Relationship between Germline MC1R Variants and BRAF-Mutant Melanoma in a North Carolina Population-Based Study

TO THE EDITOR

A few previous studies have examined the relationship between germline melanocortin-1 receptor (MC1R) status and somatic BRAF mutations in melanoma. Two publications reported strong associations in three independent populations (two from Italy and one from San Francisco) (Fargnoli et al., 2008; Landi et al., 2006), while a more recent publication found no association in an Australian population-based study (Hacker et al., 2009). We report our finding of no significant association between MC1R status and BRAF-mutant melanomas in a population-based study of malignant melanoma in North Carolina.

Participants in this study were 219 cases with first primary invasive cutaneous melanoma from North Carolina, one site in the population-based Genes, Environment, and Melanoma (GEM) study (Begg et al., 2006). The study protocol was approved by the Institutional Review Board of The University of North Carolina at Chapel Hill. The Declaration of Helsinki protocols were followed, and patients gave their written, informed consent. The participants were interviewed regarding their risk factors (Thomas et al., 2007). The subjects were asked to have the nevi on their backs counted by a family member or friend, using a glossy colored guide to aid in differentiating between nevi and other skin lesions.

One dermatopathologist (KB) reviewed the tumors for standard histologic features. The tumors were scored for chronic sun damage using a 0 to 3+ multipoint scale and dichotomized into melanomas with little histologic evidence of chronic sun damage (non-CSD) (level 0 to 2−) and those with histologic evidence of chronic sun damage (CSD) (level 2 to 3+), as in Landi et al. (Landi et al., 2006). BRAF mutations in and around codon V600 in melanomas were determined as previously described (Thomas et al., 2007).

We sequenced MC1R from buccal DNA (Kanetsky et al., 2006), and six cases were excluded because their germline DNA did not amplify. MC1R alleles were classified as “R” and “r” according to nomenclature developed by Duffy et al. (Duffy et al., 2004). MC1R “R” variants were defined as (D84E, R142H, R151C, R160W, and D294H) due to their strong association with the red hair phenotype in a recent meta-analysis (Raimondi et al., 2008). The MC1R c.86_87insA variant was also included as an “R” variant because it results in a truncated, non-functional receptor, and patients reported with red hair have been compound heterozygous for MC1R stop or frameshifts along with an “R” allele (Beaumont et al., 2008; Kennedy et al., 2001). All other non-synonymous MC1R variants were classified as “r” (V60L, A64T, 92M, D117V, I155T, V156L, R163Q, R223W, and K278E).

BRAF-mutant tumors were defined as those found to have mutations in and around the V600 codon in exon 15. ORs and accompanying 95% confidence intervals (CIs) were calculated in logistic regression models in SAS (SAS Institute, Cary NC., 1989). Age was included as a continuous variable. All significance tests were two-sided and a P value of < 0.05 was considered statistically significant. Potential interactions were assessed using likelihood ratio tests with an a priori alpha of 0.2 (Selvin, 1996). The likelihood ratio tests were performed in STATA version 10.

The 213 study cases included in the analyses had a mean age at diagnosis of 52.1 years and were 54.9% male. The study included a high percentage (55.9%) of thin (Breslow thickness <0.75 mm) melanomas. The percentages of histologic subtypes were 77.5% superficial spreading melanoma, 4.2% nodular melanoma, 10.3% lentigo maligna melanoma, 0.9% acral lentiginous melanomas, and 7.0% other or unclassifiable melanoma. The anatomic distribution of the melanomas was 15.0% head or neck, 50.2% truncal, 18.8% upper extremities, and 16.0% lower extremities. The percentage of non-CSD melanomas was 70.2%.

The presence of any non-synonymous germline MC1R variant demonstrated no association with BRAF-mutant melanoma (odds ratio [OR] 0.9, 95% confidence interval [CI] 0.4 to 2.3) after adjusting for age at diagnosis, ability to tan, and back nevus count (Table 1a). Finer MC1R categorization did not reveal an association of BRAF-mutant melanoma with carriage of one MC1R (r/wt or R/wt) [adjusted (adj.) OR 1.2, 95% CI 0.5 to 3.0] or two MC1R (r/r, R/r, or R/R) (adj. OR 0.7, 95% CI 0.2 to 1.7) variants.

