Skip to content

Klarin D, Devineni P, Sendamarai AK, et al. Genome-wide association study of thoracic aortic aneurysm and dissection in the Million Veteran Program. Nat Genet. 2023;55:1106-1115. PMID 37308786

One-paragraph summary

Prior understanding of the genetic determinants of thoracic aortic aneurysm and dissection (TAAD) came almost entirely from rare, Mendelian forms of disease. This study tested approximately 25 million DNA sequence variants in 8,626 participants with and 453,043 without TAAD in the Million Veteran Program, with replication in an independent sample of 4,459 cases and 512,463 controls drawn from six cohorts. Twenty-one TAAD risk loci were identified, 17 of which had not been previously reported. Multiple downstream analytic approaches were applied to nominate causal TAAD risk genes and the cell types in which they act. The authors used the resulting genetic architecture to argue that TAAD is a non-atherosclerotic aortic disorder, distinct from other forms of vascular disease, and concluded that the genetic architecture of TAAD mirrors that of other complex traits and "is not solely inherited through protein-altering variants of large effect size."

Key findings

  • 21 TAAD risk loci from 8,626 cases, 17 novel — the first well-powered case-control GWAS of the clinical disease rather than of aortic size.
  • Independent replication in 4,459 cases and 512,463 controls across six cohorts, which is what separates this from earlier underpowered aortic GWAS attempts.
  • Human genetic evidence that TAAD is non-atherosclerotic and genetically distinct from coronary and peripheral arterial disease — a claim with direct therapeutic implications.
  • Causal gene and cell-type prioritisation via downstream analyses, connecting GWAS signal to the cell populations characterised in the single-cell atlases (PMID 33017217, 36172868).
  • The explicit reframing that TAAD's genetic architecture "mirrors that of other complex traits" — i.e. the Mendelian-gene model that has governed clinical genetic testing describes only part of the risk.

Limitations

  • Veteran population. The Million Veteran Program is predominantly male and older than the general TAAD population, and Marfan, Loeys-Dietz and other syndromic aortopathies present decades earlier; case ascertainment is therefore skewed toward degenerative and bicuspid-valve-related disease.
  • Phenotype from clinical codes. TAAD case definition relies on electronic health record diagnosis rather than adjudicated imaging, so misclassification — particularly the conflation of aneurysm with dissection, and of thoracic with abdominal disease — is a real risk at this scale.
  • Aneurysm and dissection pooled. The two are combined into a single phenotype despite being different events with different determinants; loci for aneurysm formation and loci for wall failure cannot be distinguished in this design.
  • Not reconciled with diameter GWAS. The relationship between these 21 disease loci and the 82 + 47 aortic-diameter loci (PMID 34837083) or the 79 segment-specific loci (PMID 35902171) is not formally worked out — arguably the most important open question the paper creates.
  • Prioritised genes are nominations, not proofs. Causal gene assignment rests on statistical and expression-based inference; experimental validation is not part of the study.

Why it matters

This paper does for aortic dissection what large-scale case-control GWAS did for coronary disease: it establishes that a condition long modelled as a set of rare monogenic syndromes has a substantial common-variant component. Two consequences propagate outward. First, clinically: negative gene-panel testing in a patient with familial aortic disease no longer means "no genetic contribution", and polygenic scores become a legitimate object of study alongside single-gene diagnosis (see the genetics and biomarkers pages). Second, therapeutically: the finding that TAAD is genetically separable from atherosclerotic vascular disease is the strongest available argument against importing abdominal aortic aneurysm and coronary therapeutics — statins, metformin, colchicine — into thoracic aortic trials without dedicated testing. Every registered metformin aneurysm trial retrieved is in the abdominal aorta (NCT04500756, NCT04224051, NCT03507413), and this paper is the reason that gap should be treated as a scientific question rather than an oversight.

Cited by wiki pages

  • omics and emerging science
  • clinical trials landscape