Accuracy of dermoscopic criteria for discriminating superficial from other subtypes of basal cell carcinoma

Background The management of basal cell carcinoma (BCC) depends, among other factors, on its histopathologic subtype. Although dermoscopic criteria of BCC have been investigated, the possible role of dermoscopy in predicting the tumor subtype remains unclear. Objectives We sought to assess the diagnostic accuracy of dermoscopic criteria for differentiating superficial BCC (sBCC) from other BCC subtypes. Methods Dermoscopic images of histopathologically confirmed BCCs were retrospectively evaluated for the presence of predefined criteria. Univariate and adjusted odds ratios were calculated. Discriminant functions were used to plot receiver operating characteristic curves. Results In all, 77 sBCCs and 258 non-sBCCs were included. Maple leaf-like areas, short fine superficial telangiectasia, multiple small erosions, and shiny white-red structureless areas were potent predictors of sBCC, each making its diagnosis over 5-fold more likely. Conversely, the presence of arborizing vessels, blue-gray ovoid nests, and ulceration gave 11-fold, 15-fold, and 3-fold increased possibility for the diagnosis of non-sBCCs, respectively. Based on the results of the multivariate analysis, we propose a diagnostic algorithm that can predict the diagnosis of sBCC with a sensitivity of 81.9% and a specificity of 81.8%. Limitations The retrospective design and the inclusion of only Caucasian patients are limitations. Conclusion Dermoscopy is reliable in differentiating sBCC from other BCC subtypes. © 2013 by the American Academy of Dermatology, Inc.