Etiological and molecular determinants of atrial endomysial fibrosis: results from the CATCH ME consortium

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – EU funding. Main funding source(s): CATCH ME: Characterizing Atrial fibrillation by Translating its Causes into Health Modifiers in the Elderly Background We recently demonstrated that in human atria, endomysial fibrosis causes conduction disturbances during atrial fibrillation (AF), while overall connective tissue content (over-all fibrosis) has no effect. Etiological and molecular determinants of endomysial fibrosis are largely unknown. Methods We quantified over-all and endomysial atrial fibrosis using staining with WGA in left (LA, n=95) and right (RA, n=76) atrial appendages sampled from a European cohort (CATCH ME) of patients undergoing cardiac surgery. The contributions of AF, heart failure (HF), sex, age, and 4 principal components accounting for confounding clinical characteristics to over-all and endomysial fibrosis were determined in a multivariate model. RNA sequencing was performed to explore biological pathways associated with the two types of fibrosis. Results Over-all and endomysial fibrosis were moderately correlated (LA: r=0.69, RA: 0.61, both p<0.001) and more pronounced in RA than in LA samples (p<0.001). In LA, persistent AF, female sex, and HF were independently associated with endomysial fibrosis, while over-all fibrosis was only associated with sex and HF. Likewise, in RA, persistent AF was associated with endomysial fibrosis, while over-all fibrosis was not. Expression of 18 genes was univariately associated with endomysial fibrosis in LA but only 1 in RA. There were no genes associated with over-all fibrosis. After correction for relevant clinical traits, BMP10 showed the strongest association with endomysial fibrosis amongst coding genes (p=3.16E-04, LA). In an independent validation cohort (RACE V Tissue Bank Project, n=162 pts undergoing cardiac surgery), out of 11 biomarkers reflecting inflammation, fibrosis, vascular dysfunction, or ischemia, BMP10 plasma levels were most strongly associated with endomysial fibrosis (p<0.001), but not with over-all fibrosis. Conclusions Taken together, female sex, HF, and AF are the main drivers of fibrosis in human atria. However, endomysial fibrosis is independently associated with AF whereas over-all fibrosis is not. BMP10 is a specific marker of endomysial fibrosis. Our findings suggest that distinct mechanisms are involved in the development of over-all versus endomysial fibrosis in human atria and may provide an explanation for the recently reported predictive value of BMP10 biomarkers levels for AF recurrences after AF ablation and for stroke.