Abstract 1806: Notifiable Genetic Variants on Commercially Available SNP Arrays: Implications for Research Participants in Genome-wide Association Studies for Cardiovascular Disease
In recent years, genome-wide association studies (GWAS) for cardiovascular disease (CVD) have been conducted with increasing sample sizes using dense arrays of single nucleotide polymorphisms (SNPs). While criteria exist for determining if incidental genetic findings should be reported back to research participants, a comprehensive survey of the extent of potentially notifiable variants on genotyping arrays now widely applied in GWAS and in direct to consumer genetic tests has not been reported. We hypothesized that while genotyping arrays are generally designed to measure common variants with modest clinical impact a comprehensive survey might reveal some lower frequency genetic variants with more substantial clinical impact that could warrant participant notification regarding genotype status. Using bioinformatics techniques and multiple public genome databases, we surveyed 18 commercial SNP arrays and HapMap, searching for variants tested for genetic disorders in ≥ 2 CLIA-approved US labs within the GeneTests database. There were 298 mutations identified from GeneTests, encompassing 56 disorders. Of these 298 mutations only 88 could be identified as known SNPs in genomic databases after we employed multiple bioinformatics search strategies. We found that 18 of these 88 SNPs are present on commercial SNP arrays, or in the HapMap, meaning they could be directly genotyped or imputed in the course of GWAS. These variants were in genes including F5, HFE, CYP21A2, MEFV, SPINK1, BTD, GALT, and G6PD. Not surprisingly some of these genes underlie diseases other than CVD. In a survey of the Framingham Heart Study GWAS genotyping results (n=8,410) we found rare homozygotes for some of these SNPs. None of the variants were deemed to rise to the level of being clearly notifiable based upon prior established criteria, but our findings support the need for further clinical review or review by independent panels of experts for these and other variants that are routinely tested in research studies. We suggest there is a growing need to establish consensus reporting procedures for potential notifiable incidental genetic findings as population genetics is rapidly advancing with new discoveries and an increase in personal genome sequencing.