Phototesting in Lupus Erythematosus

53S Phototesting in Lupus Erythematosus Peter Kind, Percy Lehmann, and Gerd Plewig Ultraviolet (UV) irradiation is a major factor in the pathogenesis of certain variants of cutaneous lupus erythematosus. Photosensitivity constitutes one of the criteria of the American Rheumatism Association for the diagnosis of systemic lupus erythematosus, which further emphasizes its importance. The pathomechanism of UV-induced lupus erythematosus remains unknown. The characterization of photosensitive subacute cutaneous lupus erythematosus (SCLE) by Gilliam and Sontheimer has led to a new approach. Through the development of standardized test methods it has became possible to reproduce cutaneous lesions in the UV-A and UV-B spectrum. These standardized test methods allow a better definition of photosensitivity than clinical history does. Recent clinical data show that besides SCLE another variant, lupus erythematosus tumidus, also reveals pronounced photosensitivity. In this review article phototest procedures, phototest results, and clinical correlations in different subgroups are discussed. J Invest Dermatol 100:53S 57S, 1993 Kaposi observed that it was mostly outdoor workers that developed lupus erythematosus (LE) and that cold weather, heat, and fire led to exacerbations of the disease [1]. At the end of the last century several other groups had already realized that environmental factors played a central role in the induction of the disease [2 5]. As early as 1921 Nobl observed a patient who developed lesions after artificial irradiation [6]. The first controlled phototesting with artificial ultraviolet (UV) irradiation was performed by Epstein et al [7]. They studied 21 patients with systemic lupus erythematosus (SLE) and four patients with DLE. Using a hot quartz lamp as a UV-B source, they could reproduce clinically and histologically abnormal reactions that lasted up to three months in five cases by the application of doses equivalent to that which would produce a grade 2 to 3 sunburn erythema. Because they could not induce lesions beyond 320 nm, they suggested that the spectrum was in the UV-B range. Baer and Harber studied 23 DLE and five SLE patients, and one patient who would have been classified today as SCLE [8]. They applied one to six times the minimal erythema dose (MED)-UV-B on multiple test sites as single exposures. Only the SCLE patient showed a pathologic reaction with a reduced minimal erythema dose and persistence of the erythema. Freeman tested several patients in the 300-nm spectrum with an equivalent to eight times the erythema dose [9]. Cripps and Rankin determined the erythema action spectra and tried to induce lesions with monochromatic light in the range of 250 330 nm [10]. The dose range was between eight and 13 times the MED. In their study the first immunoglobulin deposits could be detected 8 weeks after UV irradiation. Gilliam and Sontheimer s description of subacute cutaneous lupus erythematosus (SCLE) marks a major step forward in the photobiology and the characterization of photosensitive variants [11]. Department of Dermatology, Ludwig-Maximilians University Munich, Munich; and Department of Dermatology, Heinrich-Heine University Düsseldorf, Düsseldorf, FRG Reprint requests to: Peter Kind, Department of Dermatology, Ludwig- Maximilians University Munich, Frauenlobstrasse 9-11, D(W)-8000 Munich 2, FRG. Abbreviations: DLE, discoid lupus erythematosus; SCLE, subacute cutaneous lupus erythematosus; LET, lupus erythematosus tumidus; LEP, lupus erythematosus profundus; BLE, bullous lupus erythematosus; SLE, systemic lupus erythematosus; CLE, cutaneous lupus erythematosus; ADCC, antibodydependent cellular cytotoxicity; LE, lupus erythematosus; MED, minimal erythema dose; IPD, immediate pigment darkening; MTD, minimal tanning dose; UV, ultraviolet Here SCLE was defined by clinical history, the SSA/Ro-antibody, and the HLA-B8 and -DR3 phenotype. Again photosensitivity was defined by clinical history. Further experiments evolving from these clinical observations have led to the concept that keratinocytes damage by antibody-dependent cellular