EPID-16. GENOMIC ANCESTRY-RELATED RISK RATIOS FOR GLIOBLASTOMA INCIDENCE AMONG GLOBAL POPULATIONS AND RACIAL/ETHNIC GROUPS WITHIN THE UNITED STATES

Abstract Glioblastoma (GBM) incidence varies globally, elevated rates in Europe and lower rates in Asia, Latin America, and Africa. While partially attributable to national differences in cancer diagnosis and reporting practices, analogous patterns exist within diverse Western nations. Ancestry-associated differences in GBM risk allele frequencies (RAFs) could contribute, a hypothesis supported by our prior multiethnic genome-wide association studies (GWAS). Leveraging variant effect estimates from large GBM GWAS datasets and ancestry-stratified allele frequency data from diverse reference populations (ALFA, GNOMAD, 1000 Genomes Project), we estimate the extent of GBM incidence differences across populations that can be attributed to underlying variation in RAF at the seventeen known GBM risk loci. Aggregate differences in RAF between global populations (Europe, West-Africa, East Asia) were associated with a >20% reduction in the normalized GBM risk-ratio (RR) in Africans compared to Europeans (RR=0.78) and a >50% reduction in East Asians compared to Europeans (RR=0.47). The frequency of GBM risk alleles was generally highest in Europeans, but four alleles found most commonly in Africans are at loci regulating telomere maintenance (TERC, TERT, OBFC1/STN1, RTEL1). We repeated analyses using RAF data from U.S. populations (African-Americans, Latinos, and non-Hispanic Whites). Compared to non-Hispanic Whites, aggregate differences in RAF were associated with a 15% reduction in GBM risk among African-Americans (RR=0.85) and a 28% reduction among Latinos (RR=0.72). We conclude that a sizable proportion of the reduced GBM incidence in Latinos and Asians, relative to non-Hispanic Whites, can be accounted for by population-level differences in RAF at known loci. However, registry-based analyses consistently report a two-fold lower incidence of GBM in African Americans relative to non-Hispanic Whites which is poorly explained by RAF differences at known loci. Further work is needed to identify structural, environmental, and biological factors contributing to unresolved differences in GBM incidence.