A network meta-analysis of randomized trials and observational studies on left ventricular assist devices in adult patients with end-stage heart failure.

Objectives: The use of left ventricular assist devices (LVADs) is an approved treatment option for end-stage heart failure. Several devices have been developed over the years, including 2 newer ones (HeartMate 3 and HeartWare), but an overall comparative analysis has never been performed. We conducted a network meta-analysis of randomized trials on LVAD for adults with end-stage heart failure.

Methods: Pertinent studies were searched in several databases. Selected outcomes were extracted, including death, stroke and bleeding. Incident relative risks were computed with network meta-analysis with 95% confidence intervals (CIs) and P-scores (with highest values indicating the best therapy).

Results: Four randomized clinical trials and 4 observational studies were identified, totalling 2248 patients. Using HeartMate XVE/VE as the benchmark, all LVADs provided a significant better outcome for survival rate in comparison with medical therapy, without significant differences among newer LVADs. The relative risk for death was 0.79 (95% 0.60-1.04; P-score 0.89) for HeartMate II, 0.85 (95% CI 0.62-1.17; P-score 0.64) for HeartWare, 0.88 (95% CI 0.59-1.31; P-score 0.60) for HeartMate 3 and 1.48 (95% CI 1.21-1.80; P-score 0.01) for medical management. While appraising other outcomes, new generation devices (HeartMate 3 and HeartWare) proved better than older generation devices for bleeding, device thrombosis, hepatic dysfunction, renal dysfunction, respiratory dysfunction, right ventricular failure and sepsis with significant differences among them.

Conclusions: In the management of end-stage heart failure, LVADs provided significant improvement in terms of survival rate compared to medical therapy, but no significant differences exist among LVADs. Despite the reduction of adverse events over time, further technological refinements will be crucial to improve this technology to better address decision-making and to improve clinical outcomes.