cytotoxicity might be a mechanism in photosensitive cutaneous lupus erythematosus (CLE) [12 14]. Although the UV spectrum was believed to be in the UV-B range, several authors observed exacerbations of LE after artificial irradiation in sun-tanning studios where the radiation was more UV-A [15,16]. Furthermore, animal models and in vitro data give evidence for a possible role of UV-A in UV-induced LE [17 22]. For this reason we decided to do a study that, since 1986, has involved extensive phototesting of every newly diagnosed LE patient in our clinic. This review article will discuss the clinical phototest results and compare them with recent data from literature. PHOTOTESTING PROCEDURE Patients with lupus erythematosus were routinely tested at the department of dermatology, University of Düsseldorf. Clinical examination and laboratory tests did not reveal any features of systemic disease at the time of testing. The light sources used included an UVASUN 3000 high-pressure metal halide lamp (330 460 nm, Mutzhas, FRG) for UV-A testing, and an UV 800 lamp equipped with fluorescent bulbs Philips TL 20 W/12 (285 350 nm, Waldmann, FRG) for UV-B testing. Determination of the minimal erythemal dose UV-B (MED UV-B), threshold dose for immediate pigment darkening (IPD), and minimal tanning dose (MTD) was carried out according to standard procedures. Test reactions were read immediately following and 24 h after irradiation. For the provocative phototest two areas (5 8 cm) of uninvolved skin on the back or on the extensor surface of the arms were irradiated with 100 J/cm 2 UV-A or 1.5 MED UV-B on three consecutive days (Table I). Other test areas such as the buttock did not give similar results and lesions could not be reproduced in this area. Test areas were evaluated 24,48, and 72 h and up to 3 weeks after the last irradiation [23]. Criteria for a positive provocative phototest were induced lesions clinically resembled LE, histopathology compatible with LE, and skin lesions developing slowly and persisting for several days. The patients tested were asked if exposure to sunlight had any effect on their disease, in order to evaluate the correlation between phototest results and medical history. Statistical analysis to the evaluation of photosensitivity and antibody profile was performed by contingency coefficient C. 0022-202X/93/$06.00 Copyright & 1993 by The Society for Investigative Dermatology, Inc.

54S KIND ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY Figure 1. UV-A and UV-B induced lesions in an SCLE patient at day 10 after irradiation. Lesions could be induced in the UV-A (left side) and the UV- B (right side) spectrum. Clinically lesions were compatible with LE. UV-Ainduced lesions showed small erythematosus plaques, whereas UV-B induced lesions were more papulosquamous. PROVOCATIVE PHOTOTEST The development of positive phototest reactions in LE was considerably slower and persisted longer than phototest reactions in other photodermatoses, i.e., polymorphous light reaction, persistent light reaction, photoallergy, hydroa vacciniforme, and solar urticaria. Lesions developed in general 48 96 h (range 24 h to 3 weeks) after irradiation and lasted about 1 to 3 weeks (Fig 1). In some cases, induced lesions persisted for several months. Only a few cases revealed hypopigmentation of the lesions, a feature that can be observed in SCLE patients. In our initial study, patients with SCLE showed the highest frequency of positive provocative phototest reactions [23]. In 64% of patients with SCLE, in 42% with DLE, and in 25% with SLE, skin lesions characteristic for LE could be induced. Definitely positive test results were not, however, found in 13% of all patients. About 53% of the patients in whom LE lesions were induced reacted to both UV-B and UV-A, 33% reacted only to UV-B, and 14% reacted only to UV-A. Thus in our experimental system lesions could be reproduced in the UV-A as well as the UV-B spectrum. Determination of the MED UV-B revealed that LE patients do not exhibit a decreased threshold for UV-induced erythema compared to controls with normal reactions. Although we did not study the