Aude Benigne Ikuzwe Sindikubwabo

Decoding the Genetic Architecture of Pregnancy Loss and Adverse Outcomes: Insights into Rare, Common, and Structural Variants from Large-Scale Biobanks

Aude Benigne Ikuzwe Sindikubwabo1, Lannawill Caruth1, Lindsay Guare1,Pauline Gachanja1,Craig Teerlink2, Julie Lynch5, Maja Bucan3, Ziyue Gao3, Courtney Schrieber4, Regeneron Genetics Center, Penn Medicine BioBank, Shefali S. Verma1 1Department of Pathology and Laboratory Medicine, Perelman School of Medicine, Philadelphia, PA, United States 2Departments Primary - Internal Medicine, Adjunct - Family & Preventive Medicine, University of Utah, UT, United States 3Department of Genetics, Perelman School of Medicine, Philadelphia, PA, United States 4Department of Obstetrics and Gynecology, Perelman School of Medicine Philadelphia, PA, United States 5VA Informatics and Computing Infrastructure, VA Salt Lake City Healthcare System (VINCI), Salt Lake City, UT, USA

Pregnancy Loss (PL) is the most common pregnancy complication worldwide. While chromosomal abnormalities account for 60% of cases, the remaining 40% classified as Unexplained Pregnancy Loss (UPL) - lack a defined cause. The etiology of UPL and other adverse pregnancy outcomes (APOs), such as recurrent PL (RPL) and preterm birth (PTB), remains poorly understood, particularly with respect to genetic contributions from rare, common, and structural variants. To elucidate the mechanisms underlying PL and other APOs, we conducted exome-wide and genome-wide association studies across genetically diverse cohorts from four major biobanks (PMBB,AllofUs,MVP,eMerge). We performed rigorous meta-analyses across these datasets, enabling well-powered discovery of genetic associations. EXWAS meta-analyses revealed 17 genes that reached Bonferroni significance. We identified loss-of-function and damaging missense variants in ARNT (p = 9.91×10⁻⁷) and TWSG1 (p = 4.91×10⁻⁷) associated with RPL, loss-of-function variants in LPA (p = 1.72×10⁻⁶) were linked to PTB, and variants in HSPB9 (p = 1.24×10⁻⁶) to be associated with UPL. Among AFR, we observed two additional significant associations with both RPL and PTB for variants in KLH35 (p = 2.55×10⁻⁶), AUH (p = 1.73×10⁻⁶), and VAC14 (p = 2.46×10⁻⁶). Our GWAS meta-analysis revealed locus in TRAPPC12 (rs4514905, p = 3.42×10⁻⁸) to be associated with UPL while EYS (rs11754133, p = 5.6×10⁻⁹) was associated with RPL, across ancestries. Structural variant analysis in the AllofUs dataset identified CFAPG1 (p = 7.6×10-6) as associated with UPL.This study uncovers a spectrum of genetic contributors to UPL and related outcomes, integrating statistically significant associations with functional relevance.Many identified genes are involved in embryonic development, immune regulation, and reproductive processes-supporting their potential roles in pregnancy outcomes. Beyond confirming known loci, we reveal novel ancestry-specific candidates, highlighting diverse genetic risk factors. These findings advance understanding of pregnancy loss mechanisms and suggest new approaches for precision diagnostics and therapeutics.