccompanying package insert, neurologicalside effects such as depression and dizzinessare well known. There is, therefore, littledoubt that in the aged population in whompharmacokinetics is often unpredictable, thelikelihood of greater systemic absorption anddistribution may well lead to neuropsychia-tric phenomena. Consequently, we believe

that CBS was not the cause of the complexvisual hallucinations experienced by thesepatients but may be attributed to a rarerside effect of brimonidine, which shouldnow be included in the patient informationleaflet.

I Rahman, B Fernando, M HarrisonManchester Royal Eye Hospital, Lister Centre, Nelson

Street, Manchester M1, UK

Correspondence to: Mr Imran Rahman, ManchesterRoyal Eye Hospital, Lister Centre, Nelson Street,

Manchester M1, UK; [email protected]


References

1 Tomsak RL, Zaret CR, Weidenthal D. CharlesBonnet syndrome precipitated by brimonidinetartrate eye drops. Br J Ophthalmol2003;87:917.

2 Menon GJ, Rahman I, Menon SJ, et al. Complexvisual hallucinations in the visually impaired: theCharles Bonnet syndrome. Surv Ophthalmol2003;48:58–72 (major review).

3 Gold K, Rabins PV. Isolated visual hallucinationsand the Charles- Bonnet syndrome: a review ofthe literature and presentation of six cases. ComprPsychiatry 1989;30:90–8.

4 Podoll K, Osterheider M, Noth J. Das CharlesBonnet syndrom. Fortschr Neurol Psychiat1989;57:43–60.

5 Schwartz TL, Vahgei L. Charles Bonnet syndromein children. J AAPOS 1998;2:310–3.

6 Damas-Mora J, Skelton-Robinson M, Jenner FA.The Charles Bonnet syndrome in perspective.Psychol Med 1982;12:251–61.

7 Dodd J, Heffernan A, Blake J. Visualhallucinations associated with Charles Bonnetsyndrome—an ever increasing diagnosis. Ir Med J1999;92:344–5.

8 Griffiths TD. Musical hallucinosis in acquireddeafness: phenomenology and brain substrate.Brain 2000;123:2065–76.

9 Kim DD, Bay G. A case of suspected alphagan-induced psychosis. Arch Ophthalmol2000;118:1132–3.

Periocular corticosteroid therapy:commentsI read with great interest the article by Okadaet al,1 reporting the efficacy and complicationsof trans-Tenon’s retrobulbar infusion oftriamcinolone acetonide for posterior uveiticinflammation. The authors have to be com-mended for the excellent description of thisnovel technique.

The efficacy of various methods of corti-costeroid injection has always been a matterof debate with different studies giving differ-ent results. McCartney et al2 showed that themajor route of penetration of steroids aftersubconjunctival injection was directlythrough the adjacent sclera, choroid, andretina. In addition, the authors describedmethods to inject steroids in the sub-Tenon’sspace and concluded that the injectionsshould be placed immediately adjacent tothe site of intraocular inflammation that wasunder treatment. In contrast, in a study onrabbit eyes, Wilson et al3 have elegantlydemonstrated that injection of corticosteroidsby the sub-Tenon’s route does not show asignificant effect on the blood-retinal barrierowing to inadequate penetration. The authorsanalysed the severity of blood-retinal barrierbreakdown following panretinal photocoagu-lation, using rapid sequential magnetic reso-nance image with contrast. Of note, in thisstudy the authors have taken particular careto ensure the accurate placement of the

needle in the sub-Tenon’s space. A similarresult was obtained in a study Jennings et al4

on the efficacy of posterior sub-Tenon’sinjections in patients with cystoid macularoedema caused by uveitis. The authors foundthat the injection of steroids by the sub-Tenon’s route did not consistently affect theblood-retinal barrier permeability in suchpatients and that there was no diffusion ofthe steroids into the eye in therapeuticallymeaningful concentrations. This is of parti-cular concern since it is the breakdown in theblood-retinal barrier that leads to influx ofserum/serum components leading to macularoedema, epiretinal membrane, and othersequelae.

Sub-Tenon’s injections when compared tointravitreal injections have the disadvantageof probably a decreased and difficult drugpenetration through the sclera and choroidand a rapid removal of the drug by thechoroidal circulation after penetration withthe resultant shortened duration of action.This is probably the reason why the sub-Tenon’s route of injection of steroids has notbecome popular in diabetic macular oedemain contrast with the increased popularity ofintravitreal steroid injections.

