The STENT Registry: A Real-World Look at Sirolimus- and Paclitaxel-Eluting Stents

What is the STENT Registry? The Strategic Transcatheter Evaluation of New Therapies (STENT) Registry is a multi-center group that was founded in 2003. The original purpose of the effort was to have a large, multi-center registry with late clinical follow up planned on all patients in order to assess outcomes for new clinical devices coming into the cath lab after FDA approval. We started recruiting sites in 2002, in preparation for prospectively enrolling percutaneous coronary intervention (PCI) patients by May 2003. At that time, we had four sites, and the STENT Registry has now grown to nine coronary interventional centers, mostly in the southeast (Table I). Most clinical devices, like the Cypher drug-eluting stent, get approved after a fairly small patient sample size in a randomized clinical trial. For example, the Cypher stent was approved based on results from the SIRIUS trial, which was a randomized study of bare-metal stents versus the Cypher stent. There were only approximately 1100 patients in the SIRIUS trial, and only half of those received a drug-eluting Cypher stent. In addition, patients are hand-picked for these randomized clinical trials. They do not represent the broad spectrum of patients we treat in the cath lab. We are aiming for a minimum enrollment of 80% of all coronary interventions at participating sites, with a goal of over 90% clinical follow up at 9 months. Our primary endpoint is 9-month clinical follow up, with the purpose of rapidly assessing outcomes like restenosis, death and myocardial infarction (MI), and major adverse cardiac events (MACE) the composite endpoint in patients receiving new devices. We timed the STENT Registry to start with the Cypher stent approval because it was the first upcoming big new device in the cath lab marketplace, and marked a very important phase shift in coronary stenting. Our October 2005 TCT presentation covered the time period of May 2003 through September 2004, which included all patients who were eligible for 9-month follow up through June 30, 2005. At that point, the patient population that we had was 6,336 patients. That number includes all procedures, not just drug-eluting stents. The registry also includes balloon angioplasty and atherectomy patients, as well as laser, cutting balloon, and so on it’s all procedures. Out of those 6,336 patients that were in the registry and followed out to 9 months by June 2005, 60% were pure DES cases. Of those, we had cases that were purely Taxus and cases that were purely Cypher cases. Those are the two data sets that we compared. There are patients in the registry that ended up getting a mixture of DES -they might get a Cypher in one vessel and get a Taxus in another vessel, or they might get a bare metal non drug-eluting stent in one lesion, for some reason or another. Those patients were removed from this particular analysis because we wanted to look at pure Cypher versus Taxus cases. In our TCT presentation, we had 2282 Cypher patients and 1476 Taxus patients out to 9-month follow-up, representing 95% clinical follow-up in both groups. We looked at the 9-month MACE endpoints (death, MI, TVR) for both patient populations, and we also looked at the baseline characteristics of both groups, so we could adjust for any differences that affected their outcomes. A Cox proportional hazards risk adjustment was performed between the two groups, allowing for a head-to-head comparison of the two stents. The final results basically showed that the stents performed equally (Figure 1). The point estimate for the comparison showed that Taxus looked a little more favorable for both TVR and MACE, but it was not statistically significant. The Kaplan Meier curves show freedom from event curves for the two stents (Figure 2). Freedom from MACE is one plot; the other plot is freedom from target vessel revascularization, or restenosis. The stents are practically identical. The event rates over time, from day zero out to nine months, are basically the same for both stents. The numerical rates for adverse events like restenosis and MACE were slightly lower for Taxus than for Cypher. However, the standard deviations overlap, so there is no statistical difference between the two stents. If you go back to some of the earlier studies and look at the angiographic measurements of the stents at late follow up, it’s well-accepted that there’s a little more tissue ingrowth, when you quantitate it, in the Taxus stent as compared to the Cypher stent at follow up. That’s not necessarily a bad thing, but it is a measurement of how much the drug inhibits intimal regrowth inside the stent. Thus, despite the fact that late loss is a little higher for the Taxus than the Cypher stent, it doesn’t seem to correlate with any differences in patient outcomes. They have similar rates of restenosis, mortality, and risk of myocardial infarction, at least out to 9 months in our registry. What are some of the most unique aspects of the STENT Registry? Our registry is the first time in the U.S. that these two stents have been looked at head-to-head, post-approval, in a prospective multi-center study. The STENT Registry is multi-center as opposed to single-center, so it cuts across several levels with smaller and larger centers, and it has a very high rate of enrollment. We had 83% of all patients enrolled up front, with 95% clinical follow up. Because of the high percentage rate of follow up, it is very reliable data. I think this process is going to become a more important part of evaluating products. With carotid stenting, the FDA is actually mandating that for each of the carotid stent products that companies must produce a certain number of patients after approval and show their clinical outcomes. We need to have some sort of post-approval registry to look at the outcomes of any new devices in real-world practice, in a complex group of