SURVEY OF OPHTHALMOLOGY VOLUME 55 � NUMBER 2 � MARCH–APRIL 2010

CLINICAL CHALLENGESPETER SAVINO AND HELEN DANESH-MEYER, EDITORS

Glazed (Vision) and ConfusedHeather E. Moss, MD, PhD,1 Grant T. Liu, MD,1 and Josep Dalmau, MD, PhD2

1Division of Neuro-Ophthalmology, Departments of Neurology and Ophthalmology; and 2Department of Neurology,the University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA

� 201All rig

(In keeping with the format of a clinical pathologic conference,the abstract and key words appear at the end of this article.)

Case Report

A 60-year-old man presented for neuro-ophthal-mic consultation for progressive blurred vision over2 weeks associated with black spots in all visualfields. He denied double vision or positive visualphenomena. He had had a normal screeningophthalmologic examination 4 months before.

He had been well until 3 months earlier when hehad flu-like symptoms of fever, fatigue, and malaisefor several days followed by the onset of severefrontal headaches without phonophobia or photo-phobia. One month after the headaches started hedeveloped depressed mood and anxiety. Threemonths after the headaches started he becameconfused and forgetful. He forgot the code for hishouse alarm and how to use a cellular phone. Hehad an episode when he awoke confused and thenhad brief loss of consciousness with his eyes rolledback. He had no recollection of this event. Thevisual symptoms started 1 week later. Review ofsystems included a 45-pound weight loss attributedto medication, occasional dizziness, diffuse pain,and trouble walking due to hip pain.

He was evaluated by a neurologist, endocrinolo-gist, otorhinolaryngologist, and psychologist prior

169

0 by Elsevier Inc.hts reserved.

to his neuro-ophthalmology referral. He was admit-ted to the hospital twice: once for evaluation ofheadache and once after the episode of losingconsciousness.

Laboratory testing during his illness includednormal complete blood count, basic metabolicpanel, liver studies, B12, folate, urinalysis, fastinglipid panel, and iron studies. Lead and arsenic wereundetectable. Thyroid studies were normal. Testos-terone was 132 (normal 241--827). Lyme and rapidplasma reagin (RPR) testing were negative. Lumbarpuncture 1 month after the start of headaches didnot provide any relief. Cerebrospinal fluid (CSF)examination revealed 0 white blood cells(wbc)/mL,40 red blood cells(rbc)/mL, 76 mg/dL protein, 64mg/dL glucose, negative venereal disease researchlaboratory, no cryptococcal antigen, and no malig-nant cells. Electroencephalogram (EEG) performedafter his transient loss of consciousness did notreveal epileptiform activity. Echocardiogram wasunremarkable.

Magnetic resonance imaging (MRI) of the braintwo weeks after the onset of headaches demon-strated an opacified left frontal sinus and noparenchymal abnormalities. Repeat MRI of the

0039-6257/10/$--see front matterdoi:10.1016/j.survophthal.2009.03.008

Fig. 1. Bilateral optic disk swelling; the vitreous haze accounts for the poor view.

170 Surv Ophthalmol 55 (2) March--April 2010 MOSS ET AL

brain three months into his symptoms was un-changed. Computed tomography (CT) angiogramof the head was unremarkable 3 months after theonset of headaches. Chest X-ray showed no acutedisease and was stable when compared with a studyfrom 6 years prior.

He was started on testosterone gel for lowtestosterone 2 months into his illness without anyimprovement. Therapeutic trials for his headachesand psychiatric symptoms included antibiotics, non-steroidal anti-inflammatory agents, metoclopromide,topiramate, valproic acid, methylprednisolone, oxy-codone, clonazepam, and escitalopram. None pro-vided significant symptom relief.

He had no known drug allergies. Medicationsincluded levothyroxine, testosterone gel, flutica-sone/salmeterol inhaler, fluticasone nasal spray,montelukast, clonazepam, escitalopram, multivita-mins, folic acid, B-vitamin complex, fish oil, andibuprofen as needed. Past medical history includeda benign spinal cord tumor that was removed inJune 2003, hypothyroidism, and sinus surgery.

He was married. He was self-employed in salesuntil 3 months into his illness when he becameunable to work due to his cognitive impairments. Hehad a 25-pack an year smoking history. He quitsmoking 14 years prior to presentation. He deniedillicit drug use. Family history was remarkable forcardiac disease in both parents who died at 78 and86 years of age and an aunt with epilepsy.

