Re: Association between endothelin receptor B nonsynonymous variants and melanoma risk.

The known genetic risk factors for malignant melanoma, besides CDKN2A and CDK4 mutations, include variant alleles of the melanocortin-1 receptor (MC1R) gene ( 1 ) . Recently, Soufi r et al. ( 2 ) in a study that was based on 137 malignant melanoma patients and 131 ethnically matched control subjects reported an association between nonsynonymous variants in the endothelin receptor B (EDNRB) gene and increased risk of melanoma. They reported six nonsynonymous EDNRB variants in 15 patients compared with two variants in four control subjects. EDNRB promotes migration and proliferation of melanocyte precursors during embryonic development ( 3 ) . The overexpression of EDNRB in most human melanomas and the shrinkage of tumors in immunocompromised mice and induction of apoptosis upon inhibition of EDNRB suggest a role in disease progression ( 4 , 5 ) . Mutations in the gene are associated with several genetic disorders. The most common variant detected by Soufi r et al. in 12 melanoma patients and three control subjects, S305N, has previously been reported in Hirschsprung disease ( 2 , 6 ) . To extend the study on the role of EDNRB variants in melanoma, we investigated the occurrence of the S305N variant in 462 case patients and 334 ethnically matched control subjects. The case patients were unrelated German melanoma patients (252 men and 210 women; mean age = 57.0 years, range = 7 – 92 years). The control subjects were healthy blood donors (228 men and 106 women, mean age = 41.0 years, range = 18 – 78 years). Patients gave informed consent, and the study was approved by the institutional review board of University Hospital, Mannheim, Germany. Exon 4 of the EDNRB gene was amplifi ed using DNA extracted from the blood samples of case patients and control subjects and was screened for the S305N (G > A) and the more common L277L (G > A) variant by DNA sequencing. In addition, the entire MC1R gene was screened for variants. Ageand sexadjusted odds ratios (ORs), 95% confi dence intervals (CIs), and P values for risk associated with variant genotypes were assessed by using logistic regression. All statistical tests were two-sided, and statistical analyses were carried using SAS version 9.1 software (SAS Institute Inc, Cary, NC). The frequency of heterozygote genotype for the S305N polymorphism in the EDNRD gene was 0.02 in both case patients and control subjects. The variant allele frequency of 0.01 in both case patients and control subjects was similar to that reported by Soufi r et al. ( 2 ) for control subjects in their study. No case patient or control subject was homozygous for the variant allele. The adjusted odds ratio for risk in S305N polymorphism carriers was 0.8 (95% CI = 0.3 to 2.5; P = .76; Table 1 ). The frequency of variant A allele for L277L synonymous polymorphism was 0.39 and 0.41 in case patients and control subjects, respectively. The adjusted odds ratio for associated risk in L277L synonymous polymorphism carriers was 0.8 (95% CI = 0.6 to 1.1; P = .20) compared with noncarriers. The presence of one or more MC1R variants was associated with an increased risk of melanoma (OR = 1.7, 95% CI = 1.2 to 2.5; P = .005). In contrast to the results of Soufi r et al. ( 2 ) , our data from a large study did not show any association between melanoma risk and the S305N variant in the EDNRB gene. However, our results did confi rm the risk associated with variant alleles in the MC1R gene. We cannot rule out different study populations as a cause of observed differences in results. Moreover, we do not preclude that the association of EDNRB with melanoma occurs through another gene variant or via a different mechanism.