Impact of Sirolimus-Eluting Stents on Outcome in Diabetic Patients
A SIRIUS (SIRolImUS-coated Bx Velocity balloon-expandable stent in the treatment of patients with de novo coronary artery lesions) Substudy
Background— Randomized clinical trials have shown that a sirolimus-eluting stent significantly reduces restenosis after percutaneous coronary revascularization. Diabetic patients are known to have a higher risk of restenosis compared with nondiabetic patients. The purpose of this analysis was to determine the impact of sirolimus-eluting stents on outcomes of diabetic compared with nondiabetic patients.
Methods and Results— The SIRIUS (SIRolImUS-coated Bx Velocity balloon-expandable stent in the treatment of patients with de novo coronary artery lesions) trial is a randomized, double-blind study that compared sirolimus-eluting and bare metal stent implantation in 1058 patients with de novo native coronary artery lesions. Diabetes mellitus was present in 279 (26%) patients (diabetes mellitus group, 131 patients received sirolimus-eluting stents and 148 patients received bare metal stents) and was absent in 778 patients (no–diabetes mellitus group, 402 patients received sirolimus-eluting stents and 376 patients received bare metal stents). At 270 days, target lesion revascularization was reduced in diabetic patients from 22.3% with bare metal stents to 6.9% with sirolimus-eluting stents (P<0.001) and in nondiabetic patients from 14.1% to 2.99% (P<0.001), respectively. Major adverse cardiac events were reduced in diabetic patients from 25% with bare metal stents to 9.2% with sirolimus-eluting stents (P<0.001) and from 16.5% to 6.5% (P<0.001) in nondiabetic patients, respectively.
Conclusions— Implantation of sirolimus-eluting stents compared with bare metal stents in de novo coronary lesions reduces major adverse cardiac events in patients with and without diabetes mellitus. However, among patients receiving sirolimus-eluting stents, there remains a trend toward a higher frequency of repeat intervention in diabetic patients compared with nondiabetic patients, particularly in the insulin-requiring patients.
Received February 18, 2004; revision received March 23, 2004; accepted March 23, 2004.
With increasing prevalence, diabetes is rapidly growing into a global public health problem.1 Cardiovascular disease is a major consequence of this chronic condition,2 and a critical issue facing physicians worldwide today is the choice of coronary revascularization procedures in treating these patients. The importance of aggressive medical therapy to achieve recommended glycemic control targets and the management of usual risk factors in these patients cannot be overemphasized, regardless of the revascularization therapy selected.3 Outcomes after percutaneous coronary interventions (PCI) and CABG are worse in patients with diabetes mellitus (DM) compared with those without DM.4–8
However, the role of PCI has been evolving over the last 2 decades. Coronary stent implantation has improved clinical outcomes of diabetic patients compared with balloon angioplasty, primarily as the result of a reduction in acute complications and restenosis rate.9 However, recurrence in diabetic patients undergoing stenting remains a problem, particularly in the insulin-using patients.10–13 The introduction of drug-eluting stents has reignited enthusiasm that PCI would again be at the forefront as a contender in revascularization strategies for diabetic patients. Sirolimus (rapamycin), a cytostatic macrocyclic lactone with both antiinflammatory and antiproliferative properties, when delivered from a polymer encapsulated stent, has been shown to significantly reduce restenosis in several prospective, multicenter, randomized trials.14–16 The purpose of the current analysis was to evaluate outcomes of diabetic patients in the SIRIUS trial to understand whether the overall positive results for sirolimus-eluting stents could be extended to patients with diabetes mellitus.
Study Design and Patient Eligibility
The SIRIUS trial is a randomized, double-blind study that was approved by the United States Food and Drug Administration. All 53 investigational sites received approval from their hospital institutional review boards, and written informed consent was obtained from all patients.