duration of UV-B erythema in particular, a prolonged erythematous response was not a conspicuous feature. However, we could sometimes observe induction of LE in test areas, in which the MED UV-B was determined. As these reactions remain much longer than UV-B erythema, previous investigators may have interpreted incipient UV-induced LE lesions as persistent UV-B erythema. Similar to MED UV-B the pigmentary responses IPD and MTD of LE patients were within the normal range compared to a control group. In our study photosensitivity in LE patients seems to be confined to exacerbation or induction of specific lesions by UV irradiation. About 50% of all patients were aware of an adverse effect of sunlight on their condition. Of these patients 66% showed pathologic test reactions. In contrast, pathologic reactions were induced in 46% of those patients who denied any effect of sun exposure on their disease. In the meantime we have extended our first study and looked more closely at the possible link between the SSA/Ro-antibody and photosensitivity. Furthermore we have included other subgroups of cutaneous LE, such as lupus erythematosus tumidus (LET) [24]. Lupus erythematosus tumidus is a variant of cutaneous lupus erythematosus which is not yet well defined. Clinically, erythematous plaques without epidermal involvement are seen on the face and trunk, especially the back. In Figure 2. Correlation between photosensitivity (measured by a standardized testprotocol) [24,26], Ro/SSA-antibodies, and antinuclear antibodies in DLE and LET. UV-negative LET patients showed more frequently antibodies than UV-positive patients. Statistically there was no association between the parameters (contingency-coefficient C). contrast to DLE, little epidermal involvement is seen histologically. There is a superficial and deep perivascular and periadnexal lymphocytic infiltrate with predominantly T-helper cells. Mucin between collagen bundles can be present. In the absence of underlying systemic symptoms it is sometimes difficult to differentiate from the infiltrate of Jessner lymphocytic infiltration and polymorphous light eruption. LET is further characterized by a slight male predominance and photosensitivity. Of 32 patients with LET, 22 had a positive test reaction. Of these, 14 reacted only to UV-A, 17 only to UV-B, and nine to both. In our studies there was no association between frequency of the SSA/Ro-antibody and photosensitivity in any of the subgroups (Figs 2 and 3) [25]. Only 3% of the photosensitive DLE patients were Ro-positive. These data are more obvious in LET. Not a single patient of this UV-sensitive variant (81% UV positive) has the Ro-antibody. In SCLE (Fig 3), UV-positive and UV-negative patients have the same frequency of the Ro-antibody. Thus our data again show no relation between photosensitivity and Ro-antibody. HISTOPATHOLOGY AND IMMUNOHISTOCHEMISTRY As the test procedure permits us to study the development of CLE chronologically, biopsies were taken at different time intervals from DLE and SCLE patients and studied by routine histopathology and immunohistochemistry [26]. (Teikemeier G, Goldermann R, Lehmann P, Goerz G, Plewig G, Kind P: Chronological immunohistological studies of UV-Aand UV-B-induced lesions of cutaneous lupus erythematosus (paper in preparation).) HISTOPATHOLOGY OF EARLY UV-INDUCED LESIONS All biopsies taken up to day ten were defined as early lesions. Both variants studied, DLE and SCLE, revealed a superficial perivascular

VOL. 100, NO. 1, SUPPLEMENT, JANUARY 1993 PHOTOTESTING IN LUPUS ERYTHEMATOSUS 55S Figure 4. Histopathology of an early UV-A induced lesion (day 5) from a SCLE patient. No alteration of the dermo-epidermal junction. Sparse superficial perivascular lymphocytic infiltrate with slight epidermotropism. (Hematoxylin and eosin, magnification 40.) Figure 3. Correlation between photosensitivity (measured by a standardized test protocol) [24,26], Ro-antibodies, and antinuclear antibodies in SCLE and SLE. Statistically there was no significant association. lymphocytic infiltrate with few interstitial cells in UV-A-and UV-Binduced lesions. Vacuolar alteration of the