Predictors of malignancy in melanocytic lesions presenting as new lesions compared to baseline total body photography: A case-control study.

Abstract

Only a small per cent of new melanocytic lesions developing in adults are expected to represent melanomas. Total body photography (TBP) has been widely incorporated in clinical practice, especially for follow-up of high-risk individuals with multiple naevi. However, dynamic changes detected with TBP need to be interpreted with caution to avoid unnecessary excisions. To identify clinical and dermoscopic predictors of malignancy in melanocytic lesions presenting clinically as new lesions on TBP. Melanomas and melanocytic naevi excised from a high-risk cohort and presenting as new lesions on TBP were retrospectively included. Naevi were arbitrarily collected up to approximately twice the number of melanomas. Melanomas were categorized as 'unequivocal' or 'borderline' on histopathology review. Sixty melanomas and 110 naevi were included. Median age (range) of cases (55; 27-83) was 9 years older than controls (46; 24-77) (p < 0.0001). Median diameter (IQR) of naevi was 2.6 mm (1.8-3.8) and of melanomas 4.2 mm (2.7-7.0) (p < 0.0001). On histopathology, 40% of the melanomas were 'borderline'. A positive 7-point checklist was reported in 12.5% of 'borderline' melanomas and 33.3% of 'unequivocal' melanomas (p = 0.005), while 18.3% of melanomas were completely featureless. Blue-whitish veil, atypical vascular pattern and shiny white lines were exclusively found in melanomas. The main predictors of malignancy were (OR; 95% CI) regression structures (7.13; 1.88-27.06; p = 0.004); hypo/amelanotic colour (6.00; 1.17-30.73; p = 0.03); irregular pigmentation (3.89; 1.36-11.13; p = 0.01); asymmetrical peripheral dots/globules (3.50; 1.11-11.00; p = 0.03); and asymmetry in pattern and/or colour (2.5; 1.3-4.9; p = 0.007). All invasive melanomas detected in patients younger than 50 years presented at least one dermoscopic predictor of malignancy. Melanomas presenting as new lesions are frequently featureless or feature poor on dermoscopy and difficult-to-diagnose on histopathology. In high-risk patients, the presence on any of the dermoscopic predictors of malignancy identified should prompt excision; however, the remaining lesions should be closely monitored.