SAT-177 - Dissecting PCOS genetic architecture: Insights from a Multi-Ancestry GWAS Meta-Analysis

Multi-ancestry genome-wide association studies (GWAS) in polycystic ovary syndrome (PCOS) are essential for extending genetic discovery beyond the European (EUR) and Chinese (East Asian, EAS) cohorts in which robust signals have been established, and for fine-mapping causal variants within known loci. African (AFR) populations harbor the greatest human genetic diversity and represent the ancestral baseline for interpreting evolutionary history of disease variants. Including AFR and Hispanic/Latin American (AMR) cohorts enables identification of truly shared genetic architecture versus population-specific signals, which is critical for understanding PCOS biology and developing broadly applicable risk prediction tools. PCOS cases were either deeply phenotyped or identified using validated ICD codes in biobanks at University of Colorado, University of Pennsylvania, Vanderbilt University, and the All of Us program. We performed GWAS and meta-analysis across 5 AFR, 4 AMR, 1 EUR, and 1 EAS cohorts; the EUR and EAS cohorts were phenotyped. The final analysis consisted of 6,539 PCOS cases (3,344 AFR, 1,794 AMR, 971 EUR, 430 EAS) and 32,319 controls (16,121 AFR, 12,386 AMR, 2,886 EUR, 926 EAS). We identified 1 novel [rs74939114 (p=3.6E-08) near gene INSYN1] and 6 known genome-wide significant loci [rs8191518 (p=2.1E-09) near gene NEIL2, rs10821415 (p=7.1E-10) near gene C9orf3, rs10513424 (p=2.6E-13) near gene DENND1A, rs11031002 (p=8.0E-13) near gene FSHB, rs1672714 (p=8.3E-09) near gene ZBTB16 and rs8043701 (p=1.4E-09) near gene TOX3], implicated in neuroendocrine, reproductive and inflammation pathways. All significant loci exhibited concordant directions of effect across all ancestries. DENND1A and TOX3 were robust signals with low heterogeneity, achieving at least nominal significance in all 4 ancestries, indicating shared genetic architecture rather than population-specific effects. Whether these variants have been maintained by drift, balancing selection, or pleiotropic constraints remains to be determined. The novel locus INSYN1 was contributed only by EUR and AMR cohorts and was not present in AFR and EAS, indicating it cannot be considered ancestral. C9orf3 and FSHB had significant heterogeneity and were mostly contributed by EUR, reflecting higher allele frequency. ZBTB16 was contributed by non-EUR cohorts despite similar allele frequencies, likely reflecting smaller EUR sample size and differences in LD structure. NEIL2 was mostly contributed by EUR and AMR, reflecting lower allele frequency in AFR, and was not present in EAS. There was overrepresentation of EUR ancestry signals, with 75% (22/29) of previously reported EUR loci associated with multi-ancestry PCOS compared to only 39% (26/56) of EAS loci, which may reflect ascertainment bias. DENND1A and TOX3 emerge as core determinants of PCOS shared across ancestries; INSYN1 is a novel locus.

*Unless otherwise noted, all abstracts presented at ENDO must not be released to the press or the public until the date and time of presentation. For oral presentations, the abstracts are embargoed until the session begins. The Endocrine Society reserves the right to lift the embargo on specific abstracts that are selected for promotion prior to or during ENDO.*