Clonal hematopoiesis and risk of mortality after incident ischemic stroke: The reasons for geographic and racial differences in stroke Study

Abstract

Abstract Introduction: Clonal hematopoiesis of indeterminate potential (CHIP) is associated with an increased risk of stroke and all-cause mortality. A recent study reported that CHIP at a variant allele frequency (VAF) ≥1% was associated with a higher risk of recurrent stroke, myocardial infarction and all-cause mortality after ischemic stroke (PMID: 36441964). However, this study included only White individuals with a short follow up (3 years). The association between CHIP and mortality in diverse population groups with stroke followed long-term remains unstudied. We leveraged a large prospective study, the REasons for Geographic and Racial Differences in Stroke (REGARDS) to address this knowledge gap. Methods: REGARDS study recruited 30,239 self-identified Black and White individuals ≥45 years across the US between 2003 and 2007 and collected extensive health data by telephone and in-person exams. Participants also provided blood samples at study enrollment from which DNA was extracted. Details on identification and adjudication of stroke and other comorbidities were previously reported (PMC3595534). Participants were contacted by the study staff every 6 months to ascertain possible stroke, and this was confirmed by review of medical records. The current analysis included participants who developed incident ischemic stroke and had DNA samples available for CHIP testing. Participants with history of cardiovascular disease at baseline were excluded. Error-corrected targeted deep sequencing using Twist bioscience hybrid capture technology on the Illumina platform was used to detect coding sequences of 24 genes that identify >95% of CHIP mutations. VAF ≥2% was used to define CHIP. Cox proportional hazards models were used to calculate the hazard ratios (HRs) and 95% confidence intervals (CI) of mortality associated with CHIP, adjusting for clinically relevant variables that were significant with a p-value <0.1 in univariate analysis. We also adjusted for time between study enrollment and stroke to account for the known exponential growth pattern for CHIP mutations. We calculated HRs at different VAFs for the dominant mutation (≥2-10, >10%) and for the number of mutations (1, 2+). Stratified analysis was performed by age, sex and race due to known associations of CHIP with older age, male sex and White individuals. Results: Of the 1,580 participants who developed incident ischemic stroke in REGARDS (adjudicated by September, 2021), 1,112 had DNA samples available for CHIP analysis. Median age at study enrollment was 68 years (standard deviation [SD]: 8.8); 49.9% were male and 53.6% Black individuals. CHIP was present at study enrollment in 23.6% participants at VAF ≥2%. The majority (69.1%) had a single mutation; DNMT3A (57.6%) and TET2 (24.8%) were the most common mutations. Median time from study enrollment to stroke was 6 years and median time from stroke to last follow up for those alive was 5.5 years. Adjusted for age, hypertension, diabetes, dyslipidemia, smoking, atrial fibrillation and time from study enrollment to stroke, baseline CHIP (yes vs. no)was not associated with increased mortality overall (HR=1.02, 95% CI: 0.85-1.22). However, increased mortality risk was seen with multiple CHIP mutations (≥2 mutations vs. no CHIP) and high VAF CHIP (>10 vs. ≤10%) in specific demographic groups: 1) participants ≤65 years with ≥2 mutations (HR=2.63, 95% CI: 1.37-5.05) and with high VAF CHIP (HR=1.93, 95% CI: 1.05-3.55), and 2) men with ≥2 mutations (HR=1.52, 95% CI: 1.04-2.20) and with high VAF CHIP (HR=1.57, 95% CI: 1.13-2.19). For participants older than 65 years, overall CHIP (yes vs. no; HR=0.95, 95% CI: 0.78-1.17), ≥2 mutations (HR=0.89, 95% CI: 0.66-1.19) or high VAF CHIP (HR=0.92, 95% CI: 0.68-1.24) were not associated with mortality risk. High VAF CHIP was associated with lower mortality in women (HR=0.62, 95% CI: 0.40-0.97). Conclusion: In this study, CHIP was associated with increased risk of mortality after ischemic stroke only in individuals ≤65 years and men with multiple mutations or high VAF CHIP. Results may be used in future to refine personalized risk stratification after stroke and design interventions for vulnerable populations.