On examination he was normotensive weighing228 pounds. Visual acuity was 20/80 �1 ODimproving to 20/60 �1 with pinhole and 20/50 �2OS improving to 20/30�2 with pinhole. He saw only2 of 8 Ishihara color plates with the right eye, and 6of 8 with the left. Visual fields to confrontation,pupils, eyelids, and eye movements were all normal.Slit-lamp examination showed no cells in the anterior

chamber. He had bilateral optic disk swelling andbilateral vitreous haze due to vitritis (Fig. 1).

On neurological examination he was agitated, withdecreased attention, poor effort, and labile affect. Hewas not oriented to city or year, he did not know thepresident, and he could not recall three objects after5 minutes. Spontaneous speech was perseverative andcontained many curses. He could not follow embed-ded commands. Naming and repetition were intact.He had normal strength, sensation, coordination,gait, and deep tendon reflexes.

What are the diagnostic possibilities?

How would you proceed?

Comments

COMMENTS BY JOSEP DALMAU, MD, PHD

This is the case of a middle-aged man withsubacute cognitive decline, visual changes, and lossof conciousness starting after a febrile illness. Basedon this history a central nervous system (CNS)process with brain parenchymal and/or leptome-ningeal involvement causing functional impairmentand possibly seizures is suggested. A negative work-up at this juncture should not distract from thislocalization. A normal interictal EEG does not ruleout seizures unless an event is captured andparenchymal inflammation causing clinical neuro-logical compromise may be below the resolution ofMRI and CSF sampling, particularly early in thepresentation of many leptomeningeal processes.

Visual acuity changes and color vision loss suggestan optic nerve process with or without involvement ofthe retina; identification of disk swelling confirmedinvolvement of the optic nerve. The vitritis implies aninflammatory, infectious, or neoplastic etiology. The

GLAZED AND CONFUSED 171

differential diagnosis of a combined CNS and opticnerve process with vitritis includes CNS lymphoma,herpes simplex virus, syphilis, vasculitis, sarcoidosis,and paraneoplastic syndromes.

Considerations for further work-up include re-peating the lumbar puncture and MRI brainimaging as well as serological testing for markersof inflammation and paraneoplastic antibodies.Vitritis offers the possibility of tissue diagnosis withminimal morbidity.

Fig. 2. Axial FLAIR image from brain MRI showing T2hyperintensity in bilateral hippocampi (left greater thanright).

Case Report (Continued)

The patient was referred to a retinal specialist,who confirmed the fundus findings. Fluoresceinangiogram did not reveal vasculitis or retinitis. Diskstaining was consistent with edema. A vitreousbiopsy OD showed mixed T and B lymphocytes withpossible aberrant expression of CD 43, consistentwith a B-cell proliferative disorder.

During the week following his initial evaluationthe patient clinically declined with worseningvision, more confusion, and severe nausea andvomiting. He started complaining of episodicstrange odors. He was admitted to the hospital forfurther work-up.

Laboratory data included normal complete bloodcount, metabolic profile, liver function tests, eryth-rocyte sedimentation rate, c-reactive protein, andB12. Thyroid stimulating hormone and free T4 werenormal, anti thyroglobulin antibodies were elevatedat 641 units/mL, and antithyroperoxidase antibodies(anti-TPO) were elevated atO3000 units/mL. Anti-nuclear antibodies were positive at 1:1280 ina pattern resembling perinuclear anti-neutrophilcytoplasmic antibody although lacked specificity formyeloperoxidase on further testing. Sjogren anti-bodies (a and b) were negative. Serum proteinelectropheresis showed no evidence of a paraprotein.Human immunodeficiency virus antibody testing wasnegative. Lumbar puncture had an opening pressureof 16.8 cm H2O, and the CSF had 17 wbc/mL, 1 rbc/mL, 55 mg/dL glucose, 110 mg/dL protein, nomalignant cells on cytology, and no monoclonalpopulation on flow cytometry.

EEG showed diffuse slowing without epileptiformactivity.

Brain MRI showed bilateral (left greater than right)hippocampal T2 hyperintensity with mild associatedrestricted diffusion and no enhancement (Fig. 2).These findings were not seen on direct comparisonwith the MRI performed 2 weeks earlier.

Serological testing for paraneoplastic antibodiesrevealed the presence of antibodies to collapsin-response mediator protein 5 (CRMP5).

What is your most likely diagnosis?

What are the next therapeutic and diagnostic steps?

