BACKGROUND: Drug-induced toxic epidermal necrolysis (TEN) is a rare life-threatening disease characterized by the extensive destruction of the epidermis contrasting with the discreteness of lymphoid cell infiltration. The precise pathomechanism of the disease remains unclear. METHODS: Skin specimens and peripheral and thoracic lymph nodes (LNs) were collected from 2 TEN patients. They were examined by conventional histology and immunohistochemistry using antibodies directed to factor XIIIa (dermal dendrocytes), CD1a (Langerhans cells), CD15 (granulocytes), CD20 (B lymphocytes), CD45RO (activated T lymphocytes), CD68 (macrophages) and the proliferation marker Ki-67. LNs from patients with mycosis fungoides, from subjects dead from acute cardiac failure or traumatic injuries, as well as metastasis-free sentinel LNs of cutaneous melanoma served as positive and negative controls. RESULTS: TEN LNs showed absence of germinal centers but a moderate hyperplasia of the paracortical T cell zone. Immunohistochemistry did not reveal any distinctive aspect in LN cellular densities between TEN and other control conditions except for the factor-XIIIa+ dendritic cells which were dramatically reduced in numbers in TEN LNs. This rarefaction in LNs contrasted with the great number of these cells in the skin of the same patients. CONCLUSION: The structure of TEN LNs rules out the involvement of an antibody-mediated response. As dendritic factor-XIIIa+ cells are involved in most cutaneous T-cell-mediated responses, their depletion in TEN LNs could explain the sparse lymphoid cell infiltrate in lesional TEN skin. Copyright 2003 S. Karger AG, Basel
BACKGROUND: Drug-induced toxic epidermal necrolysis (TEN) is a rare life-threatening disease characterized by the extensive destruction of the epidermis contrasting with the discreteness of lymphoid cell infiltration. The precise pathomechanism of the disease remains unclear. METHODS: Skin specimens and peripheral and thoracic lymph nodes (LNs) were collected from 2 TEN patients. They were examined by conventional histology and immunohistochemistry using antibodies directed to factor XIIIa (dermal dendrocytes), CD1a (Langerhans cells), CD15 (granulocytes), CD20 (B lymphocytes), CD45RO (activated T lymphocytes), CD68 (macrophages) and the proliferation marker Ki-67. LNs from patients with mycosis fungoides, from subjects dead from acute cardiac failure or traumatic injuries, as well as metastasis-free sentinel LNs of cutaneous melanoma served as positive and negative controls. RESULTS: TEN LNs showed absence of germinal centers but a moderate hyperplasia of the paracortical T cell zone. Immunohistochemistry did not reveal any distinctive aspect in LN cellular densities between TEN and other control conditions except for the factor-XIIIa+ dendritic cells which were dramatically reduced in numbers in TEN LNs. This rarefaction in LNs contrasted with the great number of these cells in the skin of the same patients. CONCLUSION: The structure of TEN LNs rules out the involvement of an antibody-mediated response. As dendritic factor-XIIIa+ cells are involved in most cutaneous T-cell-mediated responses, their depletion in TEN LNs could explain the sparse lymphoid cell infiltrate in lesional TEN skin. Copyright 2003 S. Karger AG, Basel