Interestingly, Freeman et al5 have postu-lated that the lack of therapeutic response tosub-Tenon’s corticosteroids may be becauseof placement at a site relatively far from thetarget zone. They determined the location ofrepository corticosteroid after sub-Tenon’sinjection by echography and showed thatthe steroid was deposited within the sub-Tenon’s space over the macula in only 11 of24 cases. They hence concluded that thetherapeutic response manifested by improve-ment in macular function may be related tothe proximity of the corticosteroid to themacular area. The impressive efficacyreported in the study by Okada et al1 couldprobably be the result of reliable drugplacement thanks to the visual confirmationof cannula entry into the sub-Tenon’s space,as the authors speculate. However, it isimportant to note that most of the patientsin this study continued to receive topicalsteroid drops. Whether these drops had anadditive effect is unclear.

It would probably be worthwhile to con-sider a planned, primary intravitreal injectionof corticosteroids under aseptic conditionsthat has the distinct advantage of gettingdistributed into a much larger volume forselected conditions. There would be no casesof ‘‘therapeutic failures’’ that are seen afterinjection of steroids into the sub-Tenon’sspace and the resultant confusion as towhether the unsatisfactory response is sec-ondary to the disease process or failure toinject the steroid into the sub-Tenon’s spaceor the debated lower efficacy of this route ofinjection. The procedure is simpler than thedescribed trans-Tenon’s retrobulbar infusion(no special cannula is required), but the riskof endophthalmitis is daunting.6

V VedanthamAravind Eye Hospital and Postgraduate Institute of

Ophthalmology, 1 Anna Nagar, Madurai, TamilNadu, India; [email protected]


References

1 Okada AA, Wakabayashi T, Morimura Y, et al.Trans-Tenon’s retrobulbar triamcinolone infusionfor the treatment of uveitis. Br J Ophthalmol2003;87:968–71.

724 PostScript




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

2 McCartney HJ, Drysdale IO, Gornall AG, et al.An autoradiographic study of the penetration ofsubconjunctivally injected hydrocortisone intonormal and inflamed rabbit eyes. InvestOphthalmol 1965;4:297.

3 Wilson CA, Berkowitz BA, Sato Y, et al. Treatmentwith intravitreal steroid reduces blood-retinalbarrier breakdown due to retinalphotocoagulation. Arch Ophthalmol1992;110:1155–9.

4 Jennings T, Rusin MM, Tessler HH, et al. Posteriorsub-Tenon’s injections of corticosteroids in uveitispatients with cystoid macular edema.Jpn J Ophthalmol 1988;32:385–91.

5 Freeman WR, Green RL, Smith RE. Echographiclocalization of corticosteroids after periocularinjection. Am J Ophthalmol 1987;103(Pt1):281–8.

6 Benz MS, Murray TG, Dubovy SR, et al.Endophthalmitis caused by Mycobacteriumchelonae abscessus after intravitreal injection oftriamcinolone. Arch Ophthalmol2003;121:271–3.

Transcaruncular approach for themanagement of frontoethmoidalmucoceles: a commentWe read the article by Lai et al1 with interest.

The authors report a modification of thenon-obliterative external procedure that wasfirst described by Lynch in 1921.2 The Lynch-Howarth procedure2–4 involved transnasalstenting to prevent medial-ward collapse ofthe orbit obstructing drainage from thefrontal sinus into the nose. Although thetranscaruncular procedure uses a differentexternal approach, it nevertheless ofteninvolves removal of part of the laminapapyracea for access to the sinuses. Hence,as with the Lynch approach, prolapse oforbital contents into the defect may occur,increasing the risk of re-stenosis. In addition,the cells in the frontal recess are not formallycleared and thus drainage into the nasalcavity is not assured. Stenting of sinusopenings results in a significant fibroticreaction in a proportion of patients, andclosure of such a previously stented openingis likely. Furthermore, the follow up period inthis study is too short to confirm the successor failure of this technique as recurrenceoften takes years to manifest.4

Endoscopic management of mucocelesprotruding into the other sinuses or nasalcavity has been an accepted treatment foryears.5–9 Frontoethmoidal mucoceles are typi-cal of such mucoceles where the bony wallsurrounding the mucocele is thin and there-fore easily accessible transnasally. The endo-scopic procedure creates a large area clear ofcells which allows the greatest possiblemarsupialisation of the mucocele. No stent-ing is required. Har-El9 reported the largestseries of 108 mucoceles with a median follow

up of 4.7 years with a recurrence rate of only0.9%. Therefore, we would recommend anendoscopic approach for frontoethmoidalmucoceles as the integrity of the laminapapyracea is maintained and the largestpossible opening is created into the mucocele,which in turn minimises the chances ofrecurrence.

