Melanoma-Associated Retinopathy a Paraneoplastic Autoimmune Complication

CLINICAL SCIENCES Melanoma-Associated Retinopathy A Paraneoplastic Autoimmune Complication

Ying Lu, MD, PhD; Lin Jia, MS; Shirley He, MD, MS; Mary C. Hurley, MS; Monique J. Leys, MD; Thiran Jayasundera, MD; John R. Heckenlively, MD

Objectives: To study 11 patients with melanoma- patients (9 of 11) had a strong family history of autoim- associated retinopathy (MAR) to clarify the reliability of mune disorders. Any type of melanoma (cutaneous, cho- various methods of diagnostic testing, to determine the roidal, ciliary body, or choroidal nevi) may be associ- underlying antigenic retinal proteins, and to study the ated with this paraneoplastic autoimmune reactivity. MAR clinical histories and types of associated melanomas. may precede or follow the diagnosis of melanoma. Pa- tients with MAR have the same antigenic retinal pro- Methods: Clinical data were obtained from patients with teins that have been associated with cancer-associated reti- melanoma who developed marked visual problems. Test- nopathy. In addition, 2 new antigenic retinal proteins, ing included electroretinography, kinetic visual fields, aldolase A and aldolase C, were found. comparative studies of Western blots, and indirect immunohistologic examination to detect antiretinal anti- Conclusions: There was a high prevalence of positive bodies, as well as proteomic studies to identify underly- family histories of autoimmune disease in patients with ing antigenic retinal proteins. MAR. To confirm the disorder, multiple clinical and serum diagnostic techniques (Western blot or indirect im- Results: Patients with MAR typically have rapid onset of photopsias, scotomata, and loss of central or paracen- munohistologic examination) are needed. Two newly ob- tral vision. Ophthalmoscopy seldom shows significant served antigenic retinal proteins, aldolase A and aldolase changes early, but electroretinograms are abnormal. Re- C, are associated with MAR. sults of Western blots and immunohistologic examination can show antiretinal antibodies but not always. Most Arch Ophthalmol. 2009;127(12):1572-1580

UTOIMMUNE RETINOPATHY with carcinomas,1 melanomas,8 tetrato- (AIR) was first recognized mas,2 and even lymphomas.9 Most cases of as a paraneoplastic disor- AIR occur in patients without a history of der as early as 1976 in pa- tumors, but occasionally a history of head tients with carcinoma,1 and trauma or preceding intraocular inflamma- subsequentA studies in the literature showed tory disease may be found.10-12 Investiga- that AIR could occur with benign2 and ma- tions of cancer-associated retinopathy lignant3 tumors. Melanoma-associated reti- (CAR) have found that the tumors in these nopathy (MAR) was initially reported by patients aberrantly express proteins nor- Gass4 in an atypical case with vitellirup- mally exclusive to retinal tissue, leading to tivelike yellow retinal lesions, but most MAR the production of antibodies directed cases have diffuse retinal atrophy without against these retinal antigens.13 There are pigment deposits. A broader view of 51 cases no reports in which primary melanomas 5 Author Affiliations: of MAR was published by Keltner et al sum- have been examined to investigate the pres- Department of Ophthalmology marizing many of the features and some of ence of retinal proteins, although mela- and Visual Sciences, Kellogg the controversies. Except for 2 case re- noma cell cultures have demonstrated their Eye Center (Drs Lu, He, ports of choroidal melanoma,6,7 only MAR presence.14 The antiretinal antibodies in pa- Jayasundera, and Heckenlively associated with cutaneous melanomas has tients with MAR are associated with pro- and Ms Jia), and Michigan been reported.5 gressive panretinal degeneration that fre- Proteome Consortium Paraneoplastic development of antireti- quently results in legal blindness.15 Patients’ (Ms Hurley), University of Michigan, Ann Arbor; and nal antibodies with resulting retinal dis- early symptoms typically manifest as Department of Ophthalmology, ease is an intriguing pathologic process and sparkles and shimmers (photopsias) and 16 University of West Virginia, remains poorly understood. There have blind spots in their vision. On clinical ex- Morgantown (Dr Leys). been reports of retinopathies associated amination, early stages of retinopathy may

