We appreciate the comment by Dr Adams regarding the strength of the association between pathological assessment of skin lesions of toxic epidermal necrolysis (TENS) and prognosis.1 Toxic epidermal necrolysis is a rare, potentially life-threatening disease provoked by ingestion of various medications. As highlighted by the accompanying editorial,2 both nosological and pathophysiological issues remain unanswered for TENS. The conventional histological viewpoint for skin biopsy specimens from patients with TENS is pan-epidermal necrosis of keratinocytes accompanied by a paucicellular inflammatory infiltrate.3 The aim of our study was to define the histological spectrum (degree of mononuclear cell infiltrates) in TENS for patients with more than 30% surface area of sloughing and to determine if there were clinically useful correlations once subsets of patients were identified based on sparse, moderate, or extensive inflammatory infiltrates. Since the clinically based SCORTEN system (a validated predictor of TEN mortality) value is fairly reliable in predicting survival, our focus was on ascertaining whether the pathological-based scoring of the degree of mononuclear cell infiltration was as good as SCORTEN. Indeed, the statistical analysis of our data revealed a fairly similar predictive value when comparing the SCORTEN value with the extent of mononuclear cell infiltration.
Sinacore J, Gamelli R, Nickoloff BJ. A Consideration of Statistical Association—Reply. Arch Dermatol. 2005;141(12):1604-1605. doi:10.1001/archderm.141.12.1604-c