What If PCI Is Just A Sham?

–A small study might upset assumptions about the benefits of stent.

What if PCI for stable ischemic heart disease (SIHD) is just a big lie? That is, what if it’s no better than a sham procedure? This may seem like a crazy proposition, but there’s at least an outside chance that a small trial coming out later this fall may seriously raise that possibility.

Despite significant limitations (discussed below), ORBITA (Objective Randomised Blinded Investigation With Optimal Medical Therapy of Angioplasty in Stable Angina) may deliver the best evaluation of the relative value of stents versus optimal medical therapy. Results of the trial won’t be presented until November 2 at the TCT meeting, but it may be worth it to start thinking about this trial now.

You may be surprised to learn that PCI, which recently celebrated its 40th anniversary and is the cornerstone of interventional cardiology, has never been tested in a thoroughly rigorous manner. Along with many other device-related procedures, PCI is inherently difficult to assess in an objective fashion in clinical trials, since it is nearly impossible to eliminate the impact that knowledge of treatment assignment might have on both patients and physicians. ORBITA appears to be the first study to correct for this bias by performing sham procedures on patients in the control group, so that neither patients nor physicians knew whether or not a real procedure had been performed.

The study design means that the actual effect of PCI, at least in the short term, may be measured for the first time. 200 patients with stable ischemic heart disease (there is broad agreement that PCI is beneficial in patients with unstable disease) will be randomized to PCI or a sham procedure. Participants in the study will undergo a battery of tests prior to their procedure and 6 weeks after. The primary outcome measure is exercise time on a treadmill at 6 weeks. Patients and their physicians will be unblinded after 6 weeks, so the trial will not assess the long term effects of PCI, but PCI for stable ischemic disease is now thought to mainly benefit patients by improving their symptoms and quality of life, effects that should be fully evident at 6 weeks.

One reason to take this study seriously is that the group leading the study has a good track record uncovering the unpleasant truth that the benefits from some procedures derive not from reality but from the collective delusion of experts. Darrel Francis (Imperial College London), the ORBITA study chair, led an earlier paper that was the first to point out, in the midst of nearly overwhelming hype, that renal denervation would not be nearly as effective as all the hypertension and interventional cardiology experts were assuming. And his group was also instrumental in pointing out the undeserved hype surrounding cardiac stem cell therapy. ORBITA, however, is the first time they have performed their own clinical trial testing an established procedure like PCI using sham controls.

Expert Perspectives

In an interview Judith Hochman (NYU) said that it was “really admirable and important of them to use a sham procedure” but said she was eager to learn more about the details of the trial. She wondered, for instance, about the use of dual anti platelet therapy (DAPT) in the trial. “Did they use DAPT in both groups?” If they did not use DAPT in both groups then that would unblind the trial. If they did then they unnecessarily exposed the sham patients to the risks of DAPT.

Hochman was also interested to learn the level of angina at baseline. She speculated that most cardiologists would be reluctant to randomize patients with severe angina. “I would be very surprised if they enrolled patients with angina refractory to antianginal meds and there was no difference.” On the other hand, she said she would not be surprised if there was little or no difference in a patient group with only mild symptoms.

Hochman is the chair of the upcoming ISCHEMIA trial. She pointed out that ORBITA only addresses the angina and exercise time parameters while the far larger ISCHEMIA trial is an event-driven trial. There is no established connection between short-term symptomatic relief and long-term outcomes, she said.

I asked a number of other experts for their thoughts about the trial. All praised the novel trial design and said the results— whatever the outcome— were sure to be interesting. But, they also agreed, the small size and other aspects of the trial design mean that the trial will never be able to establish a definitive answer to the question of whether or not PCI is beneficial. They were in agreement, however, that the trial will likely have a major influence on future directions of research in the field. Here are their comments:

Dan Mark (Duke):

“This certainly has some interesting features, particularly the sham procedure which has not been done in CAD surgical therapy since Cobb’s 1959 sham IMA ligation study.

“The problem is that there is, as far as I can determine, no detailed protocol for the study available to examine. Patients will be selected after a run in medical therapy phase and the trial endpoints are determined at 6 weeks. Sample size was 300 but reduced in 2015 to 200. No idea what the hypothesis is the authors are using to drive that sample size choice. It’s really a pilot study, perhaps showing that this approach is feasible, but if they are having enrollment problems its not clear that objective will be met.

“Primary endpoint is time on the treadmill. Even if the trial is statistically ‘positive’, the meaning of a small shift in exercise time will be hard to discern and unlikely to alter clinical practice. The problem with an endpoint like this is it assumes patients are like machines that are limited in their physical performance by one dominant malfunction and that by correcting this malfunction with PCI you can uncover the true performance. The real world is rather messier than that unfortunately. The ability to respond completely to a real physiologic improvement in myocardial blood flow may be limited by numerous extraneous factors and result in a lack of precision of the outcomes that gets misinterpreted as “play of chance” or no effect.

“The trial involves use of FFR and I am unclear whether decision to enroll patients requires the FFR to be low or not.

