Analytical quantification of the inflammatory cell infiltrate and CD95R expression during treatment of drug-induced toxic epidermal necrolysis.

The treatment of drug-induced toxic epidermal necrolysis (TEN) remains unsatisfactory. Intravenous immunoglobulins (IVIg) and intravenous cyclosporin A (CsA) have shown some efficacy in short series of patients. We assessed the effects of IVIg and CsA on TEN lesional and apparently uninvolved skin using standard histology and immunohistochemistry. Cutaneous biopsies were taken from necrotic and clinically uninvolved TEN skin at admission (D1) before any treatment, and after a 5-day treatment (D5). Two IVIg-treated patients (0.75 g/kg/day), two CsA-treated patients (5 mg/kg/day) and two control patients only receiving supportive care were compared. Biopsies were examined by standard histology and immunohistochemistry using antibodies directed to CD68 antigen (macrophages), CD45R0 antigen (activated T lymphocytes), Factor XIIIa (dermal dendrocytes) and the CD95 receptor (apoptosis marker). The different cell densities were evaluated by computerized image analysis. The clinical outcomes with the different treatments were also recorded. There was no obvious difference in the duration of hospitalization in intensive care unit between the three groups but one patient passed away in each of the IVIg- and CsA-groups. At D5, no differences were found between the three groups in the histological and clinical rate of re-epithelialization, and in the evolution of T lymphocyte, macrophage and dendrocyte densities in the epidermis and dermis. However, the expression of the CD95 receptor was similarly and strongly abated at D5 in the epidermis of IVIg- and CsA-treated patients, while it was conversely increased in the two patients under supportive care only. Such a difference was found both in necrotic and uninvolved sites. IVIg and CsA treatments thus appeared to exert no obvious effect on the inflammatory infiltrate, but both abated the expression of the CD95 receptor in the skin of TEN patients. This effect did not seem sufficient to fully reverse the clinical evolution of the disease. It is inferred that IVIg and CsA do not completely abate the TEN process.