Recent research suggests that cutaneous melanomas may arise through 2 distinct pathways, characterized by chronic sun exposure on one hand and nevus-prone phenotype of the host on the other. Two histological characteristics of melanoma consistent with these divergent origins are dermal elastosis in adjacent skin and neval remnants contiguous with the tumor, respectively. To further explore causal heterogeneity in melanoma, we compared sun exposure histories and phenotypic characteristics among a population-based sample of patients newly diagnosed with cutaneous melanoma with and without contiguous neval remnants or dermal elastosis. Tissue blocks were obtained for 141 patients: 53 with superficial spreading melanoma (SSM) of the back, 42 with SSM of head and neck (H & N), and 39 and 7 with lentigo maligna/lentigo maligna melanoma (LM/LMM) of the H & N and back, respectively. Melanomas of the H & N were less likely than those on back to have neval remnants (adjusted OR 0.6, 95% CI 0.3-1.4), but were significantly more likely to have dermal elastosis (adjusted OR 9.3, 95% CI 3.5-25). In site-specific analyses, we found that H & N melanomas with neval remnants were more likely than those without neval remnants to arise in people with more than 60 nevi (adjusted OR 2.1, 95% CI 0.3-14.3), but were less likely to arise in those with more than 20 actinic keratoses. Less marked associations were observed for melanomas of the back. High levels of sun exposure strongly predicted dermal elastosis for H & N melanomas (OR 22.5, 95% CI 2.1-245), but not for melanomas of the back (OR 2.1, 95% CI 0.4-11). We conclude that melanomas with different histologic characteristics have different risk factor profiles, particularly on the head and neck. These data accord with the hypothesis that melanomas arise through different causal pathways. ' 2006 Wiley-Liss, Inc.

The pituitary prohormone proopiomelanocortin gives rise to melanocortins of alpha, beta, and gamma primary structure in addition to corticotropin. Melanocortins have a variety of actions in mammals, and each is natriuretic. In particular, gamma-melanocyte-stimulating hormone has been shown to mediate reflex natriuresis after acute unilateral nephrectomy. We examined whether this peptide could play a role in longer term adjustments in sodium balance by measuring plasma gamma-melanocyte-stimulating hormone and corticotropin concentrations, as well as pituitary proopiomelanocortin mRNA abundance, in Sprague-Dawley rats ingesting either a low (0.07% NaCl) or high (7.5% NaCl) sodium diet. One week after the high sodium diet, plasma gamma-melanocyte-stimulating hormone concentration was double the value seen in rats on the low sodium diet (158 +/- 5 [SE] versus 76 +/- 9 fmol/mL, P < .001), a change that was accompanied by a fivefold increase in plasma atrial natriuretic peptide concentration but no change in plasma corticotropin. Whole pituitary proopiomelanocortin mRNA abundance, measured with a probe to exon 3 of the rat proopiomelanocortin gene, was significantly increased after 1 week of the high sodium diet compared with the low sodium diet and increased further at 2 and 3 weeks. This increase occurred primarily in the neurointermediate lobe as demonstrated by in situ hybridization; the content of gamma-melanocyte-stimulating hormone immunoreactivity was also increased in this lobe, but not the anterior lobe, after 1 week of the high sodium diet. These results demonstrate that high dietary sodium intake increases neurointermediate lobe proopiomelanocortin mRNA abundance compared with a very low sodium diet and also suggest that proopiomelanocortin is preferentially processed into gamma-melanocyte-stimulating hormone rather than corticotropin. These observations consequently raise the possibility of a role for this peptide hormone system in the adjustments to a high salt diet.