135 An improved ejection fraction parameter capable of representing cardiac function regardless of heart morphology to distinguish hfpef from normal hearts

Background Ejection Fraction (EF) has been an important parameter describing cardiac function, because of earlier work demonstrating its correlation with outcomes. However, during conditions such as Heart Failure preserved ejection fraction (HFpEF), EF fails to distinguish HFpEF from healthy patients. There are further reports that EF can be skewed by the geometry of the heart, and Heart Failure (HF) can present a wide variety of cardiac morphologies due to remodelling. The reason for the poor performance of EF and its dependency on geometry is unclear, and it is further unclear if such geometric changes from HF remodelling affects cardiac function. We strive to address this here, and derive an improve and simple EF parameter to resolve this. Methods We developed a simple discretized numerical model to relate incompressible myocardial strains to stroke volume. We used data from two porcine animal models of heart failure, one for HFpEF and one for HF reduced EF (HFrEF), and literature clinical measurements to inform our model. We used the model to test the effects of geometric changes relevant to HF on the ability of the heart to convert myocardial strains to flow function. Results Our animal models showed that cardiac dilation and wall thickening are primary features relevant to HF. Further investigation via our numerical model showed that wall thickening with no change to strain artificially increased EF, while cardiac dilation with no change to strain artificially decreased EF, demonstrating that EF can be skewed by geometric changes during HF remodelling, and is an inaccurate representation of cardiac function. We further showed that this is because EF is calculated using the endocardial boundary rather than the mid-wall layer, because a corrected EF parameter (CEF) that uses the mid-wall layer for quantification resolves this shortcoming. This CEF became independent of cardiac geometry, and could successfully distinguish HFpEF and healthy heart in our animal models, where EF could not. The CEF can be calculated easily with measurements of EF, wall thickness, and LV inner diameter. We further showed that the myocardial strains at the endocardial and epicardial boundary deviated significantly from each other and from the strains at the mid-wall layer, and this magnitude of deviation depended on the cardiac geometry, suggesting that any quantification of cardiac function using the epi- or endo- boundaries can be skewed by geometric changes to the heart, and will be ineffective. Finally, we tested specific types of HF morphologies, namely, eccentric hypertrophy, concentric hypertrophy, and concentric remodelling with our numerical model. We found that the concentric remodelling morphology is the most inefficient in converting myocardial strains to stroke volume, while eccentric hypertrophy is the most efficient. This may explain why HFpEF hearts, which are likely have wall thickening similar to concentric remodelling, have exercise intolerant, and our results suggest that cardiac dilation during HFrEF is providing advantages to help with flow function. Conclusion We demonstrated that geometric changes during HF remodeling can significantly impact cardiac function, and can skews functional parameters like EF. We further showed that the reason for EF’s shortcoming is due to a flawed reliance on quantification using the endocardial boundary of the heart rather than the mid-wall layer. We proposed a new CEF parameter to replace EF that can resolve the shortcoming, and that can distinguish HFpEF from healthy hearts. Conflict of Interest None to declare