dermoepidermal junction was not observed (Fig 4). HISTOPATHOLOGY OF LATE UV-INDUCED LESIONS Biopsies taken after day ten were defined as late. UV-B-induced lesions of SCLE and DLE showed a similar pattern. Parakeratosis, few necrotic basal keratinocytes, and vacuolar alteration of the dermoepidermal junction were the dominant features (Fig 5). A lichenoid infiltrate was found in the upper dermis. Mucin between collagen bundles was rarely detected. In contrast, UV-Ainduced lesions were characterized by fewer epidermal changes. Slight smudging of the dermo-epidermal junction was observed in half of the cases. A perivascular lymphocytic infiltrate was localized in the upper and middermis. Patients who were clinically photosensitive in the UV-A and the UV- B range revealed abundant mucin between collagen bundles. SSA/Ropositive patients showed histologically more pronounced epidermal changes such as parakeratosis, atrophy, necrosis, and vacuolar alteration. IMMUNOHISTOCHEMISTRY OF EARLY UV-INDUCED LESIONS UV-A UV-B induced lesions of both subgroups showed predominantly T-helper cells at the dermo-epidermal junction. Suppressor/cytotoxic T cells, B cells, and Langerhans cells were not observed in this region. HLA- DR expression of epidermal keratinocytes was observed rarely. The dermal infiltrate of early lesions rarely showed IL-2 expression. IMMUNOHISTOCHEMISTRY OF LATE UV-INDUCED LESIONS The dermo-epidermal junction of UV-A- and UV-B-induced DLE lesions consistently showed T-helper cells. In both genuine and in UV-A and UV-B induced SCLE lesions an increased number of suppressor/cytotoxic Figure 5. Histopathology of a late UV-B induced lesion (day 14) in an SCLE patient. Thin epidermis with slight vacuolar alteration and smudging at the dermoepidermal junction. Few necrotic keratinocytes. Lichenoid infiltrate in the papillary dermis with epidermotropism. (Hematoxylin and eosin, magnification 40.) T cells was observed. B cells were not detected in this region. The number of Langerhans cells was reduced. In genuine lesions of both subsets, HLA- DR expression on basal keratinocytes was increased. This was especially observed in SCLE lesions, both UV induced and genuine. The dermal infiltrate of both subtypes was primarily composed of T cells. B cells and macrophages were rarely observed. The deeper dermal infiltrate (predominantly DLE lesions) revealed a helper/suppressor ratio of 2:1. B cells, when present at all, were observed only in genuine and in late UV-induced DLE. (Teikemeier G, Goldermann R, Lehmann P, Goerz G, Plewig G, Kind P: Chronological immunohistological studies of UV-Aand UV-B-induced lesions of cutaneous lupus erythematosus (paper in preparation).) DIRECT IMMUNOFLUORESCENCE None of the UV-induced lesions showed immunoglobulin or complement deposits at the dermo-epidermal junction or around vessels. Specifically, UV-challenged skin of SCLE patients did not contain a pattern of epidermal dust-like particulate IgG deposition. This is in accordance with Cripps, who detected the first immunoglobulin deposits 6 weeks after

56S KIND ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY Table I. Testprotocol for the Reproduction of Cutaneous Lesions [24,26] Location Upper back or arm Size of testfield Lightsource UV-dose Reading 5 8cm UVASUN 3000 (330 460 nm) Mutzhas UV 800, Philipps TL 20 W/12 (285 350 nm), Waldmann 3 1.5 MEDUV-B 24, 48, 72 h up to three weeks Figure 6. Photosensitivity in different subgroups of LE measured by our standardized test protocol [24,26]. Photosensitivity was most pronounced in SCLE and LET. irradiation [10]. We did not study biopsies that were taken later than 6 weeks after irradiation. DISCUSSION The role of UV irradiation in the pathogenesis of LE has been the subject of many in vitro and in vivo studies over the last years [12 14,17 20,27]. To delineate the effects of different wavelengths, previous investigators tried to reproduce LE lesions by exposure to artificial sources of UV irradiation [7,8]. Due to different radiation sources, differing test protocols, and the limited number of patients, conflicting results were gained from these studies [7]. Nevertheless, most