Brachytherapy and Drug-eluting Stents

Defining the anatomy. A randomized study published in the Journal of the American College of Cardiology last year involved low-risk patients who underwent stress testing and coronary angiography.¹ Naturally, it found that coronary angiograms detected more disease than stress tests. That point, however, was not really the heart of the study. The recidivism figures in this study were very important: 31% for those patients who only underwent stress testing versus 9% in the catheterization group. Thus, once the patient is in the emergency room with chest pain, it behooves the physician to define the anatomy, if only to keep the negatives at home! In-stent restenosis. Fortunately, with the advent of stenting, we have narrowed the problem down by eliminating remodeling and stent recoil and are now dealing solely with intimal hyperplasia. In-stent restenosis is predominantly diffuse and simply does not yield well to device therapy. If there is proliferative disease or a total occlusion, the restenosis rates range from 50–80%. Yet despite our ministrations with all types of ablative techniques, re-stenting, and angioplasty, there has thus far been no mechanical solution to this problem. Brachytherapy for in-stent restenosis. Treating diffuse in-stent restenosis was frustrating until the advent of brachytherapy, which was the first effective therapy aimed at eliminating intimal hyperplasia. Three commercially approved brachytherapy systems are currently available: 1) the CheckMate gamma system by Cordis; 2) the beta-emitting beta-cath system by Novoste; and 3) the Guidant afterloading system which conforms to the high-dose rate treatments. Ron Waksman has helped to pioneer this system. Brachytherapy has proven to be a highly effective therapy and is the first to really have a measurable impact on reducing subsequent events in patients who suffer from in-stent restenosis (Figure 1). Ron Waksman and Patrick Serruys realized, however, that a dark side of this therapy had emerged. In retrospect, it made perfect sense that inhibiting neointimal hyperplasia in these stents might actually cause late thrombosis. The early trials, including SCRIPPS and GAMMA-I, featured the Cordis CheckMate system and showed thrombosis rates ranging from 3% to as high as 9%.^2–4 Importantly, these were late events; they were unpredictable and occurred up to several months after the treatment, unlike subacute thrombosis in stenting where this problem was essentially eliminated after about thirty days. The patient was subjected to risk eternally. Gamma therapy nearly went by the wayside because the payback for treatment of the restenosis was, in retrospect, an unacceptable rate of thrombosis and an increase in both large infarction and mortality rates. Also, out-of-hospital subacute thrombosis is not a benign event, as it has phenomenally high mortality and Q-wave myocardial infarction rates for a MACE of 70–80% (Figure 2). Fortunately, further analyses helped to define and eliminate the problem. If a new stent was not placed in these patients, the actual late occlusion rate was no different between radiation versus placebo. That fact that a problem occurred when new stents were placed made perfect sense in retrospect. Also, when these studies were launched, we were strongly encouraged to perfect the result and line these vessels with stents. In more complete analyses of multivariate predictors, symptomatic late thrombosis was predominantly predicted by both long lesions and new stent use. In response to this, two major trials were launched: SCRIPPS III, conducted by Scripps and Lenox Hill, and the WRIST trial in Washington, D.C., which attempted to address late thrombosis by limiting the use of new stents and extending antiplatelet therapy to six months with no new stents and at least one year of therapy with new stents. The SCRIPPS trial, involving approximately 500 patients, resulted in an acute thrombosis in one patient, a subacute thrombosis rate of 0.4% — which appears to be a normal rate — and only one late thrombosis case documented on recent follow-up. Importantly, we were able to discontinue the drug in only a handful of patients because they did not want to stop taking the drug. Ron Waksman’s data from the WRIST PLUS trial supported this, extending clopidogrel therapy to six months.⁵ He found that the late thrombosis rate was considerably reduced, comparable to placebo. A recently published article by Ron Waksman in Circulation,⁶ provided an intriguing look at an extension of clopidogrel therapy to twelve months that showed a significant reduction in MACE. From the standpoint of brachytherapy, it is safe to say that the safety of six months of clopidogrel therapy has been established if a new stent is not being implanted. However, based on currently available data, if a new stent has been implanted, indefinite treatment is the only recommendation that can be made from a safety standpoint (Figure 3). In terms of clopidogrel and surgery and early loading, we are headed toward an era where surgery will no longer be necessary! Radiation therapy was a blunderbuss approach, but with the progress being made in cellular biology, there are much more elegant solutions for the inhibition of the cell’s cycle. Drug-eluting stents allow for the targeting of specific sites and for the inhibition of the cell’s cycle at specific points to inhibit the proliferative process (Figure 4). The Bx Velocity™ drug-eluting stent (Cordis J&J), which currently has the largest data set, is coated with two polymers and sirolimus and has been used in several trials. Two-year follow-up data are now available and three-year follow-up data will soon become available on a relatively small but intensively studied cohort of patients. A three-year-old study involving 30 patients from Sao Paolo shows a persistent inhibition of restenosis, the maintenance of MLD, and an absolute inhibition of neointimal hyperplasia in the stent (Figure 5).