The TAXUS ATLAS Trial: Workhorse lesions and direct stenting with the Libert© paclitaxel-eluting stent

It has a new five-fold balloon design that will likely decrease any of the previous withdrawal resistances recognized in the previous platform by some operators. The uniform cell geometry of the Libert© stent hopefully allows for more enhanced and uniform drug delivery versus alternating macro and micro elements of the Express2 stent. It also has thinner struts, a lower profile (0.041 vs. 0.044) and three separate designs: a small vessel stent with two cells, which treats 2.25 to 2.5 mm vessels, a workhorse-size stent with three cells for 2.75 to 3.5 mm vessels, and a larger vessel stent, also with three cells, for 4 to 5 mm vessels. These are the main characteristics that differentiate the Libert© from the Express2 stent. Libert© has the same paclitaxel drug dosing and the same translute polymer. How was the ATLAS trial constructed? ATLAS has four separate, distinct trials, all of which have completed enrollment and are in the follow-up phase. The first trial looked at the TAXUS Libert© stent in a workhorse population, of de novo lesions, with 871 patients enrolled. The clinical program has three other expansion arms that ultimately will be used for product indication: a small vessel study (261 patients), a long lesion study (150 patients), and a direct stenting study of the TAXUS Libert© stent versus a predilation population from the ATLAS workhorse trial. Two of the ATLAS studies, the workhorse population and the DIRECT STENT population, have been presented. The workhorse 12-month data was presented at the 2006 TCT and the DIRECT STENT 9-month data was also presented at TCT. We will have long-term follow up on all of these trial arms, and in all likelihood, we will present the small vessel data at PCR in 2007 and long lesion at ACC 2007. ATLAS showed non-inferiority between the two stent platforms in the workhorse population. The workhorse trial was designed as a noninferiority trial. The TAXUS Libert© stent was then compared to a historical control, which was a matched patient population from TAXUS IV and V. What was quite interesting was that despite operators following strict inclusion criteria, the patient population enrolled in the TAXUS Libert© group was exceedingly more complex than the patient population enrolled at the time of TAXUS IV and V. The Libert© group treated smaller vessels, greater diameter stenoses, longer lesion lengths, a greater percentage of tortuosity and calcification, as well as more branch vessel disease. There was also a higher percentage of B2 and C type lesions in the TAXUS Libert© group. Despite that increase in complexity, the data clearly showed that the TAXUS Libert© stent was noninferior to the control population. In addition, a significant and interesting finding was that procedure times were actually shorter with the TAXUS Libert© group. There were also other procedural characteristics that were statistically significantly improved in the Libert© group versus the control group. Libert© may provide better deliverability, better access to more distal lesions. These positive characteristics may allow us to say that the TAXUS Libert© is not only noninferior, but may have certain advantages over the TAXUS Express2 platform. How much of the improvement can be attributed to increased experience by various operators involved in the trial? Excellent question. The control and the active arm were performed in different times. Certainly it is a possibility that in the time period between when TAXUS IV and V were performed and the TAXUS ATLAS trial was performed, we simply refined our procedural skills, and that accounts for the difference. We will never know the true answer to that question. However, the advantage of decreased procedure times, despite the increased complexity of patients, does lend itself to the idea that there may be something more than just a change in experience over the years and more related to a technology transfer. Dr. Gregg Stone, commenting on the ATLAS trial, warned about efficacy drift, referring to the fact that we no longer have comparison trials between bare metal stents and drug-eluting stents. To what he was referring? One of the key things as we move forward is going to be the question of how future DES trials will be designed. What Dr. Stone and all of us have really tried to think about as we move forward, especially with some of the safety concerns, is how we are going to adequately assess the safety and efficacy of a drug-eluting stent platform, yet be able to get these newer generation devices through the regulatory pathways and to market, so we can use some of these new and possibly improved devices in our patients. It is a key question. TAXUS ATLAS used a historical control. The pluses of a historical control are that you can use fewer patients and we can leverage an excellent database of patients from prior TAXUS trials as the historical control. In ATLAS, we are not evaluating a new drug or new polymer, just a different and improved stent design. For that type of product, do we really need to have a true head-to-head comparison between a bare metal stent and the TAXUS Libert©, or between the TAXUS Libert© and an active control arm? In addition, at least here in the United States, it would be extremely difficult for us to do a trial comparing any drug-eluting stent to a bare metal stent platform in de novo-type lesions. I think it is unlikely we are going to go back to that, but in all likelihood, we will see more historically controlled trials and more trials that actually move the angiographic endpoint further out, beyond the primary endpoint of the trial. Right now, we do angiographic endpoint trials for drug-eluting stents, usually at 8 or 9 months. There is clearly a tremendous amount of oculo-stenotic reflex triggered when those patients come back for the 8- to 9-month follow-up. In the real world, of course, patients are followed clinically. Either the patient comes to us saying that they have specific symptoms, suggesting that there is a recurrent problem either in the target vessel or in a separate distribution, or they have a provocative stress test that leads us to perform a repeat cardiac catheterization. The oculo-stenotic reflex falsely elevates some of the angiographic endpoints, especially if we are speaking about treating an increasingly complex patient population with more and more disease in the target vessel proximal or distal to where we treat the target lesion. Registry evaluation also plays a role, especially large registry trials that do a good job of collecting data. Lower