Volume 9 Number 3 September, 1983
phoid cells in x-irradiated and nonirradiated mice. J Natl Cancer Inst 23:53-73, 1959. 7. Parkman R, Mosier D, Umansky I, et al: Graft-versushost disease after intrauterine and exchange transfusions for hemolytic disease of the newborn. N Engl J Med 290: 359-363, 1974. 8. Schwarzenberg L, Mathe G, Amiel JL, et al: Study of
factors determining the usefulness and complications of leukocyte transfusions, Am J Med 43:206-213, 1967, 9. Holley TR, Van Epps DE, Harvey RL, et al: Effect of high doses of radiation on human neutrophil chemotaxis, phagocytosis and morphology. Am J Pathol 75:61-72, 1974.
Photosensitive recurrent erythema multiforme* James E. Fitzpatrick, Major, MC, U S A , Paul B. T h o m p s o n , M . D . , John L. Aeling, M . D . , and Clark Huff, M . D .
Aurora, CO A case report of photosensitive recurrent erythema multiforme occurring in a 31-year-old man is presented herein. Lesions clinically and histologically resembling erythema multiforme were reproduced with light testing. The patient's lesions appeared to respond to both oral prednisone and hydroxychloroquine. (J AM ACAD DERMATOL 9:419-423, 1983.)
E r y t h e m a m u l t i f o r m e occurring in a photodistribution has b e e n infrequently reported, only six case reports having appeared in the English medical literature. 1'2 Recurrent e r y t h e m a multiforme occurring in a photosensitive distribution has never been reported. In 1954, Anderson et al reported an epidemic of " j u v e n i l e spring e r u p t i o n , " a photosensitive dermatitis occurring in school age children that clinically r e s e m b l e d erythema multiforme, although the diagnosis was not substantiated with b i o p s i e s ) W e wish to report a case of recurrent e r y t h e m a multiforme in which we were able to reproduce clinical lesions with light testing, and to c o m m e n t on our treatment o f this disorder.
CASE REPORT The patient was a 31-year-old white man who initially presented to another clinic in the summer of 1975 From the DermatologyService, Fitzsimons Army Medical Center. Reprint requests to: Major James E. Fitzpatrick, DermatologyService, FitzsimonsArmy Medical Center, Aurora, CO 80045. *The opinions or assertions contained herein are the views of the authors and are not to be considered as reflecting the viewsof the Department of the Army or the Departmentof Defense.
with a pmritic, burning, annular dermatitis on the backs of both hands. An initial diagnosis of tinea corporis was made, and the patient was begun on oral griseofulvin and topical antifungal therapy. He was referred to the Dermatology Clinic at Fitzsimons Army Medical Center in the spring of 1976 because of the failure of antifungals to cure the dermatitis. History revealed that the lesions usually appeared in crops on sun-exposed areas. Individual lesions lasted 2 to 3 weeks and healed without scarring. The patient had not previously been aware that the lesions were related to sunlight exposure. He noted that occasional lesions arose in non-sun-exposed areas, such as the palms or the scrotum. He had never noticed the appearance of lesions following sunlight exposure through window glass. Family history revealed that he was of Irish and Welsh extraction. He specifically denied previous personal or family history of photosensitive dermatitis. He denied any medical problems except for seasonal allergic rhinitis, atopic dermatitis, and a vague history of "cold sores" on his lip, which he could not correlate with his dermatitis. He was on no medication except for various antihistamines, which he took intermittently for his allergic rhinitis. Physical examination revealed erythematous papules and plaques with some typical target lesions or 419
Fitzpatrick et al
Journal of the American Academy of Demlatology
Fig. 1. Typical for plaques and target lesions or erythema multiforme on the upper arm. Lesions stop at the upper border of a desquamating sunburn. Fig. 2. Typical target lesions of erythema multiforme restricted to areas of sunburn on patient's back. Fig. 3. Erythematous papules compatible with erythema multiforme developing in phototest site 72 hours after exposure. erythema multiforme. The lesions were in a photodistribution on the extensor surfaces of both arms and the tips of the ears (Fig. 1). The following laboratory studies were performed and were within normal limits: Chest roentgenogram, electrolytes, liver function tests, complete blood cell count with white blood cell differential, antinuclear antibody, SSA (anti-Ro) antibody, and urinalysis. A Tzanck preparation from the crusted center of one lesion was negative for viral giant cells. A 4-mm punch biopsy of an early lesion from the dorsum of the left arm was stained with hematoxylin-eosin stain
for light microscopic study. The specimen demonstrated a superficial perivascular lymphohistiocytic infiltrate, papillary edema, endothelial swelling, vacuolar alteration of basal keratinocytes, and necrotic keratinocytes. A diagnosis of photosensitive erythema multiforme was made. The patient showed some improvement on treatment with sun avoidance, topical sunscreens, and topical corticosteroids, but he continued to have multiple recurrent episodes of erythema multiforme following sunlight exposure. In August, 1978, an evaluation in
