The Future of Renal Denervation Post-SYMPLICITY HTN-3
Editor's Note: Commentary based on Bhatt DL, Kandzari DE, O'Neill WW, et al. D'Agostino R, Flack JM, Katzen BT, Leon MB, Liu M, Mauri L, Negoita M, Cohen SA, Oparil S, Rocha-Singh K, Townsend RR, Bakris GL, for the SYMPLICITY HTN-3 Investigators. A controlled trial of renal denervation for resistant hypertension. N Engl J Med 2014; 370:1393-1401.
Uncontrolled hypertension is a major and growing problem worldwide. A large part of this is due to under treatment, and patient non-adherence is a significant contributor. Even when patients appear to be adherent to lifestyle recommendations and to taking multiple medications, there is a proportion of patients whose blood pressure remains poorly controlled. These patients with medication-resistant hypertension are a clinically challenging group without good options. It appears that sympathetic nervous system overactivity is one of the underlying causes of resistant hypertension. In that context, catheter-based renal artery denervation had been developed as a potential treatment for resistant hypertension.
Prior small studies had demonstrated that renal artery denervation reduced office systolic blood pressure by 25-30 mm Hg. The results appeared durable out to at least three years of follow up. Other improvements were also noted, including regression of left ventricular hypertrophy and improvements in glycemic control. In addition to beneficial effects on blood pressure, renal artery denervation appeared to simultaneously treat other cardiovascular risk factors. Interest in renal artery denervation grew, and several devices were approved in over 80 countries, though the procedure remains investigational in the United States and is therefore not FDA-approved for any indication.
These prior studies were carefully done, but there were several methodological limitations. None were blinded or sham-controlled. Only one was randomized and it had a small sample size. Thus, the provocative data for renal artery denervation really needed to be reproduced in a rigorously designed randomized clinical trial. This is what led to SYMPLICTY HTN-3, which was meant to be the pivotal trial for FDA approval of this novel technology.
The trial design is depicted in Figure 1. Patients underwent two screening visits to confirm an office systolic blood pressure greater than or equal to 160 mm Hg. Ambulatory blood pressure monitoring for 24 hours was also performed and the average systolic blood pressure needed to be greater than or equal to 135 mm Hg in order to be eligible. This was done to exclude patients with white coat hypertension. Patients needed to be on at least three maximally tolerated doses of antihypertensive medications of complementary classes, one of which had to be an optimally dosed diuretic. Patients kept home diaries to document medication adherence.
There was a strict screening process. A total of 1441 patient were screened for eligibility. Of these, 535 patients were randomized. Many patients were not eligible because their resistant hypertension was not in fact resistant to expert medical therapy. Those who made it through the screening process were randomized after renal angiography as long as there were no angiographic exclusion criteria, such as renal artery stenosis greater than 50%. Patients were randomized to renal artery denervation or to sham control, which consisted of the renal angiogram. There were other specific measures taken to make sure the patients, as well as subsequent blood pressure assessors, were blinded.
The average baseline office systolic blood pressure in both arms of the trial was 180 mm Hg. Patients in both arms were on five antihypertensive medications at baseline, many at maximally tolerated doses.
The primary safety endpoint was all-cause mortality, end-stage renal disease, embolic events resulting in end-organ damage, renal artery or other vascular complications, hypertensive crises through 30 days, or new renal artery stenosis within six months. The rate was 1.4%, which was significantly less than the pre-specified historical control of 9.8% (p<0.001). Thus, the procedure was found to be very safe.
However, there was no significant difference in the primary efficacy endpoint, which was a comparison of the office systolic blood pressure change from baseline to six months in the renal artery denervation arm compared with the change from baseline to six months in the sham control arm (Figure 2). Interestingly, within both the renal denervation and sham control arms, there were significant differences from baseline to six months (14 and 12 mm Hg, respectively, p<0.001). Thus, if there had not been a control arm, it would have appeared that renal artery denervation produced a significant 14 mm Hg drop in blood pressure. Of note, the reduction in blood pressure in the renal denervation arm was smaller than what had been reported previously. Also of note, the reduction in the sham control arm was larger than what had been reported in the control arm of the one prior randomized renal denervation trial (though that trial had not been blinded).
There was also no significant difference in the powered secondary efficacy endpoint which was the same comparison as the primary efficacy endpoint except it substituted mean 24-hour ambulatory systolic blood pressure for the office systolic blood pressure (Figure 3).
A number of pre-specified subgroups were examined. None were significant if adjustment for multiple comparisons was done and none of the reductions were greater than 10 mm Hg. Even in patients in the highest tertile of baseline systolic blood pressure, there was no significant reduction for renal denervation versus control.
While SYMPLICITY HTN-3 was by far the most rigorous trial done to date in renal artery denervation, several limitations must be acknowledged. Drug adherence was not measured by blood or urine levels, so it is possible that patients' reports in their diaries were not accurate. If there were differential rates of drug adherence in the two arms that we could not capture, that could have affected the results. Medication changes did occur during the trial and this could have diluted the results of renal artery denervation, but the results were unchanged even when these patients were censored. In theory, the duration of follow up for ascertainment of the primary endpoint may have been too short, though prior studies had found large blood pressure reductions by six months. An operator learning curve is always a possibility in device trials, but we looked for and found no relationship with procedural volume within the trial. Perhaps most importantly, biological confirmation of denervation did not occur, as there is no accepted measure, so it is not possible to state that we achieved sufficient denervation.
There are three important lessons from this trial. First, rigorous trial design matters. To the extent possible in evaluations of medical devices, randomization, blinding, and sham controls should be employed. Of course, logistical and ethical reasons sometimes preclude such approaches in all device or strategy trials, but consideration must be given to using these established clinical trial approaches. Second, collaborative hypertension centers really improved the care of patients with seemingly resistant hypertension. Much like TAVR created the heart team concept, SYMPLICTY HTN-3 has created the hypertension team concept, and that has value regardless of the specific fate of renal artery denervation. Third, while this trial did not find any significant benefit of renal artery denervation, it did demonstrate safety, and further research should continue in the field – for resistant hypertension, and for other states marked by sympathetic nervous system overactivity, such as heart failure. It may be that the "dose" of renal denervation delivered was insufficient, though safe, and that testing of higher doses with this or other devices may yet prove efficacious.
Several renal artery denervation devices remain available for clinical use outside the United States. Operators in these countries will need to decide if the patients they are treating are dissimilar from the ones enrolled in SYMPLICTY HTN-3. For example, patients who are not adherent to their medications and present frequently to the emergency department with hypertensive crises would not have been enrolled into the trial, but it may not be unreasonable to consider treating such a patient, as the next presentation could be fatal. However, as a routine treatment for resistant hypertension or lesser degrees of hypertension, it would seem to be prudent to wait until further randomized clinical trials before continuing to offer the procedure. Several trials with various devices were in the planning stages and hopefully these trials, potentially with some modification of their designs, will still proceed and offer further insight into any possible role of renal artery denervation in clinical practice.
Keywords: Blood Pressure, Blood Pressure Determination, Hypertension, Renal Artery, Sympathetic Nervous System
< Back to Listings