The ARRIVE Registry: Paclitaxel-Eluting Stents in the Real World

Background ARRIVE I (Taxus Peri-Approval Registry, A U.S. Multi-Center Safety Surveillance Program) has enrolled nearly 2,600 consecutive patients at 50 sites in the United States and features Boston Scientific Corporation’s Taxus® Express2 paclitaxel-eluting coronary stent system. (A peri-approval registry includes patients who are enrolled before and after a product is approved.) ARRIVE was initiated in cooperation with the U.S. Food and Drug Administration (FDA) and is the first Taxus registry in the United States. The six-month clinical findings from ARRIVE demonstrated safety based on a site-reported Taxus-related cardiac adverse event rate of 4.3 percent (106/2,477), including site-reported Taxus-related cardiac death, myocardial infarction and reintervention of the target vessel. The registry reported a stent thrombosis rate of 1.7 percent (angiographically confirmed thrombotic events and all deaths less than six months without obvious cause). ARRIVE reported a 95 percent follow-up at six months. Consecutive enrollment in ARRIVE yielded a very diverse patient population, including patients receiving multi-vessel stenting (17 percent), left main stenting (3 percent), stenting of total occlusions (2 percent), stenting of bifurcated lesions (8 percent), stenting of saphenous vein grafts (6 percent) and stenting of in-stent restenotic lesions (7 percent). Ten percent of patients presented with an acute myocardial infarction. For the diabetic population overall (31%), the Taxus-related cardiac adverse event rate was 5.4 percent and the need for repeat procedures was 3.1 percent. How does the ARRIVE Registry add to the current body of knowledge on paclitaxel drug-eluting stents, as compared to the data from clinical trials? The ARRIVE registry is an FDA-mandated registry being used as a post-market surveillance survey. That’s a lot of terminology for basically wanting to know how the stent performs in an all-comers population. As a registry, ARRIVE has some unique components: It is a consecutive registry. That means, during that day or that week or that month, the operators involved in the registry are to take everybody who comes into the lab into the registry, whether it’s an acute myocardial infarction (MI) patient, a cardiogenic shock patient, or whether it is someone unfortunate enough to have terminal cancer. All of these people are to be captured in the registry to see what the overall event rates would be in a total cath lab population. It’s also a way of getting around the issues of bias that can occur if operators have a choice of taking elective cases during the day and not recruiting some of the other, more difficult, patients. This registry is really meant to include emergencies that come in the middle of the night, and in that way, it gives us a real-world experience. ARRIVE is being overseen by and reported to the FDA, so if there are any problems that come up, they will know about them fairly quickly. The registry has been extended into both low-intermediate and high-volume centers. Most trials, and even many registries, depend exclusively upon experienced operators in high-volume centers. In this registry, we wanted a better feel for what it’s like to put the stent in a variety of different patients by blue-collar operators, those actually in the trenches. Can you discuss the ARRIVE Registry endpoints? The endpoints are what everybody is most interested in: firstly, whether there is a problem with stent thrombosis. We know that stent thrombosis does exist, and it exists for all stents: bare metal stents, sirolimus-eluting Cypher stents, the paclitaxel-eluting Taxus stents, and presumably every other stent that we’ll be testing in the next year or two. We need to get a handle on what exactly the stent thrombosis rate is, and a step toward doing that is to get data from all-comers. When you’re looking at clinical trials, you’re often looking at some of the best cases around, and therefore, these patients tend to have lower event rates, particularly lower sub-acute thrombosis rates. However, in a registry, when you include patients with acute MIs, people in shock, bifurcation stenting, multi-vessel disease with diabetes¦these are all recognized risk factors for having a higher subacute thrombosis (SAT) rate. In fact, the SAT rate might be two- to three-fold higher in some of these subsets. Other events include the traditional major adverse cardiac events (MACE), which includes death, MI and the need for revascularization. ARRIVE gives us a broad-based comparison to some of the other trials that have taken place. Subacute thrombosis for the paclitaxel-eluting stent was 1.3% in the TAXUS trials and in the ARRIVE registry, that figure was 1.7%. Yes, as far as the SAT rate, it’s going to increase to 1.6 or 1.7%. What we have seen in evaluating SAT rates in a lot of these trials is a splay that can range anywhere from 0.2 up to 2%, so the ARRIVE registry data falls within that realm. The reason is because once you include a population that has 15-17% acute MIs, then throw in bifurcation stenting, diabetes and multi-vessel disease, your SAT rates are going to go up. Furthermore, anybody who should die from sudden cardiac death also counts as an SAT, so that number is actually falsely elevated within the realm of general practice in regard to what we’ve seen in bare metal stents. What is the mortality rate for subacute thrombosis? It’s usually between 20-40%. How does the ARRIVE registry data compare with TAXUS trial data? The ARRIVE registry data compares very favorably. The need for ischemic-driven revascularization is about 4.3%. In some cases, that’s the same or even