Assessment of the predictive value of right ventricular longitudinal strain and functional trends in right ventricular function of patients after LVAD implantation.

Abstract

Right heart failure (RHF) remains a major determinant of morbidity and mortality following left ventricular assist device (LVAD) implantation. Right ventricular free wall strain (RVFWS) and global longitudinal strain (RVGLS) have emerged as promising echocardiographic markers, yet their long-term prognostic utility and suitability to assess changes in RV function remain poorly defined.In this single-center retrospective study, 57 of 95 consecutive HeartMate 3 LVAD recipients (mean age 57 ± 11 years; 79% male) with adequate echocardiographic quality were analyzed. Patients were stratified by RHF phenotype as no RHF (n = 28) or RHF (n = 29; comprising early acute and early postoperative RHF). RVFWS and RVGLS were measured preoperatively and at 2 weeks, 6 months, 12 months, and 36 months post-implantation using speckle-tracking echocardiography. Predictive value was assessed using logistic regression and ROC analysis; longitudinal change using linear mixed-effects models (LMM); and outcomes using Kaplan-Meier, Cox regression, restricted mean survival time (RMST), and competing-risks analyses.Preoperative RVFWS was lower in patients who developed RHF (-6.0% (IQR-9.5, -4.1) vs. -10.5% (IQR-12.6, -6.8); p = 0.037). Preoperative RVFWS predicted RHF with an AUC of 0.66 (95% CI 0.52-0.81), an optimal cut-off of -6.3%, 55% sensitivity and 79% specificity; no other echocardiographic parameter (RVGLS, TAPSE, RVEDD/LVEDD, FAC) discriminated RHF, and no parameter was an independent predictor on multivariable analysis. On LMM, RVFWS declined early in the no-RHF group (Δ + 3.3 at 2 weeks, p = 0.005; Δ + 3.0 at 6 months, p = 0.022 vs. baseline) and then stabilized; this pattern was robust in complete-case and random-slope sensitivity analyses. RVGLS improved in the RHF group by 12 months (Δ-3.3, p = 0.010). No between-group difference was detected at any single time point. One- and two-year survival was 85% and 80% in the no-RHF group versus 61% and 57% in the RHF group; restricted mean survival time to 2 years was 169 days shorter in the RHF group (p = 0.018), reflecting significantly higher first-year mortality (HR 3.2; p = 0.044) without a later difference. Transplant incidence was similar between groups and competing-risks analysis was concordant.Preoperative RVFWS provides modest discrimination between RHF and non-RHF after LVAD implantation. Serial strain captures dynamic RV changes - an early postoperative decline in RVFWS in patients without RHF and a later recovery of RVGLS in those with RHF - and RHF is associated with worse early survival. Serial RV strain may therefore complement, rather than replace, established risk assessment.