[en] The molecular mechanisms involved in the pathogenesis of toxic epidermal necrolysis (TEN) remain not fully understood. We report a unique case of antibiotic-induced TEN developed in a patient who also suffered from prolonged severe methotrexate-induced pancytopenia. The objective of the study was to explore the nature of the cutaneous inflammatory infiltrate and the density in dermal dendrocytes (DD). Immunohistochemistry was used to identify activated T lymphocytes (CD45R0), monocyte-macrophages (Mac 387, CD68), DD (Factor XIIIa), and Langerhans cells (CD1a). The proliferation marker (Ki67) and the antibody to Fas receptor (CD95R) were also used to assess the distribution of the germinative pool of keratinocytes and the FAS-related apoptotic process, respectively. Numerous Factor XIIIa+ DD were present in the papillary dermis with only sparce perivascular CD45RO+ T lymphocytes and scattered CD68+ or Mac 387+ macrophages. Double immunostainings revealed that a minority of Factor XIIIa+ DD co-expressed the CD68 glycoprotein (a marker of phagocytic activity). No cells co-expressed factor XIIIa and Mac 387 immunoreactivities. CD45RO+ T lymphocytes, CD68+ and Mac 387+ macrophages were absent in the epidermis. The expression of CD95R was present although restricted to the basal keratinocytes, while the L1-protein (Mac 387+) was diffusely present in the epidermis. Langerhans cells (CD1a+) were sparce, but normal in distribution. The presence of a great number of Factor XIIIa+ DD without any possible recent recruitment from bone marrow suggests that these cells differentiated from resident cells of the skin. Indeed, there was no co-expression of Factor XIIIa and L1-protein, thus showing the absence of recruitment from monocytes. The simultaneous over-expression of Factor XIIIa and CD68 in some DD indicates some phagocytic activity. In view of the absence of inflammatory cells in the epidermis, keratinocytes appeared responsible for their own destruction through CD95-mediated and/or calcium-dependent apoptotic pathways. This finding entails that TEN treatments should target the keratinocyte metabolism rather than the circulating inflammatory cells which presumably play a limited role, if any, in the epidermal destructive process.