Clonal Hematopoiesis and Venous Thromboembolism in the UK Biobank

Background: Patients with myeloid malignancies have an increased risk of thrombosis, but less is known about the risk of thrombosis in clonal hematopoiesis of indeterminate potential (CHIP). We evaluated genotype-specific associations between CHIP and both prevalent and incident venous thromboembolism (VTE) in the largest such analysis to date of the UK Biobank.

Methods: We examined 425,573 individuals without a prevalent hematologic malignancy diagnosis from the UK Biobank. Whole exome sequencing and single nucleotide polymorphism array data were analyzed to identify individuals with CHIP and/or mosaic chromosomal alterations (mCA), respectively. Prevalent and incident VTE cases were ascertained with respect to the time of DNA sampling. VTE events were identified based on ICD10 codes, including but not limited to, pulmonary embolism, phlebitis/thrombophlebitis of deep veins of lower veins, portal vein thrombosis, and Budd Chiari syndrome. Median follow-up time for incident VTE was 11.8 years. Associations of CHIP and VTE were assessed using logistic regression for prevalent analyses and Cox proportional hazards for incident analyses. The association analyses were adjusted for age at the time of DNA sampling, sex, Caucasian ancestry, genetic principal components, ever smoked status, and body mass index (BMI) as covariates.

Results:

Prevalent VTE and CHIP:

While CHIP overall was not associated with an increased risk of prevalent VTE (odds ratio (OR) 1.06; 95% CI 0.87-1.28, P= 0.56) (Figure 1), JAK2-mutant CHIP and JAK2 CHIP with loss of heterozygosity at the JAK2 locus (LOH 9p) were powerfully associated with increased risk of thrombosis, OR = 7.49 (95% CI 3.02-18.56, P=1.41e-05) and 6.74 (95% CI 1.62-28.08; P=8.71e-03). Having both CHIP and autosomal mCA was associated with a nominally increased risk of VTE (OR 1.76; 95% CI 1.02-3.06, P= 0.04). Otherwise, non-JAK2 CHIP, ASXL1 CHIP, TET2 CHIP, DNMT3A CHIP were not associated with increased prevalent VTE, regardless of clone size.

Incident VTE and CHIP:

CHIP was modestly associated with incident VTE with a hazard ratio of 1.15 (1.03-1.28; P=0.01), and CHIP with a larger clone size (VAF > 10%) was associated with a hazard ratio of 1.22 (95% CI 1.08-1.39; P= 0.001) (Figure 2). Consistent with our analysis of prevalent VTE, JAK2-CHIP and JAK2-CHIP with LOH 9p were strongly associated with incident VTE with hazard ratios of 3.76 (95% CI 1.79-7.9, P= 0.0004) and 4.89 (95% CI 1.84-13.04; P=0.0015). Other associations, including overall CHIP, CHIP with VAF >10%, CHIP together with autosomal mCA, autosomal mCA, non-JAK2 CHIP, TET2 CHIP with VAF >10%, showed modest increased hazard ratios between 1 and 2.

Conclusions: In this study of risk of thrombosis in the UK Biobank, CHIP was not significantly associated with prevalent VTE and was modestly associated incident VTE. Among CHIP genotypes analyzed, only JAK2 mutations were significantly associated with both prevalent and incident VTE risk (OR 7.49 and HR 3.76, respectively). JAK2-mutant CHIP may be an undetected contributor to cases of VTE and may have therapeutic implications.