Comments (Continued)

COMMENTS BY DR. DALMAU

Although the identification of a mixed popula-tion of T- and B-cells with possible expression ofCD43 in the vitreous biopsy may have suggested anocular lymphoma with CNS involvement, none ofthe subsequent studies, including CSF analysis andflow cytometry, supported this diagnosis. Moreover,expression of CD43 has been reported in non-neoplastic inflammatory disorders. Detection ofthyroglobulin and TPO antibodies is often iden-tified in patients with Hashimoto thyroiditis(a frequent cause of hypothyroidism), other thyroidand autoimmune disorders, as well as in a smallnumber of normal individuals. Our patient did nothave thyroid dysfunction, and the ophthalmologicand MRI findings did not suggest Hashimotoencephalitis, an ill-defined steroid-responsive en-cephalitis that occurs in association with thyroglob-ulin and thyroperoxidase antibodies. In contrast,the asymmetric T2 hyperintensity involving themedial temporal lobes in association with lympho-cytic pleocytosis in the CSF and rapidly progressivebehavioral change and memory deficits are highlysuggestive of limbic encephalitis.6 This disorder is

Fig. 3. CT of the chest showing enlarged lymph nodes(closed arrows) and tumor (open arrow).

Fig. 4. Axial FLAIR image from brain MRI 5 monthsfollowing diagnosis showing persistent, though improved,hippocampal signal.

172 Surv Ophthalmol 55 (2) March--April 2010 MOSS ET AL

frequently a paraneoplastic manifestation of anoccult neoplasm, which, in a patient with historyof smoking, should raise the suspicion of lungcancer. The detection of antibodies to CRMP5 (alsocalled CV2) strongly supports the paraneoplasticetiology of this disorder and enhances the suspicionfor lung cancer.4,8 Therefore, the next step would beto perform a comprehensive tumor search using CTand positron emission tomography (PET) studies.

Case Report (Continued)

CT of the chest revealed a 2-cm left upper lobelung mass and multiple enlarged (2--4 cm) lymphnodes in the hilar, mediastinal, and paratrachealregions (Fig. 3). CT of the abdomen and pelvis didnot show lymphadenopathy or other masses.

Paratracheal needle biopsy revealed small-celllung cancer (SCLC). Bone scan and PET scan didnot reveal metastatic disease. He was diagnosed withparaneoplastic limbic encephalitis, optic neuritis,and vitritis.

Comments (Continued)

COMMENTS BY DR. DALMAU

Antibodies to CRMP5 have been identified inpatients with paraneoplastic neurological and visualsymptoms associated with SCLC, thymoma, uterinesarcoma, thyroid papillary carcinoma, and renal cellcarcinoma.1,5,8 The most common symptoms arecognitive dysfunction, memory loss, and cerebellar

ataxia; however, these symptoms do not help tonarrow the differential diagnosis as they areassociated with many other neurologic and para-neoplastic syndromes. The clinical features areheterogeneous, with optic neuritis,3 vitritis, chorea,7

and sensorimotor axonal neuropathy1 being themost helpful in distinguishing it from other para-neoplastic syndromes. In addition to CRMP5 anti-bodies in the serum and CSF, the most frequentlaboratory findings are CSF lymphocytic pleocytosisand elevated protein concentration. Less frequentfindings include an abnormal brain or spinal cordMRI and an abnormal EEG. The underlying tumoris frequently identified at the time of neurologicalsymptom presentation.5

CRMP5 immunoreactivity has been demonstratedin small-cell lung carcinoma lines as well as in thecytoplasm of neurons,1 including those in retina,optic nerve,3 and peripheral nerves,1 supporting thehypothesis of CRMP5 as the responsible pathogenicantigen. However, given the intracellular location ofthe antigen it is unlikely that the antibodies arepathogenic. Rather, the accompanying cytotoxicT-cell immune response is likely the effector of theneuronal damage. This is suggested by the frequentinfiltrates of T-cells in affected optic nerves, mesialtemporal lobes, cerebellum, brain stem, spinal cord,and peripheral nerves.3,8 It is important to note thatantibodies to CRMP5 may occur in 10% of patientswith SCLC without paraneoplastic syndromes, and

GLAZED AND CONFUSED 173

are also frequently detected in association withother paraneoplastic antibodies, such as anti-Hu orZic4.2 Neither of these 2 antibodies were identifiedin our patient.

