Search

• Background.— Toxic epidermal necrolysis is a potentially severe mucocutaneous affliction whose cause is usually drug related. To further characterize the nature of the dermal mononuclear infiltrate as well as the epidermal alterations observed by standard microscopy, we studied five cases of toxic epidermal necrolysis using labeled monoclonal antibodies. Results.— On clinically involved areas of skin, the following occurs: (1) the dermal infiltrate is composed mainly of activated T lymphocytes, with a predominant helper phenotype; (2) the number of Langerhans cells is decreased; and (3) keratinocytes express HLA-DR molecules, normally absent on their surface. Conclusions.— These findings, although not specific, are consistent with an immune cellular reaction, but they could also be linked to an inflammatory reaction initiated by epidermal damages whatever its primary mechanism. ( Arch Dermatol. 1992;128:50-53)

In Reply.— We thank Dr Koff for pointing to some differences between the results of several recent studies on the immunopathology of toxic epidermal necrolysis (TEN). As did other investigators, we observed the abnormal expression of HLA-DR by keratinocytes, a dermal infiltrate composed of T lymphocytes, mainly CD4 + , and the absence of cells with a phenotype of natural killer cells. 1 The main difference concerns the nature of the epidermal infiltrate. We observed a similar number of CD4 + and CD8 + mononuclear cells within the epidermis, while other investigators reported that all or most of the cells in the epidermis were CD8 + . 2,3 We suggest that these divergent findings resulted less from different case definition than from the different timing of the biopsies. We obtained biopsy specimens from "early" erythematous lesions, distinct from the areas of epidermal sloughing. Our pathologic samples exhibited little or no morphologic alteration of epidermal cells, in contrast

• Blister formation in skin graft donor or recipient sites is uncommon. We describe a 49-year-old female patient with bullae in sites of grafts used in the treatment of toxic epidermal necrolysis. Generalized loss of skin developed 3 weeks after she had ingested phenobarbital. Sixty days after the beginning of the toxic epidermal necrolysis, the reepidermization was only 80% and skin grafts were placed on lower-extremity and abdominal wounds using the first healed sites as donor sites. Several bullae and erosions were noted on grafted areas 3 weeks later. Skin biopsy specimens revealed separation at the dermoepidermal junction, and no autoantibodies were detected by direct and indirect immunofluorescence. Electron microscopy demonstrated that the blister was formed through the basal keratinocytes and that the dermoepidermal junction, including hemidesmosomes and anchoring fibrils, was normal. Immunofluorescence mapping was performed using polyclonal antibodies from the serum of patients with bullous pemphigoid and epidermolysis bullosa acquisita and monoclonal antibodies against GB3 antigen and collagen type VII. All but the bullous pemphigoid serum gave positive results; only faint and focal staining of the dermoepidermal junction was observed with bullous pemphigoid serum. These findings are the same as those encountered in hereditary epidermolysis bullosa simplex. A biopsy performed 1 year later in the same site as the first one revealed that bullous pemphigoid antigen was normally expressed. Keratinocytes autografted in the treatment of toxic epidermal necrolysis may become transiently, functionally abnormal because of the alteration of recipient sites. ( Arch Dermatol. 1991;127:1369-1374)