studies state that the action spectrum of LE is confined to wavelengths shorter than 320 nm. However, we have shown that the action spectrum extends into the UV-A range [23]. These data have been recently confirmed by other authors [22,31,32]. The development of positive phototest reactions in LE has been considerably slower than in other photodermatoses [23]. Because of the latency period it may be difficult for the patients to link sun exposure to skin lesions. In our studies, the MED UV-B was not decreased and a prolonged erythematous response was not observed. Other investigators observed a lowered MED in the majority of their SLE patients and in part of their DLE patients [32]. A long-lasting erythematous response was observed in SLE, SCLE, and DLE patients [32]. These differences may be due to differing test protocols. Time of reading might be essential because the development of lesions often takes weeks. In our study, photosensitivity in LE patients seems to be confined to exacerbation or induction of specific lesions by UV irradiation. There was a weak correlation between a positive history for UV sensitivity and phototest reactions [23]. Normal test reactions were found in 34% of patients who felt that they react abnormally to sunlight. Wolska et al also found a positive correlation with sun sensitivity [32]. These authors used a quite different test protocol. Induction of lesions was observed after a single UV exposure. From our experience, however, single UV exposure is unlikely to induce clinical lesions. This discrepancy may be explained by inaccurate medical histories given by the patients, by photosensitivity not being a permanent feature, or by provocative phototesting still missing some of the UV-inducible cases. In an ongoing study we have tested other subgroups of CLE patients. Interestingly lupus erythematosus tumidus patients were even more photosensitive than SCLE patients (Fig 6). These findings must be delved into further, because of the difficult differential diagnosis. The large number of photosensitive DLE patients in other studies might be due to an overlap with the LET group. Antibody-dependent cellular cytotoxicity is currently the most important theory for photosensitive CLE [12,14]. For this patho-mechanism an antibody such as the SSA/Ro-antibody in SCLE patients has to be present in the serum. As the standardized test procedure allows a much better definition of photosensitivity, we have looked at the link between SSA/Ro-antibody and photosensitivity. In our study the SSA/Ro-antibody is a marker for SCLE, but there is no association between photosensitivity and frequency of the SSA/Ro-antibody in any of the subgroups studied (Figs 2 and 3) [25]. These findings are confirmed by other groups [32,33]. Without phototesting, Sutey observed photosensitivity in 4% of black SSA/ Ropositive patients and 55% in black SSA/Ro-negative patients. For white SSA/Ro-positive patients, 87% were photosensitive; for SSA/Ro-antibody negative patients only 57%. Today most studies show no correlation between photosensitivity and SSA/Ro-antibody [32 34]. Detection of the Ro-antibody is dependent on the technique used, which could explain these different frequencies of the antibody. Other reasons could be variations in the titer, a different patient population, or a different classification [35]. The most likely explanation is that there are additional factors responsible for photosensitivity [24,27,28]. Biopsies of induced lesions taken at different time intervals also make it possible to study the chronologic development of the lesions. Early lesions of all groups showed a discrete perivascular T-helper cell infiltrate. Velthuis could also demonstrate a superficial perivascular T-helper cell infiltrate in the upper dermis [22]. Later on the T-helper cells have more lichenoid features and in SCLE lesions more suppressor/cytotoxic T cells were observed. Lichenoid features are observed in a group of graft-versus-host-like diseases such as lichen planus, graft-versus-host disease, and LE. For these diseases it has been suggested that the epidermal cell damage is the result of