frequency events are likely identified and it appears that most large registries tract randomized trial results. With that said, especially when talking about safety, everyone would love to see that one, very large randomized trial that would hopefully answer the safety question. The FDA is likely to require longer term data for safety, initiation of post approval registries as a condition of approval and larger patient numbers. Trial design is critical. How did the Libert© fare regarding rates of stent thrombosis in the workhorse trial? The ATLAS workhorse trial and DIRECT STENT trial demonstrated clear safety of the TAXUS Libert© stent. The overall cumulative stent thrombosis rate out to 12 months for the TAXUS Libert© workhorse group was 0.9%, broken down as follows: 0“30 days = 2 patients (0.2%); 31“180 days = 4 patients (0.5%); 181“365 days = 2 patients (0.2%). In the Express2 control group, overall cumulative stent thrombosis rate out to 12 months was 0.7% (7 patients). P-value between the Express2 control and the Libert© was 0.63. Patients will be followed out to 5 years. How long were patients on clopidogrel? Six months was the protocol-defined dual anti-platelet therapy requirement. Slightly more than 60% of the patients were on duel anti platelet therapy out to 12-months. Recent data from Kaiser Colorado and Duke suggests that 3 or 6 months of duel anti-platelet treatment beyond current indications may reduce death and MI, although this is still up for debate. Let's discuss the ATLAS DIRECT STENT trial. What are the advantages and potential challenges of direct stenting? The advantage of direct stenting, especially in a drug-eluting stent population of patients, is that you possibly avoid geographic miss. Geographic miss includes areas that are not covered by the DES but injured by balloon predilation. With direct stenting, a stent is directly placed at the target lesion without predilation. Patient selection is the key. Direct stenting is not something that is going to work in a very distal lesion or in an angulated, calcific-type vessel. The risk is failure to deliver the stent, possible stent embolization, embolization of debris downstream and possible further vessel injury in trying to deliver the stent to the target lesion. In general, those patients that receive direct stenting are going to be less complex than the patient populations that receive balloon predilation. Who was the control group for DIRECT STENT? The control group for direct stenting was from the TAXUS ATLAS workhorse patient population that received the TAXUS Libert© stent. In the TAXUS ATLAS workhorse trial, predilation was mandatory. There were 247 patients treated with direct stenting versus 543 patients in the predilation control group. What was important about the DIRECT STENT trial? The TAXUS Libert© stent was delivered in 97.6% of the patients where it was attempted. Ultimately, there was successful delivery in 100% of patients, meaning a few patients needed some predilation in order to deliver the stent. The key number is the percentage of patients who were able to receive direct stenting without needing any predilation. If you look at other direct stenting trials, our percentage is actually a very high number. What about TVR and TLR in direct stenting with the Libert©? Very interesting findings. Compared to the predilation control, the TLR rates were reduced by 63% with direct stenting. Obviously because of the large reduction in TLR, the overall TVR was also markedly reduced compared to the control (5% versus 11.2%, p=0.005). Overall MACE was also reduced significantly, occurring in 9.1% of patients in the direct stent group and 14.7% of patients in the predilation control, for a 38% reduction. Keep in mind that all direct stent patients received mandatory IVUS-guided therapy. There were no stent thromboses out to 9 months noted in the direct stent population versus 0.9% in the predilation control. Could the zero thrombosis rate at nine months be a result of less injury to the vessel by direct stenting? I think it could be a number of issues. First, maybe we are not allowing for any geographic miss; that is, any technical issues where we may get little edge dissections that are not fully covered by the stent segment. Second, to be honest, the direct stent population was a somewhat less complex patient population than the control group, because of the patient characteristics and angiographic characteristics that are needed to be successful in a direct stent population. The other important thing to keep in mind is there is a possible benefit to using IVUS in this patient population. IVUS-guided therapy helped us actually to identify any problems and be sure that we had good stent apposition prior to leaving the cath lab. We are also looking at a small number of patients and follow-up is still early at 9 months. How prevalent is direct stenting in the U.S.? I think it varies lab to lab, but I would probably say that somewhere around 30% of the patients that are treated with stents are treated with direct stenting. IVUS isn't necessarily automatically paired with direct stenting in real-world practice. That's correct, but IVUS is very important for good technical results in direct stent patients. Does IVUS contribute cost to the procedure? We looked at direct stenting from the standpoint of procedure times (patients in the predilation control that received IVUS, and those patients in the direct stent population, all of whom received IVUS). The procedure times for the direct stent population were significantly less (p-value = 0.02). The IVUS subset from the predilation group was 57.5 minutes, versus 52.8 minutes for the direct stent patients. There was more post-dilation performed in the direct stent population. Overall, I do think that the lower procedure times and the lack of needing a predilation balloon ultimately may lead to a cost-benefit for direct stenting versus predilation. Certainly IVUS adds cost and time to the procedure, but this may be offset by improved outcomes. What are your plans for the future? You mentioned the small vessel ATLAS trial might be presented at PCR in May 2007. Yes, we still have to present the other two expansion trials, the small vessel and long lesion. Right now we have a manuscript pending review in JACC for the TAXUS ATLAS workhorse population and we are in preparation to put the DIRECT STENT data into manuscript format. At future meetings we will be presenting small vessel and long lesion data, and then longer-term data with regards to the DIRECT STENT and workhorse population. Dr. Turco can be contacted at mturco (at) adventisthealthcare. com

NULL