Blood Pressure (Jul 2017)

Sham or no sham control: that is the question in trials of renal denervation for resistant hypertension. A systematic meta-analysis

  • Fadl Elmula M. Fadl Elmula,
  • Ying-Mei Feng,
  • Lotte Jacobs,
  • Anne C. Larstorp,
  • Sverre E. Kjeldsen,
  • Alexandre Persu,
  • Jan A. Staessen,
  • for the European Network COordinating research on Renal Denervation (ENCOReD)

DOI
https://doi.org/10.1080/08037051.2017.1311769
Journal volume & issue
Vol. 26, no. 4
pp. 195 – 203

Abstract

Read online

Background: Studies of renal denervation (RDN) in patients with apparent treatment resistant hypertension have been hampered by a number of patient and physician related confounders on blood pressure (BP) including poor drug adherence. It remains uncertain whether RDN lowers BP. We aimed to investigate whether the use of sham control is essential in RDN studies or whether systematic use of 24-hour ambulatory BP provides enough information thereby making an invasive sham control redundant. Methods: We meta-analyzed randomized controlled trials of the BP response to RDN on top of continued or optimized antihypertensive drugs in patients with resistant hypertension. On top of the randomized trials reviewed earlier, we additionally included three studies, one conducted in Spain (24 patients, RDN vs. spironolactone), one conducted in Denmark (69 patients, sham controlled) and one conducted in Netherlands (139 patients, RDN vs. continued treatment). We analyzed 24-hour ambulatory BP in 3 sham controlled studies vs. 7 no sham controlled studies. Results: The updated meta-analysis of 10 studies showed 3.6 mmHg (p = .45) and 1.0 mmHg (p = .54) reductions in office and in 24-hour systolic BP, respectively. Meta-analysis of 24-hour systolic BP in the 3 sham-controlled studies showed a reduction of 2.18 mmHg (95% confidence intervals (CIs) −4.70 to 0.33 mmHg, n = 396 vs. 230, p = .07). For the 7 no sham controlled studies there was no difference in 24-hour systolic BP (+0.38 mmHg; 95% CIs −5.29 to 6.04 mmHg, n = 215 vs. 245, p = .90). The test for sub-group heterogeneity showed no significant interaction (p = .69). Removing one trial at a time produced confirmatory results. Conclusion: The overall meta-analysis of 10 randomized and controlled studies showed no significant effect on BP of RDN in resistant hypertension. Moreover, our analysis does not support the use of sham control but rather suggests extensive use of 24-hour ambulatory BP in studies of RDN in resistant hypertension

Keywords