Right ventricular free wall longitudinal strain as a predictor of myocardial recovery in patients with advanced cardiogenic shock supported with the Impella device

Abstract

Abstract Background Right ventricular (RV) dysfunction is a known independent factor influencing the outcome of patients in cardiogenic shock (CS). The aim of this monocentric retrospective study was to investigate the impact of left ventricular (LV) unloading on the right heart function and its predictive role regarding weaning following Impella device implantation in patients with profound CS due to LV failure. Methods We performed a retrospective analysis of 41 patients admitted to intensive care unit over a period of 3 years with CS due to LV dysfunction requiring support with Impella. Clinical and echocardiographic data were collected and correlated with weaning success. Finally, a multivariate analysis to identify independent predictors of weaning success was performed. Results After Impella implantation, the following was observed: right ventricular fractional area change (RVFAC) increased from a mean value of 28% to 38% (p=0.01), the tricuspid annular systolic velocity (TASV) increased from 9 cm/s to 10 cm/s (p=0.01), the velocity time integral right ventricular outflow tract (VTI RVOT) increased from 10 to 13 cm3/stroke (p=0.01), the right ventricular free wall longitudinal strain (RVFWLS) increased from -12% to -21% (p< 0.001). However, following LV unloading, a decrease in both RVFAC from 38% to 30% (p 0.01) and in VTI RVOT from 13 to 11 cm3/stroke (p<0.01) was observed. Moreover, there was a significant decrease in RVFWLS from -21% to -14% (p < 0.001). Notably, the percentage of change in RVFWLS between maximum and minimum (Δ%RVFWLSPmax-Pmin) support with Impella during the weaning process was significantly lower in weanable patients (34.5% in weaned vs 86% in non-weaned; p=0.038). Investigating the predicting ability of these parameters, Δ%RVFWLSPmax-Pmin emerged showing an AUC of 0.826 (0.690-0.962) in identifying non-weaned patients with a cut-off value of 38%. According to the multivariate analysis, Δ%RVFWLSPmax-Pmin in non-weaned patients remained the only significant parameter even considering possible confounders. Conclusions LV unloading with Impella in advanced CS patients improved RV function echocardiographic parameters. RVFWLS increased significantly after Impella implantation and its percentage of change between maximum and minimum support was significantly lower in weanable compared to non-weanable patients. Our findings suggest that RVFWLS could be used as a predictor of myocardial recovery allowing to identify weanable patients with adequate accuracy.