Pulmonary artery pulsatility index predicts prolonged inotrope/pulmonary vasodilator use after implantation of continuous flow left ventricular assist device

Objective: Predictors of right ventricle (RV) dysfunction after continuous‐flow left ventricular assist device (CF‐LVAD) implantation in children are not well described. We explored the association of preimplantation Pulmonary Artery Pulsatility index (PAPi) and other hemodynamic parameters as predictors of prolonged postoperative inotropes/pulmonary vasodilator use after CF‐LVAD implantation.
Design: Retrospective chart review.
Setting: Single tertiary care pediatric referral center.
Patients: Patients who underwent CF‐LVAD implantation from January 2012 to October 2017.
Interventions: Preimplantation invasive hemodynamic parameters were analyzed to evaluate the association with post‐CF‐LVAD need for prolonged (>72 hours) use of inotropes/pulmonary vasodilators.
Measurements and main results: Preimplantation cardiac catheterization data was available for 12 of 44 patients who underwent CF‐LVAD implant during the study period. Median (IQR) age and BSA of the cohort were 15.3 years (10.2, 18) and 1.74 m2 (0.98, 2.03). Group 1 (n = 6) included patients with need for prolonged inotropes/pulmonary vasodilator use after CF‐LVAD implantation and Group 2 (n = 6) included those without. Baseline demographic parameters, cardiopulmonary bypass time, and markers of RV afterload (pulmonary vascular resistance, PA compliance and elastance) were similar among the two groups. PAPi was significantly lower in group 1 compared to group 2 (0.96 vs 3.6, respectively; P = .004). Post‐LVAD stay in the intensive care unit was longer for patients in group 1 (46 vs 23 days, P = .52). Brain natriuretic peptide was significantly higher at 3 months after implantation in group 1; P = .01.
Conclusions: The need for inotropes/pulmonary vasodilators in the postoperative period can be predicted by the preimplantation intrinsic RV contractile reserve as assessed by PAPi rather than the markers of RV afterload. Further investigation and correlation with clinical outcomes is needed.