Blind Trust: Analysis Predicts Smaller BP Drops in Pivotal Renal-Denervation Trials

September 12, 2013

LONDON— Studies of renal-denervation therapy to date have fueled widespread enthusiasm for the invasive therapy in patients with drug-resistant hypertension based primarily on pronounced changes in office-based blood-pressure measurements. That's led the authors of a new meta-analysis to predict a more modest but realistic treatment effect in future trials with more refined methodology[1].

So far, the group contends, none of the studies have been both randomized and blinded, with at least the patients unaware of treatment assignment. As well, they say, the reliance of most studies on office-based rather than ambulatory blood-pressure monitoring to gauge treatment effects introduces different kinds of "measurement bias" that could have favored renal denervation.

According to the meta-analysis by Dr James P Howard (National Heart and Lung Institute, Imperial College London, UK) and associates, soon to be published in the journal Heart, declines in blood pressure have been on average 5.6 mm Hg greater (p<0.0001) with office-based vs ambulatory blood-pressure measurements in hypertension drug trials that aren't randomized or blinded.

In blinded randomized trials, however, blood-pressure measurements by the two methods were about the same (p=0.45).

There haven't been any randomized, blinded renal denervation trials; in unblinded trials in the meta-analysis, renal denervation dropped office pressures by 26.6 to 27.6 mm Hg, depending on trial design, the group notes. But if the same pattern seen in the drug-trial experience holds up for the device trials, a raft of upcoming randomized, controlled denervation trials with at least single-blinding may produce less positive results than their predecessors.

In other words, say the authors, "the office and ambulatory blood-pressure reductions may converge." That is, reductions of up to 30 mm Hg with renal denervation foretold by unblinded studies using office-based measurements, "may decrease to values seen on ambulatory blood-pressure monitoring; that is, about 10 to 15 mm Hg," according to the group.

 
The whole issue is dead in drug trials, and yet it's come back to life, like a zombie, in denervation.
 

"Most of that is bias," senior author on the new analysis, Dr Darrel P Francis (National Heart and Lung Institute) said to heartwire , about the enticing but alleged blood-pressure declines of 30 mm Hg. These results have created a misconception in some that oversells the magnitude of renal denervation's likely blood-pressure effect, he said.

"Unblinded office-based [blood-pressure measurements] I think should be struck from the record of denervation," according to Francis. "No sane person" would give much weight to a hypertension drug trial that isn't blinded, especially one based on office pressures. "The whole issue is dead in drug trials, and yet it's come back to life, like a zombie, in denervation. Every single denervation study you see out there, without exception, is unblinded."

Francis is a lead investigator with the ongoing randomized, double-blinded Renal Artery Denervation in Chronic Heart Failure (REACH) trial of renal denervation in patients with chronic systolic heart failure.

"Expectations That May Be Unrealistic"

The results of Howard et al's study provide some new fuel to the debate over whether the positive results of small, early studies will hold up in the larger trials, and the real world. But just how much the analysis actually informs the discussion is also up for debate.

"I thought the paper was rather speculative," Dr Michael Weber (State University of New York, Brooklyn) said to heartwire about the analysis. "The authors depended a great deal on a kind of meta-analysis," he observed. "The problem was that the data they needed were almost certainly not available from the published literature. If you're going to compare blood-pressure effects measured in the office with [those from] ambulatory monitoring, clearly it has to be in the same patients." And some of the included trials had one type of blood-pressure reading for their patients but not the other.

Many in the field have already noticed the "apparent discrepancy" between office-based and ambulatory blood-pressure measurements, Weber added. "Almost all of the new major clinical trials that are looking at renal denervation are using ambulatory blood-pressure monitoring in the entry criteria, so we'll be certain that white-coat hypertension is eliminated and that the office readings are in fact robust and meaningful."

One such trial is the ongoing SYMPLICITY HTN-3 , which has office blood-pressure changes as its primary outcome measure, but ambulatory blood-pressure changes as its secondary outcome measure.

In fact, concerns raised in the report by Howard et al regarding past renal denervation studies "were things that we thought of and considered when designing SYMPLICITY-3," a trial coprincipal investigator Dr Deepak Bhatt (Brigham and Women's Hospital, Boston, MA) said.

