Pulmonary arterial elastance after LVAD implantation as a predictor of late right ventricular failure beyond six months of support
E Karapedi, J Guzman Bofarull, G Gallone, V A R Thirupathirajan, A Bollas Becerra, E Synowiec, G Edwards, M Monteagudo Vela, D Garcia Saez, F Fiorelli, O Dar, A Morley-Smith, F Riesgo GilAbstract
Introduction
Late right ventricular failure (LRVF)—either de novo beyond the early postoperative period or persistent from earlier phases—remains a major complication after left ventricular assist device (LVAD) implantation and is associated with increased morbidity and mortality compared with resolved early RVF. The right ventricle is highly sensitive to afterload, determined by the resistive and pulsatile components of the pulmonary circulation. Pulmonary arterial elastance (PEa) integrates these components and reflects pulmonary arterial stiffness and RV–pulmonary artery coupling. Pulmonary arterial compliance (PAC), inversely related to PEa, represents the pulsatile component of RV afterload and the pulmonary vasculature’s ability to accommodate stroke volume. However, the prognostic value of post-implant PEa and PAC for LRVF remains unclear.
Purpose
To determine whether PEa and PAC measured six months after LVAD implantation predict LRVF.
Methods
We retrospectively analysed all consecutive patients undergoing durable LVAD implantation at a single centre (2010–2025) who survived at least six months and underwent right heart catheterisation (RHC) with PEa and PAC assessment. The primary outcome was LRVF (≥6 months) defined by MCS-ARC criteria. The secondary outcome was a composite of LRVF, death, urgent transplantation, or urgent LVAD explantation. Time-to-event analyses were performed from the date of the 6-month RHC.
Results
Among 314 patients who underwent LVAD implantation, 103 with complete haemodynamic data at six months were included; 31 subsequently developed LRVF. PEa, PAC, PAPi and coupling differed significantly between patients with and without LRVF. Across resolved, de novo and persistent RVF groups, PEa and PAPi differed significantly. Among patients who did not require haemodynamic optimisation at six months, PEa remained higher in those who subsequently developed LRVF. In univariable analysis, PEa, RV coupling and PAPi were associated with late RVF. In multivariable analysis, PEa remained independently associated with LRVF. In a sensitivity analysis restricted to patients who did not meet RHC criteria triggering optimization at 6 months, PEa remained independently associated with LRVF. None of the haemodynamic parameters were associated as predictors with the composite outcome.
Conclusion
PEa at six months post-LVAD implantation emerged as the strongest haemodynamic predictor of LRVF beyond PAPi, supporting its assessment to identify patients with persistent pulmonary vascular load. LRVF after LVAD may reflect progressive pulmonary vascular disease and RV–afterload mismatch rather than merely progression of pre-existing RV dysfunction.