slightly lower than some of the subsets that have been looked at in the randomized TAXUS trial data. To the casual observer, that sounds paradoxical, but remember that many of the trials have angiographic components and the more angiography you do, the higher your need for revascularization will be, because of the occulo-stenotic reflex. You see a region that you previously didn’t know was causing a problem, but now that it’s visualized, you feel compelled to do something about it. It’s quite different when you do a registry, which in this case has no mandatory angiography endpoint. It is totally symptom-driven or ischemic-driven, and therefore you do as you would do with all of your patients,target lesion revascularization. That’s a partial explanation as to why the ischemic-driven revascularization number is low if not lower than what you see in the randomized TAXUS trials. The bottom line is that we are pleased that the paclitaxel-eluting stent performed very well and consistently in all of these critical subsets: acute MIs, bifurcations, diabetic states, people with in-stent restenosis. Each subset has been looked at independently and the target lesion revascularizations are within a percent or two of each other. What do you think about angiographic versus clinical endpoints? Well, there are various levels in which one can perform these quantitative analyses. There’s the basic total lesion revascularization (TLR), which is ischemic-driven or symptom-driven. TLR is going to be the lowest number. If you go by angiography, of course it starts to go up because binary restenosis, defined as greater than 50%, will be determined by the angiogram. You’ll be picking up people who are sort of in that borderline range of 50-70%, and who may or may not have symptoms. Once you see it, you feel compelled to do something. Compare that to things like late loss, which is even more sensitive, or intravascular ultrasound (IVUS) determined volumetric obstruction. Then you get to very sensitive measures. All of these things are interesting from the trial perspective. They shed more light on the mechanisms and effectiveness of the programs at which we’re looking. In general, however, what the interventionalist really wants to know is, will my patient come back or not? Therefore, the thing that is most germaine to that person is TLR. How did ARRIVE diabetic patients fare? The U.S. has now developed a reputation for having the heaviest people on earth, and recent reports have showed that about 62% of people are either overweight or morbidly obese. Diabetes will tend to track that number, and therefore, in the majority of U.S. studies, you see 30+% diabetes, which is also exactly what we see in the ARRIVE registry. The results were very positive. In essence, the TLR was within a tenth of a percent of the group as a whole in terms of event rates. In contrast to studies like ISAR-DIABETES, which would suggest that a slightly higher TLR rate (actually binary restenosis was higher than Cypher, but the TLR was not significantly different between the Taxus and Cypher studies) ARRIVE would suggest that diabetics seem to do as well as the general population. That consensus has been generally supported by all of the previous TAXUS randomized studies. TAXUS II through VI have all suggested that people with diabetes also attain benefit by receiving the Taxus stent. Could you talk about the current Express2 stent platform and the upcoming Liberté stent platform? The Liberte© is clearly an improvement in terms of trackability. What we can certainly tell is that the Liberte© platform dramatically enhances flexibility because of the relatively thin struts, making it more flexible. One of the trade-offs is that a thin-strut stent is not quite as visible radiographically. In the subsequent generations of the Liberte© stent, tentatively named Barracuda, it will have a platinum enhancement to improve radio-opacity. How much thinner are the stent struts between the two generations? Liberte© is .0038 and the Express2 is .0052. Is it true that the thinner the stent strut, the less chance there is for restenosis? No, that’s not true at present. There is a weak restenosis and the stent struts. The majority of the improvement in thinning the struts is in the enhanced deliverability because of increased flexibility. The Liberte© also has an interesting design pattern which brings the struts closer together without actually welding more to one another. This means the dose density of the delivered drug can potentially be improved while still allowing access to side branches. In addition, the way the balloon re-wraps on deflation will reduce some of the concerns over balloon retention. Will Liberte© be used in the ATLAS trial? Yes, this is the platform that’s being used in the ATLAS trial. That data will hopefully be used in submission to the FDA to gain approval of the Liberte© stent with paclitaxel coating. The patient population of ARRIVE seems reflective of the ongoing trend in invasive cardiology toward more difficult patients and lesions. I think there’s no doubt that we have generally been able to extend the limits of the types of patients we do with regard to technical challenge and feasibility. We’re also treating people who have many more blockages in the same sitting. The capabilities of drug-eluting stents will only be improved with subsequent generations. That will really allow us to get there, deliver the package, and get a good angiographic result. Because restenosis is reduced, we can do more than one artery in a sitting, more than one site in an artery, and we can do it safely, with a good long-term outcome. Was there anything particularly surprising in the ARRIVE results? I think