Volume 9 Number 3 September, 1983
the Allergy Clinic disclosed roentgenographic evidence of maxillary and ethmoidal sinusitis, and significant positive reactions to various trees, grasses, weeds, and house dust with prick and intradermal testing. At this point, antihistamine therapy was intensified and he was begun on hyposensitization shots to grasses, trees, weeds, and house dust. His allergic rhinitis improved and his sinusitis resolved without noticeable improvement of his recurrent erythema multiforme. In April, 1979, he returned to the Dermatology Clinic with recurrent erythema multiforme limited to areas of a sunburn received on the back of his arms and lower portion of his back 24 to 36 hours earlier (Fig. 2). Following resolution of these lesions it was elected to phototest his back, and he was given one 10-second exposure without window glass, using a Burdick UV800 hot quartz lamp which emits both ultraviolet A and ultraviolet B (UVA and UVB). The patient developed erythematous papules within 36 hours in the site of the light-tested area (Fig. 3). Some of the lesions were similar to his primary lesions but many were smaller. A 4-ram punch biopsy was taken from a lesion that was 48 hours old and was bisected for microscopic study using hematoxylin-eosin stain and direct immunofiuorescence. A 6-ram punch biopsy was taken from a lesion that was 72 hours old for repeat hematoxylineosin stain. Both histopathologic specimens demonstrated a superficial perivascular lymphohistiocytic infiltrate with papillary edema, endothelial swelling, vacuolar alteration of basal keratinocytes, sparse necrotic keratinocytes, and slight spongiosis. Both biopsies were compatible with erythema multiforme, with the older biopsy being the more characteristic. Direct immunofluorescence studies for IgG, IgA, IgM, C3, and fibrinogen were read as negative. Serum drawn 48 hours after light testing demonstrated increased cryoprecipitable protein and was positive for IgG, IgM, and IgA antibodies to type I herpes simplex antigens by enzyme immunoassay. At this time the patient was begun on 40 mg of oral prednisone per day; this resulted in prompt clearing of his disease and prevented further recurrences for the next 6 weeks as the dose was gradually tapered. In June, 1979, the prednisone was discontinued and the lesion recurred within 3 days of the final dose. He subsequently received separate clinical trials of sulfapyridine and dapsone without improvement. The patient continued to develop recurrent photosensitive erythema multiforme, and in August he was begun on 200 mg of oral hydroxychloroquine per day. Within 7 days, all existing lesions cleared and no new lesions developed. Three weeks later, while still on hy-
Photosensitive et),thema multiforme 421
droxychloroquine, he was again light-tested on his back with 10-, 20- and 30-second exposures, with and without window glass. No clinical lesions were reproduced. Four weeks later, the hydroxychloroquine was discontinued and the patient remained free of lesions until February, 1980, when he developed erythema multiforme on his forearms following a sunburn. He was again light-tested with 10- and 20-second exposures with and without window glass, but no clinical lesions were reproduced. The patient continued to develop flares of his disease until May, 1980, when he was again begun on 200 mg of oral hydroxychloroquine per day. He remained free of disease throughout the summer of 1980 while on this medication. In the summer of 1980, the patient was discharged from the Air Force and was no longer eligible for military medical care. His hydroxychloroquine was stopped at this time. He has continued to have flares of his cutaneous disease associated with excessive sun exposure. DISCUSSION Erythema multiforme is infrequently found in such a striking photodistribution as was seen in this case but is usually found in a symmetric distribution in areas that are subject to trauma, such as skin o v e r joints, palms, genitalia, and extensor surfaces o f extremities, z Recent studies of the pathogenesis suggest that erythema multiforme is the result of circulating immune c o m p l e x e s which become lodged in the walls of the vessels in the papillary dermis.4-~ It has been shown that factors such as histamine, which cause vasodilatation and increased permeability, will result in the deposition of immune complexes in normal skin of patients with active erythema multiforme. 4 It is tempting to speculate that the photo-induced lesions in this case resulted from the release of ultraviolet radiation-induced mediators, such as kinins, 7 prostaglandins, s or histamine, :~ which in turn increase vascular permeability and promote the deposition of immune complexes. The failure to demonstrate immune c o m p l e x e s by direct immunofluorescence in experimentally produced lesions was probably due to the fact that the biopsy was taken from a lesion more than 24 hours old, at which time immune complexes are usually no longer demonstrable. 4 The demonstration of increased cryoprecipitable protein is indirect evidence that immune c o m p l e x e s were present.
