Inability to Achieve Adequate Left Ventricular Unloading Predicts Hospitalization or Death after Invasive Optimization in LVAD Patients

Abstract

BackgroundA left heart catheterization optimization protocol was recently developed at our institution and several physiologic indices, including LV end diastolic pressure (LVEDP) and trans-aortic gradient (TAG), previously only available in computational or animal models, were evaluated in patients supported on LVAD therapy. We sought to evaluate the prognostic significance of these markers.HypothesisInability to achieve adequate ventricular unloading will result in worse heart failure hospitalization-free survival.MethodsAll patients undergoing combined left and right heart catheterization protocol at our institution between 2015 and 2018 were evaluated. Patients whose left ventricles were inaccessible or who underwent non-standard protocols were excluded. Comprehensive clinical data were obtained. Primary endpoints included a composite outcome of hospitalization or death and overall survival.ResultsThirty one patients were analyzed from the cohort. Optimization resulted in normalization of hemodynamic parameters on average (RA 11 mmHg, PA 25 mmHg, PCWP 13 mmHg, LVEDP 12 mmHg, cardiac index 2.53 L/min/m2), which was significantly improved from baseline for all variables [p ≤ 0.05]). On univariate modeling, only LVEDP was associated with the primary endpoint (HR 1.2 per 1 mmHg increase, CI[1.1-1.3], p<0.01). After inclusion of LVAD speed, transaortic gradient,and cardiac index in a multivariate model, the association between LVEDP and the composite endpoint remained significant (HR 1.4 per 1 mmHg, CI [1.2-1.4], p<0.001).ConclusionsLeft ventricular offloading as measured by left ventricular end-diastolic pressure was a significant marker of poor outcomes over time. Further research should focus on pathophysiology corresponding to poor unloading. A left heart catheterization optimization protocol was recently developed at our institution and several physiologic indices, including LV end diastolic pressure (LVEDP) and trans-aortic gradient (TAG), previously only available in computational or animal models, were evaluated in patients supported on LVAD therapy. We sought to evaluate the prognostic significance of these markers. Inability to achieve adequate ventricular unloading will result in worse heart failure hospitalization-free survival. All patients undergoing combined left and right heart catheterization protocol at our institution between 2015 and 2018 were evaluated. Patients whose left ventricles were inaccessible or who underwent non-standard protocols were excluded. Comprehensive clinical data were obtained. Primary endpoints included a composite outcome of hospitalization or death and overall survival. Thirty one patients were analyzed from the cohort. Optimization resulted in normalization of hemodynamic parameters on average (RA 11 mmHg, PA 25 mmHg, PCWP 13 mmHg, LVEDP 12 mmHg, cardiac index 2.53 L/min/m2), which was significantly improved from baseline for all variables [p ≤ 0.05]). On univariate modeling, only LVEDP was associated with the primary endpoint (HR 1.2 per 1 mmHg increase, CI[1.1-1.3], p<0.01). After inclusion of LVAD speed, transaortic gradient,and cardiac index in a multivariate model, the association between LVEDP and the composite endpoint remained significant (HR 1.4 per 1 mmHg, CI [1.2-1.4], p<0.001). Left ventricular offloading as measured by left ventricular end-diastolic pressure was a significant marker of poor outcomes over time. Further research should focus on pathophysiology corresponding to poor unloading.