Long-Term Safety and Efficacy of Renal Denervation: 24-Month Results From the SPYRAL HTN-ON MED Trial

Abstract

Background: Six-month results from the SPYRAL HTN-ON MED trial (SPYRAL HTN-ON MED Study of Renal Denervation With the Symplicity Spyral Multi-Electrode Renal Denervation System) demonstrated that renal denervation (RDN) reduced office blood pressure (BP), and not 24-hour ambulatory systolic BP, compared with sham control in hypertensive patients. In this prespecified analysis of the ON MED trial, long-term changes in BP, antihypertensive drug use, and safety outcomes through 24 months are compared between RDN and sham control groups.

Methods: SPYRAL HTN-ON MED is a prospective, randomized, sham-controlled, blinded trial enrolling 337 patients globally from 56 clinical centers. Eligible patients had an office systolic BP of 150 to 180 mm Hg, a diastolic BP ≥90 mm Hg, and a 24-hour ambulatory systolic BP of 140 to 170 mm Hg. Patients were randomized to RDN or a sham control procedure and were prescribed a stable regimen of 1 to 3 antihypertensive medications through 6 months. After 6 months, patients and physicians were unblinded with permitted changes to antihypertensive therapy, and control patients were permitted to cross over. Crossover patients had their last observations carried forward as part of the control group. Statistical analyses were conducted on the population as randomized.

Results: At 24 months, the RDN group experienced significantly greater mean reductions in ambulatory systolic BP (-12.1±15.3 mm Hg [n=176] versus -7.0±13.1 mm Hg [n=33]; difference: -5.7 mm Hg; P=0.039) and office systolic BP (-17.4±16.1 mm Hg [n=187] versus -9.0±19.4 mm Hg [n=35]; difference: -8.7 mm Hg; P=0.0034) compared with sham controls. At 24 months, antihypertensive medications increased significantly more in the sham group (1.7 versus 2.7) compared with the RDN group (1.8 versus 2.4; P=0.046). Sensitivity analyses accounting for missing sham patient BP values due to crossover yielded consistent results in favor of RDN for 24-hour ambulatory (P=0.023) and office systolic BP (P<0.0001). Clinically adverse events were rare, with no instances of renal artery stenosis through 24 months.

Conclusions: RDN produced significantly greater ambulatory and office systolic BP reductions at 24 months compared with sham control, despite higher antihypertensive medication use in the control group.

Registration: URL: https://clinicaltrials.gov; Unique identifier: NCT02439775.

Keywords: antihypertensive agents; blood pressure; denervation; hypertension.

Figure 1.

ON MED patient flowchart through 24 months. After a 6-month primary end point ascertainment in the SPYRAL HTN-ON MED trial (SPYRAL HTN-ON MED Study of Renal Denervation With the Symplicity Spyral Multi-Electrode Renal Denervation System), sham control patients were permitted to cross over to undergo renal denervation (RDN) irrespective of blood pressure (BP) control. *At 24 months, 23 sham patients were followed via televisits permitted per protocol. Overall follow-up compliance was 97% at 24 months, including televisits. Percentages represent patients with observed outcomes. ABPM indicates 24-hour ambulatory blood pressure measure; and ITT, intention-to-treat.

Figure 2.

Antihypertensive (AH) medication and systolic blood pressure (BP) changes through 24 months. The number of AH medications and medication burden (based on the number, class, and dose) are plotted (A) at 12 and 24 months for renal denervation (RDN; blue) and sham control patients (gray) based on drug testing if available, otherwise prescribed information is used. In (B), the changes in the 24-hour ambulatory (right) and office systolic BP (left) from baseline through 24 months are plotted with the number of patients with available measures indicated below from 6 to 24 months. Follow-up and treatment difference P values are analysis of covariance-adjusted for baseline values.

Figure 3.

The change in systolic blood pressure (SBP) measured at 24 months from baseline between renal denervation (RDN) and sham control groups. The 24-hour ambulatory systolic, morning (7–9 am), daytime (9 am–9 pm), nighttime (1–6 am), and office SBP changes are plotted for RDN (blue) and sham control groups (gray) in patients with available follow-up at 24 months. Comparisons of blood pressure (BP) measures are analysis of covariance-adjusted for baseline BP.

Figure 4.

Hourly ambulatory systolic blood pressure (BP) at baseline and 24 months. Hourly ambulatory systolic BPs at baseline and 24 months in the renal denervation (blue shades) and sham control groups (gray shades).

Figure 5.

Patient-level blood pressure (BP) changes from baseline to 24 months. Individual changes from baseline to 24 months in (A) 24-hour ambulatory and (B) office systolic BP (SBP) for renal denervation (RDN) patients (blue) and sham control patients (gray) in descending order of baseline SBP. Horizontal line depicts reference BP values.

Figure 6.

Summary of sensitivity analyses comparing renal denervation (RDN) and sham control group treatment differences at 24 months. Forest plot of sensitivity analyses comparing systolic blood pressure (SBP) reductions between RDN and sham control groups at 24 months. *Treatment difference comparisons are analysis of covariance (ANCOVA)-adjusted for baseline BP. †Treatment difference comparisons are ANCOVA-adjusted for baseline BP and the change in medication burden. ‡Hypertensive patients are categorized based on original trial eligibility criteria for SBP (office SBP≥150 mm Hg and 24-hour ambulatory SBP≥140 mm Hg). §Stabilized inverse probability censoring weights were calculated using age, body mass index, sex, baseline BP, medication number, and medication burden.