Journal of the American Academyof Dermatology
Fitzpatrick et al
The source o f antigen stimulus in this patient is not clear. The most common antigen associated with recurrent erythema multiforme is derived from recurrent herpes simplex infections, '° This patient had a vague history of herpes labialis, but no lesions have ever been observed during the many visits over a 4-year period. Subclinical recurrent herpetic infection is a theoretical source of antigen in this case. The demonstration of IgG, IgM, and IgA antibodies to type I herpes simplex antigens suggests that this may be the source of antigen. Sinusitis with possible infection was considered as a potential source of antigen, but symptomatic and radiographic clearing of our patient's sinusitis did not correlate with improvement. The differential diagnosis in a patient who presents with recurrent papules, plaques, target lesions, and annular lesions in a photodistribution includes photoallergic drug reactions, polymorphous light eruption, and lupus erythematosus. In this case the diagnosis of erythema multiforme is supported by clinical course, presence of typical target lesions, biopsy findings, and the failure to detect antinuclear and SSA (anti-Ro) antibodies. Our inability to reproduce lesions experimentally on subsequent attempts prevented us from determining the action spectrum in this case; this may have occurred because the ultraviolet radiation exposures did not coincide with the presence of circulating immune complexes. The treatment of recurrent erythema multiforme consists of conservative topical or systemic corticosteroids. Anecdotal reports suggest that systemic corticosteroids may shorten the course and reduce the severity of the disease, although double-blind studies have never been done to substantiate this. This patient appeared to respond dramatically to systemic corticosteroids given over a 6-week period, but the patient's rapid weight gain was an unacceptable side effect. The treatment of recurrent erythema multiforme occurring in a photodistribution must include patient education regarding sun avoidance and the use of topical sunscreens. Antimalarials have been reported to be effective in the treatment of other photosensitive dermatoses, such as polymorphous light eruption,
solar urticaria, and lupus erythematosus. Although the exact mechanism by which antimalarials are effective in some photosensitive diseases is not known, investigators have thus far failed to demonstrate that they function by blocking the penetration of ultraviolet rays. It has been shown that chloroquine interferes with conversion of arachidonic acid to PGF2; if the release of ultraviolet radiation-induced prostaglandins does indeed play a role in the pathogenesis, it may explain the apparent success of antimalarials in our patient. Recent studies have also suggested that antimalarials may block antibody-antigen binding and, to some extent, may dissolve preformed immune complexes." If the latter mechanism is correct, it would suggest that antimalarials might also be effective in the treatment of non-sunlight-induced erythema multiforme. This patient appeared to respond dramatically to hydroxychloroquine during both summers in which it was used, suggesting that it may be of some benefit. However, because the course of erythema multiforme is unpredictable, no definite conclusions regarding its efficacy can be made. We feel that it is worthy of further clinical trials in some cases of recurrent photosensitive erythema multiforme not responding to more conservative management. Although we have seen other patients who have had skin lesions of erythema multiforme occurring predominantly on sun,exposed skin, this patient's photodistribution was exceptionally dramatic. It is our clinical impression that erythema multiforme can be aggravated by ultraviolet light exposure, particularly in patients who have recurrent lesions secondary to herpes simplex infections. We feel it is important to recognize this association when managing patients with this frustrating problem and advise them to use appropriate sunscreens and sun avoidance in conjunction with other recommended therapies. We thank Dr. James Patterson for critically reviewing this manuscript. REFERENCES
l. Keil H: Erythema multiforme exudativum (Hebra): A clinical entity associated with systemic features. Ann Intern Med 14:449-494, 1941.
