Sildenafil and Kidney Function in Heart Failure with Preserved Ejection Fraction

KEY POINTS: Sildenafil induced an acute effect on eGFR without change in the overall eGFR slope after 24 weeks in a heart failure with preserved ejection fraction (HFpEF) cohort. N-terminal pro–brain natriuretic peptide levels and baseline diuretic use were most strongly associated with eGFR decline in this HFpEF cohort. Long-term studies are required to determine sildenafil's influence on kidney function and outcomes in HFpEF. BACKGROUND: CKD worsens the prognosis for people with heart failure with preserved ejection fraction (HFpEF). In the Phosphodiesterase-5 Inhibition to Improve Clinical Status and Exercise Capacity in Diastolic HFpEF (RELAX) trial, sildenafil decreased eGFR compared with placebo despite favorable kidney effects in preclinical models. Since acute eGFR decline precedes long-term kidney benefits for select medications, we assessed the influence of sildenafil on acute and chronic eGFR slopes. METHODS: The RELAX trial randomized 216 participants to placebo or sildenafil and assessed 24-week changes in cardiopulmonary exercise testing, cardiovascular imaging, and laboratory data. We applied linear mixed modeling to calculate the total, acute (0–12 weeks), and chronic (3–24 weeks) eGFR slopes by treatment. Using regression modeling, we assessed respective associations between eGFR slope and baseline data and clinical end points. We repeated the analyses using a binary outcome on the basis of a substantial (≥20%) decline in eGFR. RESULTS: The mean baseline eGFR was 60.8 ml/min per 1.73 m(2), and the mean eGFR slope during follow-up was −3.21 ml/min per 1.73 m(2) per year. Sildenafil did not alter total eGFR slope compared with placebo (difference +0.47 ml/min per 1.73 m(2) per year, 95% confidence interval [CI], −6.63 to 7.57 ml/min per 1.73 m(2) per year). Sildenafil users tended to experience a more negative acute eGFR slope (difference −3.15 ml/min per 1.73 m(2) per year) and more positive chronic slope (+2.06 ml/min per 1.73 m(2) per year) compared with placebo, but neither difference reached statistical significance. Baseline N-terminal pro–B-type natriuretic peptide and loop diuretic use were associated with worse eGFR trajectory regardless of treatment. Substantial eGFR decline was associated with increase in endothelin-1 and a greater risk of hospitalization or death (HR, 2.34, 95% CI, 1.21 to 4.53, P=0.01). CONCLUSIONS: Sildenafil induced an acute effect on eGFR without change in the overall eGFR slope after 24 weeks in an HFpEF cohort, suggesting lack of long-term risk related to early reduction in eGFR after initiating treatment. Long-term studies are needed to determine the effect of sildenafil on kidney function in HFpEF.