Quantitative Assessment of Angiographic Restenosis After Sirolimus-Eluting Stent Implantation in Native Coronary Arteries
Background— Sirolimus-eluting stents (SESs) reduce angiographic restenosis in patients with focal, native coronary artery stenoses. This study evaluated the usefulness of SESs in complex native-vessel lesions at high risk for restenosis.
Methods and Results— Angiographic follow-up at 240 days was obtained in 701 patients with long (15- to 25-mm) lesions in small-diameter (2.5- to 3.5-mm) native vessels who were randomly assigned to treatment with SESs or bare-metal stents (BMSs) in the SIRIUS trial. Quantitative angiographic measurements of minimal lumen diameter and percent diameter stenosis were obtained within the treated segment, within the stent, and within its 5-mm proximal and distal edges. Patients treated with SESs had lower rates of binary (>50% diameter stenosis) angiographic restenosis within the segment (8.9% versus 36.3% with the BMS; P<0.001) and within the stent (3.2% versus 35.4% with the BMS; P<0.001). SESs were associated with significantly less late lumen loss within the treated segment, within the stent, and within its 5-mm proximal and distal edges (all P<0.001). The reduction of restenosis with the SES was consistent in patients at risk for restenosis, including those with small vessels, long lesions, and diabetes mellitus. The frequency of late aneurysms was similar in the 2 groups.
Conclusions— Compared with BMSs, SESs reduced angiographic late lumen loss within the stent and its adjacent 5-mm margins in patients with complex native-vessel lesions.
Received September 24, 2003; revision received August 2, 2004; accepted August 11, 2004.
Despite its benefit over balloon angioplasty in patients with native-vessel coronary artery disease,1,2 coronary stenting using balloon-expandable stents is limited by the occurrence of angiographic restenosis in 10% to 35% of patients, highest in patients with smaller vessels, longer lesions, and diabetes mellitus.3–5 The stent-versus-stent equivalency trials have shown similarities in angiographic restenosis rates with a number of balloon-expandable bare-metal stent (BMS) designs with a consistent degree of late lumen loss (0.80 to 1.05 mm), presumably because of intimal hyperplasia within the stent and its margins.6–8 Only 1 coiled stent had more restenosis than the Palmaz-Schatz stent, owing to both immediate recoil and late lumen loss within the coiled stent.8 Excessive late lumen loss occurs most often within the central portion of the stent,9 although, in some cases, a diffuse pattern of stent develops within the stents and extends into its nonstented margins.10 Isolated margin renarrowing, particularly involving only the proximal or distal edge of the stent, is an uncommon cause of stent restenosis with the BMS.
Marked reductions in angiographic restenosis have been shown in patients with focal lesions undergoing stent implantation by use of a single balloon-expandable sirolimus-eluting stent (SES).11,12 One randomized trial of 238 patients found no cases of binary angiographic restenosis (>50% diameter stenosis) 6 months after the procedure in patients treated with the SES, compared with a 26.6% restenosis rate in patients undergoing BMS implantation.11 This benefit was because of a near obliteration of late lumen loss (−0.01 mm) caused by intimal hyperplasia within the SES by use of both angiographic and intravascular ultrasound measurements.11 These effects have persisted up to 2 years after the index procedure.12
The purpose of this study was to compare the angiographic outcomes between SESs and BMSs in patients with complex lesions at high risk of restenosis, with regard to (1) the lumen changes within the stent and its 5-mm nonstented edges; (2) the degree and incidence of angiographic renarrowing in patients with smaller vessels and longer lesions; (3) the pattern of angiographic restenosis; and (4) the occurrence of late untoward angiographic findings, including the frequency of aneurysms.
A substudy of 850 patients was prespecified as the angiographic follow-up cohort within the 1101-patient SIRIUS trial, a 53-center, randomized, double-blind study comparing the angiographic and clinical outcomes of patients treated with an SES or BMS for complex native-vessel coronary disease13 (see Data Supplement Appendix). Lesions were suitable for inclusion in the study if they were between 15 and 30 mm in length and located in a visually estimated reference vessel between 2.5 and 3.5 mm in diameter. Patients were excluded with evidence of a recent (<24 hours) myocardial infarction, markedly reduced left ventricular function (left ventricular ejection fraction <25%), a totally occluded target lesion, or a target lesion in an ostial or bifurcation location. All investigational sites received approval from their local hospital Institutional Review Boards. The primary objective of this substudy was to contrast angiographic restenosis parameters in patients treated with an SES or BMS 8 months after the procedure.