But, he continued, "I think the points the authors raise are valid points. That is, ambulatory blood pressure provides a very different perspective from office-based blood pressure" because it is less subject to measurement bias, he told heartwire . And he does worry about "the almost hyping" that's occurred in the renal denervation field. "There's a set of expectations that may be unrealistic in terms of what the procedure can perform. I think if people are expecting to see a reduction in systolic blood pressure of 30 mm Hg or more, that's a lot to ask for."

In fact, Bhatt said, "a much lesser blood-pressure reduction would still be clinically very significant, especially for people maxed out on medical therapy. I do have some concerns if that's not what we observe in SYMPLICITY-3. There's been so much in the way of expectations that have been set up that physicians, and maybe even patients, will be disappointed. But they really shouldn't be, because that initial expectation was a bit exaggerated."

Still, "I think the data to date are exciting, scientifically," he said. "I share the enthusiasm that, as far as I can tell, the vast majority of the academic community has--interventionalists in particular, but even noninvasive hypertension experts."

Whence Bias Favoring Renal Denervation?

Blinded trials wouldn't be as prone to biases that can crop up when measuring blood pressure in the office, even in the course of normal, appropriate clinical practice, Francis and his colleagues observe. For example, baseline blood pressures from a single office measurement can be overestimated in trials because "any patient has a better chance of meeting inclusion criteria on a day when their blood pressure is above their own long-term average."

Then there is underestimation of office-based pressures due to observer bias, which "might be an understandable natural extension of the common clinical practice of repeating measurements that seem to be erroneously high," notes the group.

And in unblinded trials, patients who are aware they have received denervation therapy might be less inclined to comply with drug therapy, possibly blunting the overall treatment effect.

"I suspect strongly that the major landmark [renal-denervation hypertension] trials, even though [providers] followed a very strict protocol on how to measure blood pressure, didn't quite measure it that way," Francis said. Outlier readings thought to be mistakes are often tossed in favor of measurements that seem more accurate, all in good clinical faith.

"But no one would do that for ambulatory pressure, because they'd know that's fraud. So we have a situation where what people consider to be good practice in clinical measurements actually causes office pressures to be more unrepresentative and yet causes ambulatory pressures to be more representative."

SYMPLICITY-3 is addressing such concerns. Treatment is single-blinded, and controls receive a sham renal-denervation procedure. Critically, trial eligibility is defined by ambulatory blood pressures.

"I predict that the effects on office blood pressure, if you define the effects as the active-therapy-arm change minus the placebo-arm change, will be identical to the ambulatory blood-pressure effect," said Francis, who added he had no role in the trial's design, nor has he learned anything about its results.

Bhatt agreed that when "measuring blood pressure in the office, even if it's done in a prescribed protocol-driven way, there can be unintentional measurement bias that creeps into it." And the current report's hypothesis, "that the degree of blood-pressure reduction seen in a more rigorously designed trial will be less than what's been seen before," is legitimate. "I think there's a reasonable likelihood of that being true."

Bhatt said he has not seen unblinded SYMPLICITY-3 data--enrollment has completed, follow-up continues, and results are expected next year. "I'm waiting to see what they show. Then we'll see if their points are accurate."

Weber, for his part, remains resistant to the idea that unblinded trials necessarily overestimate renal-denervation therapy benefits. "I'm not sure about that," he told heartwire .

SYMPLICITY-3, he notes, entered patients with resistant hypertension and systolic pressures of 160 mm Hg and higher. With denervation, "I'd be disappointed if they didn't see a fall of 25 mm Hg, certainly by office [measurement], and probably a corresponding [fall] of 18 or 20 mm Hg by ambulatory blood-pressure monitoring."

In future denervation trials in patients with more moderate hypertension, such as systolic blood pressure 150 to 160 mm Hg, Weber said, "Obviously, we aren't going to see falls of 30 mm Hg" and wouldn't want to. But a drop of 15 mm Hg "to me would be a terrific result. If we got someone from, say, 150 mm Hg down to 135 mm Hg, after we've done everything we could with medicines, I would be thrilled to pieces."

The analysis was funded by a British Heart Foundation fellowship. Francis discloses consulting for Medtronic, which makes devices for renal-denervation therapy.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
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.
Post as:

processing....