Patient Selection, Physician Experience, and Anti-platelet Therapy Testing are Critical

Michael J. Alexander, Professor and Clinical Chief, Department of Neurosurgery,
August 03, 2012

The response from the Stenting and Aggressive Medical Management for the Prevention of Recurrent Ischemic Stroke (SAMMPRIS) trial principal investigators (PIs) is greatly appreciated. Healthy debate helps us weed out the details that are important in these studies and future trial design. Although the PIs label some of the editorial comments as "inaccuracies", one could disagree based on the information listed below. So let's look at the numbers and the facts.

First, they state the SAMMPRIS population was the right population to test, based on the Warfarin versus Aspirin for Symptomatic Intracranial Disease (WASID) trial data. However, they base this statement on the criteria of the degree of stenosis, and some standard, but relatively artificial epidemiological criteria, not focusing on some other important clinical criteria. For example, the mean time of enrollment from qualifying event to enrollment was 17 days in the WASID trial and 7 days in the SAMMPRIS trial, so these are not exactly the same patient populations. The patients who were stented in SAMMPRIS within 7 days of their qualifying event did remarkably worse than the patients stented after day 7, indicating that either these patients had unstable plaque or some other factor that increased their risk for complications when treated early. The Wingspan safety trial that led to FDA approval of the Wingspan stent did NOT include patients who had stroke less than 7 days, so to stent patients at less than 7 days from the event was another variable introduced by the SAMMPRIS trial design that was not a part of the original Food and Drug Administration (FDA) approval of the stent.

Second, they state that the Wingspan stent was not used off-label. This is not exactly true. The Investigational Device Exemption (IDE) use of the Wingspan stent in the SAMMPRIS trial was used for expanded indications compared to the Humanitarian Device Exemption (HDE) FDA approval. In fact, this was exactly one of the reasons that an IDE approval was necessary for the trial. Interestingly, if one analyzes the presenting symptoms, time to treatment, medical failure, and other factors, only 8% of the patients enrolled in SAMMPRIS would have qualified for the original Wingspan safety study upon which the FDA granted approval (FDA Panel Review, March 2012, Baltimore, MD). Therefore, practices such as treating patients before day 7 post-event, allowing TIA versus stroke only as an entry criteria, and allowing patients who had not failed medical therapy into the trial, were markedly expanded indications for the stent in this trial. For this reason, one could argue that this trial was not designed for stenting to succeed. Typically, every device trial is initially designed for the on-label indication, and only later are there considerations for expanded indications. This was not the paradigm used in SAMMPRIS. In the subgroup of the 8% of patients in SAMMPRIS that actually would have qualified for the original Wingspan safety trial, there was no statistical difference in clinical outcomes between the stenting and the medical therapy groups.

Third, the argument that even the subgroup of patients who previously failed anti-platelet therapy did worse with stenting than medical therapy is easily predictable, if you look how stented patient were managed in the study. Many major centers test patients for anti-platelet therapy resistance prior to stent placement, since there is a documented 15% relative clopidogrel resistance in the population and 10% aspirin resistance. Patients who are resistant to anti-platelet therapy are more likely to have a thromboembolic event with a fresh stent placement, as a nidus for platelet aggregation than someone with a chronic atherosclerotic stenosis. The SAMMPRIS trial failed to account for possible anti-platelet therapy resistance as many coronary stenting trials current address. In fact, it was a protocol violation in SAMMPRIS even to measure the anti- platelet therapy resistance.

Finally, the argument that interventionalist experience did not play a role in the results is very short-sighted and is based on poorly stratified data. It would be difficult to believe that anyone would argue that less experienced stenters could have equivalent clinical results than more experienced stenters. The average number of Wingspan stents placed by interventionalists to qualify as a stenter in the study was 10 Wingspan stents. This low number, as a distinction between low and high volume stenters, would not be acceptable in any current stenting trial: coronary, carotid, peripheral, etc. So technically, both "low volume" and "high volume" groups analyzed by the PIs were low volume users.

The goals for the future should be to focus our attention on defining which population will benefit most with revascularization, what is the best timing for the procedure, and what medical therapy regimen is best for patients with newly placed intracranial stents, which may not be the same as the best therapy for those without stents. SAMMPRIS gives a significant amount of information, but it is a starting point, not an end to this therapy.

Conflict of Interest:

Dr. Alexander was an investigator in the SAMMPRIS clinical trial. He is a device proctor and consultant for Styker Neurovascular which manufactures the Wingspan stent.

Conflict of Interest

None declared