Renal denervation proceeds as U.S. trial’s flaws emerge

Even though it remains unclear why blood pressure reduction was so pronounced among the African Americans in the sham-control group, the impact of this unexpected effect substantially upended the trial’s endpoints. Among the 49 African Americans randomized to sham treatment, office-measured systolic pressure dropped by an average of 17.8 mm Hg, far exceeding the 8.6–mm Hg decline seen among the non–African Americans in the control arm and even exceeding the average 15.5–mm Hg drop in office systolic BP among African Americans treated with renal denervation.

"The absolute reduction in blood pressure by renal denervation in African Americans was identical to non–African Americans." The problem that arose "related more to what happened in the sham-control group of African Americans, who had a nearly 18–mm Hg reduction in blood pressure," said Dr. Kandzari, chief scientific officer and director of interventional cardiology at Piedmont Heart Institute in Atlanta.

The low rate at which patients assigned to receive renal denervation actually received the type of treatment spelled out in the study’s protocol may have been the biggest problem of all, although Dr. Kandzari stressed that, in his opinion "no single factor led to the neutral efficacy seen in the study."

The supplementary methods section of the SYMPLICITY HTN-3 report published in April explicitly called for patients to receive "4-6 ablations" per side, delivering them in a spiral, circumferential pattern starting distally in each renal artery. That meant each patient was to receive a minimum of eight total ablations.

But analysis of data recorded independently by the research nurse and by the proctor during each procedure, as well as cineangiography films made and submitted by the operator for each ablation, clearly showed that many patients did not receive the treatment that the protocol spelled out. Synthesis of the data collected by the three methods showed that about half of the 364 patients randomized to renal denervation received at least eight ablations, while the other half did not receive this minimum number.

The three separate sets of ablation records also contained information on whether ablations occurred in the anterior, posterior, superior, or inferior quadrants of each renal artery. Full circumferential ablation, what the protocol prescribed, required an ablation in at least one of each of these quadrants per side. What actually happened was that 253 patients (70%) received no circumferential ablations, 68 patients (19%) received circumferential ablation on just one side, and 19 patients (5%) received the bilateral circumferential ablations that the protocol called for. Data for the remaining 24 patients treated with renal denervation were not amenable to analysis for this parameter.

As might be expected, greater ablation number and completeness strongly linked with a robust blood pressure effect.

Among patients who received at least eight ablations, office systolic pressure fell by an average 13.1 mm Hg. But among the nine patients who received 16 or more ablations, the average systolic BP reduction at 6 months was 30.9 mm Hg. Among the 18 patients who received at least 15 ablations, the average systolic pressure reduction was 25.4 mm Hg. A very similar relationship occurred for BPs measured by ambulatory monitoring (see graphic), and the data also suggested a positive link between an increasing number of ablations and an increased effect on heart rate. The consistency of the association across all three measures lent further support to this as a real relationship, Dr. Kandzari noted.

Circumferentiality of the ablations showed a similar pattern. The average office systolic pressure fall in patients with no circumferential ablations was 14.2 mm Hg, and it was 16.1 mm Hg in patients who received just one circumferential ablation. But in the 19 patients who received circumferential ablations bilaterally, the average office systolic pressure reduction was 24.3 mm Hg, with a similar pattern seen for ambulatory measures as well as for home-based BP measurements.

"All patients randomized to renal denervation received renal denervation, but they may not have received it in a fashion that seemed to translate into a greater blood pressure reduction," Dr. Kandzari concluded.

Who to treat, where to treat, how to treat

"One result of the neutral HTN-3 result was a call to revisit the basic science behind renal denervation. The clinical enthusiasm had exceeded the science behind renal denervation," Dr. Kandzari observed.

Renal denervation’s many European advocates seem to agree, and have begun the process of determining characteristics of the best patients to receive renal denervation and where and how ablations are best delivered within the renal artery to achieve interruption of sympathetic innervation, although the targeting information they have right now is rudimentary.