Eligible patients had a history of stable or unstable angina and signs of myocardial ischemia. A single de novo target lesion in a native coronary artery between 15 and 30 mm in length was treated. Eligible patients were randomly assigned (per site) in a 1:1 double-blind manner to either a sirolimus-eluting stent (SES) or a bare metal stent (BMS) (Bx VELOCITY, Cordis Corp, Johnson & Johnson). Randomization blocks were created and were stratified by clinical site. Other details on inclusion and exclusion criteria can be found in previous publications.15
Coronary Stent Procedure
Before and after the index procedure, all patients received oral aspirin (325 mg daily) and oral clopidogrel (300 to 375 mg loading dose, 24 hours before and 75 mg daily for 3 months). During the procedure, intravenous heparin boluses were administered. The use of intravenous glycoprotein IIb/IIIa inhibitors was at the operator’s discretion. Lesions were with standard interventional techniques, including mandated balloon predilation before stent placement. One or two of the assigned stents were used to treat the target lesion. The sirolimus-eluting and bare metal stents (available in 8- and 18-mm lengths and 2.5-, 3.0-, and 3.5-mm diameters) were identical in appearance. The sirolimus-eluting stent contained 140 μg of sirolimus per square centimeter of stent surface area within a 5- to 10-μm-thick copolymer matrix designed to release ≈80% of the total sirolimus dose in 30 days. Stent assignment was blinded to both the physician operator and the patient.
Data Collection, Follow-Up, and Core Laboratory Analyses
All data were submitted to a data coordinating center (CDAC/Harvard Clinical Research Institute, Harvard Medical School, Boston, Mass). Clinical follow-up was obtained in all patients by the research coordinators at each site at 30 days, 3 months, 6 months, and 9 months. All clinical end points were adjudicated by an independent and blinded clinical events committee.
Coronary angiograms, obtained at baseline, at the completion of the stent procedure, and at follow-up (8 months) were submitted to the angiographic core laboratory (Brigham and Women’s Angiographic Core Laboratory, Boston, Mass) and were analyzed through the use of a computer-based system (Medis). Binary restenosis was defined as a >50% diameter stenosis of the target lesion. Late lumen loss was defined as the difference between the minimum lumen diameter at the completion of the stent procedure and during follow-up. Quantitative angiographic measurements of the target lesion were made at the “in-stent” zone (only the stented segment) and at the “in-segment” zone (the stented segment as well as the margins 5 mm proximal and distal to the stent).
Study End Points
The primary end point of this study was target vessel failure, defined as the occurrence of any of the following up to 9 months after the index procedure: cardiac death, Q-wave and non–Q-wave myocardial infarction, or target vessel revascularization (emergency or elective CABG or repeat PTCA). Secondary clinical end points included all-cause death, target lesion revascularization (TLR, clinically driven CABG or repeat PTCA for restenosis or closure of the target lesion), and stent thrombosis. All major adverse cardiac events (MACE) were determined for the time periods in-hospital, out-of-hospital, and cumulative for 270 days after stent placement.
The effectiveness analysis and the safety evaluation were performed on an intention-to-treat population that includes all randomly assigned patients, except for those who were deregistered. The treatment group differences were evaluated with the ANOVA or Wilcoxon rank sum scores for the continuous variables when appropriate. The Cochran-Mantel-Haenszel statistics, controlling for site groups, were used for the categorical variables. To identify factors that might be related to angiographic restenosis and TLR, logistic regression models were used. All statistical analyses were performed with the use of the SAS system (version 6.12), and reported probability values were 2-sided.
For the purpose of this analysis, patients were divided into 2 groups: the DM group (279 patients, 131 patients/lesions received sirolimus-eluting and 148 patients/lesions received bare metal stents) and the no-DM group (778 patients, 402 patients/lesions received sirolimus-eluting and 376 patients/382 lesions received bare metal stents).
Diabetic Versus Nondiabetic Patients: Baseline Characteristics and Procedural Variables
Diabetic patients were more often female (39% versus 25%, P<0.001), had a higher prevalence of hypertension (79.5% versus 63.4%, P<0.001) and triple-vessel disease (19.7% versus 12.7%, P=0.006), and had a trend toward lower ejection fraction (55±11% versus 56±10%, P=0.089). Also, diabetic patients had smaller vessels compared with nondiabetic patients (2.75±0.44 mm versus 2.82±0.47 mm, P=0.03), but there were no significant differences in lesion length (14.52±6.07 mm versus 14.36±5.69 mm, P=0.67) and other baseline quantitative and qualitative angiographic variables.
