Right Ventricular Longitudinal Strain in Patients Having Heart Failure With Preserved Ejection Fraction

Abstract

Heart failure with preserved ejection fraction (HFpEF) is quite common and associated with excess hospitalizations and mortality. Currently, there is no clinically effective therapy to reduce the excess mortality associated with HFpEF. A hallmark of HFpEF is heterogenous profiles and phenotypes in HFpEF patients, which may somewhat account for challenges in finding optimal treatment strategies. Regardless of phenotype, right ventricular dysfunction is common in HFpEF. Experts dispute the consistency of these findings because; (1) load dependent conditions such as pulmonary hypertension are common in HFpEF and conventional methods may inaccurately estimate RV dysfunction due to known load and angle-limitations; (2) there is limited evidence from more robust RV measurements such as RV myocardial strain or 3D echocardiography. I aimed to investigate RV myocardial strain in patients with HFpEF enrolled in the recently terminated PARAGON-HF trial. I screened a total of 1087 echocardiograms. 834 had at least one conventional RV systolic function measurement; tricuspid annular plane systolic excursion (TAPSE) or RV fractional area change (RVFAC). Of these, I conducted RV myocardial deformation analysis in 527 patients with HFpEF enrolled in PARAGON-HF and the rest excluded due to poor quality images, or no RV focused views. Myocardial strain parameters measured in this study were right ventricular global longitudinal strain (RVGLS) and right ventricular free wall longitudinal strain (RVFWLS). In chapter 1, we aimed to (1) assess abnormalities in RV function estimated by both conventional echocardiographic and myocardial strain measurements, and the extent to which RV dysfunction was underestimated by conventional methods (2) Between measurements of RV myocardial strain, how accurate could they predict abnormal conventional measurements and (3) association between RV myocardial strain and cardiovascular hospitalization and all-cause mortality. We found that RV abnormalities were frequent when estimated by RV myocardial strain compared to conventional echocardiographic methods. We also observed that even in the presence of normal conventional echocardiographic measurements, abnormal RV longitudinal strain was present in at least one half of corresponding participants. RVGLS was more sensitive but less specific to detect abnormal RVFAC and TAPSE. RVFLWS was less sensitive, more specific and was better calibrated to predict abnormal TAPSE and RVFAC. RVFWLS had better correlation to RV function estimates while RVFWLS was better correlated to LV function. We observed LV systolic function correlation to longitudinal strain peaked in septal wall segment. Comparing group with normal to abnormal strain, groups defined by abnormal RVFWLS (>-19) showed significantly higher mortality, while this was less apparent in the groups defined by abnormal RVGLS (>-17%). As a continuous association, RVFWLS was associated with increased incidence and risk (hazard) of all-cause death. This trend was also less apparent for every percentage increase in RVGLS. Chapter 1 concluded with the findings that (1) conventional measurements underestimate RV dysfunction, by at least 50% (2) RVFWLS is a much more accurate, specific for RV function and less influenced by LV function (3) Incidence of outcomes were more apparent with RVFWLS. Overall, RVFWLS is a better surrogate of RV dysfunction and may have prognostic value. More studies are needed to validate and confirm these observations. In Chapter 2, we aimed at investigating the relationship between RVFWLS and primary outcomes in PARAGON-HF trial. Specifically, we aimed at (1) assessing association of RVFWLS as a continuous function (2) Effect modification by stratifying the RVFWLS into tertiles (<-21.9%, -21.4 to 15.9%, >-15.9%). We found some differences in baseline characteristics and echocardiographic features between subgroups. Patients in group with RVFWLS >-15.9% were more likely to be male, have diabetes and atrial fibrillation. Patients with RVFWLS > 15.9% also had higher body mass index, heart rate, NTproBNP and creatinine but similar estimated mean glomerular filtration rate. There was a linear increase in incidence of total (first or recurrent) heart failure hospitalizations and cardiovascular death with every 1% increase in RVFWLS even after accounting for pulmonary pressures and baseline differences. In segmental and regional strain analysis, there was a similarly increasing trend for adverse events with every increase in apical and basal segment longitudinal strain. Patients with RVFLWS >-15.9% had at least two-times the risk of primary endpoint compared to patients with normal RVFWLS (<-21.4). Abnormalities in displacement of basal segment (TAPSE) have been reported to have prognostic value. This suggests that beyond basal free wall abnormalities, anomalies in the apical regional abnormalities may also be frequent in HEFpEF and contributive to poor outcomes. Overall, RVFWLS is independently associated with adverse outcomes in HFpEF, risk is about two-times higher if your RVFWLS falls in range of >-15.9%. This association may be driven by anomalies in basal and apical segments of the RV. In conclusion, RVFWLS may be clinically useful to (1) identify early or subclinical RV dysfunction (2) identify and stratify patients at risk of adverse cardiovascular events (heart failure hospitalizations and cardiovascular mortality (3) RV segmental or regional wall longitudinal strain is useful in increasing our understanding of RV dysfunction in HFpEF.