Pre-operative proteinuria in left ventricular assist devices and clinical outcome

Abstract

This study evaluated the association of pre-operative proteinuria before continuous flow left ventricular assist device (CF-LVAD) implantation in relation to mortality and the need for renal replacement therapy (RRT) during the first year of follow-up. This retrospective, multicenter cohort study evaluated all patients (n = 241) who underwent CF-LVAD implantation in the 2 participating tertiary referral centers. Patients were included if a urine dipstick was performed within 7 days before CF-LVAD implantation. Proteinuria was defined as trace or higher. In total, 173 patients (72%) were included (78% men; mean age, 52.3 ± 13.3; mean estimated glomerular filtration rate, 60.1 ± 25.9 mL/min/1.73 m2), and 42 patients (24%) had pre-operative proteinuria. The observed survival in patients with proteinuria vs without proteinuria was 57% vs 86% at 3 months and 52% vs 78% at 1 year (log-rank p < 0.001), respectively. In addition, during the first year after implantation, 32% of the patients with proteinuria and 15% of the patients without proteinuria required RRT (log-rank p = 0.02). Multivariate Cox regression analysis confirmed that pre-operative proteinuria was an independent predictor for mortality (adjusted hazard ratio, 2.09; 95% confidence interval, 1.10-3.80, p = 0.02) and for the need of RRT during the first year (adjusted hazard ratio, 2.23; 95% confidence interval, 1.13-4.84; p = 0.02). Proteinuria, which was present in 25% of all tested LVAD patients, predicts worse outcome in all-cause mortality and need of RRT in patients with a CF-LVAD. This warrants the use of proteinuria in risk stratification when selecting patients for CF-LVAD therapy.