(REPRINTED) ARCH OPHTHALMOL / VOL 127 (NO. 12), DEC 2009 WWW.ARCHOPHTHALMOL.COM 1572

©2009 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 be difficult to see clinically, but an electroretinogram (ERG) mobilized pH gradient buffer). The solubilized protein was loaded (a standardized evoked response that measures photore- overnight onto 11-cm (pH 3-10) immobilized pH gradient strips ceptor and inner retinal responses) will clearly show reti- using active rehydration at 50 V. Isoelectrofocusing was per- nal dysfunction and aids in diagnosis of this condition with formed at 40 000 V/h using isoelectric focusing Cell (BioRad In- otherwise minimal retinal changes.17 dustries, Hercules, California). The focused IPG strips were equili- brated in buffers containing sodium dodecyl sulfate (SDS)-Tris Conventional confirmatory evidence for MAR has in- (2-carboxyethyl)phosphine hydrochloride and SDS- cluded examination of serum for antiretinal antibodies iodoacetamide. The second-dimension gels were run on 4% to by Western blot or measurement of antibody staining of 12% acrylamide Bis-Tris gels (BioRad Industries). After electro- donor retina, particularly bipolar cells, by indirect im- phoresis, duplicate gels were inmmunoblotted or stained with munohistologic studies.18 However, there has been con- SYPRO Ruby (Molecular Probes, Eugene, Oregon). troversy about the best diagnostic tests, with data sug- Separated proteins were transferred onto nitrocellulose mem- gesting that Western blots may be unreliable for detecting branes and incubated with serum from patients having MAR antiretinal antibodies in patients with MAR.8 Milam et and with horseradish peroxidase–conjugated mouse antihu- al8 recommended indirect immunohistologic investiga- man immunoglobulin as a secondary antibody. Immunoreac- tions for retinal inner nuclear layer bipolar cell staining tive spots on each membrane were compared visually and by using a computerized system. Each of the spots was matched as a reliable technique for confirming MAR. To date, no to an equivalent spot on staining gels. studies have been performed in patients with MAR com- paring Western blot with indirect immunohistologic de- IN GEL ENZYME DIGESTION tection of antiretinal antibodies. In addition, the spe- AND MASS SPECTROMETRY cific antibodies (and their antigenic retinal proteins) involved in MAR have not been systemically examined. Spots from 2-dimensional electrophoresis–stained gels were ex- We investigated 11 patients with MAR who were ini- cised and in gel digested (Trypsin Gold; Promega, Madison, Wis- tially seen with visual loss and photopsias and who un- consin). Peptides were extracted from the gel plugs in 30 µL derwent clinical examinations, electrophysiological and of 2% acetonitrile, 1% formic acid. Five microliters of alpha- psychophysical testing, Western blots, indirect immu- cyano-4-hydroxycinnaminic acid (5 mg/mL in 50% acetoni- nohistologic investigations, and proteomic analysis to trile, 0.1% trifluoroacetic acid [TFA], and 2mM ammonium ci- identify the antigenic retinal proteins. Careful family his- trate) matrix was added to the digested peptides. The extracts tories were taken for autoimmune disorders to clarify the were evaporated to dryness and then dissolved in 5 µL of 60% genetic autoimmune backgrounds of patients with MAR. acetonitrile, 0.1% TFA. A 0.5-µL volume of this solution was spotted on a 192-well matrix-assisted laser desorption ioniza- tion target and allowed to dry. METHODS Mass spectra were acquired using tandem time-of-flight mass spectrometry(4800 Proteomics Analyzer; Applied Biosys- CLINICAL EXAMINATIONS tems, Foster City, California). Database searching was performed using commercially available software (GPS Explorer Eleven patients with histories of melanoma and subsequent reti- version 3.6, with Mascot version 2.1 against International Pro- nopathies of unknown origins and 8 patients with CAR and AIR tein Index human version 3.32 using the following para- (included for comparison) were examined at the retinal dys- meters: 50-ppm mass tolerance, 1 missed cleavage, carbami- trophy clinic at the Kellogg Eye Center, University of Michi- domethyl (C) fixed modification, and variable modifications gan, Ann Arbor. The patients were evaluated using standard- pyroglu (N-term Q) and oxidation (M). Spectra were acquired ized kinetic visual fields, electroretinographic testing, fundus in tandem mass spectrometry 2-kV–positive mode. photography, and clinical examinations. Serum samples were obtained after informed consent and were stored at −80°C un- IMMUNOBLOTS AND SDS–POLYACRYLAMIDE til studied. As control subjects, 9 healthy volunteers with nor- GEL ELECTROPHORESIS mal ERGs and no family history of autoimmune diseases were recruited. Informed consent was obtained from all patients and Retinal extract (20 µg of proteins), 0.1 µg of aldolase A, 0.1 µg control subjects. The experimental protocol was approved by of carbonic anhydrase II (CAII), 0.4 µg of aldolase C, 0.2 µg of the University of Michigan institutional review board. recoverin, 0.2 µg of S-arrestin, 0.4 µg of S-transferase (GST) fusion ␣-enolase, and 0.2 µg of heat shock protein 60 (HSP60) were 2-DIMENSIONAL ELECTROPHORESIS used. Samples were separated on 10% SDS–polyacrylamide gel WESTERN BLOT ANALYSIS electrophoresis gels. After electrophoretic run, immunoblot analysis was performed as described previously. We used GST fusion ␣ Reagents and materials were obtained (from GE Healthcare, -enolase (Abnova, Taipei, Taiwan) and aldolase A and HSP60 Uppsala, Sweden, unless otherwise indicated) for 2-dimen- (Sigma-Aldrich). Bovine recombinant recoverin was purified as previously described.19 Human recombinant CAII was purified sional electrophoresis experiments. Recent postmortem normal 20 human retina (Michigan Eye Bank, Ann Arbor) was lysed (Cel- as previously described. Lytic MT, 5mM Tris[2-carboxyethyl]phosphine hydrochloride), 1% protease inhibitor cocktail [Sigma-Aldrich, St Louis, IMMUNOHISTOCHEMISTRY Missouri]) and centrifuged at 10 000g for 10 minutes at 4°C, and the supernatant was pooled. Protein (300 µg) was precipitated Frozen sections were obtained from recent postmortem nor- from the lysate (Amersham Clean-Up Kit, GE Healthcare). The mal human retina (Michigan Eye Bank) embedded in opti- pellets were incubated for 60 minutes at 30°C in 220 µL of iso- mum cutting temperature compound (Tissue-Tek; Miles Inc, electric focusing rehydration buffer (7M urea, 2M thiourea, 1% Elkhart, Indiana). Sections were blocked for nonspecific pro- amidosulfobetaine 14, 1% Triton X-100, 1% 3-[(3-cholamido- tein binding with 5% goat serum in a phosphate-buffered sa- propyl)dimethylammonio]-1-propanesulfonic acid, and 1% im- line (PBS) solution at room temperature for 1 hour and then

(REPRINTED) ARCH OPHTHALMOL / VOL 127 (NO. 12), DEC 2009 WWW.ARCHOPHTHALMOL.COM 1573

kDa 50 50 50 50 kDa

kDa 37 37 kDa 37 37

25 25 25 20 25 20

250 123 12 3 322 250 471 150 150 100 100 45 6 7 8910 75 45678 910 75 75 75 50 kDa 50 50 50 75 kDa kDa 37 37 kDa 37 25 25 25 20 25 20 20

