MECHANISMS OF INFLAMMATORY RESPONSE IN SYMPATHETIC OPHTHALMIA AND VKH SYNDROME

NARSING A. RAG Los Angeles, California

SUMMARY sympathetic ophthalmia. Moreover, the inciting

Although the iucitiug events in the pathogenesis of event for the two entities is different: the develop­ sympathetic ophthalmia and Vogt-Koyanagi-Harada ment of intraocular inflammation in sympathetic (VKH) syndrome are different, these two forms of ophthalmia requires a penetrating ocular injury, but bilateral granulomatous uveitis share several clinical, no traumatic event is required for the development histopathological and immunohistochemical features, of VKH syndrome. The latter usually follows including their association with HLA types and in their prodromal symptoms and signs suggestive of a in vitro T-cell response to retinal antigens. These systemic viral infection. In addition, although it clinical and immunopathological features indicate that appears that extraocular manifestations such as there is an underlying T -cell-mediated autoimmunity to vitiligo, alopecia, poliosis, meningismus and dysacu­ uveaVretinai antigens in the development of these sis are relatively more frequent in VKH syndrome, forms of uveitis. Both forms exhibit preservation of there are studies revealing the presence of similar the choriocapillaris and retina despite extensive inflam­ manifestations in sympathetic ophthalmia?,4 matory cell infiltration in the choroid. Recent experi­ The typical histopathological features seen in the mental studies suggest that this preservation of early phases of both sympathetic ophthalmia and choriocapillaris could be the result of anti-inflammatory VKH syndrome include a granulomatous inflamma­ products secreted by the retinal pigment epithelium, tion that primarily involves the choroid, with a including transforming growth factor-beta and a novel similar albeit milder inflammatory infiltrate that protein called retinal pigment epithelial protective involves the iris and ciliary body. However, the protein that is known to suppress the phagocyte choroid is the site where several characteristic generation of superoxide. Such suppression of the features are observed, including: (1) diffuse lympho­ oxidant release in the choroidal inflammation could cytic infiltration interrupted at multiple sites with help protect the uvea from necrotic change and collections of epithelioid cells and few multinu­ preserve the choriocapillaris from inflammatory cell cleated giant cells; (2) the presence of pigment in infiltration. the epithelioid cells and giant cells in the absence of apparent choroidal necrosis; (3) sparing of the Although sympathetic ophthalmia and Vogt-Koya­ choriocapillaris from inflammatory cell infiltration; nagi-Harada (VKH) syndrome are two distinct and (4) preservation of retinal pigment epithelium clinical entities, they share virtually identical histo­ (RPE) and retina except at the sites of Dalen-Fuchs' pathological features, fluorescein angiographic find­ nodules and other foci where RPE junctions are ings and association with HLA antigens of DR4, disrupted. These disrupted RPE cells are commonly DRw53 and Bw54.1 However, whereas VKH syn­ detected by fluorescein angiography as focal leaks at drome has a predilection for darkly pigmented races the level of RPE. - Asians, Hispanics, Native Americans and Asian Sympathetic ophthalmia and VKH syndrome also Indians2 - no racial predilection is observed in share several immunopathological features, such as infiltration of primarily T-lymphocytes in the chor­ From: A. Ray Irvine, Jr, Ophthalmic Pathology Laboratory. Doheny Eye Institute, University of Southern California, School oid. In both entities, most of these infiltrated cells of Medicine, Los Angeles, California, USA. exhibit the markers of helper and suppressor/ Correspondence to: Narsing A. Rao, MD, Doheny Eye cytotoxic cells, along with class II MHC molecules.5,6 Institute, 1450 San Pablo Street DVRC-211, Los Angeles, CA 90033-1088, USA. Similar molecules are also noted on the activated

