Proliferative nodules in congenital melanocytic nevi: a clinicopathologic and immunohistochemical analysis.
Congenital melanocytic nevi (CMN) occur in 1% to 2% of newborns, and the risk of malignant melanoma is increased in patients with large CMN. Appearance at birth or later of a nodular or hyperpigmented area within a CMN simulates malignant melanoma and prompts biopsy. Although their clinical and pathologic features seem ominous, proliferative nodules (PNs) typically are benign and may regress, although atypical features cause greater concern. Here we report clinical and pathologic findings with outcome in 10 children who had multiple biopsies of large CMN with PNs. We reviewed 78 separate samples from the 10 patients and classified the 60 PNs according to published criteria. A subset of 30 samples containing both the CMN and a PNs was analyzed for immunohistochemical reactivity for melanocytic (S-100 protein, HMB45, melan-A), lymphocytic (CD45), cell-cycle/proliferative (Mib-1, p16, p21, p27, c-Myc), apoptotic (p53, Bax, c-kit, CD95), and anti-apoptotic (bcl-2) markers. Both CMN and PNs had similar expression of melanocytic, lymphocytic, and most cell-cycle/proliferative and apoptotic markers, including Mib-1, p16, p21, p27, c-Myc, Bax, CD95, and bcl-2. A greater proportion of PNs than CMN were reactive for p53 (67% vs. 30%, P < 0.0098) and c-kit (97% vs. 3%, P < 0.0001). p53 and p21 expression in CMN and all types of PNs were inversely correlated. When ordinary and atypical PNs were compared, the atypical PNs more frequently expressed p53, Mib-1, Bax, and bcl-2, but less frequently expressed p21. The c-kit expression in nearly all PNs and its absence in nearly all CMN is potentially useful for recognition of PN, suggests a delayed melanocytic maturation process in proliferative nodules, and may be likely indicative of their benign nature. p53 reactivity in concert with a lack of p21 up-regulation by immunohistochemistry suggests that a p53 mutation may be present in PN, although the immunohistochemical findings alone cannot exclude possible overexpression of wild-type p53. Regressive, involutional, or maturational changes were observed in sequential samples from 4 patients. No patient developed malignant melanoma or another melanocytic nevus-associated malignancy during the follow-up period. These findings underscore the similarities between PNs and the underlying CMN and suggest that maturational, proliferative, and apoptotic processes are involved in their clinical evolution.