Diabetic patients in the BMS and SES groups had similar baseline vessel size (2.77±0.45 mm versus 2.73±0.44 mm, P=0.46), lesion length (14.93±6.64 mm versus 14.07±5.36 mm, P=0.24) and attained similar postprocedure results. Likewise, nondiabetic patients in the BMS and SES groups also had also similar vessel size, lesion length, and postprocedure results. However, diabetic patients assigned to the SES group had a longer stent-to-lesion ratio than those assigned to the BMS group (1.7±0.6 versus 1.6±0.5 mm, P=0.054).
Angiographic Outcome at Eight Months
Eight-month follow-up angiography was performed in 186 of 279 diabetic patients (67%) and in 516 of 778 nondiabetic patients (66%). Late loss (in-stent and in-segment) was significantly lower in the SES compared with the BMS group in both diabetic and nondiabetic patients (Table 1). This observation was further confirmed in diabetic patients at follow-up intravascular ultrasound, which demonstrated significantly lower percent volume obstruction in the SES group compared with the BMS group (1.84% versus 35.3%, P<0.0001).
In-stent binary angiographic restenosis was significantly reduced in the SES compared with the BMS group in both diabetic and nondiabetic patients (Table 2). Similarly, in-lesion binary angiographic restenosis was also significantly reduced by the SES in diabetic patients (relative reduction 65%) and in nondiabetic patients (relative reduction 80%). Restenosis pattern was predominantly focal in the SES arm, both in diabetic and nondiabetic patients (Table 2).
Clinical Outcome: In-Hospital, Thirty Days, and Nine Months
No statistically significant differences were observed in in-hospital events and in early and late stent thrombosis among treatment groups in both diabetic patients and nondiabetic patients (Table 3).
As shown in Table 4, the primary end point of target vessel failure was significantly reduced in the SES group in both diabetic and nondiabetic patients. This reduction was primarily due to the lower rate of clinically driven repeat revascularization of the target lesion seen in the SES group in both diabetic and nondiabetic patients.
The event-free survival analysis demonstrates that diabetic patients in the SES group had significantly higher event-free (TLR and MACE) survival rates than did diabetic patients in the BMS group (Figure 1). However, compared with nondiabetic patients, diabetic patients in the SES group had a trend toward lower TLR-free survival, but this difference did not reach statistical significance (Figure 2). As shown in the multivariate logistic regression model (Table 5), diabetes mellitus continues to be an independent predictor of the need for repeat revascularization at 9 months.
Insulin-Requiring Versus Non–Insulin-Requiring Diabetic Patients
As shown in Table 6, diabetic patients assigned to the SES and BMS groups had similar baseline vessel size and lesion length and attained similar postprocedure results in both the insulin-requiring and non–insulin-requiring patients. Binary in-stent angiographic restenosis was significantly reduced by the SES in both groups. However, Although in-lesion restenosis was also significantly reduced in the non–insulin-requiring patients (49.3% versus 7.7%, P<0.001) this reduction was not statistically significant in the insulin-using patients (50% versus 35%, P=0.38) because of an unexpectedly high frequency of peristent restenosis (Table 6).
The findings of this diabetic subanalysis of the SIRIUS trial demonstrate (1) safety is preserved with implantation of the sirolimus-eluting stent in diabetic patients as well as in nondiabetic patients, as reflected by the low rate of in-hospital MACE and stent thrombosis; (2) sirolimus-eluting stents reduced the relative incidence of in-lesion angiographic restenosis in diabetic patients by 65% (from 50.5% to 17.6%) and in nondiabetic patients by 80% (from 30.7% to 6.1%); (3) sirolimus-eluting stents transformed angiographic recurrence from diffuse to focal pattern in both diabetic patients and nondiabetic patients; (4) sirolimus-eluting stents led to a 63% relative reduction of MACE in diabetic patients (25% versus 9.2%, P<0.001) compared with a 61% relative reduction in nondiabetic patients (16.5% versus 6.5%, P<0.001). Therefore, the angiographic and clinical outcomes of the SIRIUS trial support the concept that sirolimus-eluting stents significantly reduce the added risk for restenosis in diabetic patients after PCI.