“So is a trial that is almost certainly underpowered for any changes likely to persuade us to alter clinical practice for these patients. Given the complexity of the problem being studied and the likelihood that results can vary substantially depending on which patients actually get enrolled in the study plus the filtering effect of the run in phase plus the likely filtering of clinicians who would be uncomfortable enrolling patients they are worried about into a trial like this, you have a small treatment effect applied to a cohort with a low propensity to benefit and the results are most likely going to be officially negative but more properly ambiguous. Will this promote more tolerance for sham procedures in future trials of stable angina/CAD? Hard to predict without more details but that is going to be a harder sell than in the renal arteries because of what is perceived to be at stake if things go wrong.”

Robert Yeh (Beth Israel Deaconess Medical Center):

“I think this is a really important study both for its unique design as well as the importance of the question that it seeks to answer. Randomizing symptomatic patients with significant coronary lesions potentially to sham procedure despite PCI being a commonly accepted procedure is very bold and important. It also is the source of the greatest risk for the trial – namely, whether it is able to successfully enroll patients who have very severe angina or critically stenosed coronary lesions. If the investigators enroll truly symptomatic patients despite medications, then my expectation would be better exercise time and fewer symptoms with PCI compared to medication. Certainly, a null finding despite the enrollment of such patients would be a real eye-opener for the field.

“My greatest fear though is a null trial after enrollment of patients with mild symptoms, if physicians refuse to randomize those that might benefit the most from treatment. That type of result would be very polarizing, with those in one camp emphasizing the absence of benefit of PCI and the other side claiming the trial didn’t enroll the right people. It doesn’t help our field.

“I like the design very much. It is the execution that is going to drive the study’s importance to the field.

“I think it takes the COURAGE/ISCHEMIA trial to the next level by adding a sham arm, as well as making sure patients are all on medications prior to randomization. Obviously it is a much smaller trial, but I commend the investigators for the challenging and important undertaking.”

John Spertus (University of Missouri):

“It is a very interesting concept and the sham procedure adds an important element that is rarely used in clinical trials of non-pharmacological treatments. It was certainly illuminating in knee surgery, LIMA ligation and percutaneous transmyocardial revascularization. It is a shame that they have such short follow-up, as I would expect that any placebo effect that they are trying to demonstrate would not be attenuated without longer follow-up.”

Sanjay Kaul (Cedars Sinai):

“I have always been suspicious of angioplasty yielding a placebo/sham effect! If there is no difference, and it is replicated in other trials, it is going to shake the world of interventional cardiology.

“The vast majority of PCIs in the US are done for symptomatic relief of angina; and if this trial shows that it can all be attributed to a sham effect, it will question the need for the procedure to begin with. Or maybe not, perhaps sham effect is therapeutic.”

Eric Velasquez (Duke)

“The sham component is quite novel and may be important, especially as it relates to the upcoming ISCHEMIA study results. For example, if ORBITA shows no difference perhaps due to unanticipated improvements in sham group this would raise some questions regarding how to interpret a benefit of PCI v med strategy if found later in ISCHEMIA.”

Ajay Kirtane (Columbia University):

“To my knowledge, this trial is unique in that it employs a sham-controlled design in examining the effects of PCI + very optimized medical therapy (as per the investigators) compared with very optimized medical therapy alone. The study investigators really should be commended for undertaking a very challenging trial, which in some cases may have resulted in 3 procedures for study patients (initial angiography, randomized procedure, and potential crossover after endpoint ascertainment in cases of failed medical therapy)! The uniqueness of the trial is no doubt why it was rated so highly by the late-breaking trials selection committee at TCT, despite the small sample size.

“For me, the major questions that I’m waiting to see revealed relate to the following: 1. The 70% angiographic entry criteria into the study (rather than angiography with physiology prior to randomization); 2: The quality-of-life assessments, particularly as all patients are aggressively managed with medical therapy prior to randomization and given that symptomatic patients frequently prefer less medications, rather than more; and 3. The basis of the sample size/power calculation (e.g. prior trials of antianginal therapies such as the CARISA trial of ranolazine have employed larger sample sizes for a similar effect size in treadmill exercise duration). I feel slightly better about the sample size concern given the experience of the study investigators, but all of these details of the study will be of key interest, especially concerning the interpretability of the findings to contemporary practice.”

Related reading:




  1. I wonder how they know that the “optimal medical therapy” is optimal. Do they mean a therapy that would be applied identically to each patient, or do they mean a custom-made design of optimal?

  2. According to TheNNT,the result of randomised trials using coronary stenting has shown that, In summary, none of the patients were helped (life saved, heart attack prevented, symptoms reduced), See at http://www.thennt.com/nnt/coronary-stenting-for-non-acute-coronary-disease-compared-to-medical-therapy/ .

  3. blah blah blah says

    Why isn’t anyone concerned that 30% of the patients in the study had an FFR >0.8? The trialists showed that patient’s who shouldn’t get PCI shouldn’t get PCI. Big deal.

Speak Your Mind