DNMT3A-Mutant Clonal Haematopoiesis Is Associated with Worse Recovery after ST-Elevation Myocardial Infarction Via Increased Inflammation

Clonal Haematopoiesis (CH) is an age-related phenomenon where haematopoietic stem cells (HSCs) acquire somatic mutations in genes related with blood cancers in individuals without development of overt blood disease. As mutant HSCs differentiate, mature circulating blood cells, including immune cells, carry mutations which may alter their function. In addition to associations with increased risk of blood cancers, CH has been linked to increased incidence and worse outcomes of cardiovascular disease (CVD) in retrospective studies. There are ~750K and ~30K cases of ST-elevation myocardial infarction (STEMI) annually in the USA and UK respectively, and ischemic heart disease is the most common cause of heart failure (HF). Despite current treatment with evidence-based therapy, 2-year mortality for HF remains as high as 23%. Given the high frequency of CH in older individuals, we aimed to characterise the impact of CH on recovery of left ventricular ejection fraction (DLVEF, median interval 4.5 months after acute STEMI) in a large prospective cohort of STEMI patients admitted to King’s College Hospital, UK (n=243, mean age 64 years, 21% female).

97 STEMI patients (39.9%) had a detectable CH mutation (VAF ≥ 1%) across 33 genes in peripheral blood. Of those with CH, 41 had at least one other mutation indicating high rates of clonal complexity. Reflecting the distribution of CH in the wider population, the majority of mutations occurred in DNMT3A (n=43) and TET2 (n=27). The only clinically significant difference in the baseline characteristics was that CH patients were older than non-CH (68 vs. 61 years, p<0.001). Interestingly, neither clone size nor number of mutations were associated with worse LVEF recovery. However, our key finding was that patients with DNMT3A CH had significantly lower DLVEF vs. non-CH controls, whereas TET2-CH were no different from controls.

Clone size was highest in flow-sorted peripheral blood monocytes in DNMT3A- (n=7) and TET2-mutant patients (n=6) compared with B (p<0.0001) and T cells (p<0.0001). Monocytes from DNMT3A-CH had significantly higher proportions of pro-inflammatory CD14+ classical monocytes vs. non-CH (p<0.05) and significantly higher expression of a monocyte activation marker HLA-DR (by FACS) that mediates antigen presentation to helper T cells, compared with TET2-CH (p<0.05). These findings suggest monocytes from DNMT3A-mutant STEMI patients are more pro-inflammatory during acute STEMI.

We performed scRNA-seq on PBMCs from 23 STEMI patients (10 non-CH, 10 DNMT3A-CH and 3 TET2-CH). We found upregulation of Alarmins (e.g. S100A8, S100A9 and S100A12) in DNMT3A-CH vs. both non-CH and TET2-CH. Pathway enrichment analysis confirmed significant upregulation of the receptor for advanced glycation end products (RAGE) gene set in classical and intermediate monocytes. RAGE mediates a damage-associated molecular pattern (DAMP) response to inflammatory molecules released from necrotic cells.

To further examine the inflammatory milieu of STEMI patients, we tested plasma from n=93 patients for cytokines. We found significant upregulation of multiple inflammatory cytokines in DNMT3A-CH compared with non-CH. However, these cytokine levels were largely similar between DNMT3A and TET2-CH. Considering that TET2-CH patients do not have inferior recovery post-STEMI, we focused on potential further immune differences between DNMT3A and TET2-CH. Quantifying T cell frequencies using our scRNAseq data, TET2-CH individuals had lower proportions of CD4+ T effector memory and TH17 cells compared with DNMT3A-CH. Furthermore, versus TET2-CH, these T cell populations in DNMT3A-CH upregulate IFITM1, S100A9 and IL7R which suggests heightened T cell responses to inflammatory stimuli.

While some inflammation is vital in repairing infarcted myocardium, aberrant inflammation is linked to worse remodelling and clinical outcomes. In dissecting the immune cell populations in STEMI patients, our data highlights multiple layers of hyperinflammation in DNMT3A-CH associated poorest recovery and highest risk of HF development post-STEMI that may be amenable to novel anti-inflammatory therapies in the future.