Efficacy and safety of sacubitril/valsartan according to frailty in heart failure with preserved ejection fraction: a post hoc analysis of the PARAGON-HF trial

Abstract Introduction Frailty is an increasingly common problem and frail patients are less likely to receive pharmacological therapy because the benefit/risk profile is perceived to be less favorable than in non-frail patients. Purpose We investigated the prevalence of frailty, the relationship between frailty status and outcomes, and the efficacy of sacubitril/valsartan, compared with valsartan, according to frailty status in patients with heart failure (HF) with preserved fraction (HFpEF) randomized in PARAGON-HF. Methods Patients aged ≥50 years with a left ventricular ejection fraction ≥45%, structural heart disease, and elevated natriuretic peptide were enrolled in PARAGON-HF. Using the Rockwood cumulative deficit approach, a 41-item Frailty Index (FI) was constructed, and a FI score was calculated, with higher scores indicating greater frailty. The primary endpoint was a composite of total HF hospitalizations and cardiovascular death. Results Of the 4,796 patients randomized in PARAGON-HF, a FI was calculable in 4,795. Mean FI was 0.227 (standard deviation, 0.091; range, 0.061–0.537). In total, 2,165 (45.2%) patients had class 1 frailty (FI <0.210, i.e., not frail), 2.084 (43.5%) had class 2 (FI 0.211–0.310, i.e. more frail), and 546 (11.4%) were in class 3 frailty (FI >0.311, i.e. most frail). There was a graded relationship between FI class and the primary endpoint, with a significantly higher risk associated with greater frailty (class 1, reference; class 2, rate ratio 2.19 [95% CI, 1.85–2.60]; class 3, 3.29 [95% CI, 2.65–4.09]). The effect of sacubitril/valsartan versus valsartan on the primary endpoint from lowest to highest FI class (as a rate ratio) was: 0.98 [95% CI, 0.76–1.27], 0.92 [95% CI, 0.76–1.12], and 0.69 [95% CI, 0.51–0.95]), respectively (P for interaction=0.23) (Table). When FI was examined as a continuous variable, the interaction with treatment was significant for the primary outcome (P for interaction 0.002) and total HF hospitalizations (P for interaction <0.001) with those most frail deriving greater benefit (Figure). Adverse reactions and discontinuation of trial treatment were not more frequent with sacubitril/valsartan than valsartan, in frailer patients. Conclusions Frailty was common in patients with HFpEF in PARAGON-HF and associated with worse outcomes. There was a greater reduction in total HF hospitalizations with sacubitril/valsartan, compared with valsartan, in the frailest patients. Funding Acknowledgement Type of funding sources: Other. Main funding source(s): The PARAGON trial was sponsored by Novartis