Can early malignant melanoma be differentiated from atypical melanocytic nevi by in vivo techniques?: Part I. Clinical and dermoscopic characteristics

Background/aims: Atypical melanocytic nevi (AMN) share some or all of the clinical features of malignant melanoma (MM). The clinical sensitivity for diagnosing MM by physicians experienced in evaluating pigmented lesions is reported to range from 73% to 89%. This study attempted to determine whether clinical sensitivity can be increased by dermoscopy (epilumnescence microscopy, dermatoscopy).

Methods: A total of 72 melanocytic neoplasms were histologically diagnosed as either AMN or early MM (<1 mm Breslow thickness). Prior to excisional biopsy, each lesion was photographed without oil being applied to its surface (clinical photo); subsequently it was photographed with oil applied (dermoscopic photo). Each 35 mm color transparency, visualized on a rear-view projector, was analyzed by two experienced and two less experienced observers. Each of them recorded one clinical, one "overall" dermoscopic, and one "scored" dermoscopic diagnosis: either AMN or MM for each.

Results: Histologically, 21 lesions were diagnosed as MM and 51 as AMN. The observers, except one of the less experienced, had an increase in sensitivity ranging from 5 to 15% for diagnosing MM with dermoscopy. The increase was higher for the scored dermoscopy than for overall dermoscopy. Diagnostic Accuracy (DA) changed from -20% to +11%. Specificity (SP) changed from -23% to +24% with scored dermoscopy. Three of the 21 MMs were missed by one or more observers using all three in vivo diagnostic methods.

Conclusions: In this study dermoscopy, when used by experienced dermatologists, increased diagnostic sensitivity and index of suspicion but decreased specificity and diagnostic accuracy for diagnosing MM. Therefore, dermoscopy may result in an increased number of biopsies of benign lesions (AMN), but would decrease the probability of missing MM.