123 250 123 886 1066 250 150 150 100 4 5 678 9 10 100 75 456789 10 75 75 75

kDa 50 50 kDa 50 37 50 kDa 37 kDa 37 25 37 20 25 25 15 25

12 3 12 3 250 250 1099 150 2052 150 100 100 456 78910 45 6 7 8910 75 75 75 75

kDa 50 50

kDa 50 37 50 kDa 37 37 kDa 25 37 20 25 25 15 20 20 25

1 2 3 123 250 250 150 150 100 S49 3361 100 75 75 45 6 7 8910 45 678910 75 50 75

kDa 50 37 50 50 kDa 37 37 kDa 37 kDa 25 25 25 20 25 20 20

Figure 1. Western blots of serum samples from patients with melanoma-associated retinopathy (patients 175, 317, 322, 471, 886, 1066, 1099, 2052, S49, and 3361) showing immunoreactivity of autoantibodies to (1) human, (2) mouse, and (3) bovine retinal extract. Bands found on human Western blots were isolated and analyzed by mass spectrometry, and results were confirmed by immunoblotting with the following candidate proteins: (4) aldolase A (39 kDa), (5) carbonic anhydrase II (29 kDa), (6) aldolase C (39 kDa), (7) recoverin (23 kDa), (8) S-arrestin (45 kDa), (9) glutathione S-transferase fusion ␣-enolase (74 kDa), and (10) heat shock protein 60 (60 kDa). Molecular size standards are shown on the left.

incubated with serum from patients with MAR (dilution, 1:500) RESULTS overnight at 4°C. After washing with PBS plus 0.2% Triton X-100 (3 times at room temperature for 10 minutes each), the sections were incubated with Alexa Fluor 488–conjugated anti- IDENTIFICATION OF RETINAL AUTOANTIBODIES goat IgG secondary antibody (dilution, 1:2000) (Molecular IN SERUM SAMPLES FROM PATIENTS WITH MAR Probes). They were then counterstained with 4Ј,6Ј-diamidino- 2-phenylindole (0.3µM) before observation with a fluores- Proteomic analysis was performed of antigenic retinal pro- cence microscope (Olympus, Tokyo, Japan). teins identified on Western blots. The identified MAR-

(REPRINTED) ARCH OPHTHALMOL / VOL 127 (NO. 12), DEC 2009 WWW.ARCHOPHTHALMOL.COM 1574

©2009 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 Table 1. Mass Spectroscopy Results for Each Western Blot Spot and Prevalence of Autoantibodies Found Against Each Retinal Protein in Melanoma-Associated Retinopathy (MAR), Cancer-Associated Retinopathy (CAR), and Control Serum

Recurrence of Autoantibodies Protein Protein in Serum, No./Total No. (%) Retinal Molecular Isoelectric Start-End Calculated Protein Weight, Da Point Residue Best Peptide Sequence Mass, Da MAR CAR Control Aldolase A 45 688.3 8.48 98-111 LQSIGTENTEENRR 1646.81 2/11 (18) 1/8 (13) 0/9 142-153 ADDGRPFPQVIK 1342.71 166-188 GVVPLAGTNGETTTQGLDGLSER 2272.14 Aldolase C 36 614.8 7.60 44-57 LSQIGVENTEENRR 1644.83 4/11 (36) 0/8 0/9 88-96 DDNGVPFVR 1652.93 112-134 GVVPLAGTDGETTTQGLDGLSER 2273.13 158-173 TPSALAILENANVLAR 1652.93 216-230 YTPEEIAMATVTALR 1681.85 Carbonic 29 400 6.63 172-181 SADFTNFDPR 1169.60 2/11 (18) 0/8 1/9 (11) anhydrase II ␣-Enolase 47 481.4 7.01 33-50 AAVPSGASTGIYEALELR 1804.94 5/11 (45) 0/8 0/9 163-179 LAMQEFMILPVGAANFR 1939.98 184-193 IGAEVYHNLK 1143.62 240-253 VVIGMDVAASEFFR 1556.78 270-281 YISPDQLADLYK 1425.73 359-372 LAQANGWGVMVSHR 1541.76 Heat shock 61 187.4 5.70 97-121 LVQDVANNTNEEAGDGTTTATVLAR 2560.25 11/11 (100) 7/8 (88) 5/9 (56) protein 60 251-268 ISSIQSIVPALEIANAHR 1919.07 430-446 AAVEEGIVLGGGCALLR 1684.91 S-arrestin 45 262.5 6.14 25-33 SVTIYLGNR 1022.56 3/11 (27) 1/8 (13) 0/9 61-70 VYVTLTCAFR 1229.63 71-84 YGQEDIDVIGLTFR 1625.82 180-193 KVQHAPLEMGPQPR 1603.84 281-292 TLTLLPLLANNR 1338.81 337-358 LTVSGFLGELTSSEVATEVPFR 2339.21 370-384 ESYQDANLVFEEFAR 1817.83 Recoverin 23 229.7 5.06 44-55 ITQQQFQSIYAK 1454.77 2/11 (18) 1/8 (13) 0/9 140-151 LLPDDENTPEKR 1426.72

related antigenic retinal proteins are the same ones that have ods, with 9 of 11 patients having positive Western blots, been identified in many CARs and AIRs cases, namely, an- 8 of 11 patients having positive immunohistologic reac- tibodies against recoverin, ␣-enolase, S-arrestin, CAII, and tivity, and 7 of 11 patients having both. Some patients with HSP6010,14,21 (Figure 1 and Table 1). We found 2 new MAR have antibodies that are reactive with retinal cells antigenic retinal proteins on Western blots (aldolase C and other than bipolar cells, so the usefulness of bipolar cell aldolase A), which have not been reported in MAR, CAR, staining as the sole basis for diagnosing MAR is limited to or AIR, to our knowledge. Immunoblot analysis of serum cases in which it occurs. Inner nuclear layer staining was samples from patients with MAR using human, mouse, or common in our patients with MAR, but colocalization stud- bovine retinal extract, as well as purified proteins, found ies found bipolar cell staining in only 4 of 11 patients, while autoantibodies to recoverin (2 of 11), CAII (2 of 11), other inner nuclear layer retinal cells had focal immuno- S-arrestin (3 of 11), ␣-enolase (5 of 11), aldolase A (2 of reactivity to patient serum (Figure 2). More detailed stud- 11), aldolase C (4 of 11), and HSP60 (11 of 11). Immuno- ies will be needed to elucidate the specificity of this im- blot analysis of 8 serum samples from patients with CAR munoreactivity and to investigate other immunologic and 9 control serum samples using purified proteins was variables, as a few patients with MAR did not have reac- also performed for comparison. Antibodies to recoverin (1 tive antiretinal antibodies by this method. of 8), S-arrestin (1 of 8), aldolase A (1 of 8), and HSP60 (7 of 8) were identified in serum samples from patients with CLINICAL FEATURES OF MAR CAR, while antibodies to CAII, ␣-enolase, and aldolase C were not identified in serum samples from patients with To date, except for 2 case reports of choroidal mela- CAR. Antibodies to CAII (1 of 9) and HSP60 (5 of 9) were noma,6,7 only MAR associated with cutaneous melano- identified in control serum samples, while antibodies to re- mas has been reported.5 We were surprised to find that coverin, S-arrestin, ␣-enolase, aldolase A, and aldolase C any form of melanoma may potentially autosensitize the were not identified in control serum samples. patient and lead to MAR. We examined 11 patients hav- Some serum samples from patients with MAR did not ing MAR with the following types of melanoma: cutane- show immunoreactivity on Western blots using solubi- ous (6 patients), choroidal tumors (2 patients), flat cho- lized normal retinal proteins, which led to development roidal nevi (2 patients), and ciliary body (1 patient). Flat of the indirect immunohistologic method.8 Our investi- choroidal nevi usually are regarded as benign with little gations demonstrated immunoreactivity using both meth- malignant potential unless there is demonstrated growth.22

