DNMT3A-mutant clonal haematopoiesis and the recovery of patients with acute ST-elevation myocardial infarction

Abstract

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): British Heart foundation Background Haematopoietic stem cells can acquire mutations and produce mutant blood cells. When this occurs in the absence of overt blood disease, it is clonal haematopoiesis of indeterminate potential (CHIP). CHIP is a novel, independent risk factor for ischaemic heart disease and heart failure (HF)(1). However, the molecular mechanisms are poorly understood. Myocardial infarction (MI) is a leading cause of HF due to ventricular remodelling and scar formation(2). We postulate CHIP mutations result in aberrant inflammation that contributes to maladaptive ventricular remodelling following MI. We investigated clinical outcomes and immune cell biology in CHIP and non-CHIP patients (pts) with acute ST-segment elevation MI (STEMI). Purpose Identify and quantify mutations with variant allele frequency (VAF)≥1% in STEMI pts. Assess leukocyte mutational burden. Investigate transcription profiles and leukocyte function in CHIP and non-CHIP STEMI pts. Methods 119 pts diagnosed with STEMI were prospectively recruited (REC19/SC/0362). We collected clinical data, including echocardiogram (echo) parameters and isolated plasma and peripheral blood mononucleated cells (PBMCs) from blood collected 48-96 hours post onset of chest pain. PBMCs were screened for mutations and analysed by flow cytometry. We assessed CHIP VAF with digital droplet PCR in flow-sorted monocytes (Mono), T and B-cells. We performed single-cell RNAseq (10x Genomics) on PBMCs. Luminex or ELISA kits were used to measure plasma proteins. Results 59/119 recruited pts had CHIP (VAF≥1%). CHIP pts were significantly older than non-CHIP pts. DNMT3A was the most commonly mutated gene (28%), followed by TET2 (17%) and ASXL1 (5.3%). 53 pts had impaired LV ejection fraction (LVEF) at admission and a follow-up Echo at median 4.6 months (2.5-8 IQR) post STEMI. They were classified as responder (R) n=24, Δ LVEF≥5% or non-responder (NR) n=29, ΔLVEF<5%. Pts with DNMT3A CHIP were 12 times more likely to be NR (p=0.02). Baseline LVEF was similar in CHIP and non-CHIP pts, but ΔLVEF was significantly reduced in DNMT3A vs non-CHIP pts (p<0.01). DNMT3A mutations were highest in Mono (mean VAF 6.3%), present in B cells (3%) but rare in T cells (1.7%). Within the Mono population, DNMT3A mutant pts had significantly higher proportions of classical Mono compared to non-CHIP (p=0.02). Using scRNAseq, we computed Mono trajectories and differential gene expression. Compared to non-CHIP, DNMT3A-mutant samples upregulated receptor for advanced glycation end-products (RAGE) and toll-like receptor signaling gene signatures. Despite this, we did not observe raised leukocyte count or detect increased levels of inflammatory cytokines in DNMT3A-mutant pts plasma samples. Conclusions DNMT3A CHIP is associated with poor LVEF recovery post-STEMI due to a dysregulated inflammatory response to acute myocardial injury. Furthermore, our data suggests that this may occur via different mechanisms than non-CHIP NR.