Abstract
Drug-induced toxic epidermal necrolysis (TEN) is a rare but potentially lethal bullous disease whose complex pathomechanisms remain uncertain. The aim of the study was an exploratory attempt to assess TEN pathobiology using a combination of immunohistochemistry and proteomics. Skin blister fluid (BF) and serum were collected in a patient in the early TEN stage at day (D) +4 of evolution and three days later (D +7). Intravenous cyclosporine A (CsA) treatment was initiated since D +4. Immunohistochemistry was performed on skin blister biopsies. In addition, proteomic analyses compared the BF protein content before and at the issue of the 3-day CsA treatment. Proteins were selected according to their prominent differential abundance in BF between D+4 and D+7, when influenced by lesional skin cells, but not in serum. Among 300 proteins, four were considered. Glutathione transferase π was related to oxidative stress in TEN epidermis. The monocyte differentiation antigen CD14 and myeloperoxidase indicated macrophage activation. The proinflammatory S100-A8 protein probably originated from activated keratinocytes and/or macrophages. These proteomic findings were in line with immunohistochemistry and supported the prominent involvement of keratinocytes and macrophages in TEN pathomechanism. As opposed to CD14, other proteins were mainly present in BF at D+7, confirming that CsA expressed little effect, if any, on the activity of keratinocytes and macrophages in the present TEN patient. Of note, the present exploratory study using proteomic analyses in a single TEN case supports a pathogenic hypothesis without establishing any firm conclusion.
Keywords: Toxic epidermal necrolysis, proteomics, immunohistochemistry, macrophage, keratinocyte, Histopathological, biopsy, pancytokeratin, Protein, granulocytes