(REPRINTED) ARCH OPHTHALMOL / VOL 127 (NO. 12), DEC 2009 WWW.ARCHOPHTHALMOL.COM 1575

PRL PRL

ONL

INL INL

GCL

Healthy subject Metastatic melanoma 471 Choroidal melanoma 317

50 µm

PRL

INL INL

GCL Serum + α = PKC

Choroidal melanoma S49 Cutaneous melanoma 175

Figure 2. Immunostaining analysis of serum from patients with melanoma-associated retinopathy (patients 471, 317, S49, and 175) on human retina, rod outer segments, cone photoreceptor cells, and bipolar cells. Red indicates serum; blue, 4’,6’-diamidino-2-phenylindole nuclear stain; and green, ␣–protein kinase C (␣-PKC) (as a bipolar cell marker). Sample identification numbers correlate with the patients and immunohistologic findings summarized in Table 2. GCL indicates ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; PRL, photoreceptor layer; and RPE, retinal pigment epithelium. Arrows indicate staining on the retina.

However, the carcinogenic potential of flat choroidal mela- monly have personal and family histories of autoim- nomas seems to differ from their ability to serve as anti- mune disease, suggesting a special susceptibility for the genic foci in autoimmune-susceptible individuals. Both of putative retinal antigens released by their melanomas our patients with choroidal nevi (patients 1066 and 3361 (Table 2). The latency from melanoma diagnosis to rec- in Table 2) had central scotomata, antiretinal antibod- ognition of MAR ranged from 6 months to 14 years (me- ies, and severe ERG changes similar to the other patients dian, 2 years) preceding the diagnosis. The trigger that with MAR. These 2 patients also had obvious degenera- set off the MAR response in the patients with longer la- tion of the retina at the margins of the nevi, and patient tencies is unknown. One particularly intriguing indi- 1066 showed asymmetry of disease, which was worse in vidual with MAR (patient 175) with malignant mela- the eye with the large nevus (Figure 3). Four individu- noma of an amputated toe did not manifest MAR until als (patients 322, 1066, 2052, and S49 in Table 2) with after she had knee surgery in the same leg 5 years after different forms of melanoma are shown as representative amputation of the toe. examples in Figure 3. Clinically, the patients with MAR have similar findings, with rapid onset that included photopsias and loss COMMENT of vision. Electroretinograms show retinal dysfunction (Figure 3) and aid in diagnosis of this condition with oth- Typically, malignant melanoma is a highly invasive tu- erwise minimal retinal changes.17 Patients often have nega- mor derived from neuroectodermal melanocytes, which tive electroretinographic waveforms (a-waves and poor share a lineage with retinal cells.23 The molecular mecha- b-waves not reaching the isoelectric point) on dark- nisms underlying the development of MAR are un- adapted maximal stimulation and central scotomata on known. Investigations of patients with MAR, CAR, and visual field testing (Figure 3). In more advanced cases, AIR frequently show positive family histories for auto- patients develop retinal vessel attenuation and areas of immune diseases, and many patients have autoimmune diffuse atrophy. Retinal pigmentary changes are absent genetic templates that increase their susceptibility to au- to minimal. We also found that patients with MAR com- toimmune diseases.11 Examination of tumors in pa-

