Impact of Clonal Hematopoiesis in Ischemic and Nonischemic Heart Failure

Michael J. Rauh, MD, PHD


J Am Coll Cardiol. 2021;77(14):1760-1762. 

Normal hematopoiesis is polyclonal, with circulating blood cells arising from a pool of approximately 50,000 to 200,000 hematopoietic stem cells (HSCs) in the bone marrow.[1] However, HSCs commonly acquire mutations with age, some of which are advantageous, particularly when they occur in the epigenetic regulator genes, DNMT3A and TET2. Expansion of mutant HSCs and their progeny in blood can be detected in at least 15% to 20% of older adults by conventional massively parallel sequencing (MPS). This phenomenon is called clonal hematopoiesis (CH) or, specifically, clonal hematopoiesis of indeterminate potential (CHIP), when the variant allele frequency (VAF) in blood reaches at least 0.02 (i.e., 2% of alleles or 4% affected blood cells, because CHIP mutations are usually heterozygous).[2] CHIP is a pre-malignant state, with normal blood cell counts or only subtle changes, associated with an overall risk of transformation to hematologic malignancy of approximately 0.5% to 1% per year.

Although CHIP is opening the door to hematologic malignancy surveillance and potentially preventive strategies, readers of the Journal will also be familiar with the unexpected relevance of CHIP to cardiovascular disease risk.[3–5] This is related in part to circulating and infiltrating cells of the monocyte/macrophage lineage that carry CHIP mutations and exacerbate inflammation, best characterized in Tet2-mutant murine models.[6–9] Inactivating Tet2 mutations worsens atherosclerosis and cardiac function in ischemic heart failure (HF) with reduced left ventricular ejection fraction (LVEF) and in aged mice.[6,7,9,10] In keeping with this, CHIP is associated with adverse outcomes in patients with ischemic HF with reduced LVEF.[11,12]

In this issue of the Journal, Pascual-Figal et al.[13] extend our understanding of the adverse clinical impact of CHIP in patients with chronic HF with ischemic and nonischemic etiology. They studied a cohort of 62 patients with HF who were older than 60 years, with an LVEF of <45% and at least 3 years of follow-up after blood sampling at inclusion. In this HF cohort, 52% of patients had nonischemic etiology, and the majority (66%) exhibited dilated cardiomyopathy. CHIP was detected in 39% of patients, with no difference between ischemic and nonischemic cases and an expected increase in prevalence with age.[2] More than one-half of CHIP cases were associated with mutations in DNMT3A or/and TET2 (henceforth referred to as D/T-CHIP), and most of the subsequent analysis was focused on these genes. The major findings included statistically significant associations between D/T-CHIP and lower survival time, the composite outcome of all-cause death or HF hospitalization, and adverse HF outcomes when considering competing risk analysis and adjusted models. Importantly, D/T-CHIP remained associated with adverse HF progression in patients with both ischemic and nonischemic disease in competing risk analyses, and no significant interaction was observed between ischemic HF etiology and the existence of CHIP for any included outcomes.

This study may represent the first evidence that clonal hematopoiesis is associated with accelerated progression of HF in the absence of ischemic heart disease. Its strengths include the rigorous characterization of patients and long-term outcomes, especially related to HF progression, and the sensitive (high-depth, error-corrected) MPS strategy to detect CHIP. The main limitation of this study, acknowledged by Pascual-Figal et al.,[13] is its modest sample size. This is somewhat offset by providing greater sensitivity to detect CHIP beyond studies using conventional MPS, by rich clinical characterization and long-term follow-up, and by robust findings after adjustment for relevant confounders. However, larger studies, potentially including younger patients and more women, will be required to validate their findings, especially the effect of CHIP on the progression of nonischemic HF.

The senior author, Fuster, and others have published evidence for increased pro-inflammatory cytokine signaling, particularly the NLRP3 inflammasome/interleukin (IL)-1β/IL-6 axis, in experimental murine models of Tet2-mutant atherosclerosis and HF.[6,7,9] Studies of human CHIP have revealed the strongest association with elevated circulating levels of IL-6,[14,15] and Pascual-Figal et al.[13] reproduced this finding for D/T-CHIP in their HF cohort. Murine models of Dnmt3a-mutant CHIP are limited but demonstrate some convergence of Dntm3a/Tet2 inactivation on dysregulated macrophage gene expression and cardiac dysfunction.[16] Furthermore, the groups of Dimmeler and Zeiher[17,18] have applied single peripheral blood mononuclear cell RNA sequencing in patients with ischemic HF with or without D/T-CHIP and matched control individuals to demonstrate transcriptomic changes in CHIP consistent with increased NLRP3/IL-1β/IL-6 activation and augmented monocyte interactions with endothelial and T cells. In a recent randomized clinical trial, the IL-1β inhibitor canakinumab reduced hospitalization for HF and hospitalization for HF or HF-related mortality in patients with inflammation and prior myocardial infection.[19] Taken together, these results suggest that patients with HF with D/T-CHIP may benefit from NLRP3/IL-1β/IL-6–directed therapy, including those with nonischemic etiology. However, larger prospective studies with CHIP stratification would be helpful to address the therapeutic potential.

Pascual-Figal et al.[13] also identify the lower D/T-CHIP boundary of 2% VAF as clinically relevant in the context of ischemic and nonischemic HF outcomes. A similar threshold (approximately 1% VAF) has also been defined for D/T-CHIP and worse outcome in a recent deep-sequencing study of ischemic HF.[20] Although it is not currently recommended to screen unselected patients with HF for CHIP,[4,5] these findings indicate that more sensitive MPS technologies may be required in the future to capture clinically relevant CHIP clones. Finally, these studies consider CHIP mutations to be mostly equivalent, but it is clear that they vary in their intrinsic fitness and biological effects and may be influenced by extrinsic factors,[21] which will need to be factored into future precision and preventive strategies for patients with HF.