Volume 9 Number 3 September, 1983
2. Huff C, Weston WL: The photodistribution of erythema multiforme. Arch Dermatol 116:477, 1980. 3. Anderson D, Wallace t-IJ, Howes EB: Juvenile spring eruption. Lancet 1:755-756, 1954. 4. Wuepper KD, Watson PA, Kazmierowski JA: Immune complexes in erythema multiforme and the StevensJohnson syndrome. J Invest Dermatol 74:368-371, 1980. 5. Bushkell LL, Mackel SE, Jordan RE: Erythema multiforme: Direct immunofluorescence studies and detection of circulating immune complexes. J Invest Dermatol 74:372-374, 1980. 6. HuffJC, Weston WL, Can: RI: Mixed cryoglobulinemia. I. Clq binding and skin immunofluorescence in erythema multiforme. J Invest Dermatol 74:375-377, 1980.
Photosensitive elytherna multiforme
7. Epstein JH, Winkelmann RK: Ultraviolet light-induced kinin formation in human skin. Arch Dermatol 95:532536, 1967. 8. Snyder DS, Eaglestein WH: Intradermal anti-prostaglandin agents and sunburn. J Invest Dermato162:47-50, 1974. 9. Gilchrest BA, Soter NA, Stoff JS, et al: The human sunburn reaction: Histologic and biochemical studies. J AM ACAD DERMATOL5:411-422, 1981. I0. Huff JC, Swinehart JM, Weston WL, et al: Immune complexes involving herpes antigen in erytbema multiforme. Clin Res 27:242a, 1979. 11. Logan WS: Antimalarials. Prog Dermatol 14:1-6, 1980.
Aleukemic leukemia cutis: Juvenile chronic granulocytic leukemia presenting with figurate cutaneous lesions Neil S. Heskel, M . D . , * Clifton R. White, M . D . , * Sarah Fryberger, M . D . , * * Robert C. Neerhout, M . D . , * * Mary Spraker, M . D . , * * * and Jon M. Hanifin, M . D . *
Portland, OR, and Atlanta, GA We report a 31/i-year-old girl who developed a figurate cutaneous eruption. Distinctive findings in her skin biopsies, as well as unusual red cell characteristics, in the presence of a normal peripheral smear and bone marrow biopsy, led us to suspect the diagnosis of preleukemic juvenile type chronic granulocytic leukemia. This is the first case we know of in which this diagnosis was suspected prior to abnormal findings in the peripheral smear or bone marrow. (J AM ACAD DERMATOL 9:423-427, 1983.)
O f the childhood leukemias, less than 20% are the acute granulocytic type, and f e w e r than 5% are
From the Departments of Dermatology* and Pediatrics,** Oregon Health Sciences University, and the Departmentof Dermatology, Emory University.-*** Supported by National Institutes of Health grant, Public Health Service CA 26044. Presented in part at the meeting of the Society for Pediatric Dermatology in Aspen, CO, July 15, 1982. Reprint requests to: Dr. Jon M. Hanifin, Oregon Health Sciences University, Departmentof Dermatology,3181 S.W. Sam Jackson Rd., Portland, OR 97201.
chronic granulocytic l e u k e m i a (CGL). Juvenile type chronic granulocytic l e u k e m i a ( J C G L ) is a subtype o f C G L which differs f r o m the m o r e c o m m o n adult C G L by its early age of onset, absence of the Philadelphia c h r o m o s o m e , characteristic r e d cell abnormalities, and p o o r r e s p o n s e to therapy. 1 Reported cutaneous manifestations of J C G L inelude eczematous eruptions, e r y t h e m a t o u s m a c u lopapular rashes, and indurated lesions with central clearing. X a n t h o m a s and caf6 au lait spots occur with increased frequency in J C G L . " W e report a 423