(REPRINTED) ARCH OPHTHALMOL / VOL 127 (NO. 12), DEC 2009 WWW.ARCHOPHTHALMOL.COM 1576

%a Antibodies Found in Serumb Time From Malignant Personal Mixed Melanoma or Family Response Carbonic Heat Staining Patient Diagnosis to History of Goldmann Rod- (Ratio of Anhy- S- ␣- Aldo- Aldo- Shock Location No./Sex/ Tumor Recognition Autoimmune Visual Retinal Visual Isolated b-Waves to Photopic Recov- drase Arres- Eno- lase lase Protein With Age, y Type of MAR Disease Acuity Appearance Fields Response a-Waves) Response erin II tin lase A C 60 Serum 175/F/75 Malignant MAR 2004 DM (sister), CF OU Granular 5-Degree NR OU 0.5, 100OU ...... ϩϩ...... ϩ INL inner melanoma, lupus pigmentation central Negative, plexiform R toe (daughter), of macular scotoma with 46 OD; layer amputated fibromyalgia OU enlarged 0.5, 1998, (daughter) blind spots negative, R knee OU 48 OS surgery 2003 317/F/85 Choroidal 14 y Unknown 20/25 OD, Diffuse retinal Constricted NR OD 0.5, 100OD ...... ϩϩ...... ϩ INL enucleated and retinal peripheral Negative, OS pigment fields (I-4-e 40 OD epithelial isopter atrophy OD constricted to 15 degrees) OD 322/M/58 Ciliary body Concurrent None 20/20 OU Generalized Paracentral 13 OD, 1.3, Near 29 OD, ...... ϩ ...... ϩ No staining depigmented scotoma OD, 10 OS negative, 22 OS fundus with partial ring 29 OD; focal scotoma OS 1.0, pigment negative, deposits OU 24 OS 471/M/50 Metastatic 6 mo Before Ulcerative 20/40 OU Posterior pole Full fields OU 63 OD, 1.7, 81 OD; 70 OD, ϩϩ...... ϩϩ ϩCone inner cutaneous melanoma colitis and subretinal 79 OS 1.5, 100 OS segments diagnosis arthritis exudation 87 OS INL (patient), OU asthma (mother) 886/F/72 Cutaneous 3 y RA (patient) 20/80 OD, Retinal atrophy Central 31 OD, 1.2, NR OU . . . ϩ ... ϩ ... ϩϩINL 20/100 OS at macular scotoma OU 41 OS Negative, and posterior (5-degree 36 OD; pole RPE OD, 1.3, near mottling OU 10-degree negative, OS) 44 OS 1066/M/88 Choroidal Concurrent RA (mother, 2/200 OD, Large Temporal 11 OD, NR OD; NR OD, ...... ϩϩ ϩINL GCL nevus sister), DM 20/200 OS choroidal island of 34 OS 1.1, 16 OS (father), nevus with vision only negative, thyroid surrounding OD, ring 3OS disease chorioretinal scotoma with (daughter) atrophy OD, constriction peripheral of peripheral retinal/RPE field to 50 atrophy OS degrees (IV-4-e) OS

(continued)

tients manifesting CAR has demonstrated ectopic ex- phosphoenolpyruvate and generates adenosine triphos- pression of retinal antigens, which are recognized as phate in glycolysis.28 ␣-Enolase was found to be present foreign by the immune system and lead to initiation of in central nervous system neurons28 and in human A375 an immune response.24 In this study, we identified 7 dif- melanoma cells.29 Autoantibodies against ␣-enolase are of- ferent antiretinal antibodies in our patients with MAR ten associated with progressive visual loss in patients with against S-arrestin, recoverin, ␣-enolase, aldolase A, al- CAR and noncancer (nonparaneoplastic) AIR; results of dolase C, HSP60, and CAII. We did not have access to studies21,30 suggest that they have a pathologic role in reti- the patients’ neoplasms to evaluate the presence of retinopathy. Anti–␣-enolase autoantibodies have been shown nal proteins in the tumors. to have a role not only in systemic and invasive autoim- S-arrestin expresses in melanoma cell lines at the mes- mune disorders31 but also in some cancerous diseases.32 senger RNA level, and autoantibodies against these pro- To our knowledge, this is the first report that serum from teins were detected by SEREX (serological expression of patients with MAR commonly (5 of 11) contains anti–␣- cDNA expression libraries) in the serum of patients with enolase antibodies. The present study confirms that au- melanoma.25 Recoverin, a Ca2ϩ-binding regulatory pro- toantibodies against S-arrestin, recoverin, ␣-enolase, and tein, is present in retinal photoreceptor and bipolar cells26 CAII occur in many patients with MAR.33 Anti-HSP60 an- and is considered a major antigen involved in the immu- tibody was found in patients with MAR and in control se- nologic pathogenesis of CAR.13 Recoverin is aberrantly ex- rum samples. pressed in carcinoma13 and melanoma cells.14 The func- Aldolase is a glycolytic enzyme that catalyzes the tion of these proteins in the carcinomas is unknown. In reversible conversion of fructose-1,6-bisphosphate to the retina, S-arrestin and recoverin have important roles glyceraldehyde 3-phosphate and dihydroxyacetone in phototransduction.27 Enolase is a ubiquitous enzyme phosphate. The following 3 isozymes of aldolase occur that catalyzes the conversion of 2-phosphoglycerate to in vertebrates: aldolase A (the major form in muscle),34

(REPRINTED) ARCH OPHTHALMOL / VOL 127 (NO. 12), DEC 2009 WWW.ARCHOPHTHALMOL.COM 1577

%a Antibodies Found in Serumb Time From Malignant Personal Mixed Melanoma or Family Response Carbonic Heat Staining Patient Diagnosis to History of Goldmann Rod- (Ratio of Anhy- S- Aldo- Aldo- Shock Location No./Sex/ Tumor Recognition Autoimmune Visual Retinal Visual Isolated b-Waves to Photopic Recov- drase Arres- ␣- lase lase Protein With Age, y Type of MAR Disease Acuity Appearance Fields Response a-Waves) Response erin II tin Enolase A C 60 Serum 1066/M/88 Choroidal Concurrent RA (mother, 2/200 OD, Large Temporal 11 OD, NR OD; NR OD, ...... ϩϩ ϩINL GCL nevus sister), DM 20/200 OS choroidal island of 34 OS 1.1, 16 OS (father), nevus with vision only negative, thyroid surrounding OD, ring 3OS disease chorioretinal scotoma with (daughter) atrophy OD, constriction peripheral of peripheral retinal/RPE field to 50 atrophy OS degrees (IV-4-e) OS 1099/F/78 Cutaneous 2 y Treated with 20/400 OU Scalloped Midperipheral NR OU NR OU NR OU ϩ ...... ϩϩNo hydroxy- retinal, RPE islands of staining chloroquine atrophy vision to sulfate posterior IV-4-e (patient) pole with isopter OD, peripheral midperipheral pigment OU inferonasal island of vision to IV-4-e isopter OS 2002/F/63 Cutaneous 11 y RA (patient) 20/20 OU Cresenteric Paracentral 20 OD, 2.1, Near 53 OD, ...... ϩ No depigmenta- scotoma OU 12 OS negative, 42 OS staining tion temporal 24 OD; to macular 2.1, near and negative, peripheral 24 OS retinal atrophy OU 2052/F/44 Cutaneous 3 y RA (patient) 20/80 OD, Macular retinal Paracentral NR OU 0.8, 33 OD, ...... ϩ ...... ϩ INL 20/50 OS atrophy OU scotoma OU Negative, 37 OS 6 OD; 0.8, negative, 8OS 3361/F/82 Choroidal Concurrent RA, asthma, 20/60 OD, Inferoperiph- Paracentral NR OU NR OU NR OU ...... ϩ ...... ϩ INL GCL nevus Sjögren CF OS eral nevus scotoma OD, syndrome with central (patient); surrounding scotoma OS, DM chorioretinal moderate (mother, atrophy OS, constriction father, generalized of peripheral brother) retinal/RPE field OU atrophy OU S49/M/73 Choroidal 2 y Arthritis Enucleated Generalized Central 40 OS 1.5, Near 33OS ...... ϩ Cone INL treated with OD, retinal 20-degree negative, prednisone 20/125 OS atrophy OS scotoma OS 45 OS (patient)