The therapeutic approach is aimed at immuno-suppression with high-dose steroids and treatmentof the underlying tumor. As with other paraneo-plastic syndromes associated with intracellular anti-gens (e.g., anti-Hu) response to therapy is oftenlimited to stabilization of neurological symptoms.6

Case Report (Concluded)

The patient was treated with high-dose intrave-nous corticosteroids to reduce inflammation fromhis paraneoplastic syndrome. Seven days later hisvision was 20/25 OU, he identified 6 of 6 colorplates OU, and the vitreous haze and disk swellingwere resolving. However, his cognitive deficitspersisted. He was treated empirically for presumedtemporal lobe seizures with antiepileptic medica-tions. Although the initial EEG was not epilepti-form, future studies revealed electrographictemporal lobe seizures. Agitation and hypersexualbehaviors, likely due to bilateral temporal lobeinjury, were managed with benzodiazepines, seroto-nin reuptake inhibitors, and mood stabilizingagents. He received radiation and chemotherapywith cisplatin and etoposide for his lung cancer.

Twelve months after evaluation (16 months aftersymptom onset) his visual acuity was 20/40 pinholeOD and 20/30 OS. He had no difficulties with colorperception. He had cataracts bilaterally, right morethan left. His optic disks and retinas were normal.He remained emotionally labile, but had improvedattention, was oriented to month and place andcould recall one object at 3 minutes. The abnormalsignal on brain MRI had improved (Fig. 4). Thelung mass and lymphadenopathy also had improvedradiographically.

Comments (Concluded)

COMMENTS BY DR. DALMAU

This case illustrates that a prompt diagnosiscoupled with aggressive immunotherapy and onco-logical therapy can modify the outcome of paraneo-plastic syndromes that are often considered poorlyresponsive or refractory to treatment.

References

1. Antoine JC, Honnorat J, Camdessanche JP, et al. Paraneo-plastic anti-CV2 antibodies react with peripheral nerve andare associated with a mixed axonal and demyelinatingperipheral neuropathy. Ann Neurol. 2001;49:214--21

2. Bataller L, Wade DF, Graus F, et al. Antibodies to Zic4 inparaneoplastic neurologic disorders and small-cell lungcancer. Neurology. 2004;62:778--82

3. Cross SA, Slomao DR, Parisi JE, et al. Paraneoplasticautoimmune optic neuritis with retinitis defined by CRMP-5-IgG. Ann Neurol. 2003;54:38--50

4. Honnorat J, Antoine JC, Derrington E, et al. Antibodies toa subpopulation of glial cells and a 66 kDa developmentalprotein in patients with paraneoplastic neurological syndromes.J Neurol Neurosurg Psychiat. 1996;61:270--8

5. Honnorat J, Cartalat-Carel S, Ricard D, et al. Onco-neuralantibodies and tumor type determine survival and neurolog-ical symptoms in paraneoplastic neurological syndromes withHu or CV2/CRMP5 antibodies. J Neurol Neurosurg Psychiat.2009;80:412--6

6. Tuzun E, Dalmau J. Limbic encephalitis and variants:classification, diagnosis and treatment. Neurologist. 2007;13:261--71

7. Vernino S, Tuite P, Adler CH, et al. Paraneoplastic Choreaassociated with CRMP-5 Neuronal antibody and lung carci-noma. Ann Neurol. 2002;51:625--30

8. Yu Z, Kryzer TJ, Griesmann GE, et al. CRMP--5 neuronalautoantibody: marker of lung cancer and thymoma-relatedautoimmunity. Ann Neurol. 2001;49:146--54

The authors wish to thank Dr. Darma Ie for performing theretina evaluation, biopsying the vitreous, and providing thefundus photos. The authors reported no proprietary or commer-cial interest in any product mentioned or concept discussed inthis article.

Reprint address: Heather E. Moss, MD, PhD, Department ofNeurology, Hospital of the University of Pennsylvania, 3400Spruce St., Philadelphia, PA 19104. e-mail: Heather.Moss@uphs.upenn.edu.

Abstract. A 60-year-old man presented with vitritis and optic neuropathy in the setting of headachesand behavioral changes. MRI brain revealed bilateral temporal lobe inflammation consistent withlimbic encephalitis. He was subsequently diagnosed with small cell lung cancer with a paraneoplasticsyndrome characterized by CRMP5 IgG as a cause of his symptoms. His visual symptoms improvedmarkedly after anti-inflammatory therapy and his cognitive symptoms were mildly better followingsystemic chemotherapy. The clinical presentation, pathophysiology, and therapy of CRMP5 associatedparaneoplastic syndromes are discussed. (Surv Ophthalmol 55:169--173, 2010. � 2010 Elsevier Inc.All rights reserved.)

Key words. CRMP5 � limbic encephalitis � optic neuropathy � paraneoplastic syndrome� vitritis