BACKGROUND: The earlier detection of melanomas occurring within preexisting nevi is theoretically possible using sequential dermoscopy. Characterizing the early follow-up changes of nevus-associated melanomas (NAMs) and differentiating them from those observed in de novo melanomas (DNMs) may help the earlier recognition of NAMs. METHODS: Using descriptive dermoscopic features to detect focal changes, we blindly evaluated retrospectively the baseline and follow-up images of 32 melanomas that were subsequently classified as histopathologically defined NAMs or DNMs. RESULTS: Correlates of growth, as structureless brown‐black areas or clods, complemented each other for the identification of DNMs at baseline (structureless brown‐black areas: 66.7% DNMs, 15% NAMs, P < 0.01; combined with clods, one or the other being present: 100% DNMs, 30% NAMs, P < 0.01) and when considering their baseline presence or their appearance at follow‐up (100% DNMs, 35% NAMs, P < 0.01). Correlates of fibrosis, as white lines, when considering their baseline presence or their appearance at follow-up, were associated with NAMs (60%, 16.7% DNMs, P = 0.027). CONCLUSION: Significant differences, distinguishing NAMs from DNMs, were detected particularly when considering both baseline signs and follow-up changes. Earlier identification of NAMs and their subsequent improved histological characterization will help define the subgroup of high-risk patients, for whom comprehensive image monitoring may be beneficial.
BACKGROUND: The earlier detection of melanomas occurring within preexisting nevi is theoretically possible using sequential dermoscopy. Characterizing the early follow-up changes of nevus-associated melanomas (NAMs) and differentiating them from those observed in de novo melanomas (DNMs) may help the earlier recognition of NAMs. METHODS: Using descriptive dermoscopic features to detect focal changes, we blindly evaluated retrospectively the baseline and follow-up images of 32 melanomas that were subsequently classified as histopathologically defined NAMs or DNMs. RESULTS: Correlates of growth, as structureless brown‐black areas or clods, complemented each other for the identification of DNMs at baseline (structureless brown‐black areas: 66.7% DNMs, 15% NAMs, P < 0.01; combined with clods, one or the other being present: 100% DNMs, 30% NAMs, P < 0.01) and when considering their baseline presence or their appearance at follow‐up (100% DNMs, 35% NAMs, P < 0.01). Correlates of fibrosis, as white lines, when considering their baseline presence or their appearance at follow-up, were associated with NAMs (60%, 16.7% DNMs, P = 0.027). CONCLUSION: Significant differences, distinguishing NAMs from DNMs, were detected particularly when considering both baseline signs and follow-up changes. Earlier identification of NAMs and their subsequent improved histological characterization will help define the subgroup of high-risk patients, for whom comprehensive image monitoring may be beneficial.