Genome-wide association study identifies novel susceptibility loci for cutaneous squamous cell carcinoma

Abstract

Cutaneous squamous cell carcinoma represents the second most common cutaneous malignancy, affecting 7-11% of Caucasians in the United States. The genetic determinants of susceptibility to cutaneous squamous cell carcinoma remain largely unknown. Here we report the results of a two-stage genome-wide association study of cutaneous squamous cell carcinoma, totalling 7,404 cases and 292,076 controls. Eleven loci reached genome-wide significance (P<5 × 10(-8)) including seven previously confirmed pigmentation-related loci: MC1R, ASIP, TYR, SLC45A2, OCA2, IRF4 and BNC2. We identify an additional four susceptibility loci: 11q23.3 CADM1, a metastasis suppressor gene involved in modifying tumour interaction with cell-mediated immunity; 2p22.3; 7p21.1 AHR, the dioxin receptor involved in anti-apoptotic pathways and melanoma progression; and 9q34.3 SEC16A, a putative oncogene with roles in secretion and cellular proliferation. These susceptibility loci provide deeper insight into the pathogenesis of squamous cell carcinoma.

Figure 1. Manhattan plot of stage 1 GWAS analysis of SCC.

Total stage 1 GWAS analysis included 6,579 cases and 280,558 controls from the 23andMe cohort. The y axis represents log-scaled P-values. Loci with smallest P<10⁻⁵ (via logistic regression), of which there are 14 in total, are labelled with the name of the nearest gene. Positions with P<5 × 10⁻⁸ (genome-wide significance) are shown in red. Seven SCC susceptibility loci reached genome-wide significance in stage 1, including six previously reported pigmentation loci at 6p25.3 (IRF4), 5p13.2 (SLC45A2), 16q24.3 (MC1R), 11q14.3 (TYR), 20q11.22 (RALY-ASIP) and 15q13.1 (OCA2), which are highlighted by asterisks. The seventh susceptibility locus reaching significance in stage 1, 9q34.3 (SEC16A), is novel and not pigmentation related (marked by a double asterisk). Three additional novel susceptibility loci, 2p22.3, 7p21.1 (AHR) and 11q23.3 (CADM1-BUD13), reached genome-wide significance in the overall meta-analysis, as did the previously reported locus 9p22.2 (BNC2-CNTLN) (Table 2). The remaining three loci in the figure, 3q28 (LPP), 6p21.3 (HLA-DQB1) and 8q23.3 (TRPS1), had P<10⁻⁵ in stage 1 but did not reach genome-wide significance in the meta-analysis.