Apical myocardial fibrosis burden identifies a high-risk phenotype and predicts cardiac mortality after LVAD implantation.

Abstract

Advanced heart failure is characterized by progressive myocardial remodelling, with fibrosis representing a final common pathway of chronic injury. We investigated whether quantitatively assessed apical myocardial fibrosis burden at the time of left ventricular assist device (LVAD) implantation is associated with overall and cause-specific mortality. This single-centre retrospective cohort study included 47 consecutive patients undergoing durable LVAD implantation between 2019 and 2025. Median age was 63 years and 14.9% were female. Myocardial tissue obtained from the left ventricular apical core at implantation was analysed using structured histopathologic scoring (THS-10) and quantitative digital morphometry (QuPath). Fibrosis burden was quantified as collagen-positive area relative to total myocardial area. Overall survival was assessed using Kaplan-Meier analysis and Cox regression. Cardiac mortality was evaluated using Fine-Gray competing-risk models, with non-cardiac death treated as a competing event. Median follow-up was 1513 days. During follow-up, 27 deaths occurred (57.4%), including 16 cardiovascular and 11 non-cardiac deaths. Quantitative fibrosis burden showed a strong unadjusted association with overall survival (log-rank P = .00042). Receiver operating characteristic (ROC) analysis identified a cohort-derived fibrosis threshold of 33.7% for overall mortality risk stratification (area under the curve [AUC] 0.739, 95% confidence interval [CI] 0.566-0.912). In univariable Cox analysis, fibrosis burden and THS-10 score were associated with mortality (hazard ratio [HR] 39.39, 95% CI 2.61-594.56, P = .008; and HR 1.51, 95% CI 1.14-2.01, P = .005, respectively). After multivariable adjustment for clinical severity, including INTERMACS profile and key laboratory parameters, the association with all-cause mortality was attenuated. In competing-risk analysis, high fibrosis burden independently predicted cardiac mortality (subdistribution HR approximately 4.1, P = .02). Additional exploratory ROC analysis for cardiovascular death showed an AUC of 0.752, with a cohort-derived threshold of 39.6%. Quantitatively assessed apical myocardial fibrosis burden identifies a biologically high-risk myocardial phenotype in LVAD recipients. While the association with all-cause mortality is attenuated after adjustment, fibrosis burden shows a specific relationship with cardiac mortality when competing non-cardiac risk is taken into account. These findings support the prognostic relevance of myocardial substrate and warrant external validation in larger prospective cohorts.