Table 1.

Germline variants of MC1R and tumor-specific BRAF mutations melanomas (North Carolina First Incident Primary Cutaneous Melanomas) (N=213)

MC1R1	BRAF Mut (N=88)	BRAF Wt (N=125)	OR (95%CI)	P-value	OR (95% CI)2	P-value
a. All variants
wt/wt	11 (12.5%)	18 (14%)	Reference		Reference
Any Variant	77 (87.5%)	107 (86%)	1.2 (0.5–2.6)	0.69	0.9 (0.4–2.3)	0.90
wt/wt	11 (13%)	18 (14%)	Reference		Reference
One variant ( r/wt or R/wt)	46 (52%)	53 (42%)	1.4 (0.6–3.3)	0.42	1.2 (0.5–3.0)	0.70
Two variants ( r/r, R/r, or R/R)	31 (35%)	54 (43%)	0.9 (0.4–2.2)	0.89	0.7 (0.2–1.7)	0.39
b. “R” variants
wt/wt, r/wt, or r/r	39 (44%)	58 (46%)	Reference		Reference
Any “R” variant (R/R, R/r, or R/wt)	49 (56%)	67 (54%)	1.1 (0.6–1.9)	0.76	1.4 (0.7–2.6)	0.35
wt/wt, r/wt, or r/r	39 (44%)	58 (46%)	Reference		Reference
One “R” variant (R/wt or R/r)	41 (47%)	45 (36%)	1.4 (0.8–2.5)	0.31	1.8 (0.9–3.6)	0.09
Two “R” variants (R/R)	8 (9%)	22 (18%)	0.5 (0.2–1.3)	0.18	0.6 (0.2–1.6)	0.26

Abbreviations: MC1R, melanocortin-1 receptor; BRAF mut, contains a somatic mutation in or around the V600 codon in BRAF exon 15; BRAF wt, negative for a somatic mutation in BRAF exon 15; CI, confidence interval; OR, odds ratio.

¹MC1R variants defined as “R” (D84E, R151C, R160W, D294H, R142H, and c.86_87insA). All other non-synonymous variants MC1R variants were defined as “r” (V60L, A64T, V92M, D117V, I155T, V156L, R163Q, R223W, and K278E).

²Logistic regression models adjusted for age at diagnosis, ability to tan and number of moles on the back.

The analyses were repeated considering only association with the “R” variants (Table 1b). BRAF-mutant melanoma was not associated with presence of any “R” variant (R/R, R/r, or R/wt) (adj. OR 1.4; 95% CI 0.7 to 2.6). Finer categorization of one (R/r, or R/wt) or two (R/R) “R” variants did not result in significant associations (adj. OR 1.8; 95% CI 0.9 to 3.6 and adj. OR 0.6; 95% CI 0.2 to 1.6, respectively).

We examined the potential interactions of anatomic site (head and arms vs. trunk and legs) and CSD (CSD vs. non-CSD melanoma) as proxy measures of sun exposure upon the relationship between MC1R status (any variant) and BRAF-mutant melanoma. No statistically significant interactions were found (data not shown).

In Landi et al. 2006, D84E, R142H, and c.86_87insA were defined as “r” rather than “R”. Reclassification of these variants as “r” did not significantly change the results (data not shown). For purposes of comparison with cases included by Landi et al. and Fargnoli et al., we repeated all analyses excluding two cases with melanomas on acral skin and four cases with positive or indeterminate germline CDKN2A mutations (Orlow et al., 2007). These re-analyses did not reveal any significant associations (data not shown).

Our data do not support a strong association of MC1R variants with BRAF mutations in our North Carolina population. Differences in the findings between studies may be due, in part, to differing frequencies of distinct MC1R variants between study populations, which may not all be similarly associated with BRAF mutation. Another possibility is that the risk-modifying effect of MC1R variants may vary between populations based on unidentified genetic factors. Climate differences, such as ambient sun exposure, could also influence the relationship between MC1R status and BRAF-mutant melanoma. However, despite our study being the largest to date, the sample sizes are relatively modest in all the studies, and further investigation is necessary to clarify the relationship of germline MC1R variants and BRAF–mutant melanomas among different populations.

Abbreviations

MC1R

melanocortin-1 receptor