Author reply



DOI 10.1002/cncr.11244

Public health strategies emphasize the worldwide need for the early diagnosis and treatment of melanoma. In this context, dermoscopy should allow clinicians to increase their diagnostic sensitivity– (i.e., their ability to detect the highest possible number of melanomas).

The sensitivity of dermoscopy as a diagnostic test can be measured roughly by independent clinical and dermoscopic examinations in a set of pigmented skin lesions (PSLs) that subsequently undergo histopathologic examination. Using this approach, some studies1–4 already have reported the benefits of dermoscopy; in fact, this diagnostic procedure has been demonstrated to have a sensitivity that is up to 30% higher than the macroscopic diagnosis of melanoma.

Our study, recently published in Cancer,5 was not aimed at testing the diagnostic sensitivity of dermoscopy, but rather the dermoscopic and histopathologic-interobserver agreement concerning PSLs that already had been considered to be clinically and/or dermoscopically equivocal and subsequently were excised. With reference to dermoscopy, our eight panelists were asked to assess PSLs that had been judged a priori to demonstrate no clear-cut features of benignity (i.e., equivocal melanocytic lesions). Therefore, the series we collected in our study did not reflect the daily clinical setting, in which the majority of PSLs examined do not require surgical excision.

Obviously, the final determination regarding PSLs suspected to be melanoma is best left to the pathologist. Indeed, our study on 107 lesions demonstrated a nearly excellent histopathologic interobserver agreement, with an even better Schouten kappa value compared with previously reported studies.6, 7 However, can we be satisfied when we consider that the requested histopathologic determination was not unanimous in 29 of 107 cases? There is an obvious need for better standardization and greater reproducibility of histopathologic criteria for the diagnosis of PSLs. Based on the results of our study, we would like to challenge the diagnostic monopoly of histopathology, at least for the diagnosis of dermoscopically equivocal PSLs.

Finally, regardless of the final histopathologic diagnosis, a question remains to be answered: does dermoscopy really improve the ability of physicians to detect melanoma? In this context, the most reliable method for measuring the diagnostic sensitivity of dermoscopy is not the comparison of dermoscopic evaluations with the clinical and the histopathologic diagnoses, but rather the follow-up data regarding patients who have been sent home with lesions judged to be benign; these data should provide an idea of the true rates of dermoscopic false-negative diagnoses, which to our knowledge currently are unknown.

Gerardo Ferrara M.D.*, Giuseppe Argenziano M.D.†, Lorenzo Cerroni M.D.‡, H. Peter Soyer M.D.‡, * Pathologic Anatomy and Histology Service, “G. Rummo” General Hospital, Benevento, Italy, † Department of Dermatology, Second University of Naples, Naples, Italy, ‡ Department of Dermatology, University of Graz, Graz, Austria.