Effect of mineralocorticoid antagonists on blood pressure lowering: overview and meta-analysis of randomized controlled trials in hypertension

Abstract

Objectives: Although mineralocorticoid antagonists (MRAs) have been proposed as effective fourth-line blood pressure (BP)-lowering agents in resistant hypertension, this effect in hypertension at-large is unclear. We evaluated whether MRAs-mediated BP lowering is both effective and safe against controls and whether the extent of BP lowering differs between resistant hypertension and nonresistant hypertension .

Methods: We searched Medline and the Cochrane Collaboration Library databases from 1991 to mid-September 2017 for randomized controlled trials (RCTs), in which MRAs were compared with placebo or other active drugs. Main outcomes were SBP and DBP lowering and treatment-related discontinuations. Continuous outcome variables were pooled as mean difference and the categorical variables as risk ratios, both with 95% confidence interval (CI). The risk of bias was assessed by using the Cochrane collaboration tool.

Results: We included 21 RCTs (2736 patients) of MRAs compared either with placebo or with active agents. Whenever all selected trials were analyzed together (MRAs versus controls), the resulting BP reduction was -7.6 (95% CI -10 to -5.3) mmHg for SBP and -2.5 (95% CI -4.2 to -0.8) mmHg for DBP, while limiting our analysis to MRAs versus placebo mean difference was increased by -2.1/-1.3 mmHg. We find no differential BP lowering between resistant hypertension and non-resistant hypertension. Treatment-related discontinuations were not different compared with either placebo or active comparators.

Conclusion: As the extent of BP lowering following treatment with MRAs in patients with and without resistant hypertension was similar and not accompanied by increased rate of treatment-related discontinuations compared with other active comparators, these agents may have a role in BP lowering in already treated hypertensive patients who did not reach target BP values.