Pharmacotherapies in Heart Failure With Preserved Ejection Fraction: A Systematic Review and Network Meta-Analysis.

Publication/Presentation Date



Various pharmacotherapies exist for heart failure with preserved ejection fraction (HFpEF), but with unclear comparative efficacy. We searched EMBASE, Medline, and Cochrane Library from inception through August 2021 for all randomized clinical trials in HFpEF (EF >40%) that evaluated beta-blockers, mineralocorticoid receptor antagonist (MRA), angiotensin-converting enzyme inhibitors (ACE), angiotensin receptor blockers (ARB), angiotensin receptor-neprilysin inhibitor (ARNI), and sodium-glucose cotransporter-2 inhibitors (SGLT2i). Outcomes assessed were cardiovascular mortality, all-cause mortality, and HF hospitalization. A frequentist network meta-analysis was performed with a random-effects model. We included 22 randomized clinical trials (30,673 participants; mean age = 71.7 ± 4.2 years; females = 49.3 ± 7.7%; median follow-up = 24.4 ± 11.1 months). Compared with placebo, there was no statistically significant difference in cardiovascular mortality [beta-blockers; odds ratio (OR) 0.79 (0.46-1.34), MRA; OR 0.90 (0.70-1.14), ACE OR 0.95 (0.59-1.53), ARB; OR 1.02 (0.87-1.19), ARNI; OR 0.97 (0.74-1.26) and SGLT2i; OR 1.00 (0.84-1.18)] or all-cause mortality [beta blockers; OR 0.75 (0.54-1.04), MRA; OR 0.90 (0.75-1.08) ACE; OR 1.05 (0.71-1.54), ARB; OR 1.03 (0.91-1.15), ARNI; OR 0.99 (0.82-1.20) and SGLT2i; OR 1.00 (0.89-1.13)]. The certainty in these estimates was low or very low. There was a significantly reduction in HF hospitalization with the use of SGLT2i [OR 0.71 (0.62-0.82), moderate certainty], ARNI [OR 0.77 (0.63-0.94), low certainty], and MRA [OR 0.81 (0.66-0.98), moderate certainty]; with corresponding P scores of 0.84, 0.68, and 0.58, respectively. In HFpEF, the use of beta-blockers, MRA, ACE/ARB/ARNI, or SGLT2i was not associated with improved cardiovascular or all-cause mortality. SGLT2i, ARNI, and MRA reduced the risk of HF hospitalizations.




Medicine and Health Sciences




Department of Medicine, Cardiology Division

Document Type