Lupus erythematosus with an erythema multiforme-like eruption

ARTICLE

The presence of erythema multiforme (EM)-like lesions has been reported in early subacute cutaneous lupus erythematosus (SCLE) [1], and in systemic lupus erythematosus (SLE) [2]. In 1963, Rowell et al. [3] described a distinctive subset of patients diagnosed with discoid lupus erythematosus (DLE) associated with EM-like lesions and a characteristic pattern of immunological abnormalities in the serum: 1) speckled type of antinuclear factor; 2) presence of rheumatoid factor and 3) precipitating antibodies to saline extract of human tissue (anti-Sj-T). It was also accompanied by perniotic-like lesions.

We present a patient diagnosed with lupus erythematosus (LE) in which subsequently an EM-like acute eruption appeared. This case shares many similarities with the so-called Rowell's syndrome.

Case report

In October 1997, a 27 year-old woman presented to the department of Dermatology with a two-month history of gradually appearing, small, erythematous, scaling patches on the face (Fig. 1). The lesions worsened with sun exposure. Her family and personal medical history was unremarkable. She also complained of asthenia, arthralgia, and low-grade fever. On physical examination a discrete malar erythema was also present. In addition, there were perniosis-like lesions on the fingers and on the ear lobes (Fig. 2).

Laboratory investigations showed leucopenia (3,500/mm³); VDRL (+) 1/1. TPHA (-); a mixed speckled-homogeneous antinuclear antibody (ANA) titer of 1:640; and a slight hypocomplementemia (C4: 15.4; normal range > 20). Double-stranded DNA antibodies, anti-Ro (SS-A), anti-La (SS-B) antibodies and rheumatoid factor were negative. Urinalysis, biochemical parameters, renal function studies, immunoglobulins, and chest X-ray were normal. Histological examination (Fig. 5) revealed lymphocytic superficial perivascular inflammatory infiltrates in the dermis. There were foci of vacuolar degeneration at the dermo-epidermal junction. Thickening of the basement membrane was also evident. On the basis of the American Rheumatism Association (ARA) criteria, the patient was diagnosed as having systemic lupus erythematosus. Treatment was begun with oral prednisone (tapering dose) and hydroxychloroquine sulfate. Sunscreens were also recommended.

Three months later, the patient returned with a rapidly progressive, acute, generalized eruption involving the face, trunk, upper and lower extremities. No provoking factor (drug, infection, sun) was found. On physical examination there were numerous target-like lesions (Fig. 3), that coalesced on the trunk and the extremities in a policyclic pattern (Fig. 4). Necrosis and erosions in the center of the lesions were clearly observed. There were also a few superficial ulcers in the oral mucosa. Surprisingly, the patient had good general health (moderate pruritus, low-grade fever). Laboratory studies did not show any differences from those previously mentioned. Histological examination revealed moderate vacuolar changes at the dermo-epidermal junction with frequent necrotic keratinocytes in all layers of the epidermis (Fig. 6). There were minimal dermal lymphocytic inflammatory infiltrates. Direct immunofluorescence studies in lesional and non lesional skin were negative. The biopsy shared features of EM and LE.

Treatment was started with high-dose oral prednisone and azathriopine with complete remission of the eruption in two weeks. No residual lesions remained. Gradual supression of prednisone and azathriopine was made. Hydroxychloroquine was added again without relapse. Two years later, the eruption recurred associated with arthralgias. Once more, no provocative factor was found.

Discussion

Since the first report of Rowell's syndrome [3], we have only found 14 additional cases in the literature in which the presence of EM-like lesions in LE is described [4-16] (Table I). It is a striking feature that almost all of them, as in our patient, do not fit all the diagnostic criteria of Rowell's original description, specially the presence of rheumatoid factor and anti-La antibody (this antibody is actually believed to be identical to anti-Sj-T, although some authors believe that anti-Sj-T is also similar to anti-Ro [12]). Perniotic lesions are also very rarely seen. Twelve cases have discoid lesions. Including our case there were 15 women and 4 men (ratio 3.75:1), with a clear female predominance, as in LE. Age ranged from 9 to 72 years, with a mean of 37 years. Provoking factors were suspected in five patients. Drugs may have been responsible in four (mepacrine, nitrofurantoin, doxycicline, phenolphthalein) [7, 10, 11 16] and sun exposure was claimed in one [5]. It is noteworthy that the only case that fits all the immunological criteria exposed by Rowell was the patient in whom doxycicline was suggested as the provoking factor and a "true" EM was diagnosed [11]. A speckled pattern of antinuclear factor was present in 13 cases (associated with a homogeneous pattern in two), homogeneous pattern alone was seen in two. Absence of ANA in serum was described in one patient. Rheumatoid factor was encountered only in three cases and in the four patients of the original report. Positive anti-Ro and/or anti-La antibodies were reported in four cases. Anti-Sj-T antibodies were only performed in the original report. Prognosis was relatively good in the cases diagnosed as DLE, although recurrent episodes were common. When patients had enough criteria to be classified as SLE, as happened with our patient, evolution was more variable. Mild systemic involvement was reported in six cases, prognosis was poor in two and very poor (death) in other two. With this data, it is difficult to be sure if Rowell's syndrome has any prognostic significance.

