A systematic review and meta-analysis of all sham and placebo controlled trials for resistant hypertension.

INTRODUCTION

There is a lack of consensus regarding the best add on therapy for treatment of resistant hypertension (RH). This is likely secondary to a paucity of data on the comparative effectiveness of proposed therapies for RH.

METHODS

Placebo-controlled and sham-controlled randomized clinical trials testing therapies for the treatment of RH were included in this meta-analysis. Therapies with two or more studies were included as subgroups in this meta-analysis. The primary outcomes being tested were 24-hr systolic blood pressure (SBP) and office SBP.

RESULTS

Eight studies were identified that tested mineralocorticoid receptor antagonist (MRA) including 1,414 participants. The raw mean difference (RMD) between MRA and placebo control was statistically significant for 24-hour SBP (-10.56 mmHg; 95% confidence interval (CI) -12.82 to -8.30), 24-hour diastolic (DBP) (-5.48 mmHg; 95% CI -8.48 to -2.58), office SBP (-11.97 mmHg; 95% CI -16.41 to -7.54), and office DBP (-4.14 mmHg; 95% CI -5.62 to -2.65). Six studies were identified that tested renal denervation (RD) including 989 participants. The RMD between RD and sham control was not statistically significant for 24-hour SBP (-1.84 mmHg; 95% CI -3.92 to 0.24), 24-hour DBP (-0.66 mmHg; 95% CI -1.85 to 0.54), office SBP (-1.57 mmHg; 95% CI -6.04 to 2.89), and office DBP (-1.49 mmHg; 95% CI -3.52 to 0.55). Four studies were identified that tested endothelin receptor antagonists (ERA) including 1,193 participants. The raw mean difference (RMD) between ERA and placebo control was statistically significant for 24-hr systolic (SBP) (-7.02 mmHg; 95% CI -9.15 to -4.90, 24-hr diastolic (DBP) (-6.22 mmHg; 95% CI -7.61 to -4.82), office SBP (-5.84 mmHg; 95% CI -10.08 to -1.60), and office DBP (-3.73 mmHg; 95% CI -5.87 to -1.59).

DISCUSSION

MRA lowers BP in patients with RH more than RD, which seems to have little to no effect in RH. ERAs lead to a statistically significant reduction in BP but the confidence in efficacy is limited due to the low number of studies and differences in trial population. Individual factors and their impact on treatment response in RH should be investigated in future research.