Author + information
- Received November 6, 2003
- Revision received January 6, 2004
- Accepted January 12, 2004
- Published online March 17, 2004.
- Erick Schampaert, MD*,* (, )
- Eric A. Cohen, MD†,
- Michael Schlüter, PhD‡,
- François Reeves, MD§,
- Mouhieddin Traboulsi, MD∥,
- Lawrence M. Title, MD¶,
- Richard E. Kuntz, MD, MSc, FACC#,
- Jeffrey J. Popma, MD, FACC#,
- the C-SIRIUS Investigators
- ↵*Reprint requests and correspondence:
Dr. Erick Schampaert, Hôpital du Sacré-Coeur de Montréal, 5400 Bl. Gouin O., Montréal, Québec, Canada H4J 1C5.
Objectives We assessed the safety and effectiveness of the sirolimus-eluting stent (SES) in treating single de novo long lesions in small native coronary arteries compared to an identical bare metal stent (BMS).
Background The SES was previously demonstrated to reduce restenosis significantly. However, patients with long lesions in small vessels have not been well studied and may define a group at very high risk.
Methods The Canadian Study of the Sirolimus-Eluting Stent in the Treatment of Patients With Long De Novo Lesions in Small Native Coronary Arteries (C-SIRIUS) was a multicenter, randomized, double-blind trial comparing SES versus identical BMS. The primary end point was in-stent minimal lumen diameter (MLD) at eight months. Secondary end points included angiographic restenosis at 8 months, target lesion revascularization (TLR), and major adverse cardiac events (MACE) at 270 days.
Results A total of 100 patients were enrolled at eight Canadian sites. The in-stent MLD at eight months was 2.46 ± 0.37 mm in the SES compared with 1.49 ± 0.75 mm in the BMS (a 65% increase, p < 0.001). Angiographic restenosis occurred in 1 of 44 SES patients (2.3%, with no in-stent restenosis) and in 23 of 44 BMS patients (52.3%, p < 0.001). At 270 days, there were two clinically driven TLRs in the SES (4%) and nine in the BMS (18%, p = 0.05). The Kaplan-Meier estimate of freedom from MACE at 270 days was 96.0% for SES patients and 81.7% for BMS patients (p = 0.029).
Conclusions Patients with long lesions in small vessels are at very high risk of restenosis. In these patients, the SES dramatically reduces the risk of restenosis at eight months, translating into an excellent clinical outcome at nine months.
Stent implantation has become the standard percutaneous coronary intervention (PCI) (1–3). However, in-stent restenosis (ISR) within three to eights months has continued to limit the long-term success of this therapy (4). Neointimal hyperplasia has been identified as the main mechanism of ISR (5,6). Predictors of ISR include a reference vessel diameter smaller than 3.0 mm, lesion lengths above 10 mm, and diabetes mellitus (7–9). Although intracoronary brachytherapy is available to treat established ISR (10), stents eluting pharmaceutical agents capable of suppressing neointimal hyperplasia represent a promising approach to prevent ISR (11).
Preclinical experiments of such an agent, sirolimus (rapamycin), suggested efficacy and safety (12,13). A powerful natural immunosuppressive macrocyclic lactone, it inhibits cytokine-mediated proliferation and migration of lymphocytes and smooth muscle cells (14). Incorporated into a biocompatible non-erodable stent polymer, sirolimus is released into the stented vessel segment over a period of 90 days.
Four clinical trials with the sirolimus-eluting stent (SES) have enrolled patients with de novo lesions in native coronary arteries, with a pattern of increasing lesion complexity (determined by vessel size and lesion length) in successive trials. The “First-in-Man” study (15)restricted treatment to the implantation of a single 18-mm stent in vessels 3.0 to 3.5 mm in diameter, guided by intravascular ultrasound. The subsequent Randomized Study with the Sirolimus-Coated Bx-VELOCITY Balloon-Expandable Stent in the Treatment of Patients with De Novo Native Coronary Artery Lesions (RAVEL) (16)included vessels 2.5 to 3.5 mm in diameter, but still to be covered with a single stent. The pivotal Sirolimus-Eluting Stent in De Novo Native Coronary Lesions (SIRIUS) trial (17)enrolled patients with longer lesions of 15 to 30 mm, again in 2.5 to 3.5 mm vessels. Finally, in E-SIRIUS (18), patients with a target lesion length of 15 to 32 mm in smaller vessels 2.5 to 3.0 mm in diameter, were randomized to the SES versus a bare metal stent (BMS) of identical architecture.
During the same period, in Canada we conducted a randomized, multicenter trial based on the same protocol as E-SIRIUS, targeting the same patients with long lesions, potentially requiring multiple stents, in small coronary arteries—all conditions that are known to increase the risk of restenosis. Contemporary interventional techniques, including direct stenting, were allowed.