Coronary Stent Procedure
After pretreatment with oral aspirin, clopidogrel, and intravenous heparin, patients were randomized to undergo implantation of either a bare-metal (uncoated) Bx velocity Balloon-Expandable BMS or SES (Cordis Corp, Johnson & Johnson). Stents were available in 8- and 18-mm lengths and in 2.5-, 3.0-, and 3.5-mm diameters. The concentration of sirolimus on the SES was 140 μg/cm2. Controlled release of sirolimus was facilitated by a polymeric matrix coating contained on the stainless steel stent that allowed elution of the drug up to 45 days after implantation. Predilatation with a conventional balloon was performed before stent implantation and postdilatation after the procedure was used to obtain a 0% visual residual stenosis within the stent. Angiographic success was defined as a final vessel diameter stenosis <50%.
Standard image acquisition was performed at the clinical sites using 2 or more angiographic projections of the stenosis, intracoronary nitroglycerin to provide maximum coronary vasodilation, and repetition of identical angiographic projections of the lesion at the time of follow-up angiography. Cineangiograms were then forwarded to the Brigham and Women’s Hospital Angiographic Core Laboratory for review by observers who were blinded to the treatment assignment. Repeat angiography was planned at 240±30 days in the first 850 randomized patients.
All procedural and follow-up angiograms were reviewed by use of standard morphological criteria.14,15 Lesion length was defined as the axial extent of the lesion that contained a shoulder-to-shoulder lumen reduction by ≥20%. Coronary dissections were assessed using the National Heart Lung and Blood Institute criteria.16 Restenosis patterns were qualitatively assessed using the Mehran classification system.10 Coronary aneurysms were defined as a maximum lumen diameter within the treatment zone that was 1.2 times larger than the average reference diameter of the vessel.
With the contrast-filled injection catheter as the calibration source, quantitative angiographic analysis was performed by use of a validated automated edge-detection algorithm (Medis CMS).17 Selected images for analysis were identified by use of angiographic projections that demonstrated the stenosis in an unforeshortened view, minimized the degree of vessel overlap, and displayed the stenosis in its “sharpest and tightest” view. A 5-mm segment of reference diameter proximal and distal to the stenosis was used to calculate the average reference vessel diameter at baseline, after stent implantation, and at follow-up; side branches and other anatomic landmarks were used to identify and maintain the consistency of the measurement length during the follow-up period. Minimal lumen diameters (MLDs) were measured at these same time points within the stent (in-stent analysis), within the 5-mm proximal and distal edges of the stent, and within the segment between the proximal and distal reference vessel (in-segment analysis) (Figure 1, A and B). Total occlusions were assigned an MLD=0 mm and a 100% diameter stenosis.
Angiographic follow-up was performed 6 to 9 months after the index procedure unless earlier angiography was required clinically. Binary angiographic restenosis was defined as the incidence of percent diameter stenosis >50% at the qualifying angiographic follow-up, performed from as early as 1 month for symptomatic patients to up to 9 months for asymptomatic patients. Angiographic percent diameter stenosis was defined as [1 − (MLD ÷ reference vessel diameter)] × 100. Acute gain was defined as the MLD immediately after the procedure minus the MLD before the procedure, and late loss was defined as the MLD immediately after the procedure minus the MLD at 8-month follow-up. The loss index was defined as the slope of the relationship between late loss and acute gain.
A sample size of 850 patients provided 80% statistical power to detect a 0.2-mm difference in the follow-up MLD between the SES- and BMS-treated patients with a 5% (2-sided) false-positive rate. After blinded randomization, a total of 26 patients (3% of the angiographic substudy) were deregistered and did not receive the assigned treatment. Deregistration from the study was because of either unavailability of the assigned stent at the site or discovery of exclusion criteria prohibiting further trial participation apparent only after the pretreatment angiograms. The final patient cohort included 824 patients and 824 lesions; a “per-lesion” analysis was performed for all angiographic parameters in this report.
