ASSOCIATION BETWEEN POLYGENIC RISK SCORE AND CUTANEOUS MELANOMA

In addition to the known high/medium-penetrance genes involved in melanoma susceptibility, recent Genome-Wide Association Studies (GWAS) have identified many common allelic variants (SNPs) that may affect the risk of developing the disease. Each of these variants alone weakly affects melanoma risk, but their combination, according to a Polygenic Risk Score (PRS) model, may increase the risk by three-fold, and even more in association with other risk factors (i.e. nevus count and skin phototype). Most of these previous PRS findings stem from population-based studies, while little is known about the role of PRS in risk assessment in MPM as well as in germline mutated melanoma patients. Then, the main aim of our project focuses on assessing whether PRS might explain the susceptibility to recurrent melanoma diagnoses analyzing MPM patients without germline pathogenic variants. We analyzed 259 melanoma cases by NGS using a Custom DNA Panel targeting known melanoma-predisposing genes and 64 SNPs selected on the basis of their significance of association with melanoma as reported in different and independent GWAS studies. The PRS was calculated for each individual by summing the 64 SNPs weighted for their β value, that is the per-allele log OR for melanoma associated with the SNP alternative allele. We found a significantly higher mean PRS in all melanoma cases than controls (0.423 vs 0.00, p=0.0001) and a double mean PRS in multiple vs single melanoma cases (0.54 vs 0.23, p=0.0160). In particular, an association between PRS and multiple melanoma risk was found with a per-SD OR of 1.4 (95% CI 1.06-1.79, p=0.016) corresponding to almost tripled risk (OR 2.78, 95% CI 1.18-6.61, p=0.0189) for individuals in the highest PRS quintile compared to those in the lowest quintile. These evidences highlight that individuals with a high PRS are more likely to develop one melanoma, and subsequent melanomas. We also evaluated a possible association between PRS with specific non-genetic risk factors for cutaneous melanoma finding a strong association with some individual phenotypic/behavioral characteristics. Our preliminary results suggest that the PRS can be a useful tool for selecting patients without pathogenic variants in currently known high-risk melanoma genes, who should undergo more stringent surveillance protocols. On the other hand, a second aim of the project is to investigate whether PRS might be a potential modifier of melanoma risk in patients carrying pathogenic variants in CDKN2A gene or the medium penetrance MITF p.E318K variant. This last part of the project has recently started in collaboration with Dr. Thomas Potjer and Dr. Nienke van der Stoep at the University Medical Centre in Leiden. We selected 82 (64 CDKN2A and 18 MITF p.E318K) Italian mutated samples from patients tested by NGS or Sanger (single-gene) sequencing; these samples were then added to 482 Dutch CDKN2A/MITF mutated cases recruited from Clinical Genetic Centres in the Netherlands, and were SNP genotyped at the Human Genotyping Facility (HuGe-F) at Erasmus MC in Rotterdam, using Illumina Global Screening array. For this part of the project, the analysis on the contribution of PRS in risk predictive models of melanoma development are still ongoing.