Gene-Based Rare Variant Burden Analyses Across Biobanks Identify Novel High-Risk Genes for Thoracic Aortic Disease

Background: Thoracic aortic aneurysms enlarge silently and can cause fatal aortic dissection without timely surgical repair, underscoring the need for improved approaches to identify individuals at high risk. Rare pathogenic variants in established heritable thoracic aortic disease (HTAD) genes explain only a subset of familial and fewer nonfamilial thoracic aortic disease (TAD) cases. Methods: We performed phenotype-stratified, genome-wide, gene-based rare-variant burden analyses of ultrarare damaging missense and predicted loss-of-function variants. Primary analyses focused on aortic dissection, thoracic aortic aneurysm requiring surgical repair, and their combined phenotype. Broader thoracic aortic aneurysm (TAA) was evaluated as a secondary phenotype. Discovery analyses were conducted in the UK Biobank and All of Us, followed by independent replication in the Penn Medicine BioBank, Mass General Brigham Biobank, and Million Veteran Program. Discovery and replication results were subsequently combined in an overall fixed-effect, inverse-variance-weighted meta-analysis across up to five biobanks. Implicated genes were further evaluated in additional clinically ascertained TAD cohorts and using single-cell transcriptomic data from human thoracic aortic tissue. Results: Discovery analyses identified 80 genes reaching study-wide significance across the prespecified TAD phenotypes. These included six established and two putative HTAD genes. Fourteen genes demonstrated independent replication support and reached study-wide significance in the overall meta-analysis across up to five biobanks, which included more than 10,000 cases and 880,000 controls. The eight novel candidate genes among these were FNDC3B, ROCK1, URM1, SLFN11, ENPP1, CLEC16A, CREM, and VCAN. Associations were strongest for dissection and TAA requiring surgical repair. Four novel associations were driven exclusively by missense variants. FNDC3B was observed in a family with HTAD, while additional variants were identified primarily in sporadic dissection or aortic surgery cohorts, suggesting that other genetic or physiologic factors may influence penetrance. The implicated genes showed cell-type-specific expression patterns in human thoracic aortic tissue. Conclusions: These findings expand the genetic architecture of TAD by identifying eight novel candidate genes and demonstrate the utility of phenotype-stratified rare variant burden analyses across large biobanks for gene discovery.