Angiographic restudy was obtained in 701 patients (85.1%), including 350 (85.6%) in SES-treated patients and 351 (84.6%) in BMS-treated patients. All analyses were performed on a per-lesion basis. Categorical data are expressed as rates or proportions, and continuous variables are expressed as mean±SD. Binary variables were compared by use of χ2 and continuous variables were compared by use of the Student’s t test. The angiographic data set was used to build a multivariable model of 8-month angiographic restenosis. All analyses were performed by use of SAS (Version 8.0). A 2-sided probability value of P≤0.05 was considered significant.
Baseline Clinical Characteristics
The average age was 61.7 years in patients assigned to treatment with an SES and 62.6 years in patients assigned to treatment with a BMS. Diabetes was present in 25.7% of patients assigned to treatment with the SES and 29.5% of patients assigned to treatment with a BMS. In 27.7% of patients, 2 or more stents were used, with an average of 1.4 stents per patient. The average stent length was 21.4 mm, and the stent length-to-lesion length ratio was 1.6 to 1.
Baseline angiographic findings were similar in the 2 groups (see Data Supplement Table I). The left anterior descending coronary artery was the most commonly treated vessel (42.5%), and the majority of lesions (89.4%) were found in the proximal and mid segments of the vessel. The mean lesion length was 14.4 mm, and the American College of Cardiology/American Heart Association lesion complexity was Class B or C in 93.4% of lesions. Angiographic success rates were not different between the 2 arms (99%). There were no differences between the 2 arms in the incidence of abrupt closure (0.7%), dissection (1%), or occurrence of no reflow (<1%).
Quantitative changes in the lumen dimension within the treated arterial segment, within the stent, and within its 5-mm margins are found in Table 1. The mean reference vessel diameter was 2.79±0.45 mm in the SES group and 2.82±0.49 in the BMS group. SESs were associated with significantly less late lumen loss within the segment (0.24±0.47 versus 0.81±0.67 mm with BMS; P<0.001), within the stent (0.17±0.44 versus 1.00±0.70 mm with BMS; P<0.001), and within its 5-mm proximal (0.16±0.48 versus 0.32±0.63 mm with BMS; P<0.001) and distal (0.04±0.24 versus 0.24±0.61 mm with BMS; P<0.001) edges 8 months after stent placement (Figure 2).
Relative changes in lumen dimensions are found in Table 2. Patients treated with SESs had lower rates of binary (>50% diameter stenosis) angiographic restenosis within the segment (8.9% versus 36.3% with BMS; P<0.001), within the stent (3.2% versus 35.4% with BMS; P<0.001), and at its distal edge (2.0% versus 7.5% with BMS; P<0.001). The in-segment binary restenosis rates were significantly reduced in patients treated with the SES in all clinical and angiographic subsets (see Data Supplement Figure).
Different patterns of lumen renarrowing were found in patients who developed restenosis after treatment with the SES (n=31) compared with those treated with the BMS (n=128) (Table 3). In patients who developed restenosis, a pattern of focal margin restenosis (type 1B) was found in 61.3% of restenotic patients treated with the SES but was found in only 14.1% of restenotic patients treated with the BMS (P<0.001). Conversely, a diffuse pattern of restenosis (type II-IV) was found more often in patients treated with the BMS (57.0%) than in those treated with the SES (16.2%; P<0.001). The in-stent restenosis length was also significantly shorter in patients treated with a SES (9.1 mm) than in those treated with the BMS (14.8; P<0.001). Aneurysms were uncommon (<1%) and not different between the 2 groups.
Relationship to Vessel Size and Lesion Length
Late angiographic findings categorized by the baseline reference vessel terciles are found in Table 4. In patients treated with the BMS, in-segment binary restenosis rate was lowest in the largest vessel size tercile (30.5% versus 36.0% in the middle tercile versus 42.7% in the smallest vessel tercile). In patients treated with the SES, the in-segment binary angiographic restenosis rate was also lowest in patients in the largest vessel tercile (1.9% versus 6.6% in the middle vessel tercile and 17.6% in the smallest vessel tercile). For each tercile of reference vessel diameter, there was a significant (P<0.001) reduction in binary restenosis within the segment and within the stent. In patients treated with the SES, late lumen loss within the axial length of the stent was similar with all vessel sizes, but more late lumen loss occurred within the segment in patients treated with smaller vessels.