Patients and methods
This study was a randomized, double-blind trial involving eight Canadian teaching hospitals (Appendix).Patients were at least 18 years old, with documented angina pectoris (Canadian Cardiovascular Society angina class 1 to 4), unstable angina (Braunwald classification B and C, I or II), or silent ischemia. Their PCI target had to be a singlede novo lesion in a native vessel, between 15 and 32 mm in length, and with a diameter stenosis of 50% to 99%. The vessel diameter was limited to 2.5 to 3.0 mm. All angiographic criteria were based on visual assessment. Major exclusion criteria were the same as E-SIRIUS (18). The study protocol was approved by the ethics committee at each participating center, and all patients gave written informed consent.
The control stents used were the bare metal Bx-VELOCITY stent (J&J Cordis, Miami Lakes, Florida), which is a balloon-expandable, tubular 316L stainless-steel stent pre-mounted on a monorail balloon-dilation catheter. The study stents, using the same Bx-VELOCITY platform, had a 5-μm coating consisting of a blend of 33% sirolimus and 67% of a non-erodable polymer. The drug-polymer matrix contains 140 μg of sirolimus per cm2of surface area. A drug-free polymer topcoat serves as a control drug release barrier, such that 80% of sirolimus is released within 30 days of implantation and with no residual drug by 90 days. The SES (brand name, Cypher) is visually and radiographically indistinguishable from its uncoated counterpart, allowing for the double-blind design of the trial.
As previously described (18), patients were randomly assigned either to sirolimus or control stents by means of sealed randomization envelopes. Neither the operator nor the patient knew which stent would be implanted.
According to standard care, patients were pre-medicated with 81 to 325 mg of aspirin, begun at least 12 h before the procedure, and clopidogrel, administered as a loading dose of 300 mg before or immediately after the procedure. During the procedure, intravenous boluses of heparin were administered to maintain an activated clotting time in excess of 250 s. The use of glycoprotein IIb/IIIa receptor antagonists was left to the investigator's discretion.
Stent implantation followed current accepted techniques, as described (18). One distinguishing feature of this study, compared with previous trials, was to allow “direct stenting” (without lesion pre-dilation) in centers where this was standard practice. The decision to pre-dilate or not was left to the investigator. Heparin was discontinued immediately after the procedure. Patients were discharged on a regimen of aspirin (81 to 325 mg/day indefinitely) and clopidogrel (75 mg/day) for two months only.
Patients were evaluated clinically at 30, 90, 180, and 270 days. A repeat angiographic study was scheduled after eight months in all patients.
At the outset we distinguished between “in-stent” and “in-lesion” angiographic variables, with the former referring to the vessel segment inside the stent and the latter including the 5-mm vessel segments adjacent to the proximal and distal stent edges. Late luminal loss was defined as the difference between the minimal lumen diameter (MLD) at eight months and the MLD post-procedure.
Study end points
The primary end point of this study was in-stent MLD at eight months, determined by quantitative coronary angiography (QCA). Secondary end points included: eight-month angiographic in-lesion MLD; in-stent and in-lesion angiographic restenosis (a ≥50% diameter stenosis by QCA); major adverse cardiac events (MACE)—a composite end point comprising death, myocardial infarction, emergent coronary artery bypass surgery, and clinically driven repeat target lesion revascularization (TLR)—all at nine months; and TLR at nine months. A clinically driven TLR procedure was defined as one done in response to recurrent angina and/or documented ischemia on noninvasive tests (all recorded prior to repeat angiography), with >50% diameter stenosis by QCA, or >70% diameter stenosis by QCA in the absence of symptoms.
Offline QCA at baseline, post-procedure, and after eight months was performed by an independent core laboratory (Brigham and Women's Hospital Angiographic Core Laboratory, Boston, Massachusetts). All clinical end points were adjudicated by an independent clinical events committee.
Data management and statistical methods
At each participating center, patients' data were prospectively recorded on standard case report forms. Complete data monitoring was performed by an independent clinical research consulting firm (DT Consultant, Montreal, Canada), which forwarded the completed case report forms to the study coordinating center for data entry and analysis. Treatment allocation was unblinded at Harvard Clinical Research Institute after nine-month clinical follow-up for analysis. However, individual patient assignment has remained blinded for unbiased clinical follow-up to five years. All data were held at the study coordinating center, but the authors of this report had full access to them.