⁷ These data were confirmed in the RAVEL randomized trial, in which the restenosis rate in this relatively simple cohort of patients with short lesions who received sirolimus-eluting stents was 0%, with virtually no late-loss, as opposed to a standard late-loss of 0.8. These results represent up-front de novo inhibition of the intimal hyperplasia. This also translates into a 94% event-free survival rate at one year in the treatment arm as compared to a 70% rate in the standard stent arm. Luminology, thus, does occasionally translate into positive clinical events. Tomorrow we will see the complete cohort of patients from the SIRIUS trial. The first 400 patients were discussed at the Paris meeting. The data from the SIRIUS trial are quite impressive.⁸ SIRIUS included a more complex lesion subset involving smaller vessels: 2.5 to 3.5 mm maximum; visual lesion length of at least 15–30 mm. A 90%+ inhibition rate of in-stent restenosis was once again found in the RAVEL trial. Some residual restenosis was found at the margins, yet there was an impressive 72% reduction of in-segment restenosis (Figure 6). All events were comparably reduced: target lesion revascularization was 72% and target vessel revascularization was 61% (Figure 7). In all drug-eluting stent trials, antiplatelet therapy is prolonged from two to six months. This has been incorporated into virtually every randomized trial and will certainly be incorporated into the labeling indication for these stents as they are approved for use. When looking at these trials, however, we must bear in mind that a certain amount of "background noise" has not been strongly emphasized. The HCRI’s large 6,000-patient stent database shows that beyond 30 days, there actually is a late thrombosis rate of 1–1.6% in stents.⁹ It is important to note that these are predominantly simple lesions and trial stents. The late thrombosis problem has thus been underestimated in the past. In the SIRIUS trial, the thrombosis rate in this group was not increased with three months of antiplatelet therapy.⁸ However, the effect was not uniform; one must keep in mind that there are specific anatomic subsets and, of course, specific drugs. At higher doses of a drug such as paclitaxel, there is inflammation and fibrin deposition as well. A case in point is the SCORES trial in which the six-month restenosis rate with a taxane-eluting stent was profoundly reduced.¹⁰ The trade-off, however, was a marked increase in late thrombosis rates. Importantly, in ASPECT, which is similar to the DELIVER trial, a high dose-rate stent showed a major reduction in restenosis, but in the small twelve-patient group that did not receive dual antiplatelet therapy, three patients experienced late stent thrombosis, which powerfully highlights the importance of dual antiplatelet therapy.^11,12 Despite the positive results of the SIRIUS trial, however — even with sirolimus — if the envelope is pushed too far, thrombosis can occur. For example, a patient with in-stent restenosis who had several stents placed with excellent final results, experienced a late thrombosis six weeks after stopping dual antiplatelet therapy. In conclusion, dual antiplatelet therapy will be an important component in the safety of drug-eluting stents. And as we begin to treat more complex lesion subsets, it will undoubtedly require longer courses of treatment than the simple clinical and angiographic subsets observed in the current clinical trials. DISCUSSION Elliot Rapaport: At about six months out in the angiographic substudy of RAVEL, approximately 20% of the patients showed a lack of adherence of the stent to the wall. Are there any follow-up data regarding what percentage of those patients experienced late thrombosis? Jeff Moses: Actually, you will hear more about that tomorrow and over the subsequent days, but let me just say that none of these thus far have translated into any adverse clinical events, including the SIRIUS trial. The RAVEL cohort is being followed systematically; those patients are undergoing follow-up angiograms. Furthermore, the FDA has mandated that all of these trials, including SIRIUS, conduct five-year follow-ups. The answer, therefore, is that we don’t know. It’s another argument for vigilance and, in my opinion, for prolonged therapy. Spencer King: The two-year follow-up report of the Sao Paolo study essentially states that nothing happened to any of these patients, but when I asked Eduardo about it, he said that some events had occurred — just not any related to the target lesion. Jeff Moses: First of all, these stents do not confer immortality, though we wish they did. Yes, there have been events such as one case of an apparent vulnerable plaque at the margin proximal to a stent; it was definitely patent at 12-month follow-up and at 14 months, a thrombosis occurred. Whether the drug itself contributed to the vulnerability by inhibiting collagen synthesis is a matter of speculation, but it was recorded as a thrombosis. The reason you are seeing two-year data is because they are phased in that manner. In Rotterdam, they are publishing data at eighteen months, and in Sao Paolo, at two years. They could merge these data, but it wouldn’t really change the results. Jeff can speak about that in more detail. Jeffrey Popma: This begs the question about duration of treatment. We are saying treatment should be given for three to six months, but the guidelines say that treatment should be continued indefinitely in many of these patients. Why are we even talking about stopping treatment at three months or at all? Jeff Moses: There are two very powerful issues here: protracted therapy, which has to do with plaque vulnerability, and recidivism, both in the target lesion and, of course, in other lesions — which is an entirely separate issue. In my view, the solution will be easy: every ACS patient should receive a drug-eluting stent and protracted antiplatelet therapy. I foresee this all collapsing into one simple paradigm.