Lesion length was a less important determinant of angiographic restenosis (see Data Supplement Table II). In patients treated with the BMS, the in-segment binary restenosis rate was 37.9% in the longest lesion length tercile, whereas it was 35.0% in the middle tercile versus 36.7% in the shortest lesion length tercile. In patients treated with the SES, the in-segment binary angiographic restenosis rate was 12.1% in patients in the longest lesion tercile, compared with 5.8% in the middle vessel tercile and 8.5% in the shortest lesion length tercile. For each tercile of lesion length, however, there was a 68% to 83% reduction in binary angiographic restenosis associated with the use of the SES within the treated segment (P<0.001) and an 81% to 97% reduction in binary restenosis associated with the use of the SES within the stent (P<0.001).
Multivariable Predictors of Angiographic Restenosis
Multivariable predictors of higher binary in-segment restenosis rates included treatment with the BMS (P<0.001), the presence of diabetes mellitus (P<0.001), a smaller baseline (P=0.036) or final in-stent (P=0.001) MLD, and longer total stent lengths (P=0.039). The effect of these and other known risk factors for restenosis in patients assigned to either the SES or BMS is demonstrated in the nomogram of angiographic binary restenosis (Table 5), which estimates restenosis rates on the basis of the multivariable models and levels of the risk factor chosen by tercile distribution cutoff points.
This study demonstrates that compared with treatment using a BMS, treatment with the SES reduced all angiographic parameters of restenosis in patients with complex native-vessel disease at high risk for restenosis. A 91% reduction of restenosis was found within the stent, and a 75% reduction was found within the treated arterial segment that included the 5-mm proximal and distal nonstented margins. The effect on restenosis was presumably a result of a profound inhibition of intimal hyperplasia with the SES (0.17 mm) compared with the BMS (1.0 mm) (P<0.001). The reduction in intimal hyperplasia with the SES was also shown in the smallest vessels and longest lesions. Focal restenosis at the stent margin was the most common cause of failure in patients treated with the SES, whereas central restenosis was the most common cause of failure in patients treated with the BMS. There were no late deleterious effects associated with SES placement, such as the occurrence of late total occlusion or the formation of late aneurysms within the vessel.
Restenosis After BMS Implantation
Conventional angiography cannot adequately assess the relative degree of late lumen loss because of intimal hyperplasia or arterial remodeling after balloon angioplasty or directional atherectomy,18 but angiography can estimate the magnitude of intimal hyperplasia after stent implantation, because one of the benefits of stenting is the near elimination of late arterial remodeling. A number of angiographic studies have demonstrated a relatively fixed degree of late lumen loss in patients undergoing BMS implantation, ranging from 0.80 to 1.05 mm.6–8 A fixed amount of late lumen loss has an even larger effect on restenosis in smaller vessels.19 Angiographic studies in patients treated with BMS have reported binary restenosis rates within the stent and within the treated arterial segment that included its nonstented margins.6,7 Although previous BMS stent studies did not examine the specific lumen changes at the proximal or distal nonstent margins, intravascular ultrasound studies have shown that balloon injury proximal and distal to the stent may result in lumen renarrowing resulting from arterial remodeling at the edges of the stent.9 Studies evaluating angiographic outcomes in patients undergoing radiation brachytherapy or using radioactive stents have also shown that proliferation at the stent edges may be associated with higher-than-expected restenosis rates at the stent margins (“candy wrapper effect”) because of an untoward effect of low-dose radiation at the stent margins.20,21 No previous studies have evaluated these potential effects at the margins of the SES.
