Drug-induced Stevens-Johnson syndrome/toxic epidermal necrolysis.

Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) is a rare (occurring in approximately 2 to 3 people/million population/year in Europe and the US), life-threatening, intolerance reaction of the skin. It is most often caused by drugs (most commonly sulfonamides, nonsteroidal anti-inflammatory drugs, antimalarials, anticonvulsants, and allopurinol). SJS/TEN is characterized by a macular exanthema ('atypical targets') which focusses on the face, neck, and the central trunk regions. Lesions show rapid confluence, a positive Nikolsky's sign, and quickly result in widespread detachment of the epidermis and erosions. Mucosal, conjunctival, and anogenital mucous membranes are prominently involved. Histopathology shows satellite cell necrosis in the early stages progressing to full thickness necrosis of the epidermis, contrasting with rather inconspicuous inflammatory infiltrates of the dermis. Damage to the skin is thought to be mediated by cytotoxic T lymphocytes and mononuclear cells which induce apoptosis in keratinocytes expressing drug-derived antigens at their surfaces. No guidelines for the treatment of SJS/TEN exist since no controlled clinical trials have ever been performed. The controversy over whether systemic corticosteroids should be used to curtail progression is still unresolved; while many authors agree that corticosteroids do in fact suppress progression, it is obvious that they also greatly enhance the risk of infection, the complication which most frequently leads to a fatal outcome. It appears reasonable to only administer corticosteroids in the phase of progression and to withdraw them as soon as possible, and to add antibacterials for prophylaxis. Recently, in a small series of patients, intravenous immunoglobulins were presumed to be effective by the blockade of lytic Fas ligand-mediated apoptosis in SJS/TEN. However, these results have to be confirmed by large clinical trials. Supportive treatment and monitoring of vital functions is of utmost importance in SJS/TEN, and out-patient treatment is unacceptable. Recovery is usually slow, depending on the extent and severity and the presence of complications, and may take 3 to 6 weeks. Skin lesions heal without scars as a rule, but scarring of mucosal sites is a frequent late complication, potentially leading to blindness, obliteration of the fornices and anogenital strictures. There is no reliable laboratory test to determine the offending drug; diagnosis rests on the patient's history and the empirical risk of drugs to elicit skin SJS/TEN. Provocation tests are not indicated since re-exposure is likely to elicit a new episode of SJS/TEN of increased severity.