RESUMO
BACKGROUND: Frozen section (FS) is often performed when histopathological evaluations are urgently required for implementation of therapeutic measures. In dermatology, this method is most commonly used to evaluate excision margins of tumors. FS are also routinely employed to differentiate toxic epidermal necrolysis from staphylococcal scalded skin syndrome. However, little is currently known about the performance of FS in the diagnosis of inflammatory dermatoses. OBJECTIVES: To compare histopathological diagnoses in a series of patients with a clinical diagnosis of an inflammatory dermatosis for which FS and paraffin-section (PS) specimens were obtained on the same day. METHODS: We conducted a single-center retrospective analysis of 43 cases. All histological slides were reviewed by a single dermato-pathologist. Concordance was calculated between FS and PS. RESULTS: Patients were divided into three groups according to diagnosis: papulosquamous diseases (group I), drug eruptions (group II), and a heterogeneous group (group III) that included cases of bullous vasculitis and Sweet syndrome. Among the three groups, the results of FS and of PS were discordant only in five cases (5/43, 11.6%). Compared to PS, FS had a sensitivity of 92.9% [95% confidence interval (95%CI) 64.17-99.63%] and a specificity of 100% in group I, sensitivity of 90.9% (95%CI 57.12-99.52%) and specificity of 100% in group II, and sensitivity of 83.33% (95%CI 60.78-94.16%) and specificity of 100% in group III. The degree of agreement between the results of the FS and of the PS was almost perfect (kappa = 0.95, 0.93 and 0.85 respectively). CONCLUSIONS: This study suggests that FS is a valid approach for the rapid diagnosis of inflammatory dermatoses. This method is as specific as PS, although it is less sensitive.