Abbreviations: CF, counting fingers; DM, diabetes mellitus; GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; NR, not recordable; RA, rheumatoid arthritis; RPE, retinal pigment epithelium; ellipsis, not applicable. aPercentage of mean normal (age 20-60 years) was used as a reference to grade the severity of dysfunction. Negative indicates a negative waveform in which the b-wave does not return to the isoelectric baseline, and near negative indicates that the b-wave generated is marginally above the isoelectric baseline. bPlus sign indicates presence.

aldolase B in liver,35 and aldolase C together with aldol- noma cells are characterized by a high rate of glycolysis, ase A in brain.34,36 Because aldolase A and aldolase C are which is their primary energy source. Therefore, the of brain origin, the presence of circulating anti–aldolase antialdolase antibody level in patients with MAR might A and anti–aldolase C autoantibody should serve as an be a marker for the progression of melanoma. However, indicator of blood–retinal barrier damage in patients a pathologic role for aldolase antibodies has not yet with MAR. Aldolase C is expressed in the ganglion cell been elucidated. layer and inner nuclear layer in retina37 and in mela- In summary, we identified 7 retinal proteins that are noma tissue.38 The role of aldolase A and aldolase C in immunoreactive with the serum of patients with MAR. retina is unknown at this time. Anti–aldolase A antibod- Preliminary evidence suggests that, if present, antibodies have been found in other diseases such as rheuma- ies to aldolase A and aldolase C may prove to be useful toid arthritis34 and Alzheimer disease,39 while anti– markers of MAR, and several (antirecoverin and aldolase C antibodies that cross-reacted with aldolase A ␣-enolase antibodies) are valuable as specific markers of have been found in diabetic retinopathy.40 In the AIR. The reliability of each antigen or a combination of present study, we found a high prevalence of antialdol- them for diagnostic or prognostic evaluation has to be ase antibody in patients with MAR. Cancer or mela- determined by studies that will correlate the presence

(REPRINTED) ARCH OPHTHALMOL / VOL 127 (NO. 12), DEC 2009 WWW.ARCHOPHTHALMOL.COM 1578

A MAR from cutaneous melanoma 2052 120105 90 75 60 120105 90 75 60 135 70 70 135 Rod 60 60 150 50 50 150 40 40 30 30 165 165 Mixed 20 20 10 10 180 90 80 70 60 50 40 30 20 10 10 20 30 40 50 60 60 50 40 300 20 10 10 20 30 40 50 60 70 80 90 180 10 10 20 20 195 195 30 30 40 40 Photopic 210 50 50 210 60 60 I-4-e 225 70 70 225 III-4-e 240255 270 285 300 240255 270 285 300 30 Hz Left Right IV-4-e

B MAR from choroidal melanoma S49

12010590 75 60 12010590 75 60 135 70 70 135 Rod 60 60 150 50 50 150 40 40 30 30 165 165 Mixed 20 20 10 10 180 90 80 70 60 50 40 30 20 10 10 20 30 40 50 60 60 50 40 300 20 10 10 20 30 40 50 60 70 80 90 180 10 10 Right eye enucleated 20 20 195 195 30 30 40 40 210 50 50 210 Photopic 60 60 I-4-e 225 70 70 225 240255 270 285 300 240255 270 285 300 III-4-e 30 Hz Left Right IV-4-e

C MAR from ciliary body melanoma 322 120105 90 75 60 120105 90 75 60 135 70 70 135 Rod 60 60 150 50 50 150 40 40 30 30 165 165 20 20 Mixed 10 10 180 90 80 70 60 50 40 30 20 10 10 20 30 40 50 60 60 50 40 300 20 10 10 20 30 40 50 60 70 80 90 180 10 10 20 20 195 195 30 30 40 40 210 50 50 210 Photopic 60 60 I-4-e 225 70 70 225

240255 270 285 300 240255 270 285 300 III-4-e 30 Hz Left Right IV-4-e

D MAR from choroidal nevus 1066 12010590 75 60 12010590 75 60 135 70 70 135 Rod 60 60 150 50 50 150 40 40 30 30 165 165 20 20 Mixed 10 10 180 90 80 70 60 50 40 30 20 10 10 20 30 40 50 60 60 50 40 300 20 10 10 20 30 40 50 60 70 80 90 180 10 10 20 20 195 195 30 30 40 40 210 50 50 210 Photopic 60 60 I-4-e 225 70 70 225

240255 270 285 300 240255 270 285 300 III-4-e 30 Hz Left Right IV-4-e

E Healthy subject

12010590 75 60 12010590 75 60 135 70 70 135 Rod 200 µV 60 60 I 150 50 50 150 40 40 30 30 165 165 Mixed 50 ms 20 20 10 10 180 90 80 70 60 50 40 30 20 10 10 20 30 40 50 60 60 50 40 300 20 10 10 20 30 40 50 60 70 80 90 180 10 10 20 20 195 195 30 30 40 40 Photopic 100 µV 210 50 50 210 I 60 60 I-4-e 225 70 70 225 240255 270 285 300 240255 270 285 300 III-4-e 30 Hz 50 ms Left Right IV-4-e

