BRAF, NRAS and MC1R status in a prospective series of primary cutaneous melanoma

DEAR EDITOR, Nonacral cutaneous melanomas have been suggested to develop through chronic sun exposure or modest intermittent sun exposures on a genetic host background of melanocytic instability, which is expressed as a phenotype with multiple common melanocytic naevi. In general, melanomas located at the areas of skin with chronic sun damage (CSD) differ from those without any histological evidence of CSD (nonCSD). Melanomas on non-CSD sites carry mainly BRAF mutations along with loss of CDKN2A and PTEN. Although NRAS mutations occur mainly in melanomas on CSD, mutations in the gene even occur in about 20% of melanomas on non-CSD sites. This study was designed to investigate the factors associated with BRAF and NRAS mutations in a homogeneous series of melanomas located on areas of skin without histological CSD. In a multi-institutional series of non-CSD melanomas we investigated distribution of BRAF and NRAS mutations, and their association with clinical–pathological factors and MC1R variants. Patients prospectively enrolled between January 2009 and June 2012 at five centres from the Region of Valencia, with a histologically proven diagnosis of primary invasive cutaneous melanoma localized on areas with intermittent sunexposure, and without a degree of solar elastosis in the range of CSD, were selected. The trunk and the extremities were predefined as intermittently exposed areas. However, an individual patient was excluded based on information provided by the patient about being chronically exposed as in the case of an outdoor worker. The patients included in the study gave their written informed consent and the study was approved by the ethics boards of all participating centres. Detailed methodology can be found in Supplementary Supporting Data S1 (see Supporting Information). Clinical and phenotypic data were collected at the first visit to the clinic by the patients (Table 1). MC1R genotypes were generated by direct sequencing of germline DNA. Statistical analysis was performed to determine associations between mutations and different clinical and phenotypic parameters using Pearson’s chi square and/or Fisher’s exact test. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using logistic regression models. All analyses were carried out using the statistical package SPSS version 15.0 for Windows (SPSS Inc., Chicago, IL, U.S.A.). All P-values were two sided and P-values < 0 05 were considered statistically significant. Of 230 (125 men and 105 women; median age 55 5 years) Caucasian patients with primary invasive cutaneous melanoma investigated for NRAS and BRAF mutations, MC1R genotype data was available for 191 patients. Results showed that mutations in BRAF were present in 75 (32 6%) and NRAS in 17 (7 4%) melanomas (Table 1 and Table S1; see Supporting Information). The difference in age at diagnosis based on mutational status was statistically significant (Table 1). The NRAS mutations were abundant in patients over 60 years (52 9%) and almost absent in patients 40 years or younger. In contrast, a relatively high percentage (22 7%) of the patients with BRAF mutations were ≤ 40 years old at the time of diagnosis (Table 1). NRAS mutations were more frequent in melanoma on extremities (68 8%) than on other sites (P = 0 02) (Table 1). BRAF mutations were common in nodular melanoma (NM) [29/49 (59 2%); P = 0 001] (Table 1). Tumours with BRAF mutations were thicker (mean 3 2 mm) than tumours with NRAS mutations (mean 2 6 mm) and without either mutation (1 6 mm) (P < 0 001; Table 1). BRAFmutated tumours were more frequently ulcerated (P = 0 001) and showed a tendency to have microscopic satellites (P = 0 09) (Table 1). Tumours with BRAF or NRAS mutations showed predominance of epithelioid cell type (P = 0 03) (Table 1). Mild or moderate (non-CSD) solar elastosis was more frequent in tumours without mutations than in tumours with NRAS and BRAF mutations (P = 0 04) (Table 1). The presence of preexisting naevus or the number of common and atypical melanocytic naevi was not significantly different (Table 1). We found no statistically significant association between BRAF and NRAS mutations and MC1R variants (P = 0 43) (Table 1 and Table S2; see Supporting Information). Similar results were observed when only red hair colour variants were considered (data not shown). The skin phototype also did not show any association with the mutational status (P = 0 96) (Table 1). Multivariate analysis showed that Breslow thickness > 2 mm, predominance of epithelioid cells in the tumour as well as the absence of solar elastosis were independently associated to BRAF mutations (Table 2). The association of NRAS mutations with tumour location on the lower extremities (OR 5 2; 95% CI 1 5–18 0) and Breslow thickness > 2 mm (OR 6 0; 95% CI 1 8–20 0) was also statistically significant in multivariate analysis (Table 2). Thus, in this study on a prospective series of 230 consecutive primary cutaneous invasive melanomas we show that melanomas at non-CSD sites with BRAF and NRAS mutations tend to be associated with aggressive tumour characteristics. The