a T-lymphocyte mediated attack on epidermal cell antigens [36]. Interestingly, in contrast to DLE, late SCLE lesions showed more CD8-positive cells in the upper dermis. Whereas in early lesions HLA-DR expression seems to be reduced by UV light, in later stages, especially in SCLE, keratinocytes express HLA-DR antigens on their surface. This observation suggests that HLA-DR-positive keratinocytes might stimulate epidermotropic T cells to release additional lymphokines such as IFN-y, resulting in augmentation of epidermal cell damage. Conceivably, some cell-surface molecules become immunogenic to helper T cells when la antigens are co-expressed on the membrane of keratinocytes [36,37]. Unfortunately in our system we cannot rule out that HLA-DR expression is secondary to epidermotropic cells. Thus, especially in SCLE, HLA-DRbearing keratinocytes could play an important role in the perpetuation of epidermal cell damage mediated by HLA-DR recognizing T cells. The data in the literature on direct immunofluorescence differ. In this study and in an older report no immunoglobulin or complement deposits could be detected in early UV-induced lesions [9,10], whereas in a recent study granular deposits were detected in 12 of 16 UV-induced lesions [22]. The presence of immunoglobulin deposits in the early phase of CLE is therefore still controversial. Humoral autoimmunity is somewhat unlikely, because the negative immunofluorescence and the late presence of very few B cells militate against a localized immunoglobulin production [38]. Although numerous new data have been accumulated, the pathophysiology of photosensitive CLE is still unknown. Photosensitivity in LE is in

VOL. 100, NO. 1, SUPPLEMENT, JANUARY 1993 PHOTOTESTING IN LUPUS ERYTHEMATOSUS 57S the UV-A and UV-B spectrums. It is dependent on local factors such as the site of the lesion. The presence and density of an antigen such as the SSA/ Ro antigen could vary at different body sites [39]. A serum factor is essential for the photosensitivity in some variants of CLE [27,28]. SSA/Ro-positive SCLE patients have an increased photosensitivity and show more damage to the dermo-epidermal junction. In addition, LET, a not well characterized variant, might be even more photosensitive. Thus the current clinical and experimental data suggest different but possibly overlapping mechanisms for photosensitive CLE. REFERENCES 1. Kaposi MK: Neue Beiträge zur Kenntnis des Lupus erythematodes. Arch Dermatol Syph 1:26 78, 1872 2. Hutchinson J: Harveian lectures on lupus. Br J Med 1:113 118,1888 3. Jesionek A: Richtlinien der modernen Lichttherapie. Strahlentherapie 7:41 65,1916 4. Freund H: Inwiefern 1st der Lupus erythematodes von allgemeinen Faktoren abhängig? Dermatol Wochenschr 89:1939 1946, 1929 5. Pusey WA: Attacks of lupus erythematosus following exposure to sunlight or other weather factors. Arch Dermatol Syph 388, 1915 6. Nobl G: Lupus erythematodes mit starker Reizung durch Höhensonne. Arch Dermatol 133:98 99, 1921 7. Epstein JH, Tuffanelli DL, Dubois EL: Light sensitivity and lupus erythematosus. Arch Dermatol 91:483 485, 1965 8. Baer RL, Harber LC: Photobiology of lupus erythematosus. Arch Dermatol 92:124 128, 1965 9. Freeman RG, Knox JM, Owens DW: Cutaneous lesions of lupus erythematosus induced by monochromatic light. Arch Dermatol 100:677 682, 1969 10. Cripps DJ, Rankin J: Action spectra of lupus erythematosus and experimental immunofluorescence. Arch Dermatol 107:563 567, 1973 11. Sontheimer RD, Thomas JR, Gilliam JN: Subacute cutaneous lupus erythematosus: a cutaneous marker for a distinct lupus erythematosus subset. Arch Dermatol 115:1409 1415, 1979 12. Lee LA, Weston WL, Krueger GG, Emam M, Reichlin M, Stevens JO, Surbrugg SK, Vasil K, Norris DA: An animal model of antibody binding in cutaneous lupus. Arthritis Rheum 29:782 788, 1986 13. Le Feber WP, Norris DA, Ryan SS: Ultraviolet light induces expression of selected nuclear antigens on cultured human keratinocytes. J Clin Invest 74:1545 1551, 1984 14. Norris DA, Lee LA: Antibody-dependent cellular cytotoxicity and skin disease. J Invest Dermatol 85:165S 175S, 1985 15. Stern RS, Docken W: An