Figure 3. Clinical features of melanoma-associated retinopathy (MAR). A, A 44-year-old woman (patient 2052 in Table 2) had a cutaneous melanoma excised 2 years before developing pericentral scotomata. She was initially seen with generalized retinal atrophy demonstrating no pigment deposits and with rod-cone dysfunction on an electroretinogram (ERG) showing a negative waveform. B, A 73-year-old man (patient S49) had a choroidal melanoma in the right eye requiring enucleation. His remaining eye developed a scotoma 15 months later. Examination showed generalized atrophy of the retina, cone-rod dysfunction on an ERG, and an almost negative waveform. C, A 58-year-old man (patient 322) with a 2-month history of scotomata was found to have a ciliary body mass in the right eye and an overlying conjunctival sentinel vessel. He had generalized depigmented fundus with focal pigment deposits, central scotomata in the right eye, and partial ring scotomata in both eyes. His ERG showed severe cone-rod dysfunction and negative waveforms. The ciliary mass completely regressed over 4 months without treatment, and visual function stabilized in both eyes. D, An 88-year-old man (patient 1066) had progressive deterioration of central and night vision over 10 months, with a large choroidal nevus (7.5-mm wide and 1-mm thick) and surrounding chorioretinal atrophy in the right eye. He had preexisting dry age-related macular degeneration in both eyes with central visual field loss far greater than the area of macular degeneration. Visual fields were asymmetric, with severe loss in the eye with the nevus. The ERG was nonrecordable in the right eye, and his left eye had cone-rod dysfunction with a negative waveform. E, A 40-year-old woman with no ophthalmic disease and with normal retinal appearance underwent visual field tests and an ERG.

or absence of these markers with clinical data. The in patients with MAR may have a therapeutic role in diversity of autoantibodies produced in MAR may be retarding tumor growth. the consequence of varied overexpression among retinal proteins involved in melanoma tumor development, as Submitted for Publication: April 7, 2009; final revision well as among these patients’ propensity to develop received May 18, 2009; accepted June 19, 2009. autoimmune autoantibodies.14 In general, AIR is treat- Correspondence: John R. Heckenlively, MD, Depart- able with immunosuppression therapies,11,18 but no ment of Ophthalmology and Visual Sciences, Kellogg Eye standardized protocols have been established. In addi- Center, University of Michigan, 1000 Wall St, Ann Ar- tion, it is unknown whether the antibodies that develop bor, MI 48105 ([email protected]).