Based on the hypothesis that in-stent MLD by QCA at eight months would be 1.6 mm for the control stent and 2.4 mm for the SES, with a common standard deviation of 0.7 mm, detecting this difference with an 80% power and a two-sided alpha error of 5%, would require a sample size of <60 patients (30 patients per study arm). Adjusting for an 80% compliance with eight-month angiographic follow-up, the sample size of 100 patients was judged adequate. All analyses were based on the intention-to-treat principle. Continuous variables are presented as mean value ± SD, with differences between groups assessed by the Student unpaired ttest. Discrete variables are presented as counts and percentages, with differences between groups assessed by the Fisher exact test. The Kaplan-Meier method was used to analyze the occurrence of the composite end point of MACE during the nine-month period of follow-up, with differences between event-free survival curves assessed by log-rank test. Statistical significance was assumed at the 5% level (p < 0.05).
Between November 2001 and April 2002, we enrolled 102 patients. Two patients were randomized (one in each group) but subsequently deregistered because no study stent implantation was attempted (one patient underwent coronary artery bypass grafting after failure to cross the lesion with a guidewire and one patient was withdrawn from the study after failed predilation with standard balloons). Thus, 100 patients entered the trial for end point analysis, with 50 patients receiving an SES and 50 patients receiving uncoated control BMS. Baseline characteristics of the patients (Table 1) were well matched between groups, with 24% of patients having diabetes mellitus.
Device success, defined as the achievement of <50% residual diameter stenosis with the assigned stent, was 100% in both groups.
The mean lesion length as measured by QCA was 13.6 ± 5.8 mm, in vessels with a mean reference diameter of 2.63 ± 0.33 mm. Glycoprotein IIb/IIIa inhibitors were administered to 53% of patients. The average number of stents implanted was 1.5 ± 0.7 per patient, with two or more stents implanted in 40% of patients. This resulted in a mean total stent length of 23.8 ± 8.4 mm and a stent length to lesion length ratio of 1.8 ± 0.8. Direct stenting was performed in 31% of cases. Of the stented lesions, 64% were post-dilated (using a shorter balloon in one-half the cases), with a mean maximum pressure of 17.3 ± 3.4 atm and a nominal balloon to artery ratio of 1.0 ± 0.1.
Immediate post-procedure results were similar for SES and BMS, with a mean in-stent MLD of 2.51 ± 0.29 mm, and residual in-stent and in-lesion stenoses of 5.7% and 17.9%, respectively.
Eight-month angiographic follow-up was available in 88% of patients, with 44 patients in each group. The eight-month in-stent MLD, the primary end point, was significantly greater in the SES group at 2.46 ± 0.37 mm vs. the BMS group at 1.49 ± 0.75 mm (a 65% increase, p < 0.001) (Table 2). The corresponding late luminal loss was reduced by 90%, from 1.02 ± 0.69 mm to 0.12 ± 0.37 mm (p < 0.001) (Table 2). The eight-month in-lesion MLD was also significantly improved in the SES group compared to the BMS group (Table 2). Consequently, the in-lesion late luminal loss was significantly reduced in the SES patients (Table 2), with an effect demonstrated both at the proximal and distal stent edges (Fig, 1). Angiographic restenosis occurred in 1 of 44 SES patients (2.3%), with no ISR, and in 23 of 44 BMS patients (52.3%, p < 0.001). The single SES patient with in-lesion restenosis had a 58% stenosis proximal to the SES stent. The relative reductions in eight-month binary restenosis associated with use of the SES were thus 96% within the lesion and 100% within the stent. No aneurysms were seen.
Major adverse cardiac events at 270 days for all 100 patients are listed in Table 3. There were no deaths and no Q-wave myocardial infarctions. Stent thrombosis occurred in one patient in each group (on day 8 in an SES patient after an initially successful intervention involving three stents in a small right coronary artery, and on day 57 in a BMS patient), requiring TLR in both cases. Eight additional clinically driven TLR procedures occurred in the BMS group and one such procedure in the SES group, for a total TLR rate at nine months of 18% and only 4%, respectively (p = 0.05). Only two BMS patients sustained a non-clinically driven TLR following repeat angiography at eight months. Consequently, the Kaplan-Meier estimate of freedom from MACE at 270 days was 96.0% for SES patients and 81.7% for BMS patients (p = 0.029) (Fig. 2).