exacerbation of SLE after visiting a tanning salon. JAMA 255:3120, 1986 16. Tronnier H, Petri H, Pierchalle P: UV-provozierte bullöse Hautveränderungen bei systemischem Lupus erythematodes (abstr). Zentralbl Haut 154:617A, 1988 17. Bruze M, Forsgren A, Ljunggren B: Antinuclear antibodies in mice induced by long wave ultraviolet radiation (UVA). Acta Derm Venereol 65:25 30, 1985 18. Emerit I, Michelson AM: Mechanism of photosensitivity in SLE patients. Proc Nad Acad Sci USA 78:2537 2540, 1981 19. Golan DT: Exposure to UV light in the pathogenesis of systemic lupus erythematosus (SLE). Isr J Med Sci 24:360 362, 1988 20. Golan DT, Borle Y: Increased photosensitivity to near ultraviolet light in murine SLE. J Immunol 132:705 710, 1984 21. Kocheva IE: Action spectrum and mechanisms of UV radiation-induced injury in lupus erythematosus. J Invest Dermatol 85:140S 143S, 1985 22. Velthuis PJ, Van Weden H, Baart de la Faille H: The immunopathology of light provoked skin lesions in lupus erythematosus. Acta Derm Venereol 70:93 98, 1990 23. Lehmann P, Hölzle E, Kind P, Goerz G, Plewig G: Experimental reproduction of skin lesions in lupus erythematosus by UV-A and UV-Bradiation. J Am Acad Dermatol 22:181 187, 1990 24. Kind P, Goerz G: Derkutane Lupus erythematodes (LE). In: MacherE, KnopJ, BröckerEB (Hrsg). Jahrbuch der Dermatologie, Dr. Hans Biermann Wissenschaftliche Verlagsgesellschaft, Munster, pp 85 103, 1988 25. Kind P, Lehmann P: Photobiologie des Lupus erythematodes. Hautarzt 41:66 71, 1990 26. Kind P, Teikemeier G, Goldermann R, Lehmann P, Hölzle E, Goerz G, Plewig G: Chronological, immunohistochemical studies of UV-A and UV-B-induced lesions of cutaneous lupus erythematosus (abstr). J Invest Dermatol 93:459, 1989 27. Davis TL, Lyde CB, Davis BM, Sontheimer RD: Perturbation of experimental ultraviolet light-induced erythema by passive transfer of serum from subacute cutaneous lupus erythematosus patients. J Invest Dermatol 92:573 577, 1989 28. Iacobelli D, Blanche L, Hashimoto K: Passive transfer of photosensitivity by intradermal injection of vesiculobullous lupus erythematosus serum and ultraviolet light irradiation in the guinea pig. Photodermatol 4:288 295, 1987 29. Michelson AM: Mechanism of photosensitivity in systemic lupus erythematosus patients. Proc Natl Acad Sci USA 78:2533 2540, 1981 30. Zamansky GB, Minka DF, Deal DL: The in vitro photosensitivity of SLE skin fibroblasts. J Immunol 134:1571 1576, 1985 31. Welden van H, Velthuis PJ, Baart H de la Faille: Light-induced skin lesions in lupus erythematosus: photobiological studies. Arch Dermatol Res 281:470 474, 1989 32. Wolska H, Blasczyk M, Jablonska S: Phototests in patients with various forms of lupus erythematosus. IntJ Dermatol 28:98 103, 1989 33. Sutey PG, Gear AJ, Morrison RCA, Tikly M, De Beer M, Dos Santos L, Sher R: Photosensitivity and anti-ro(ssa) antibodies in black patients with systemic lupus erythematosus (SLE). Br J Rheumatol 28:321 324,1989 34. Mond CB, Peterson MGE, Rothfield NF: Correlation of anti-ro antibody with photosensitivity rash in systemic lupus erythematosus patients. Arthritis Rheum 32:202 204, 1989 35. Purcell SM, Lieu TS, Davies BM, Sontheimer RD: Relationship between circulating anti-ro/ssa antibody levels and skin disease activity in subacute cutaneous lupus erythematosus. Br J Dermatol 117:277 287, 1987 36. Shiohara T, Moriya N, Tanaka Y, Arai Y, Hayakawa J, Chiba M, Nagishima M: Immunopathologic study of lichenoid skin diseases: correlation of HLA-DR-positive keratinocytes or Langerhans cells and epidermotropic T cells. J Am Acad Dermatol 18:67 74, 1988 37. Morris SC, Cohen PL, Eisenberg RA: Experimental induction of systemic lupus erythematosus by recognition of foreign la. Clin Immunol Immunopathol 57:263 273, 1990 38. Synowski DR, Provost TT: Characterization of the inflammatory infiltrate in Lupus erythematosus lesions using monoclonal antibodies. J Rheumatol 10:920 924, 1983 39. Müller, WE, Mayet W, Bachmann M, Meyer zum Büschenfelde KH, Pfeifer K, Diehl-SeifertB, Schöder HC: Age-correlation changes in the lupus erythematosus antigen Ro, La, Sm and RNP of the thymus gland. Compr Gerontol A 1:40 44, 1987