I was surprised, in a good way, that so many of the difficult subgroups did almost as well as the general population. I would expected to have seen more variability, if anything. I was surprised at one number in particular. When investigators did encounter restenosis of the stent itself, 88% of them put in another Taxus stent. This opens up a broader question, which is what do you do when a drug-eluting stent (DES) fails? Now that brachytherapy will really no longer be with us (within the next six months it’ll be gone), we have to think about what sort of strategies we should use for DES failures. This is my personal feeling and probably needs to be substantiated, but a lot of the focal restenosis may not be a failure of the drug itself. Certainly in edge restenosis, there is no drug at that spot, so it’s hard to call that a failure of the drug. Treating edge restenosis with a Taxus or Cypher stent (if you put that in previously) is probably appropriate. The real question is, what do you do when you have a diffuse or proliferation response, where the whole stent becomes involved? We used to use brachytherapy for that type of restenosis, but now the only thing we can do, short of going to surgery, would be to use an onocologic approach applying a different cell-cycle inhibitor. If the Taxus fails, put in the Cypher. If the Cypher fails, put in the Taxus. This requires validation, obviously. Fortunately, the diffuse restenosis would be many months out from implantation, so it’s not like you have two active drugs present at one time. Instead, you are using one drug to treat the other’s failure. Are you involved with the ATLAS trial? I’m not a participant in the ATLAS trial because of my involvement in ARRIVE. My understanding is that ATLAS is mostly wrapping up and we are just waiting to see the final results. I don’t expect any specific surprises. The ATLAS trials are looking at a wide variety of both small- and large-diameter vessels. There’s also a left main component. I believe there will be validation in all of these subsets using the new Liberte© platform for the delivery of paclitaxel. In general, this is clearly a very exciting time to be in interventional cardiology. We have looked at two very successful drugs, we’re looking at the introduction in trial form of two more this year, and a whole different type of platform being tested with the Connor stent and its non-polymer-based release of paclitaxel. I would encourage readers not to lose sight of the fact that our real mission is to find out more about the biology through these trials and determine what’s the best step to take for your patient in the future. Basically, these trials will provide more DES information to keep us going up to the end of the decade. Beyond that, we might see even more substantial improvements that will lower the restenosis rate further, provide optimal treatment for de novo and restenotic lesions, and potentially include a second drug on the stent delivery platform. These drugs might reduce platelet aggregation and thrombus formation. Although the problems with SATs may not be much different than bare metal stents right now, an SAT is a devastating event with a high mortality when it occurs in a non-hospital setting. We’d really like to dedicate our research toward lowering that number even further. We hope to lower the restenosis rate to its theoretical lowest component and begin to develop specialty DES stents that will help us treat, in a dedicated fashion, bifurcation lesions, and SVGs that move on into the periphery. Clinical Trial CliffNotes ATLAS trial results support safety of Taxus® Liberte© Paclitaxel-Eluting Stent System The thirty-day safety data from ATLAS clinical trial presented at EUROPCR 2005 meeting demonstrates the safety of the next-generation Taxus Liberte© stent system for the treatment of coronary artery disease. Mark A. Turco, MD, FACC, Director, Center for Cardiac and Vascular Research, Washington Adventist Hospital, and Co-Principal Investigator for the ATLAS study said that he has found the Liberte© stent platform to be a significant step forward in deliverability and conformability, particularly in the most challenging lesions. He further added that he also hopes to demonstrate in the ATLAS program that the homogenous stent architecture of the Liberte stent would further optimize drug delivery. The ATLAS trial is a global, multi-center pivotal study designed to support U.S. FDA approval of the Taxus Liberte© stent system. It is assessing the safety and efficacy of a slow-release dose formulation paclitaxel-eluting Taxus Liberte stent system for the treatment of coronary artery disease. ATLAS compares the Taxus Liberte© system to a matched control group from the TAXUS IV and V trials made up of patients with the Taxus Express2 system. The results presented include 871 patients at 61 sites in the United States, Canada, Australia, New Zealand, Singapore, Hong Kong and Taiwan. The primary endpoint for the study is target vessel revascularization at nine months. In addition to the ATLAS trial, the Taxus Liberte© program includes several expansion studies for long lesion stenting, small-vessel stenting and direct stenting of coronary lesions. Thirty-day MACE was 3.3 percent in Group X and 2.8 percent in Group Y (the study remains blinded through the primary endpoint at nine months). Cardiac deaths were 0.2 percent in both groups. The overall myocardial infarction (MI) rate was 3.0 percent in Group X and 2.6 percent in Group Y. The overall target vessel revascularization (TVR) rate was 0.4 percent in Group X and 0.2 percent in Group Y. Stent thromboses were also low, with a rate of 0.5 percent in Group X and 0.2 percent in Group Y.

NULL