Renal Denervation 2.0: Assessing the Failure of a Pivotal Trial

In a 2-part series, TCTMD looks at the blow delivered to the promising field of renal denervation by the failure of the pivotal Symplicity HTN-3 trial to meet its primary efficacy endpoint. Experts discuss the need to reboot the field by returning to preclinical and feasibility studies and offer ideas about alternative trial designs. They also address Symplicity HTN-3’s impact on industry and device innovation.

The shock was such that an old-hand trialist remembers where he was when he heard the news: Symplicity HTN-3, the US pivotal trial of renal denervation and a much-anticipated milestone in the global march of the innovative interventional therapy for resistant hypertension, had failed to meet its primary efficacy endpoint.

“We were lulled into a false sense of security based on a great deal of [prior] data,” Martin B. Leon, MD, of Columbia University Medical Center (New York, NY), commented in a post-trial TCTMD video roundtable. “The expectations were very high.”

Complete results from the rigorous, 535-patient trial, presented at the American College of Cardiology (ACC)/i2 Scientific Session in March 2014 and simultaneously published in the New England Journal of Medicine, did not meet the prespecified margin for superiority of denervation over a sham procedure in reduction of office systolic blood pressure (BP) at 6 months (-14.13 ± 23.93 mm Hg vs -11.74 ± 25.94 mm Hg; P = 0.26 for superiority).

Reuters reported that the finding prompted some prominent cardiologists to call for a moratorium, or at least curb, on sales of the Symplicity catheter (Medtronic; Minneapolis, MN), which is widely approved outside the United States.

After the Shock, an Effort to Understand

In a statement released shortly after ACC, Medtronic said that the Symplicity technology would remain available in countries where there is regulatory approval. Meanwhile, the company noted, it continues to explore trial data in search of explanations for the trial’s results. “We will use these post-hoc analyses to help inform future clinical and technological efforts to bring [renal denervation] to its full therapeutic potential,” the statement continued, adding that “Medtronic is in discussions with the FDA to determine our path forward for the next [investigational device exemption], which will likely be a global trial.”

The disappointing Symplicity HTN-3 findings emerged against the background of a stream of positive, albeit limited, data from outside the United States. For example, the most recent results from the randomized (but not sham-controlled) SymplicityHTN-2 trial continued to show a marked reduction in baseline office BP (33 mm Hg systolic) for denervation in a 40-patient cohort at 3 years (Esler MD, et al. Eur Heart J. 2014;Epub ahead of print), while in the Global Symplicity Registry, a cohort of patients with baseline office systolic pressure of at least 160 mm Hg experienced a reduction of roughly 20 points.

Although the systolic decline in Symplicity HTN-3 was far less robust than in earlier studies—an attenuation that some say could have been anticipated due in part to the trial’s larger sample size—the biggest surprise was the large effect in the sham-control arm, David E. Kandzari, MD, of the Piedmont Heart Institute (Atlanta, GA), told TCTMD in a telephone interview.

Notwithstanding the diminished efficacy, Symplicity HTN-3 “reminded us that the procedure is very safe, with a 1.4% rate of major adverse events at 6 months, and that has important implications for geographies where renal denervation remains commercially available,” he added.

Lessons Learned

With expectations for renal denervation brought down to earth, the Symplicity investigators as well as the many supporters of the therapy that appeared to address an unmet clinical need have entered a period of sober reassessment, said Dr. Kandzari. Michael A. Weber, MD, of SUNY Downstate College of Medicine (New York, NY), suggested in the post-trial roundtable, “We have to go back to square one and do clinical trials where we just focus on renal denervation. Let’s make sure that compared with sham we have a procedure that does something.”

Part of denervation’s initial recovery entailed dissection of the Symplicity HTN-3 data to tease out possible confounders. No single factor doomed the trial, Dr. Kandzari said, but in an exploratory analysis presented at EuroPCR in May 2014, he zeroed in on 3 variables that likely contributed to the negative outcome and have become the focal point of trial postmortems: variable drug adherence, the patient population, and procedural variability.

A Drug Problem

Especially problematic, according to Dr. Weber, was the “incredibly shifting baseline” of participants’ drug regimens, despite strict entrance criteria (maximally tolerated doses of 3 or more antihypertensives of complementary classes, including a diuretic) and injunctions not to change medications.

“We had no way of knowing whether or not [patients] were taking their medicines,” he said. Only 80% were on stable regimens for at least 6 weeks before randomization, which research suggests is about the time required for the drugs to achieve their full effectiveness. Furthermore, about one-third of patients in both arms had more than 1 medication change during the course of the trial.

In addition, certain classes of drugs were found to affect the response to denervation. Counterintuitively, aldosterone antagonists increased the BP-lowering effect, Dr. Kandzari noted, while vasodilators were associated with a negative response.

Another potential confounder stems from the composition of the study population. About one-quarter were African American, an ethnicity that was all but unrepresented in previous studies. Although African Americans experienced roughly the same BP reduction from denervation as other participants—about 15 mm Hg of office systolic pressure—they had twice the response to the sham procedure: a dramatic 17.8-point reduction. In fact, among non-African American patients, denervation produced a statistically significant difference in reduction of systolic pressure compared with that of the sham control (15.2% vs 8.6%; P = .012).

Adding to the complexity, about half of African American patients were taking vasodilators.

On-the-Job Training?

Others identified problems with procedural technique that may have compromised denervation’s effectiveness. The trial’s 111 operators at 88 centers were mostly newcomers to the procedure, with almost one-third having performed only 1 denervation. Also telling was the finding from Dr. Kandzari’s analysis that the degree of BP reduction depended on the number of ablation attempts (only 4-6 were recommended) and use of circumferential ablation. While most previous studies of renal denervation demonstrated reduction in heart rate, the failure of Symplicity HTN-3 to show a reduction has been cited as a possible signal that the denervation technique was suboptimal (Pathak A, et al. EuroIntervention. 2014;10:21-23).

It remains to be seen whether the lack of difference between the treatment and sham arms at 6 months is maintained over longer follow-up, Krishna J. Rocha-Singh, MD, of Prairie Vascular Institute (Springfield, IL), told TCTMD in a telephone interview.

“One would expect that if the treatment really worked, its benefits would be sustained and the placebo effect should, if anything, decrease,” added Symplicity HTN-3 co-principal investigator Deepak L. Bhatt, MD, MPH, of Brigham and Women’s Hospital (Boston, MA), in a telephone interview with TCTMD. He noted that 1-year results from the trial will be presented at the European Society of Cardiology Congress 2014 in Barcelona, Spain, this September.

The second half of this feature, to appear on June 19, will present ideas about what might get the field back on track, including development of a real-time test of procedural efficacy and better targeting of denervation candidates. It will also offer a European perspective, assess the impact on industry, and discuss the burden of designing a future pivotal trial.

Note: Dr. Leon is a faculty member of the Cardiovascular Research Foundation, which owns and operates TCTMD.