Microsoft Word - DRM249BF

Jean-Claude Roujeau, MD, Service de Dermatologie, Hôpital Henri-Mondor, F–94010 Créteil (France) One hundred and thirty years after the original description of von Hebra, erythema multiforme (EM) is still giving rise to many controversies on all the aspects which constitute a disease as an entity. Clinical definition and nosology, pathology, etiology, physiopathology and treatment remain the subjects of often confuse debates. Nosology In the last 30 years it became widely accepted that EM minor, EM major, ectodermosis erosiva pluriorificialis, Stevens-Johnson syndrome (SJS) and toxic epidermal necroly-sis (TEN) were all part of a single ‘EM spectrum’. That unifying concept, prevalent in major textbooks of dermatology, has probably contributed to the current confusion. In fact there is little resemblance between the two ends of the spectrum. It was recently proposed to split again the EM spectrum and to come back as close as possible to the original clinical descriptions [1]. The denomination of EM should be restricted to acrally distributed typical targets or raised edematous papules fitting the original descriptions (erythema iris, erythema papulatum). Depending on the presence or absence of mucous membrane erosions the cases could be classified as EM major or EM minor. In keeping with the original description, the denomination of SJS should be used for a syndrome characterized by mucous membrane erosions and widespread blisters, often predominant on the chest, arising on eryfhematous or pur-puric macules, clinically quite different from targets. A retrospective reclassification of 63 cases of ‘EM major/SJS’ according to the above-mentioned criteria supported the validity of that separation by showing different etiologies. EM major was mainly related to herpes simplex virus (HSV) infection and SJS to drug reactions [unpubl. data]. In my mind SJS and TEN are probably severity variants of the same drug-induced process. Anyhow until this has been demonstrated on the basis of common etiology and physiopathology, it seems preferable to keep the two denominations (1) for analytic purposes and (2) as the severity of SJS and TEN is very different. Because the extent of necrolysis is a major prognosis factor, we proposed the limits of 10 and 30% of the body surface area involved by epidermal detachment as arbitrary boundaries between SJS, ‘overlap SJS/ΓEN’ and TEN [1]. Whether all the categories proposed represent distinct etiopathologic entities will require further clinical, epidem-iologic and laboratory investigations. Anyhow that ‘splitting’ attitude is a prerequisite to such studies by creating more homogeneous subgroups. Physiopathology The pathogenetic mechanisms responsible for EM remain unknown in spite of many recent advances. In HSV-associated cases, herpes simplex antigens and DNA are frequently detected in skin lesions of EM [2]. Isolation of infective HSV is usually negative, and pathology does not show cytopathogenic effects. It is therefore postulated that the lesions are not directly caused by