Importance: Somatic mutations causing clonal expansion of hematopoietic cells (clonal hematopoiesis of indeterminate potential, CHIP) are increased with age and associated with atherosclerosis and inflammation. Age and inflammation are the major risk factors for heart failure, yet the impact of CHIP on heart failure in humans is entirely unknown.

Objective: To assess the potential prognostic significance of CHIP in patients with chronic heart failure due to ischemic origin (CHF).

Design: We analyzed bone marrow-derived mononuclear cells from 200 patients (pts) with CHF by deep targeted amplicon sequencing to detect the presence of CHIP and associated such with long-term prognosis in patients with CHF.

Results: Median age of the patients was 65 years, NYHA class 2 and left ventricular ejection fraction 31%. 47 mutations with a variant allele fraction (VAF)≥0.02 were found in 38 out of the 200 pts with CHF (18.5%). The somatic mutations most commonly occurred in the genes DNMT3A (14 cases), TET2 (9), KDM6A (4), and BCOR (3) followed by 13 other genes. None of the classical CHF baseline parameters differed significantly between CHIP and Non-CHIP carriers, except that CHIP carriers were older and suffered more frequently from a history of hypertension. During a median clinical follow-up of 4.4 years, a total of 53 pts died, and 23 pts required hospitalization for heart failure. There was a significantly worse long-term clinical outcome for pts with either DNMT3A or TET2 mutations compared to Non-CHIP carriers. By multivariable Cox proportional regression analysis including age and a history of hypertension, the presence of somatic mutations within TET2 or DNMT3A remained independently associated with adverse outcome (HR 2.1, 95% confidence interval 1.1 – 4.0, p=0.02, for death combined with heart failure hospitalization) in addition to age (HR 1.04, 95% confidence interval 1.01 – 1.07 per year, p=0.005), whereas hypertension was not independently associated in this multivariable model. Likewise, death and rehospitalization for heart failure remained independently associated with the presence of TET2 or DNMT3A mutations, when NT-proBNP serum levels and Seattle Heart Failure Model (SHFM) score were used as covariates. Importantly, there was a significant dose-response relationship between VAF and clinical outcome.

Conclusions and Relevance: Our data suggest that somatic mutations in hematopoietic cells, specifically in the most commonly mutated CHIP driver genes TET2 and DNMT3A, may be significantly associated with the progression and poor prognosis of CHF. Future studies will have to validate our findings in larger cohorts and to address whether targeting specific inflammatory pathways may be valuable for precision medicine in pts with CHF carrying specific mutations encoding for CHIP.