Naevus-associated melanomas: cause or chance?

Controversies exist regarding the true incidence and significance of the histological association between melanocytic naevi and melanomas. The current study was undertaken to determine the incidence of melanocytic naevi histologically associated with melanomas, to compare the clinicopathological profiles of melanomas associated with a naevus (MNS) and melanomas not associated with a naevus (MN-), and to verify the interobserver reliability of classifying the naevus remnants as congenital or acquired. We evaluated 131 patients with invasive melanoma less than or equal to 4 mm in thickness with a Clark's level <V. Histological evidence of an associated melanocytic naevus was found in 27 out of 131 melanoma cases (20.6%). MN+ were significantly more frequent among males compared with MN- (P = 0.002) and were predominantly located on the trunk (P = 0.001). No significant differences between MNS and MN- were found in the distribution of histotype, Clark's level and thickness. Among MN+, the naevus component showed acquired features in 14 cases (51.8%), whereas 12 cases (44.5%) had features of small superficial congenital naevi; in one case (3.7%) a confident distinction between congenital and acquired naevus could not be made. Overall, there was concordance among the three observers in the diagnosis of the type of naevus remnants in 81.4% of cases (kappa = 0.78; P<0.0001). We conclude that, although the majority of cutaneous melanomas arise de novo, in approximately 20% of cases an associated melanocytic naevus is found. Our observation that MN+ and MN- have different clinicopathological profiles suggests that, at least for some cases, a causal relationship between the two lesions may be present. However, when the naevus is histopathologically adjacent to the melanoma, the possibility of a collision event cannot be ruled out. Finally, we documented a significant concordance among examiners in the diagnosis of the congenital versus acquired nature of the associated naevus. This suggests that the histopathological criteria generally employed to distinguish between congenital and acquired naevi have good reproducibility. (C) 1999 Lippincott Williams & Wilkins.