Assessment of melanoma candidate genes in a meta‐analysis of 16 534 melanoma cases

Background: Numerous single nucleotide polymorphisms (SNPs) have been implicated with cutaneous melanoma (CM) Objective: The aim of our study was to establish novel susceptibility loci for CM by assessing candidate genes in a published meta‐analysis of 11 genome‐wide association studies (GWASs).

Methods: 760 variants were retrieved from the MelGene database and tested for association in a meta‐analysis of 11 GWASs consisting of 12,874 CM cases and 23,303 controls of European ancestry. We calculated a false discovery rate and variants that surpassed this p‐value threshold were genotyped in three independent replication samples from UK, Greece and Cyprus. We excluded variants robustly associated with CM or in linkage disequilibrium with already known loci. The results from the GWASs and replication stages were synthesized in a meta‐analysis.

Results: Based on our approach, the rs909253 polymorphism in the LTA gene (Lymphotoxin Alpha) was followed up for further consideration. The meta‐analysis of the 11 GWAS indicated a protective effect for the G allele on the risk of CM (OR: 0.921, p‐value: 2.7x10‐5). However, the estimates from the three replication samples were not statistically significant. The combined meta‐analysis of 3,660 CM cases and 350,915 cancer‐free controls provided a similar OR (0.935, p‐value: 1.95x10‐5); very large studies in the future will be required to determine if this SNP is a genuine melanoma risk SNP.

Conclusion: By assessing over 16,000 CM cases, none of the published candidate genes tested for association with CM were significant in our study. Interest should be shifted to larger GWAS.