13155 Reevaluating the ABCD criteria using a consecutive series of melanomas

Abstract

The ABCDs of melanoma were previously evaluated in studies that used lesion photographs taken due to a clinician’s suspicion for melanoma, possibly excluding more clinically subtle lesions, including amelanotic melanomas. In addition, these criteria were defined prior to the widespread use of dermoscopy, and thus may reflect a different spectrum of lesions now detected by dermatologists. In our practice, physicians routinely photograph all lesions prior to biopsy, allowing us to evaluate the ABCD criteria using a consecutive series of images of melanoma. We retrospectively reviewed 290 consecutive cases of primary cutaneous melanomas diagnosed within the University of Pittsburgh dermatology department from 01/2014 to 07/2016. In a reader study, three independent dermatologists assessed photographs for ABCD criteria. Among our sample of melanomas, 159 (55%) were in situ and 131 (45%) were invasive, with a median Breslow thickness of 0.55 mm (0.35-0.95). The prevalence of each individual ABCD criterion was 85%, 86%, 71% and 60% respectively. Assigning one point per ABCD criterion, we found that most melanomas had a score of 3 (38%) or 4 (41%). Evaluating dermatologists clinically classified 13 (4.5%) lesions as amelanotic melanomas, 6 of which were invasive, with a median Breslow thickness of 1 (0.3-1.8). Amelanotic melanomas did not differ significantly in ABCD features compared with all other melanomas within our sample. Our study validates, in the post-dermoscopy era, the ABC clinical criteria while suggesting that the D criterion should not be relied upon as heavily as a tool for educating patients and physicians on melanoma detection.