Associations between mosaic chromosomal alterations and incident cardiovascular disease in 419,000 individuals
S Flynn, A Schuermans, S Zelavat, M Uddin, T Nakao, S Murillo, P Libby, G Genovese, P Natarajan, M HonigbergAbstract
Background
Clonal hematopoiesis refers to the age-related clonal expansion of blood stem cells with genomic alterations and is linked to cancer and other adverse outcomes. These genomic alterations range from single-nucleotide variants or short insertions/deletions, termed clonal hematopoiesis of indeterminate potential (CHIP), to larger structural alterations, termed mosaic chromosomal alterations (mCAs). Certain gene-specific forms of CHIP associate consistently with cardiovascular disease (CVD), and recently loss of sex chromosome Y (LOY) has been associated with incident myocardial infarction. However, overall associations between mCAs and CVD are not well described.
Purpose
To test the association of the full spectrum of mCAs – including autosomal mCAs, LOY, and loss of chromosome X (LOX) – with incident CVD.
Methods
This study included UK Biobank participants without prevalent hematologic malignancy or CVD and with available genotype data. The co-primary exposures were autosomal mCAs, LOY, and LOX, ascertained from genome-wide genotyping. The primary outcome, ascertained by linkage to national health records, was composite CVD (coronary artery disease [CAD], atrial fibrillation, heart failure, ischemic stroke, or aortic stenosis); secondary analyses examined outcomes individually. Associations of mCAs and large mCAs (clonal fraction ≥10%) with incident CVD were tested using multivariable Cox regression, adjusted for baseline demographics, cardiovascular risk factors, and medications. Two-sided P<0.017 indicated statistical significance.
Results
Among 419,762 individuals (mean age 56.2 [SD 8.1] years; 56% female), 58,187 (13.9%) had one or more types of mCAs (12,386 [3.0%] with large mCAs). The primary CVD endpoint occurred in 41,531 individuals (9.9%) over a median of 11.1 years (IQR 10.4-11.8). CVD incidence was higher in individuals with autosomal mCAs and LOY (Figure 1). The presence of autosomal mCAs or LOY was associated with multivariate adjusted HR of composite CVD (HR, 95% CI 1.08 [1.03-1.13], p=0.002 and 1.04 [1.01-1.07], p=0.016, respectively). These relationships were stronger in large clonal fraction autosomal mCAs and LOY (HR 1.24 [1.12-1.36] and 1.12 [1.07-1.17], respectively, both p<0.001). LOX was not associated with CVD when adjusted for multiple comparisons. Autosomal mCAs were associated with incident CAD, and large autosomal mCAs with CAD, heart failure, and atrial fibrillation. LOY was associated with incident atrial fibrillation, and large LOY with CAD and atrial fibrillation, with trend towards significance for heart failure and aortic stenosis (Figure 2). Findings were unchanged in sensitivity analyses excluding individuals with CHIP or with both autosomal and sex mCAs.
Conclusion
Our findings suggest that mCAs, particularly large LOY and large autosomal mCAs, constitute a novel risk factor for CVD. Future work should investigate pathways through which specific mCAs relate to CVD.