At the present time there seems to be enough evidence to classify Rowell's syndrome within the SCLE subset, rather than as a distinct entity. Some authors even question if it really exists [16]. Rowell's patients were reported before Sontheimer's description of SCLE. In a review of 8 SCLE patient series [1], the clinical and immunological characteristics of Rowell's syndrome were not infrequently found in SCLE: discoid lesions were present in 10-35% of cases, positive ANA were found in 49-81%, anti-Ro in 28-100%, anti-La in 12-71%, and rheumatoid factor was reported in 36-48% of cases. Finally, early lesions of annular-policyclic pattern of SCLE may resemble EM.

A coincidental appearance of a true EM in patients with LE can not completely ruled out in some cases. In our patient, after an extensive search, no provocative factor was found. The only suspicious factor could be hydroxychloroquine, but it was subsequently reinstalled without relapse. EM has very rarely been reported associated with DLE and SLE [3]. The causative factor of EM in these cases is not very clear: it could be drugs, concomitant infections, or may be lupus itself. On the other hand, care needs to be exercised when diagnosing EM in LE patients, because EM-like lesions may be atypical features of LE. LE may even simulate toxic epidermal necrolysis [16], and the presence of necrotic keratinocytes in SCLE lesions is not so infrequent [17].

CONCLUSION

In conclusion, we have described a case of lupus erythematosus associated with an erythema multiforme-like eruption, something very rarely reported in the literature. Our patient should be included in the SCLE subset of lupus erythematosus. There are some similarities with the so-called Rowell's syndrome, but it seems probable that the cases included under this eponymous are a mixture of 1) LE patients with a true episode of EM, and 2) LE patients of the SCLE subset with target-like lesions.

Article accepted on 10/3/00

REFERENCES

2. Gilliam JN, Sontheimer RD. Skin manifestations of SLE. Clin Rheum Dis 1982; 8: 207-18.

3. Rowell NR, Swanson-Beck J, Anderson JR. Lupus erythematosus and erythema multiforme-like lesions. Arch Dermatol 1963; 88: 176-80.

4. Jablonska S, Blaszcyk M, Chorzelski T. Syndrome de Rowell: lupus erythemateux avec des lesions coexistantes de type erytheme polymorphe bulleux. Med Hyg 1972; 1026: 1390-3.

5. Sánchez M, Cruces J, Fonseca A, et al. Lupus eritematoso sistémico con manifestaciones cutáneas de eritema exudativo multiforme. Acta Dermo-Sifiliogr 1975; 66: 631-8.

6. Fabbri P, Panconesi E. Syndrome de Rowell: étude clinique et immunopathologique de deux cas. Ann Dermatol Syphiligr 1975; 102: 405.

7. Lawrence CM, Marshall TL, Byrne JPH. Lupus erythematosus associated with erythema multiforme-like lesions in identical twins. Br J Dermatol 1982; 107: 349-56.

8. Heyl T, Meyer T. Lupus erythematosus with erythema multiforme (Rowell´s syndrome). South Afr Med J 1984; 65: 568-9.

9. Crickx B, Crickx L, Vissuzaine C, et al. Lupus érythemateux chronique avec lésions à type d'érythème polymorphe: syndrome de Rowell. Ann Dermatol Venereol 1985; 112: 705-6.

10. Chapman JA. An unusual nitrofurantoin-induced drug reaction. Ann Allergy 1986; 56: 16-8.

11. Lewis Jones MS, Evans S, Thompson CM. Erythema multiforme occurring in association with lupus erythematosus drug therapy with doxycycline. Clin Exp Dermatol 1988; 13: 245-7.

12. Parodi A, Drago EF, Varaldo G, et al. Rowell´s syndrome. J Am Acad Dermatol 1989; 21: 374-7.

13. Fiallo P, Tagliapietra AG, Santoro G, et al. Rowell´s syndrome. Int J Dermatol 1995; 34: 635-6.

14. Fitzgerald EA, Purcell SM, Kantor GR, et al. Rowell´s syndrome: report of a case. J Am Acad Dermatol 1996; 35: 801-3.

15. Chua SH, Giam YC, Sim CS. Systemic lupus erythematosus with erythema multiforme-like lesions and histiocytic necrotizing lymphadenitis- a case report. Ann Acad Med Singapore 1996; 25: 599-601.

16. Shteyngarts, Warner MR, Camisa C. Lupus erythematosus associated with erythema multiforme: Does Rowell´s syndrome exist? J Am Acad Dermatol 1999; 40: 773-7.

17. Bielsa I, Herrero C, Font J, et al. Lupus erythematosus and toxic epidermal necrolysis. J Am Acad Dermatol 1987; 16: 1265-7.

18. Herrero C, Bielsa I, Font J, et al. Subacute cutaneous lupus erythematosus: Clinicopathologic findings in thirteen cases. J Am Acad Dermatol 1988; 19: 1057-62.