Eye (1997) 11, 213-216 © 1997 Royal College of Ophthalmologists 214 N. A. RAO macrophages along with various adhesion molecules, appears that the formation of these epithelioid cells including intercellular adhesion molecule-l.6 More­ requires the presence of activated T-cells in the over, in both entities the choroidal infiltrates inflammatory lesion. Moreover, the cytokines of T­ generate various pro-inflammatorycytokines? Over­ cells, such as the interleukins IL-2, IL-4, IL-6, IL-8 all, these immunohistological features suggest a and IL-lO, interferon gamma and others, are also delayed type of hypersensitivity (T-cell-mediated) known to modulate epithelioid cell formation and the mechanism in the induction and/or perpetuation of maintenance and resolution of the granulomatous the uveal inflammation,possibly directed to the uveal inflammation?1 Granuloma formation is also modu­ melanocytes or other antigens in the uveal tract.4,8 lated by the level of class II MHC molecule Once initiated, an immune response against a tissue expression by the macrophages. Adhesion molecules target, such as in the uvea, can lead to the further displayed on the macrophages and vascular endothe­ release of various sequestered self-antigens, against lium may also contribute to the development of which self-tolerance has never been established. granulomatous inflammation.6 Release of such antigens can lead to the perpetuation The metabolic products released at the site of of uveitis, even though the initiating trigger may be inflammation are found either to accentuate granu­ directed to a restricted response to a single specific loma development or to suppress formation of target.9,lO epithelioid cell granulomas and the associated chronic inflammatory process.12 Studies on experi­ mental granulomatous uveitis have revealed GRANULOMATOUS INFLAMMATIONS OF enhancement of the granulomatous process in the THE UVEA presence of lip oxygenase metabolic products of AA. The cellular and molecular mechanisms leading to In contrast cyclo-oxygenase products are noted to development of granulomatous uveitis are poorly suppress granuloma formation.lO Such modulation of understood. However, the role of macrophages, T­ granulomatous inflammation in the presence of AA cells and their cytokines, melanocytes, vascular metabolites could be the result of altered macro­ endothelial cells and arachidonic acid (AA) metabo­ phage expression of MHC class II molecules. More­ lites in the initiation, maintenance and resolution of over, lipoxygenase products of AA are important the granulomatous inflammation is becoming clear. inflammatory mediators with pro-inflammatory func­ In general the granulomatous uveitis can be classified tions of chemotaxis, vasodilation and induction of as either hypersensitivity type or foreign body type. lysosomal enzyme release.13 The latter is rarely observed and usually results from T-cells cause pathological changes through their traumatic events associated with retained foreign ability to produce cytokines, which can recruit material. The hypersensitivity granulomas usually additional T-cells, macrophages and other phagocytic result either from an infectious agent such as cells, including eosinophilic leucocytes. All these cells mycobacteria, or from a non-infectious aetiology, are present in the uveal tracts of patients with such as an autoimmune disorder, mediated by either sympathetic ophthalmia and all may cause tissue T-cells or immune complexes. damage if not properly regulated. Additionally, a Hypersensitivity granulomas show three charac­ subset of T-Iymphocytes, cytotoxic T-cells, may cause teristic microscopic patterns: zonal, sarcoidal or these pathological changes directly. diffuse. These three patterns may indicate an under­ lying mechanism of granuloma formation. The zonal pattern of granulomatous uveitis consists of a central TISSUE INJURY IN UVEITIS area of necrosis surrounded by zones of different Several kinds of experimentally induced uveitis in infiltrating inflammatory cells, such as epithelioid animals have revealed the importance of macro­ cell/histiocytes, lymphocytes and other inflammatory phages and other phagocytic inflammatory cells in cells, including fibroblasts. Such a zonal pattern is inducing tissue damage, although T-cells are required seen in tuberculosis and other infectious diseases. for the induction of uveitis. These phagocytes cause The sarcoidal pattern consists of a collection of necrosis of the tissue and amplify the inflammation epithelioid cells without necrosis. Typically such a by releasing various secretory products, including pattern is observed in sarcoidosis. The diffuse pattern proteases, AA metabolites and free radicals. The reveals