Discordance in the histopathologic diagnosis of melanoma at a melanoma referral center

Abstract

Histopathologic analysis remains the gold standard for the pathologic diagnosis of melanoma. Numerous histologic criteria are used to diagnose melanoma, but none alone are sufficient to establish this diagnosis. Therefore, differentiating between benign pigmented lesions and melanoma may be controversial. Although several studies have examined the interobserver variability in the pathological diagnosis of melanoma, the prevalence of discordant diagnoses of melanocytic neoplasms is unknown. We sought to examine the discordance rate of melanoma diagnoses referred to our pigmented lesion clinic, a subset of the University of California, San Francisco (UCSF) Department of Dermatology and Comprehensive Cancer Center Melanoma Center during a 2-year period. A total of 392 new patients given a diagnosis of thin melanoma (melanoma in situ, stage IA, stage IB) or benign nevus were referred in 2006 and 2007, with initial diagnoses rendered by an outside dermatopathologist or surgical pathologist. Subsequently, these specimens were re-evaluated by routine histopathologic examination at the UCSF Dermatopathology Service and a distinct diagnosis was rendered. The two available diagnoses were compared, and discordance was defined as the lack of agreement between two pathologists when rendering a benign versus malignant versus ambiguous diagnosis. The discordance rate of melanomas and nevi between the referring centers and UCSF was 14.3%. This review was limited in that there were few patients with benign pigmented lesions referred to the pigmented lesion clinic at UCSF. The level of discordance in the routine histopathologic interpretation of melanocytic neoplasms can be high.