Unadjusted Confounding in the Association Between Clonal Hematopoiesis and Heart Failure After Cancer Therapy

To the Editor A significant association between clonal hematopoiesis of indeterminate potential (CHIP) and increased risk of heart failure following cancer therapy was recently reported in the cohort study by Shyr and colleagues. In multivariable models adjusting for age, sex, race and ethnicity, cancer treatment regimen, hypertension, and low-density lipoprotein cholesterol, factors that were included in CHIP-associated heart failure (subdistribution hazard ratio [sHR], 1.26; 95% CI, 1.02-1.56; P = .03).

However, many well-known confounding factors, including diabetes, HIV infection, obesity, chronic kidney disease, and cirrhosis, were not adjusted for in their study. Moreover, no in-depth discussion or sensitivity analyses were conduct to assess the bias introduced by these unadjusted confounders.

Given that diabetes, HIV infection, obesity, chronic kidney disease, and cirrhosis are well-known risk factors for heart failure, the absence of formal assessment of unadjusted confounding substantially limited the robustness of the statistically significant findings and constrains their causal interpretability.