CARC5: Optimizing HeartMate3 LVAD Speed Using Patient-Specific Hemodynamic Analysis

Study: This study investigates the interaction between LVAD H-Q relationships and the native cardiovascular system to optimize hemodynamic performance for a cohort of 38 heart failure patients who underwent Heartmate 3 LVAD implantation at the San Camillo Hospital in Rome. Methods: An in-house developed computational hemodynamic optimization model (HOM) of the cardiovascular system with the inclusion of LVAD pressure-flow curves was utilized to perform a multi-step optimization process for each patient. Step 1: A patient-specific model was created utilizing recorded parameters via cardiac catheterization (CO, LVAD speed, and MAP). Step 2: Blood pressure (BP) management was performed in silico to meet a MAP target (77 mmHg) while maintaining a constant LVAD speed and assessing changes in CO compared to a CO target of 5 L/min. Step 3: Both BP management and LVAD speed optimization were performed to increase the chances of meeting both hemodynamic targets (MAP & CO), i.e. hemodynamic optimization. Results: 38 Patient-specific HOM were obtained by matching their catherization laboratory measured MAP, CO, and LVAD speed. BP management without adjusting LVAD speed resulted in 11 patients (29%) undershooting the flow target and 37% (14 patients) overshooting the CO target of 5 L/min. The HOM was able to optimize all patients to meet both MAP and CO targets by combining BP management and LVAD speed adjustment. After LVAD speed optimization, the HOM predicted that speed reduction was necessary for 50% of patients (n=19) to meet both MAP and CO targets, with the maximal speed reduction being 700 RPM (Figure 1b). Interestingly, most of those patients who had pre-optimization speeds greater than 5500 RPM required a speed reduction, while a majority of patients with a pre-optimization speed lower than 5500 RPM could benefit from an increase in LVAD speed to achieve hemodynamic optimization (Figure 1c). Conclusions: Optimization of LVAD patients requires MAP management in conjunction with LVAD speed adjustments to meet Hemodynamic targets. Patients with higher MAP and CO values may also benefit from medical therapy optimization together with LVAD speed optimization to minimize risk of complications such as right heart failure. The results show that not only must the LVAD H-Q curve relationships be considered but also the individual status of each patient is crucial for optimizing each patient.