Despite the highly significant reductions in late loss and restenosis in diabetic patients receiving the sirolimus-eluting stent, the absolute late loss and restenosis remain higher than that observed in nondiabetic patients in both the sirolimus and bare metal stents. However, the pattern of recurrence in diabetic patients has shifted from a diffuse pattern to a focal one that is more amenable to PCI.17
Oral- and diet-controlled diabetic patients had angiographic and clinical results similar to nondiabetic patients, with major improvement in all end points favoring sirolimus-eluting stents. Insulin-receiving diabetic patients also revealed a marked reduction in in-stent restenosis of 77.4% compared with bare metal stents, supporting the efficacy of sirolimus in suppressing neointimal hyperplasia in these patients. However, when the in-segment analysis is considered, these patients had reduced effectiveness of sirolimus-eluting stents, predominantly because of excessive peristent restenosis. The insignificant effect of sirolimus-eluting stents on in-segment restenosis reduction in insulin-receiving diabetic patients must be interpreted in the context of the following limitations: (a) the small sample size; (b) the incomplete angiographic follow-up; and (c) the high incidence of peristent restenosis in the SIRIUS trial.15 This excess in peristent restenosis was not observed in other clinical trials of sirolimus-eluting stents.14,16,18 This phenomenon may be due to either proximal balloon injury in nonstented zones or incomplete lesion coverage by the stent. However, it is also possible that the effect of sirolimus-eluting stents in insulin-using patients is biologically attenuated. Additional prospective studies are needed to better understand the optimal treatment approach in this problematic patient subset.
This study should be interpreted with the recognition that it is a nonrandomized, subset analysis, which may limit the conclusiveness of its findings.
Implantation of sirolimus-eluting stents in patients with de novo coronary lesions dramatically reduces the rate of TLR in patients with and those without diabetes mellitus. From a clinical perspective, a diabetic patient who is treated with a sirolimus-eluting stent is at lower risk for repeat revascularization at 9 months than a nondiabetic patient who is treated with a bare metal stent. Future prospective, randomized, clinical trials comparing drug-eluting stents with coronary arterial bypass surgery are needed to determine the optimal revascularization strategy in diabetic patients with multivessel disease.
This study was supported by a grant from Cordis, a Johnson & Johnson Company.
This article originally appeared Online on May 3, 2004 (Circulation. 2004;109:r101–r106).
Freeman DJ, Norrie J, Sattar N, et al. Pravastatin and the development of diabetes mellitus: evidence for a protective treatment effect in the West of Scotland Coronary Prevention Study. Circulation. 2001; 103: 357–362.
Stein B, Weintraub WS, Gebhart SP, et al. Influence of diabetes mellitus on early and late outcome after percutaneous transluminal coronary angioplasty. Circulation. 1995; 91: 979–989.
Magee MJ, Dewey TM, Acuff T, et al. Influence of diabetes on mortality and morbidity: off-pump coronary artery bypass grafting versus coronary artery bypass grafting with cardiopulmonary bypass. Ann Thorac Surg. 2001; 72: 776–780;discussion 780–781.
Niles NW, McGrath PD, Malenka D, et al. Survival of patients with diabetes and multivessel coronary artery disease after surgical or percutaneous coronary revascularization: results of a large regional prospective study: Northern New England Cardiovascular Disease Study Group. J Am Coll Cardiol. 2001; 37: 1008–1015.
Lemos PA, Arampatzis CA, Hoye A, et al. Restenosis after sirolimus-eluting stent implantation: long-term evaluation following repeat percutaneous intervention. J Am Coll Cardiol. 2004; 43: 71A.
Moses JW, Leon MB, Popma JJ, et al. Matched comparison of direct stenting to predilatation with the sirolimus-eluting Bx Velocity stent. J Am Coll Cardiol. 2004; 43: 97A.