J J Khong, P Wormald, D SelvaRoyal Adelaide Hospital, North Terrace, Adelaide,

Australia

Correspondence to: Dr Jwu Jin Khong, Royal AdelaideHospital, North Terrace, Adelaide, Australia;

jjkhong@yahoo,com


References

1 Lai PC, Liao SL, Jou JR, et al. Transcaruncularapproach for the management of frontoethmoidmucoceles. Br J Ophthalmol 2003;87:699–703.

2 Lynch RC. The technique of a radical frontal sinusoperation which has given me the best results.Laryngoscope 1921;31:1–5.

3 Ritter FN, ed. Surgical procedures on theparanasal sinuses: the frontal sinus. In: Theparanasal sinuses: surgery and technique, 2nded. St Louis: Mosby, 1978:136–45.

4 Neel HB, McDonald TJ, Facer GW. ModifiedLynch procedure for chronic frontal sinusdiseases: rationale, technique, and long-termresults. Laryngoscope 1987;97:1274–9.

5 Kennedy DW, Josephson JS, Zinreich SJ, et al.Endoscopic sinus surgery for mucoceles: a viablealternative. Laryngoscope 1989;99:885–95.

6 Schaefer SD, Close LG. Endoscopic managementof frontal sinus disease. Laryngoscope1990;100:155–60.

7 Har-EL G, Balwally AN, Lucente FE. Sinusmucoceles: is marsupialization enough?Otolaryngol Head Neck Surg 1997;117:633–40.

8 Lund VJ. Endoscopic management of paranasalsinus mucocoeles. J Laryngol Otol1998;112:36–40.

9 Har-El G. Endoscopic management of 108 sinusmucoceles. Laryngoscope 2001;111:2131–4.

NOTICES

Cataract surgeryThe latest issue of Community Eye Health (No48) discusses a solution to reduce worldwidecataract blindness, including sutureless non-phaco cataract surgery. For further informa-tion please contact: Journal of CommunityEye Health, International Resource Centre,International Centre for Eye Health, Depart-ment of Infectious and Tropical Diseases,London School of Hygiene and Tropical Medi-cine, Keppel Street, London WC1E 7HT, UK

(tel: +44 (0)20 7612 7964; email: [email protected]; website: www.jceh.co.uk). Annual subscription (4 issues) UK£28/US$45. Free to developing country applicants.

Elimination of avoidable blindnessThe 56th World Health Assembly (WHA)considered the report on the elimination ofavoidable blindness (doc A56/26) and urgedMember States to: (1) Commit themselves tosupporting the Global Initiative for the Elimi-nation of Avoidable Blindness by setting up anational Vision 2020 plan by 2005; (2) Esta-blish a national coordinating committee forVision 2020, or a national blindness preven-tion committee to help implement the plan;(3) Implement the plan by 2007; (4) Includeeffective monitoring and evaluation of theplan with the aim of showing a reduction inthe magnitude of avoidable blindness by2010; (5) To support the mobilisation ofresources for eliminating avoidable blind-ness. The WHA also urged the Director-General to maintain and strengthen WHO’scollaboration with Member States and thepartners of the Global Initiative for the Elimi-nation of Avoidable Blindness as well as aidin the coordination and support of nationalcapability.

XVth Meeting of the InternationalNeuro-Ophthalmology SocietyThe XVth Meeting of the InternationalNeuro-Ophthalmology Society will take place18–22 July 2004, in Geneva, Switzerland.Further details: Prof. A Safran, UniversityHospital Geneva, c/o SYMPORG SA, Geneva(fax: +4122 839 8484; email: [email protected]; website: www.symporg.ch).

4th International Congress onAutoimmunityThe 4th International Congress on Auto-immunity will take place 3–7 November2004 in Budapest, Hungary. The deadlinefor the receipt of abstracts is 20 June 2004.Further details: Kenes International GlobalCongress Organisers and Association Mana-gement Services, 17 Rue du Cendrier, PO Box1726, CH-1211 Geneva 1, Switzerland (tel:+41 22 908 0488; fax: +41 22 732 2850; email:[email protected]; website: www.kenes.com/autoim2004).

XVI International Congress for EyeResearchThe XVI International Congress for EyeResearch will be held on 29 August – 3 Sep-tember 2004 in Sydney, Australia. For furtherinformation, please contact: [email protected] (website: www.tourhosts.com.au/icer2004).

PostScript 725




j.com/

Br J O


pril 2004. Dow

nloaded from

http://bjo.bmj.com/

postscript - bjo.bmj.com · bicarbonate producing a metabolic acidosis— ... physiological stress...

Documents