September 16th, 2013
Realistic Expectations for Renal Denervation
Early trials of renal denervation have resulted in extremely impressive drops in systolic blood pressure approximating 30 mm Hg. These results have sparked a great deal of excitement in the hypertension community and stirred the interest of a multitude of medical device companies. Some experts have proclaimed renal denervation a potential “cure” for resistant hypertension, perhaps enabling a significant number of patients to eliminate all drug therapy. Other expanded uses of the technology in more moderate forms of hypertension and other diseases are under active consideration. Now, however, a new analysis of the available data suggests the troubling possibility that renal denervation may not be nearly as effective in cutting blood pressure as had been suggested in the earlier trials. Rather than reducing systolic blood pressure by 30 mm Hg, the new analysis suggests that a more realistic estimate of the effect of renal denervation may be a much more modest reduction of about 13 mm Hg.
The analysis, published online in Heart, demonstrates that the large reductions in blood pressure seen so far in clinical trials of renal denervation may be a consequence of certain key aspects in the design of these trials. The apparent blood pressure-lowering effect of renal denervation has been greatly magnified because the trials have been uncontrolled and unblinded and have utilized office-based blood pressure measurements rather than the far more reliable and consistent ambulatory blood pressure monitoring (ABPM).
When randomized and blinded trials do appear, starting next year with Medtronic’s SYMPLICITY HTN-3 Trial, “people are going to be severely disappointed,” predicts the senior author of the paper, Darrel Francis, because “the office and ambulatory effects are going to have to converge, and that will make them both in the 10-15 mmHg territory.”
Francis says that the real effect of a 10-15 mm Hg drop in pressure “is very worthwhile, but talking about 2 or 3 times that effect size does nobody any good, because it eliminates the scope for scientific discussion.”
A Medtronic spokesperson did not take issue with the conclusion that the blood pressure lowering effect of renal denervation is likely to be more modest than initially reported. “Frankly, it’s important that we set realistic expectations,” she said. The company expects that the ABPM reduction will be less than that seen with office measurements. “It’s not a cure, you’re not going to go off all your medications,” she said, in response to some of the hype that has accompanied the early trials.
“This is probably the most extreme case I have ever seen of inadvertent bias in measurement producing – concordantly across several studies – massive overestimation of an effect size,” said Francis.
Hypertension experts Franz Messerli and Sripal Bangalore agree that the hype over renal denervation is due for deflation:
Small wonder that the cookie is finally crumbling, bigger wonder that it actually took so long. The irrational exuberance for renal denervation in resistant hypertension was entirely driven by uncontrolled non-randomized observational office-based blood pressure studies. We showed that in the Spanish registry of more than 8000 patients about a quarter of patients with office blood pressures of above 160/100 were actually normotensive when monitored by 24-hour ambulatory blood pressure (ABPM). Further, in this same Spanish registry about one third of patients with resistant hypertension based on office blood pressure measurements had normotensive ABPMs. Thus many patients with so-called resistant hypertension by office blood pressure measurements may be pseudo-resistant only. This would indicate that the exaggerated benefits observed with office blood pressure measurements are due to the simple fact that some patients actually had normal blood pressure at trial baseline. Conceivably many of these patients have white coat hypertension which is unmasked by ABPM.
The lesson to be learned here is that all of these studies should either be placebo controlled or at least mandate ABPM at baseline and every follow up.
Renal denervation has not been approved for use in the U.S. by the FDA. Renal denervation products are available in Europe, and many more companies are working on new devices.
Measuring Blood Pressure
Why the discrepancy between office-based measurements and ABPM? The same discrepancy between office measurement and ABPM has also been seen in trials of blood pressure drugs, but disappears in randomized, double-blind trials. In those trials, ABPM measurements will generally be lower than office-based measurements, but the difference between the placebo and active arms will be consistent. The authors of the Heart paper offer a number of reasons why office measurements are often less reliable.
For one, in the office, physicians are likely to re-measure readings they think are inconsistent with their clinical impression of the patient. In an open study, this is especially problematic, since the physicians may be more likely to re-measure untreated patients with a low blood pressure or treated patients with a high blood pressure.
Another factor is that patients are more likely to enter a trial when they have a high blood pressure reading at the office. But in many cases this represents a higher-than-usual value for the patient, and a substantial treatment effect may then be observed by simply re-measuring the blood pressure at a later time point. This “regression to the mean” can easily lead to an overestimation of the treatment effect in an uncontrolled trial or an unblinded trial.
Another theory explaining the difference between office and ambulatory blood pressure is the “white coat effect” in which the higher blood pressures observed in the office signal a potentially dangerous “alerting response,” possibly mediated by the renal nerves. However, if this were true, then the blood pressure-lowering effect as measured by ABPM would be greater during the day, when the sympathetic nervous system is stimulated, than at night, when it is at rest. But this difference has not been observed in the renal denervation trials, according to the Heart authors.
A different but related problem is that some patients with apparent “resistant” hypertension may not actually be taking their medications as prescribed. The treatment effect in these patients may appear much greater than in adherent patients.