Previous Restenosis Studies Using SES
Sirolimus is a naturally occurring macrolide antibiotic and antiinflammatory agent that inhibits smooth muscle cell replication by interrupting the G1 phase in the cell cycle that prevents the downregulation of p27 before DNA replication, in addition to its antiinflammatory effects.22 A first-in-humans study of 45 patients treated with the SES showed a profound reduction in late lumen loss within the stent, an effect that persisted up to 24 months after the procedure.12,23 A subsequent randomized trial confirmed the effect of SESs in a larger series of patients with focal lesions (average lesion length, 9.6 mm), with a reported −0.01-mm late lumen loss; there were no cases of angiographic restenosis in patients treated with the SES compared with 26.6% in patients treated with a BMS.11 These results were confirmed with serial intravascular ultrasound, which showed a marked reduction in neointimal volume in patients treated with the SES.24
The results of this study demonstrate that angiographic restenosis was markedly reduced in patients with complex native-vessel disease by use of the SES. A 91% reduction in binary angiographic restenosis was shown within the stent, and a 75% reduction in binary angiographic restenosis was demonstrated within the treated segment, which included the nonstented margins. The beneficial effect of the SES on binary angiographic restenosis was attributed to reductions in late lumen loss within the SES (0.17 mm) compared with the BMS (1.00 mm), a favorable effect that extended to its proximal and distal 5-mm nonstented margins (Figure 2). These angiographic effects on late lumen loss were presumably attributable to reductions in intimal hyperplasia because of the sustained release of sirolimus into the vessel wall after stent implantation.
This study also showed that the angiographic benefit with the SES occurred in the most complex patients: those with the smallest vessels and the longest lesions and in patients with diabetes mellitus. The beneficial effects of the SES were most pronounced within the stent in smaller vessels and longer lesions but were also highly significant within the segment that included the nonstented margins. The higher restenosis rates within the segment that included the nonstented margins were attributable to lumen changes that occurred in regions that were not fully treated with sirolimus. These non-stent margin lumen changes were also found in patients treated with the BMS, but it appears that the beneficial effect of the SES on late lumen loss did extend to a lesser extent into the 5-mm proximal and distal margins of the stent, potentially because of upstream or downstream elution of the drug or because of the effect on an overall reduction in stent restenosis on the adjacent nonstented margins. No deleterious effects of the SES were found at the margins of the stent.
Patterns of Restenosis
The study also demonstrated a novel shift in the pattern of restenosis with the SES. Intimal hyperplasia after BMS implantation most commonly involves the central portion of the stent. In patients treated with a BMS in this study, 50% of patients had a diffuse pattern (>10 mm) of restenosis that was confined within the stent (35.9%) or extended from within the stent to include its proximal or distal margins (13.3%). In contrast, 61.3% of the SES patients who developed restenosis had the intimal hyperplasia localized to the margin of the stent, often sparing the axial region of the stent. The restenosis length was also shorter in patients treated with the SES. Given the finding that the edge lumen dimensions were actually larger at follow-up in patients treated with the SES, it is not certain what precise factors contributed to the margin restenosis in patients treated with the SES. The most likely cause of the margin restenosis pattern was the occurrence of injury at the margins of the stent that was not adequately covered with sirolimus. Alterations in operator technique, such as complete coverage of the lesion with adequate margins, avoidance of injury at the edges of the stent, and careful postdilatation within the stent may reduce the pattern of margin restenosis.
Coronary aneurysms are rare findings after coronary intervention, with an incidence of 0.3% to 6%, depending on the definition used for vessel expansion.25–29 Pseudoaneurysms are the most common form of trauma-induced coronary aneurysms, because of abnormal healing of an intimal dissection that occurs at the time of intervention. True aneurysms involving all 3 layers are uncommon after coronary intervention. Stents that elute antiproliferative agents have the potential to cause both forms of aneurysms, similar to potent systemic antiproliferative agents.30 With a conservative definition of 1.2 times the reference diameter, there was no difference in the incidence of late aneurysms in patients treated with the SES.
There were several limitations to this study. Although angiographic recording of all balloon inflations before and after stent implantation was recommended, it was often difficult to determine whether a significant degree of arterial injury occurred outside the margin of the stent. The margin restenosis that occurred in patients treated with the SES, although less than in those treated with the BMS, may have been a result of margin injury that was not matched with drug elution. Further studies are ongoing to determine the precise distribution of vessel injury and the degree of late lumen loss with the SES and BMS. Another limitation of this study was that the smallest stent size was 2.5 mm, and vessels as small as 2.0 mm by quantitative angiography were treated. It is not clear whether smaller stents (2.25 mm) more appropriately matched to the reference diameter would have been associated with less margin restenosis in the smallest vessels. Finally, repeat angiography was performed 8 months after stent implantation, and although there was no evidence of a delayed “catch-up” in restenosis in a pilot series of sirolimus-treated patients 2 years after the procedure,12 the very late occurrence of angiographic events will require additional follow-up studies.
