Renal Denervation for Hypertension: A Systematic Review and Meta-Analysis of Randomized,Blinded, Placebo-Controlled Trials.

Abstract

The authors performed an updated meta-analysis of randomized placebo-controlled trials of renal denervation and specifically compared the effect of renal denervation in patients taking medications and in those not taking medications. Renal denervation has now undergone several blinded placebo-controlled trials, covering the spectrum from patients with drug-resistant hypertension to those not yet taking antihypertensive medications. All blinded placebo-controlled randomized trials of catheter-based renal sympathetic denervation for hypertension were systematically identified, and a random-effects meta-analysis was performed. The primary efficacy outcome was the change in ambulatory systolic blood pressure beyond the effect of the placebo procedure. Analysis was stratified by whether there was background antihypertensive medication use. There were 7 eligible trials, totaling 1,368 patients. Denervation significantly reduced ambulatory systolic (mean difference-3.61mmHg; 95% CI:-4.89 to -2.33mmHg; P< 0.0001), ambulatory diastolic (-1.85mmHg; 95%CI:-2.78 to-0.92mmHg; P< 0.0001), office systolic (-5.86mmHg; 95%CI:-7.77 to-3.94mmHg; P< 0.0001), and office diastolic (-3.63mmHg; 95%CI:-4.77 to-2.50; P< 0.0001) blood pressure. There was no evidence that the use of concomitant antihypertensive medication had a significant impact on the effect of denervation on any of these endpoints (Pinteraction=NS for each comparison). The randomized placebo-controlled trials show consistently that renal denervation provides significant reduction in ambulatory and office blood pressure. Although the magnitude of benefit, about 4/2mmHg, is modest, it is similar between patients on background antihypertensive medications and those who are not. Denervation could therefore be a useful strategy at various points for patients who are not willing to add antihypertensive agents. Whether the effect changes with time is currently unknown.