patients. After the Cypher stent was approved, there were some reports of stent thrombosis, which occurs with non drug-eluting stents as well, but the question was, is it more frequent with drug-eluting stents? The FDA couldn’t get a good handle on it from the reports, so they required that Cordis do a PMS (post-market surveillance) registry. They had about a thousand patients in that registry, but it included only the Cypher stent. The Taxus stent had the ARRIVE registry by Boston Scientific, which only included outcomes in Taxus stents. The STENT Registry data from our multi-center organization is the first Cypher and Taxus comparison clinical data in the U.S. We are an independent study group and have received unrestricted funding from both Cordis and Boston Scientific. They are now interested in funding us to complete two-year follow up. You mentioned subacute stent thrombosis. What were some of the conclusions from the STENT registry on this issue? The two stents were equal. The stent thrombosis rate for Cypher was 0.7%, and for the Taxus stent, it was 0.5%. There is no statistical difference and there was a low percentage rate for both stents. Out there in the real world, where we are treating a lot of complex and longer lesions than in the clinical trials, as well as a wide range of clinical types of patients, restenosis and stent thrombosis rates look very favorable, very similar to the randomized trial data. There was also a high percentage of small vessels stented in the registry. Since we don’t have an angiographic, quantitative measurement in the core lab, the way we do vessel size in the registry is by visual estimate. Thus, the vessel size numbers that we have in there are the physician’s visual estimate of the vessel size that’s the best we are going to get. It is not a continuous variable, because when operators estimate vessel size, they usually will cut it no closer than a quarter millimeter. Thus, this method of measurement is not as accurate as if you had a core angiographic lab, like in the randomized trials, but it does give you a general idea. Regardless, the numbers in the STENT Registry show that vessels Can you share some of the future plans for the STENT Registry? Future plans are to continue long-term follow up at two years for the drug-eluting stents, but we are also looking at some other things. We collect all of the medications given during the procedure, so our group is going to be looking at outcomes in relation to the type of anticoagulation given during the procedure. We are particularly interested in comparing bivalirudin to the IIb/IIIa agents in real-world practice. We are also looking at Plavix loading and relationship to outcome. This is a big issue in the lab now. Patients are getting all kinds of different loading doses because there has not been one dose regimen mandated thus far. Patients are getting 300 or 600mg of Plavix, depending on what the operator wants to do at the time, and they are getting it perhaps a day before the procedure all the way up to on the table or after the procedure. We’re interested in collecting that data to see if there is any impact on outcomes. It actually has been shown in previous trials to affect outcome. Also, since we have such a large volume of patients, we are able to look at some subsets of patients that you can’t usually look at unless you have a large registry. We’ve looked at success rates in chronic total occlusions. We have a paper that we presented at TCT this year on drug-eluting stents and saphenous vein bypass graft lesions. We’re of course looking at our diabetes population. One late-breaking presentation for the American College of Cardiology Scientific Sessions 2006 will be how Cypher and Taxus stents are comparing in diabetics. One of our investigators, Dr. Barry Cheek of High Point Regional Hospital, presented a paper at TCT on Cypher and Taxus drug-eluting stents versus bare metal stents in diabetics. Early on, in 2003, there were some patients still getting bare metal stents, and we followed them up. We have an interesting comparison of diabetics versus non-diabetics, and bare metal versus drug-eluting stents. Investigator Dr. Bruce Brody, in Greensboro, North Carolina, at the LeBauer Cardiovascular Research Foundation, is looking at acute MI patients. Drug-eluting stents in STEMI is another important focus, and we’re accumulating bigger numbers so we can make some conclusions about restenosis, stent thrombosis, and so on, for acute infarct lesions. We also look forward to evaluating the other newer drug-eluting stents as they gain approval, such as the Endeavor stents, the CoStar stent and the ZoMaxx stent. Those will be rolling out in the next 1 to 2 years, and we obviously want to keep our organization active so that when we get to that point, we can start comparing the other new drug-eluting stents to the ones we have now. Do you anticipate expanding to additional centers? I think we’re going to stay where we are now. Our database coordinators are fully active at this point. The STENT Registry has plenty of patients. We’re enrolling now at about 1000 patients a month, or about 12,000 patients a year. It’s a big effort, and it wouldn’t make any sense to enlarge the registry size further. Those are big-enough numbers to investigate most of the subgroups of interest. One of the nice things about the registry is that it is a small group of centers. We are able to meet more easily than when there are 100 or 200 centers involved in a trial. We can have investigator meetings where everyone is actually involved. All the coordinators from each of the centers come into Charlotte for all-day symposia on a regular basis. We review our procedures at the different sites, as well as results, and share tips and tricks on enrollment and follow up. It’s a nice-sized group, and the people at the different centers are getting to know each other and enhance each other’s performances in the registry.

NULL