collections, primarily of lymphocytes, free radicals are potent cytotoxic agents that throughout the choroid and the anterior uvea with primarily cause peroxidation of lipid cell membranes focal collections of epithelioid cells. This pattern is at the site of their release. The free radicals that commonly seen in immune-mediated granulomas cause such tissue damage are mostly derived from such as sympathetic ophthalmia and VKH syndrome. the oxygen metabolite superoxide and its derived In granulomatous uveitis, the epithelioid cells agents such as hydrogen peroxide, hypochlorous acid represent modifiedmacrophages displaying primarily and hydroxyl radicals.14 Nitric oxide, another mole­ the morphological features of secretory functions. It cule released by the activated macrophages, may also INFLAMMATORY RESPONSE IN SYMPATHETIC OPHTHALMIA & VKH SYNDROME 215 induce tissue damage in the presence of superoxide because of the plausibility of such a role for this by formation of peroxynitrites. Release or formation novel RPE protein, we called the protein 'retinal , of all these oxidants at the site of inflammation can pigment epithelial protective protein .16,18 There are lead to uveal necrosis and amplification of the other anti-inflammatory factors produced by the inflammatory process.14,15 However, tissue necrosis RPE, such as transforming growth factor-beta; is minimal or is not apparent in sympathetic these factors could also down-regulate uveal inflam­ ophthalmia and VKH syndrome, thus suggesting mation by altering ongoing inflammatory events, the presence of protective mechanisms in the uvea such as antigen presentation. and/or in the granulomas that may minimise tissue In sympathetic ophthalmia and VKH syndrome necrosis. the histological finding of preserved choriocapillaris and retina may suggest that the RPE could offer protection against extension of inflammatory cells PRESERVATION OF CHORIOCAPILLARIS into the choriocapillaris and retina. However, RPE AND RETINA cells in culture are known to generate various pro­ A constant feature of sympathetic ophthalmia and inflammatory cytokines such as IL-1, IL-6, IL-8, VKH syndrome is the preservation of choriocapil­ platelet-derived growth factor-like protein, granulo­ laris and retina despite extensive uveal infiltrationby cyte macrophage-colony stimulating factor and var­ mononuclear cells and other phagocytes. Although ious adhesion molecules.19.20 All these molecules, if no precise reason for such preservation is known, operative in vivo, could enhance uveitis and lead to plausible explanations include one or a combination extension of inflammation into the choriocapillaris of the following mechanisms: (1) RPE cells may and the adjacent retina. Paradoxically, the enu­ release soluble and diffusible factors that down­ cleated eyes of patients with sympathetic ophthalmia regulate the uveal inflammation; (2) endothelial cells and VKH show preservation of the retina and of the choriocapillaris could produce anti-inflamma­ choriocapillaris at the site of intact RPE, even tory cytokines that could down-regulate the inflam­ though the choroid is heavily infiltrated by macro­ mation; (3) the choriocapillaris endothelial cells may phages and other inflammatory cells. Such contra­ not express adhesion molecules required for attach­ dictory findings suggest that similar to macrophages, ment and migration of leucocytes; and/or (4) immune RPE cells can modulate uveal inflammation, i.e. complexes may not deposit in the capillaries. We either enhance it by releasing various pro-inflamma­ recently made an attempt to investigate the role of tory cytokines or suppress the inflammation by RPE in the down-regulation of inflammation, parti­ releasing transforming growth factor-beta and other cularly with regard to its role in the preservation of soluble factors, including the novel retinal pigment choriocapillaris and retina in severe uveitis as seen in epithelial protective protein.18 However, further in sympathetic ophthalmia and VKH syndrome. vivo studies are required on sympathetic ophthalmia Our current in vitro studies on RPE have revealed and VKH syndrome to delineate the importance of that these cells synthesise and release a protein that RPE and its products in the modulation of uveitis. suppresses phagocyte generation of superoxide16 and that is known to dismutate into hydrogen peroxide, Supported in