Abstract Body (Do not enter title and authors here): Introduction
Clonal hematopoiesis of indeterminate potential (CHIP) has been shown to be an independent risk factor for heart failure and is associated with increased cardiomyopathy and mortality in cancer patients. Given that CHIP is associated with increased inflammatory cytokines, we sought to characterize the CHIP secretome in cancer patients and its impact on human engineered heart tissue (EHT) function.
Methods
We used machine learning to identify top cytokine features associated with CHIP in a cardio oncology cohort of 173 patients. Serum from breast cancer patients who were age, co-morbidity and medication matched between CHIP and non CHIP were used in the experiments. We used human EHTs to assess whether CHIP serum treatment affects contractility and used RNA sequencing to identify significantly differentially expressed pathways. Cytokine profiling of serum samples was done using a Luminex platform for 71 human cytokines.
Results
Machine learning identified IL-1B and CD40L as two top features of CHIP. Given DNMT3A CHIP was the most common in the cardio onc cohort, we focused on evaluating the effect of DNMT3A CHIP serum on EHT function. Compared to non CHIP serum, DNMT3A CHIP serum significantly reduced EHT contractility (p= 0.03). To evaluate whether the top cytokines identified by our machine learning model could contribute to the impairment in contractility, we co-incubated CHIP and Non-CHIP serum treated EHTs with antagonists of IL-1B or CD40L. Treatment with either IL-1B or CD40L antagonists rescued contractility defects in CHIP serum treated EHTs (p=0.015, 0.0095, respectively) but had no significant impact on non CHIP serum treated EHTs.
RNA sequencing of EHTs and pathway analysis revealed differential expression of genes and pathways in CHIP-treated and Non-CHIP serum-treated EHTs as well as in IL-1B antagonist and CD40L antagonist treated EHTs. Inflammation, metabolism pathways were top pathways differentially regulated in CHIP vs non CHIP serum treated EHTs and antagonist treatment showed modulation of these pathways.
Conclusion
Identification of IL-1B as a top feature of CHIP is consistent with prior work. CD40L, however, has not previously been implicated in CHIP and may be more unique to cancer patients with CHIP. We show that antagonizing this pathway was able to rescue the contractility defect in CHIP serum treated EHTs, which could be another therapeutic target in patients with CHIP. The effect of other CHIP variants are underway.