Continuous-flow LVAD exchange to a different pump model: Systematic review and meta-analysis of the outcomes

Despite improved outcomes of modern continuous-flow left ventricular assist devices (CF-LVADs), device exchange is still needed for various indications. While the majority of CF-LVADs are exchanged to the same model, exchange to a different pump model is occasionally warranted. In this meta-analysis, we sought to consolidate the existing evidence to better elucidate the indications and outcomes in these cases. A comprehensive systematic search of adult patient cohorts who underwent CF-LVAD exchange to a different CF-LVAD model was performed. Study-level data from 10 studies comprising 98 patients were extracted and pooled for analysis. Mean patient age was 58 (95% CI: 48-65) and 81% were male. Indication for initial CF-LVAD was ischemic cardiomyopathy in 45% (34-57). Initial device was HeartMate II LVAD (HMII) in 93 (94.9%) and HeartWare HVAD (HW) in 5 (5.1%) patients. After mean CF-LVAD support time of 18.8 (15.2-22.4) months, exchange indications included thrombosis in 71% (43-89), infection in 21% (8-47) and device malfunction in 12% (7-21). HMII to HW exchange occurred in 53 (54.1%) patients, HMII to HeartMate III (HM3) in 32 (32.7%), and HM II to either HW or HM3 in 13 (13.2%) patients. Postoperatively, right ventricular assist device was required in 16% (8-32). Overall, 20% (8-40) of patients experienced a stroke, while HW patients had a significantly higher stroke incidence than HM3 patients (HW: 21% (8-47) vs. HM3: 5% (1-24), P < .01). Overall 30-day mortality was 10% (6-17), while HW had a significantly worse 30-day mortality than HM3 (HW: 13% (7-24) vs. HM3: 5% (1-24), P = .03). Following device exchange from a different CF-LVAD model, HM3 is associated with lower stroke and higher survival when compared to HW.