Stents vs. Pills — Summer Re-run in February

Pills vs. StentI’m looking at this morning’s news and I’m seeing headlines like these: “Stents Overused in Stable Heart Patients” (WebMD), “Pills as good as stents for stable heart patients” (Reuters), “No Extra Benefits Are Seen in Stents for Coronary Artery Disease” (New York Times), and “Stents no better than pills for some heart patients” (MassDevice) — and I think I have time-traveled half a decade back to March 2007 when the results of the COURAGE study were presented at the American College of Cardiology Annual Meeting.

The reason for these headlines is a new “meta-analysis”, or “re-run” as I call it. It is titled, “Initial Coronary Stent Implantation With Medical Therapy vs Medical Therapy Alone for Stable Coronary Artery Disease“, and it appeared in yesterday’s Archives of Internal Medicine (AIM), a journal that has historically worn its anti-PCI bias on its sleeve. In my opinion, there are several significant issues/flaws in this publication that need to be addressed and here are five of them.

1. Eight Trials (Really Three) and Not “Vanilla” Patients
The AIM meta-analysis includes eight past trials that compared how well patients fared with optimal medical therapy (OMT) vs. medical therapy plus stents. This is basically the same question that COURAGE asked and the conclusion of this meta-analysis is the same as COURAGE: there was no difference in the outcomes (a conclusion that then gets mutated by the mass media into headlines that proclaim “Stents Are Useless”). That’s no surprise because COURAGE was one of the eight trials included and accounts for one-third of all the 7,229 patients looked at in this new analysis. Kind of a COURAGE re-run. All the trials that were included, with numbers of patients and years in which they were enrolled follows:

Trial
# patients
enrollment years
TOAT
66
1997-1999
Hambrecht
101
1997-2001
DECOPI
212
1998-2001
OAT
2,166
2000-2005
MASS II
408
1995-2000
COURAGE
2,287
1999-2004
JSAP
384
2002-2004
BARI 2D
1,605
2001-2005

Remember, a meta-analysis is not a study generating new information; it’s a retrospective view attempting to put various data into perspective. Unfortunately, in my opinion, it’s a flawed perspective. For example, look at the numbers in the above trials. Three of the eight trials accounted for 84% of the total patients. COURAGE, as I noted above, and then the OAT (Occluded Artery Trial) which accounted for 30% of the patients. Except that the OAT study looked at patients who had an occluded artery opened up AFTER having a heart attack. It was a very specific study and revealed important information, but it would be hard to say that this constitutes what doctors refer to as your vanilla “stable” patient. Nor is the patient population in BARI 2D — a trial that looked at PCI, CABG and OMT for patients with Type 2 diabetes — once again, not your vanilla “stable” patient.

2. Pre-or-Early Drug-Eluting Stent: The Past
Then there is the time frame. If you look at column three, you’ll note that all of these patients were enrolled pre-2005. The authors of this study prided themselves on updating the known information from previous meta-analyses that compared early OMT with balloon angioplasty, clearly outdated. Yet the trials included in the AIM study are comparing medical therapy with bare metal stents mostly. For example, only 31 out of 2,287 patients in COURAGE were treated with drug-eluting stents. What would the COURAGE results have been if drug-eluting stents were used? Or if the newest second and third generation DES were used? For example, just this week, the FDA approved the Medtronic Resolute Integrity for use in diabetic patients and the Boston Scientific Liberte and Ion for use in Acute Myocardial Infarctions.

So what, you say? Well, the Annals meta-analysis showed that vessel revascularization for stent patients was 21.4% — compare that with the results from the recent TWENTE trial, showing revascularization with the newer Resolute and Xience DES hovering around 3%!! Might the headlines in this morning’s news be different if the latest stents had been used instead?

3. What is a “Stable” Patient?
What exactly is a “stable” patient? Normally, this is a patient who has known coronary artery disease (CAD) but who only experiences angina upon exertion. Upon resting, the angina predictably goes away. In unstable angina, the pain comes on unpredictably and does not necessarily recede with cessation of activity. Unstable angina can easily progress to acute coronary syndrome (ACS) and to an acute episode, or heart attack.

4. Triaging CAD with FFR instead of Angio: The Future
A recent study, titled FAME II, previously discussed in this blog, specifically looked at patients with stable angina. But instead of measuring them with an angiogram, the physicians used Fractional Flow Reserve (FFR) during the catheterization to determine if the blockage they saw was ischemic, i.e. it restricted blood flow significantly. If the blockage was ischemic, the patient was randomized to medical therapy or stenting. So this is not your vanilla “stable” patient either. It’s a patient with stable angina, but who has been diagnosed with an ischemic lesion. So what did FAME II discover? Well, we don’t know the full story yet, because the Data Safety Monitoring Board advised the investigators to end enrollment early due to ethical concerns.

Uh oh. Stents vs. medical therapy. Study stopped early. Sounds bad for stents, right? Wrong. The study was stopped early because a significant number of patients who were randomized to medical therapy alone had progressed to “urgent revascularization” — Dr. De Bruyne, the PI for FAME II, told me that the great majority of them had moved into ACS. So, while it will take time to digest and analyze the results from FAME II, and at least a year or two to see if there will be a mortality benefit to PCI, there’s no question that in this patient population of so-called “stable” patients, stenting was definitely the preferred treatment.

5. Individualized Therapy and Patient Choice
Which just calls up a very important issue of how to treat patients. As discussed in Chapter Two of Dr. Eric Topol‘s amazing book, “The Creative Destruction of Medicine“, medicine today, and this meta-analysis is a prime example, looks at population studies instead of individualized medicine. In its quest for “evidence-based medicine”, blanket therapies are prescribed for patients, when their individual needs may be different.

For example, some patients do not react well to medications. Even today, the FDA issued a warning about statins, the centerpiece drug for lowering cholesterol. Something like 1 in 200 patients have suffered increased risk for diabetes, high blood sugar levels or memory loss. Many patients who have written into our Forum complain of muscle pain, rashes, etc. from medications. Likewise, some patients don’t do as well as expected with PCI or stenting.

The danger, however, is in the blanket statement that stenting adds nothing to medical therapy. The fine points of these studies get lost, even in medical publications: for example, MassDevice’s coverage of the AIM meta-analysis starts off by proclaiming, “Patients rushed to the hospital for a heart attack are no better off with a stent than they would be with less-costly drug therapy, according to a new study.” Absolutely wrong and not in any way what the meta-analysis says.

The other and bigger danger, however, lies in a a statement made to Reuters by co-author of the meta-analysis, Dr. David Brown:

Brown…doubts the study will sway too many doctors, but said it may influence insurance companies.

“A few practitioners might change their behavior, but third-party payers will be influenced by it and they will start by making stricter criteria for reimbursing these procedures,” he said.

Should this come to pass, medical decision-making will no longer rest with the physician, and certainly not with the patient.

Leave a Comment

Filed under Clinical Trials / Studies, COURAGE, Media Coverage, Stent

Leave a Reply

Your email address will not be published. Required fields are marked *