Allele frequency variation in people of European and admixed African ancestry in the United States do not fully explain incidence differences in adult-type diffuse glioma
Presenter: Christine Ballard, MPH Session: Etiology and Molecular Epidemiology Approaches to Decipher Cancer Disparities Time: 4/21/2026 9:00:00 AM → 4/21/2026 12:00:00 PM
Authors
Christine Ann Pittman Ballard 1 , Carol Kruchko 2 , Mackenzie Price 2 , Quinn T. Ostrom 3 1 Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, NC, 2 Central Brain Tumor Registry of the United States, Pewaukee, WI, 3 Duke University School of Medicine, Durham, NC
Abstract
Background Glioma is a heterogeneous group of tumors that make up the most common type of primary malignant brain tumor. Incidence varies globally, with highest rates in Europe and North America and lowest in Asia and Africa. Within the United States, these tumors are most common in individuals who are non-Hispanic White. Previously, genome-wide association studies (GWAS) have identified 25 single nucleotide polymorphisms (SNPs) which affect risk for glioma (European only), most of which have subtype-specific differences. SNP allele frequencies (AF) vary between continental populations. We estimated population-level incidence for adult-type diffuse glioma subtypes and calculated the difference between observed variation and variation attributable to AF differences. Methods Race/ethnicity-stratified age-adjusted incidence rates for glioma subtypes (Astrocytoma with IDH1/2 mutation [IDHmut], Astrocytoma with wildtype IDH1/2 [IDHwt], and Oligodendroglioma with IDH1/2 mutation and 1p/19q codeletion [IDHmut-codel]) from diagnosis years 2018-2022 were calculated using the Central Brain Tumor Registry of the United States, an aggregation of CDC’s National Program of Cancer Registries and NCI’s SEER. Incidence rate ratios (IRR O ) were estimated as compared to non-Hispanic White. Effect estimates for known glioma risk SNPs were extracted from a prior subtype-specific glioma GWAS (Labreche, 2018) for SNPs with p -4 . AF were extracted from the Allele Frequency Aggregator for the European and African American populations. We calculated the normalized IRR (IRR N ) for each SNP based on population AF and summed these to estimate the incidence variation attributable to AF. IRR N was calculated by multiplying the beta by the corresponding AF. Results As compared to individuals who are non-Hispanic White, incidence was decreased in non-Hispanic Black individuals for all subtypes [IDHmut IRR O =0.40, IDHwt IRR O =0.49, IDHmut-codel IRR O =0.29]. IRR N were attenuated as compared to observed rates [IDH-mut IRR N =0.78, IDH-wt IRR N =0.75, IDHmut-codel IRR N =0.64]. The IRR N is higher compared to the IRR O , with the greatest differences being observed in the non-Hispanic Black population. Conclusion IRR N fail to fully explain population differences in incidence of glioma. This may be a result of limited ancestral diversity in GWAS, which have not assessed SNPs in individuals who are African American. Sample sizes for subtype-specific GWAS have been small and underpowered. These SNPs do not capture all genetic risk for glioma (all identified SNPs explain 30% of heritable risk), and there is significant ‘missing heritability.’ These results emphasize the need for inclusion of more diverse populations in glioma genetic epidemiology to ensure accurate genetic risk estimation.
Disclosure
C. A. Ballard, None.. C. Kruchko, None.. M. Price, None.
Cited in
Control: 7465 · Presentation Id: 4748 · Meeting 21436