part by NIH grant EY10212 and by an forming highly reactive hydroxyl radicals. In inflam­ unrestricted grant from Research to Prevent Blindness Inc.,New York. matory conditions these highly reactive oxidants are primarily formed at the extracellular sites. Such Key words: Sympathetic ophthalmia, Vogt-Koyanagi-Harada oxidants can induce tissue damage at the site of their syndrome, Granulomatous uveitis, Oxygen free radicals, Retinal pigment epithelium, Protective protein. formation by various mechanisms, including perox­ idation of lipid cell membranes. Such oxidised lipids REFERENCES can amplify the inflammatory process by their chemotactic property.15 1. Davis JL, Mittal KK, Freidlin V, Mellow SR, Optican DC, Palestine AG,Nussenblatt RE. HLA associations Although the RPE is endowed with multiple and ancestry in Vogt-Koyanagi-Harada disease and antioxidant enzymes, including superoxide dismu­ sympathetic ophthalmia. Ophthalmology 1990;97: tase, such enzymes have at most a limited signifi­ 1137-42. cance in scavenging the phagocyte-generated 2. Moorthy RS, Inomata H, Rao NA. Vogt-Koyanagi­ superoxide in the extracellular sites because they Harada syndrome. Surv Ophthalmol 1995;39:265-92. 3. Rao NA, Marak GE Jr. Sympathetic ophthalmia are localised to the cytosol of the RPE.17 Moreover, simulating Vogt-Koyanagi-Harada's disease: clinico­ scavenging of phagocyte-derived extracellular super­ pathologic, study of four cases. Jpn J Ophthalmol oxide by these antioxidants has never been demon­ 1983;27:506-11. strated. It is possible that the newly discovered RPE 4. Goto H, Rao NA. Sympathetic ophthalmia and protein may offer protection against such extracel­ Vogt-Koyanagi-Harada syndrome. Int Ophthalmol Clin 1990;30:279-85. lular superoxide formation in uveitis, since this 5. Sakamoto T, Murata T, Inomata H. Class II major protein is secreted into extracellular sites.16 Primarily histocompatibility complex on melanocytes of Vogt- 216 N. A. RAO

Koyanagi-Harada disease. Arch Ophthalmol 1991; 13. Kunkel SL, Chensue SW, Mounton C, et al. Role of 109:1270-4. lipoxygenase products in murine pulmonary granuloma 6. Kuppner MC, Liversidge J, McKillop-Smith S, Lums­ formation. J Clin Invest 1984;74:514-24. den L, Forrester JV. Adhesion molecule expression in 14. Rao NA. Role of oxygen free radicals in retinal acute and fibrotic sympathetic ophthalmia. Curr Eye damage associated with experimental uveitis. Trans Res 1993;12:923-34. Am Ophthalmol Soc 1990;88:797-850. 15. Goto H, Wu GS, Gritz DC, Atalla LR, Rao NA. 7. Inomata H, Sakamoto T. Immunohistochemical studies Chemotactic activity of the peroxidized retinal lipid of Vogt-Koyanagi-Harada disease with sunset sky membrane in experimental autoimmune uveitis. Curr fundus. Curr Eye Res 1990;9(Suppl):35-40. Eye Res 1991;10:1009-14. 8. Rao NA, Wong VG. Aetiology of sympathetic 16. Wu GS, RaoNA. A novel retinal pigment epithelial ophthalmitis. Trans Ophthalmol Soc UK 1981;101: protein suppresses neutrophil superoxide generation. I. 357-60. Characterisation of the suppressive factor. Exp Eye 9. Cooke A. Autoimmunity update. Immunologist 1995; Res 1996;63:713-25. 3:241-3. 17. Rao NA, Thaete LG, Delmage JM, Sevanian A. super oxide dismutase in ocular structures. Invest 10. deSmet MD, Yamamoto JH, Mochizuki M, Gery I, Ophthalmol Vis Sci 1985;26:1778-8l. Singh VK, Shin ohara T, et al. Cellular immune 18. Wu GS, Swiderek KM, Rao NA. A novel retinal responses of patients with uveitis to retinal antigens pigment epithelial protein suppresses neutrophil super­ and their fragments. Am J Ophthalmol 1990;110: oxide generation II. Purification and microsequencing 135-42. analysis. Exp Eye Res 1996;63:727-37. 11. Boros DL. The role of cytokines in the formation of 19. Elner VM, Strieter RM, Elner SG, Baggiolini M, the schistosome egg granuloma. Immunobiology 1994; Lindley I, Kunkel SL. Neutrophil chemotactic factor 191:441-50. (IL-8) gene expression by cytokine-treated retinal 12. Rao NA, Patchett R, Fernandez MA, Sevanian A, pigment epithelial cells. Am J Pathol 1990;136:745-50. Kunkel SL, Marak GE Jr. Treatment of experimental 20. Planck SR, Dang TT, Graves D, Tara D, Ansel JC, granulomatous uveitis by lip oxygenase and cyc1o­ Rosenbaum JT. Retinal pigment epithelial cells secrete oxygenase inhibitors. Arch Ophthalmol 1987;105: interleukin-6 in response to interleukin-l. Invest 413-5. Ophthalmol Vis Sci 1992;33:78-82.