George L. Bakris, MD


April 18, 2014

This feature requires the newest version of Flash. You can download it here.

I'm Dr. George Bakris, Professor of Medicine and Director of the American Society of Hypertension (ASH) Comprehensive Hypertension Center at the University of Chicago Department of Medicine, and the senior author and co-principal investigator on the SYMPLICITY HTN-3 trial.[1] This trial was meant to delineate whether, in fact, renal denervation lowered blood pressure in people with resistant hypertension. The primary endpoint was similar to the primary endpoint of the SYMPLICITY HTN-1 and HTN-2 trials -- that is, change in systolic blood pressure at 6 months.

The difference in this trial is that there was a sham control. The study was powered to look at a 5 mm Hg difference (advantage, if you will) for renal denervation and a 2 mm Hg advantage in the ambulatory blood pressure monitoring (ABPM) level, a prespecified secondary endpoint.

Another difference between this trial and the previous trials is that we mandated that everyone have an ABPM reading to qualify for the study. We also did a 6-month 24-hour ABPM reading to see how that correlates with office blood pressure. We also measured home blood pressures, in part to assess adherence and also for corroboration with ABPM, because there is a good correlation already established in the literature between ABPM and home BP.

A third major difference is that we mandated that everyone be on maximally tolerated doses of antihypertensive medications, fundamentally the triad that is recommended for such patients -- meaning a thiazide or thiazide-like diuretic, a renin-angiotensin system (RAS) blocker (either an angiotensin-converting enzyme [ACE] inhibitor or an angiotensin receptor blocker [ARB]), and a calcium channel blocker. After that, you could add spironolactone, you could add a beta blocker, you could add whatever you wanted to add. That did not have to be at maximally tolerated doses, and certainly didn't have to be if you got the blood pressure controlled.

The people who came into the study were for the most part on those medications at maximally tolerated doses. About one fourth of the patients were on spironolactone and still failed; these are the "spironolactone failures," if you will.

There was a clear check of adherence with medication diaries and pill counts and those kinds of things. Beyond that, there really were no other major differences with the prior trials -- a few small differences, but nothing major.

What Went Wrong?

The bottom line is, why did we get the result we got? Well, there are at least 3 possibilities.

The first possibility is that everybody in the placebo or sham group decided to take all their medicines when they weren't taking them before. There is actually good precedent for that. It's been described in other studies.

In fact, there's a recent study that was just published in the Journal of Hypertension from an outpatient clinic in Germany, where they gave questionnaires to the patients and measured urine metabolites of drugs, and found that less than 50% were actually taking all the drugs they were supposed to be taking.[2] Interestingly, when the patients were confronted with the fact that they said they were taking them on the questionnaire and there was no evidence of it in the urine, they insisted they were taking the drugs. So anyway, there's definitely that in the equation.

The other issue is, did everyone do proper denervation? In other words, was the methodology used in the original studies used in this study? Everybody was proctored. Everybody was shown how to do it. These are experienced interventionalists. These are not people who are novices.

There are others that are definitely looking into whether, in fact, there could have been an issue with not everybody getting proper denervation in terms of the method that was applied. Did they get denervation? Yes, they did, but did they get it using the proper method? That's another question. So there are a lot of unanswered questions in that regard.

Then, fundamentally it could be that everything was good. Everybody got the drugs they were supposed to get. Then when you add this on, maybe you don't get that much more. That's obviously possible. If you look at antihypertensive drugs, the greatest reduction in blood pressure is with the first drug. You add the second drug, you get less. You add the third, you get less. It's not additive or synergistic; it's progressively less.

Same with statins: The greatest benefit is from the first dose. Each increment in dosing increase after that, you only get 6% more reduction.

Is Renal Denervation Dead?

I think people were enamored with the 30 mm Hg blood pressure difference seen in the early trials, but they weren't sham controlled. Who knows what medications the patients were taking.

Does renal denervation work? No question about it. It wouldn't be around today if it didn't. It's been studied for over 40 years. It definitely works, but in the absence of anything else.

Then the question is what about in the real world -- and today at the American College of Cardiology (ACC) session that I'm a panelist on, we're going to hear the Global SYMPLICITY registry data from Michael Boehm.[3] Now that's more real-world, and you're going to see that, in fact, there is a nice effect.

So wait a minute: How can you have a nice effect in a registry and not have such a nice effect here? Well, in the registry, you don't know whether people are taking the drugs or not. They certainly may not be taking them at maximal doses, and in fact, the denervation may have been done in a different way. So there are all these variables.

The question that a lot of people are asking is, is renal denervation dead? As they said at ACC opening ceremonies, what we have here is the prologue. We need to look at this and figure out what went wrong. Is it the case that everybody got denervated properly and people in the trial were taking their drugs, but in the real world they're not? Could that be the reason? Or was there a problem with the procedure?

Now the company [Medtronic] to their credit -- and they have been phenomenal about this -- is forging ahead and putting together a plan at a basic science level to get at the nerve distribution. It's been assumed to be kind of equal throughout the artery, but that's not true. It's probably highly concentrated at the hilum, and less so as you come away towards the aorta. Where did you denervate? Did you denervate deep there? Did you come back? These are all unanswered questions.

There are other questions that they're going to look at. At this time, I don't think another clinical trial should be done by anybody, because nobody knows this information. It's not like somebody knows and it's a secret. Nobody knows this, and if you go forward without knowing this information and get another negative trial, then the assumption is that it really must not work, so forget it. Then it will die.

I think a little patience and a little more knowledge would be wise at this point. That's my 2 cents.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.