Compared with previous trials (15–17), the patients enrolled in this Canadian multicenter controlled trial, along with the E-SIRIUS patients (18), had a higher clinical-risk profile for restenosis: long target lesions, small target vessels, multiple stents in 40% of cases, and a mean implanted stent length of 23.8 mm. This is reflected in the progressive increase of the in-lesion restenosis rate of the BMS group: 26.6% in RAVEL (16), 36.3% in SIRIUS (17), 42.3% in E-SIRIUS, and 52.3% in our trial. Despite the higher risk profile of our patients, the eight-month in-stent and in-lesion MLDs were well maintained in the SES patients. Late luminal loss at the proximal and distal edges of the SES was reduced by 77% and 100%, respectively, compared with the BMS, suggesting a true protective effect of the SES at the stent margins. Consequently, the angiographic restenosis rate following SES implantation was only 2.3%, with no ISR. These findings confirm the efficacy of the SES to prevent restenosis, as observed in all previous trials (15–18). Thus, the need for a clinically driven revascularization procedure, prompted by recurrent angina and/or documented ischemia, fell from 18% in BMS patients to 4% in SES patients. This means that for every 1,000 patients undergoing stent implantation for a native coronary artery lesion of the type included in this study, 140 patients may be spared from clinical restenosis and repeat intervention at nine months by initial treatment with SES.
Because the same SES was used in four randomized trials, certain observations regarding the SES patients across the trials may be relevant. The somewhat higher in-lesion restenosis rate observed in SIRIUS (8.9%), compared with RAVEL (0%), E-SIRIUS (5.9%), and C-SIRIUS (2.3%), was associated with more proximal margin restenosis: 5.8% in SIRIUS (especially in patients with the smallest vessels), compared with 0%, 2.1%, and 2.3% in RAVEL, E-SIRIUS, and C-SIRIUS, respectively. This difference could be related to subtle but more frequent or pronounced proximal edge trauma during pre-dilation, stent implantation, or post-deployment dilation, overwhelming the protective effect at the SES proximal margin. The use of direct stenting in E-SIRIUS and C-SIRIUS may also have limited proximal edge trauma and subsequent restenosis in some patients. Taken together with the observation that the total stent length to lesion length ratio of 1.8 in our trial was identical to that in RAVEL and E-SIRIUS suggests that meticulous (and generous) coverage of all the injured and diseased vessel area appears to be desirable.
The C-SIRIUS demonstrated that patients with long lesions in small vessels are at very high risk of restenosis. In these patients, the SES dramatically reduced the risk of restenosis, with no ISR at eight months, translating into an excellent clinical outcome at nine months. Ongoing follow-up will evaluate the durability of these clinical benefits over the next four years.
Investigators in the Canadian Study of the Sirolimus-Eluting Stent in the Treatment of Patients With Long De Novo Lesions in Small Native Coronary Arteries (C-SIRIUS): E. A. Cohen (Sunnybrook and Women's College Health Sciences Centre, Toronto), F. Reeves (Centre Hospitalier de l'Université de Montréal, Pavillon Notre-Dame), E. Schampaert (Hôpital du Sacré-Coeur de Montréal), D. Traboulsi (Calgary Heart Centre) L. Title (Queen Elizabeth II Health Science Centre, Halifax), D. Raco (Hamilton General Hospital), S. Plante (Hôpital Laval, Québec), and R. Mildenberger (Victoria Heart Centre). Study coordination: R. E. Kuntz (Harvard Clinical Research Institute [HCRI], Boston, Massachusetts).
☆ This study was sponsored by Cordis Canada, a Johnson & Johnson Company, the manufacturer of the study stents.
- bare metal stent
- in-stent restenosis
- major adverse cardiac events
- minimal lumen diameter
- percutaneous coronary intervention
- quantitative coronary angiography
- sirolimus-eluting stent
- target lesion revascularization
- Received November 6, 2003.
- Revision received January 6, 2004.
- Accepted January 12, 2004.
- American College of Cardiology Foundation
- Serruys P.W.,
- de Jaegere P.,
- Kiemeneij F.,
- et al.
- Kastrati A.,
- Schömig A.,
- Dietz R.,
- Neumann F.J.,
- Richardt G.
- Gordon P.C.,
- Gibson C.M.,
- Cohen D.J.,
- Carrozza J.P.,
- Kuntz R.E.,
- Baim D.S.
- Kornowski R.,
- Hong M.K.,
- Tio F.O.,
- Bramwell O.,
- Wu H.,
- Leon M.B.
- Hoffmann R.,
- Mintz G.S.,
- Dussaillant G.R.,
- et al.
- Kastrati A.,
- Schömig A.,
- Elezi S.,
- et al.
- Bauters C.,
- Hubert E.,
- Prat A.,
- et al.
- Sousa J.E.,
- Serruys P.W.,
- Costa M.A.
- Gallo R.,
- Padurean A.,
- Jayaraman T.,
- et al.
- Suzuki T.,
- Kopia G.,
- Hayashi S.,
- et al.
- Marx S.O.,
- Marks A.R.
- Sousa J.E.,
- Costa M.A.,
- Abizaid A.C.,
- et al.
- Morice M.C.,
- Serruys P.W.,
- Sousa J.E.,
- et al.