The results of this study demonstrate that treatment with the SES in patients with native-vessel coronary artery disease resulted in marked reductions of all angiographic indices of restenosis. When restenosis occurred with the SES, it developed most often at the margins of the stent. The beneficial effect was also present in the smallest vessel and longest lesions. There were no differences in late aneurysms between the 2 groups. We conclude that the use of the SES is safe and effective for the treatment of complex native-vessel coronary artery disease.
This study was supported by a grant from Cordis, a Johnson & Johnson Company.
Dr Moses has served a consultant to and speaker for Cordis, a Johnson & Johnson Company. Dr Popma received research grants from Cardis for the angiographic analysis of this study.
The online-only Data Supplement, which contains an Appendix, can be found with this article at http://www.circulationaha.org.
Holmes D, Savage MPS, Lablanche J-M, et al. Results of Prevention of REStenosis with Tranilast and its Outcomes (PRESTO) Trial. Circulation. 2002; 106: 1243–1250.
Hoffmann R, Mintz GS, Dussaillant GR, et al. Patterns and mechanisms of in-stent restenosis: a serial intravascular ultrasound study. Circulation. 1996; 94: 1247–1254.
Mehran R, Dangas G, Abizaid A, et al. Angiographic patterns of in-stent restenosis: classification and implications for long-term outcome. Circulation. 1999; 100: 1872–1878.
Sousa J, Costa M, Sousa A, et al. Two-year angiographic and intravascular ultrasound follow-up after implantation of sirolimus-eluting stents in human coronary arteries. Circulation. 2003; 107: 381–383.
Ellis S, Vandormael M, Cowley M, et al. Coronary morphologic and clinical determinants of procedural outcome with angioplasty for multivessel coronary disease. Circulation. 1990; 82: 1193–1202.
Ryan T, Faxon D, Gunnar R, et al. Guidelines for percutaneous transluminal coronary angioplasty: a report of the American College of Cardiology/American Heart Association Task Force on Assessment of Diagnostic and Therapeutic Cardiovascular Procedures (Subcommittee on Percutaneous Transluminal Coronary Angioplasty). J Am Coll Cardiol. 1988; 12: 529–545.
Mintz GS, Popma JJ, Pichard AD, et al. Arterial remodeling after coronary angioplasty: a serial intravascular ultrasound study. Circulation. 1996; 94: 35–43.
Albiero R, Nishida T, Adamian M, et al. Edge restenosis after implantation of high activity 32P radioactive β-emitting stents. Circulation. 2000; 101: 2454–2457.
Albiero R, Adamian M, Kobayashi N, et al. Short- and intermediate-term results of 32 P radioactive β-emitting stent implantation in patients with coronary artery disease: the Milan Dose-Response Study. Circulation. 2000; 101: 18–26.
Marx S, Marks A. Bench to bedside: the development of rapamycin and its application to stent restenosis. Circulation. 2001; 104: 852–855.
Sousa J, Costa M, Abizaid A, et al. Sustained suppression of neointimal proliferation by sirolimus-eluting stents: one-year angiographic and intravascular ultrasound follow-up. Circulation. 2001; 104: 2007–2011.
Degertekin M, Serruys P, Foley D, et al. Persistent inhibition of neointimal hyperplasia after sirolimus-eluting stent implantation: long-term (up to 2 years) clinical, angiographic, and intravascular ultrasound follow-up. Circulation. 2002; 106: 1610–1613.
Baumbach A, Bittl JA, Fleck E, et al. Acute complications of excimer laser coronary angioplasty: a detailed analysis of multicenter results. Coinvestigators of the U.S. and European Percutaneous Excimer Laser Coronary Angioplasty (PELCA) Registries. J Am Coll Cardiol. 1994; 23: 1305–1313.
Condado JA, Waksman R, Gurdiel O, et al. Long-term angiographic and clinical outcome after percutaneous transluminal coronary angioplasty and intracoronary radiation therapy in humans. Circulation. 1997; 96: 727–732.