(REPRINTED) ARCH OPHTHALMOL / VOL 127 (NO. 12), DEC 2009 WWW.ARCHOPHTHALMOL.COM 1579

©2009 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 Financial Disclosure: None reported. 18. Ferreyra HA, Jayasundera T, Khan NW, He S, Lu Y, Heckenlively JR. Manage- Funding/Support: This study was supported by Foun- ment of autoimmune retinopathies with immunosuppression. Arch Ophthalmol. 2009;127(4):390-397. dation Fighting Blindness and by grant R01 EY007758 19. Ray S, Zozulya S, Niemi GA, et al. Cloning, expression, and crystallization of re- from the National Institutes of Health. coverin, a calcium sensor in vision. Proc Natl Acad Sci U S A. 1992;89(13): Additional Contributions: Scott Grant, MD; Frank Gar- 5705-5709. ber, MD; Mark Johnson, MD; Thomas Mehelas, MD; Alan 20. Nair SK, Calderone TL, Christianson DW, Fierke CA. Altering the mouth of a hy- Ruby, MD; David Cooke, MD; and David Kaufman, MD; drophobic pocket: structure and kinetics of human carbonic anhydrase II mu- tants at residue Val-121. J Biol Chem. 1991;266(26):17320-17325. referred patients. James B. Hurley, PhD, provided recov- 21. Dot C, Guigay J, Adamus G. Anti–␣-enolase antibodies in cancer-associated reti- erin plasmid. Carol Fierke, PhD, provided carbonic an- nopathy with small cell carcinoma of the lung. Am J Ophthalmol. 2005;139 hydrase II plasmid. Vsevolod V. Gurevich, PhD, pro- (4):746-747. vided S-arrestin. Dean R. Tolan, PhD, provided aldolase 22. Singh AD, Mokashi AA, Bena JF, Jacques R, Rundle PA, Rennie IG. Small cho- C. Naheed Khan, PhD, performed the electroretino- roidal melanocytic lesions: features predictive of growth. Ophthalmology. 2006; 113(6):1032-1039. grams. Jill Oversier, BS, coordinated the study. 23. Arnheiter H. Evolutionary biology: eyes viewed from the skin. Nature. 1998;391 (6668):632-633. REFERENCES 24. Darnell RB, Posner JB. Paraneoplastic syndromes involving the nervous system. N Engl J Med. 2003;349(16):1543-1554. 25. Hartmann TB, Bazhin AV, Schadendorf D, Eichmüller SB. SEREX identification 1. Sawyer RA, Selhorst JB, Zimmerman LE, Hoyt WF. Blindness caused by photo- of new tumor antigens linked to melanoma-associated retinopathy. Int J Cancer. receptor degeneration as a remote effect of cancer. Am J Ophthalmol. 1976; 2005;114(1):88-93. 81(5):606-613. 26. Milam AH, Dacey DM, Dizhoor AM. Recoverin immunoreactivity in mammalian 2. Suhler EB, Chan CC, Caruso RC, et al. Presumed teratoma-associated paraneo- cone bipolar cells. Vis Neurosci. 1993;10(1):1-12. plastic retinopathy. Arch Ophthalmol. 2003;121(1):133-137. 27. Fu Y, Yau KW. Phototransduction in mouse rods and cones. Pflugers Arch. 2007; 3. Saito W, Kase S, Ohguro H, Furudate N, Ohno S. Slowly progressive cancer- 454(5):805-819. associated retinopathy. Arch Ophthalmol. 2007;125(10):1431-1433. 28. Nakajima K, Hamanoue M, Takemoto N, Hattori T, Kato K, Kohsaka S. Plasmino- 4. Gass J. Acute Vogt-Koyanagi-Harada-like syndrome occurring in a patient with gen binds specifically to ␣-enolase on rat neuronal plasma membrane. metastatic cutaneous melanoma. In: Saari K, ed. Uveitis Update: Proceedings of J Neurochem. 1994;63(6):2048-2057. the First International Symposium on Uveitis. Amsterdam, the Netherlands: Elsevier 29. Clauser KR, Hall SC, Smith DM, et al. Rapid mass spectrometric peptide se- Science; 1984:407-408. 5. Keltner JL, Thirkill CE, Yip PT. Clinical and immunologic characteristics of mela- quencing and mass matching for characterization of human melanoma proteins noma-associated retinopathy syndrome: eleven new cases and a review of 51 isolated by two-dimensional PAGE. Proc Natl Acad SciUSA. 1995;92(11): previously published cases. J Neuroophthalmol. 2001;21(3):173-187. 5072-5076. 6. Zacks DN, Pinnolis MK, Berson EL, Gragoudas ES. Melanoma-associated reti- 30. Weleber RG, Watzke RC, Shults WT, et al. Clinical and electrophysiologic char- nopathy and recurrent exudative retinal detachments in a patient with choroidal acterization of paraneoplastic and autoimmune retinopathies associated with melanoma. Am J Ophthalmol. 2001;132(4):578-581. antienolase antibodies. Am J Ophthalmol. 2005;139(5):780-794. 7. Nieuwendijk TJ, Hooymans JM. Paraneoplastic vitelliform retinopathy associ- 31. Gitlits VM, Toh BH, Sentry JW. Disease association, origin, and clinical related with metastatic choroidal melanoma. Eye. 2007;21(11):1436-1437. evance of autoantibodies to the glycolytic enzyme enolase. J Investig Med. 2001; 8. Milam AH, Saari JC, Jacobson SG, Lubinski WP, Feun LG, Alexander KR. Auto- 49(2):138-145. antibodies against retinal bipolar cells in cutaneous melanoma-associated 32. Jankowska R, Witkowska D, Porebska I, Kuropatwa M, Kurowska E, Gorczyca retinopathy. Invest Ophthalmol Vis Sci. 1993;34(1):91-100. WA. Serum antibodies to retinal antigens in lung cancer and sarcoidosis. 9. To KW, Thirkill CE, Jakobiec FA, Lessell S, Berson EL. Lymphoma-associated Pathobiology. 2004;71(6):323-328. retinopathy. Ophthalmology. 2002;109(11):2149-2153. 33. Haapasalo J, Nordfors K, Järvelä S, et al. Carbonic anhydrase II in the endothe- 10. Adamus G, Ren G, Weleber RG. Autoantibodies against retinal proteins in para- lium of glial tumors: a potential target for therapy. Neuro Oncol. 2007;9(3): neoplastic and autoimmune retinopathy. BMC Ophthalmol. 2004;4:e5. http://www 308-313. .pubmedcentral.nih.gov/articlerender.fcgi?tool =pubmed&pubmedid=15180904. 34. Ukaji F, Kitajima I, Kubo T, Shimizu C, Nakajima T, Maruyama I. Serum samples Accessed September 9, 2009. of patients with rheumatoid arthritis contain a specific autoantibody to “dena- 11. Heckenlively JR, Ferreyra HA. Autoimmune retinopathy: a review and summary. tured” aldolase A in the osteoblast-like cell line, MG-63. Ann Rheum Dis. 1999; Semin Immunopathol. 2008;30(2):127-134. 58(3):169-174. 12. Hooks JJ, Tso MO, Detrick B. Retinopathies associated with antiretinal antibodies. 35. Tsutsumi K, Ito K, Ishikawa K. Developmental appearance of transcription fac- Clin Diagn Lab Immunol. 2001;8(5):853-858. tors that regulate liver-specific expression of the aldolase B gene. Mol Cell Biol. 13. Polans AS, Witkowska D, Haley TL, Amundson D, Baizer L, Adamus G. Recov- 1989;9(11):4923-4931. erin, a photoreceptor-specific calcium-binding protein, is expressed by the tu- 36. Arai Y, Kajihara S, Masuda J, et al. Position-independent, high-level, and correct mor of a patient with cancer-associated retinopathy. Proc Natl Acad Sci U S A. regional expression of the rat aldolase C gene in the central nervous system of 1995;92(20):9176-9180. transgenic mice. Eur J Biochem. 1994;221(1):253-260. 14. Bazhin AV, Schadendorf D, Willner N, et al. Photoreceptor proteins as cancer- 37. Caffe´ AR, Von Schantz M, Sze´l A, Voogd J, Van Veen T. Distribution of Purkinje retina antigens. Int J Cancer. 2007;120(6):1268-1276. cell–specific Zebrin-II/aldolase C immunoreactivity in the mouse, rat, rabbit, and 15. Potter MJ, Thirkill CE, Dam OM, Lee AS, Milam AH. Clinical and immunocyto- human retina. J Comp Neurol. 1994;348(2):291-297. chemical findings in a case of melanoma-associated retinopathy. Ophthalmology. 38. Prasad R, Romsdahl MM, Shaw CR, Mumford DM, Smith JL Jr. Isozyme varia- 1999;106(11):2121-2125. tions in human malignant melanoma. Cancer Res. 1974;34(6):1435-1438. 16. Chan JW. Paraneoplastic retinopathies and optic neuropathies. Surv Ophthalmol. 39. Mor F, Izak M, Cohen IR. Identification of aldolase as a target antigen in Alzhei- 2003;48(1):12-38. mer’s disease. J Immunol. 2005;175(5):3439-3445. 17. Heckenlively JR, Aptsiauri N, Holder G. Autoimmune retinopathy, CAR and MAR 40. Ahn BY, Song ES, Cho YJ, Kwon OW, Kim JK, Lee NG. Identification of an anti- syndromes. In: Heckenlively JR, Arden GB, eds. Principles and Practice of Clini- aldolase autoantibody as a diagnostic marker for diabetic retinopathy by immu- cal Electrophysiology of Vision. Cambridge, MA: MIT Press; 2006:691-698. noproteomic analysis. Proteomics. 2006;6(4):1200-1209.

(REPRINTED) ARCH OPHTHALMOL / VOL 127 (NO. 12), DEC 2009 WWW.ARCHOPHTHALMOL.COM 1580