Utility of Sirolimus-Eluting Cypher‚Ñ¢ Stents to Reduce 12-Month
Target Vessel Revascularization in Saphenous Vein Graft Stenoses

Approximately 1 million patients are now treated annually in the United States with drug-eluting stents (DES). Usage of these stents is substantiated in many patient and lesion subsets by their documented relative safety and reduction in clinical restenosis in large-scale randomized clinical trials.^1,2 Such trials cannot be feasibly performed in all groups of patients. The pathology of atherosclerosis in saphenous vein grafts (SVGs) differs somewhat from that in native coronary arteries, and restenosis after bare-metal stenting (BMS) tends to occur more frequently in the former than the latter.³ Although generally considered inferior to well-designed randomized clinical trials, well-designed cohort or case-control studies have recently been found to provide similar results to randomized trials in a variety of clinical situations.^4–6 In addition, they are usually considerably less costly to perform.

To date, only one very small randomized clinical trial (75 patients, 96 lesions) has been performed to evaluate the performance of DES in patients with diseased SVGs.⁷ This showed a dramatic reduction in clinical restenosis relative to BMS, but a heightened risk of late death. The results of somewhat larger nonrandomized comparisons are quite varied, with some studies showing no benefit at all.^8-10 Therefore, we undertook to perform a carefully controlled multicenter casecontrol study evaluating the potential benefit in terms of reduction in target vessel revascularization (TVR) of the Cypher^™ DES (Cordis Corp., Warren, New Jersey) compared to a variety of BMS in clinical use during the 2000s.

Methods

Patient population. Eligibility criteria: (1) De novo SVG lesion; (2) suitable for 3.0–3.5 mm stent (can be expanded to approximately 4.5 mm); (3) target stenosis 50–99% with Thrombolysis and Myocardial Infarction (TIMI) 2–3 flow pretreatment; (4) left ventricular ejection fraction > 30%; (5) no evidence of myocardial infarction (MI) < 24 hours, dialysis, BMS placed during the same procedure as a DES, debulking prior to stent placement. All angiographic assessments were by visual inspection or caliper measurement at the treatment site.

Cypher stent patients. Consecutively treated patients receiving Cypher stents who met the eligibility criteria were recruited (retrospectively and prospectively once this study had begun).

Matching. Patients were matched on the expanded stent diameter (≤ 3.0, 3.1–3.5 and > 3.5 mm) stent length(< 10, 10–19.9, 20–29.9 and > 30 mm), diabetes and number of stents placed. From a registry of all patients receiving BMS for SVG lesions at each contributing center from 2000–2003, the patient treated most recently who matched on all elements was selected, without regard to clinical outcomes. Matching was performed at the local clinical site. If a Cypher patient could not be matched with patients treated at the same institution, matching was performed at the Cleveland Clinic Data Coordinating Center with a patient from another institution. If multiple BMS patients were equally eligible for matching, one was selected at random.

Data management and analysis. Dedicated case report forms were developed, and source-documented data were entered directly onto these forms at the individual clinical sites. Outcomes were obtained by the local sites using a combination of chart review, telephone and mail contact. Data were reviewed for completeness and consistency at the Data Coordinating Center at the Cleveland Clinic, and then entered into a central data repository. Analyses were prespecified by the investigators and performed using SAS for Windows 9.1 software.

Clinical endpoints. The primary study endpoint was determined a priori to be freedom from TVR at 12 months. Other secondary endpoints to be evaluated included in-hospital major adverse cardiac events (MACE) and total mortality, target lesion revascularization (TLR) and stent occlusion at 12 months.

Statistics. Data are presented as mean ± 1 standard deviation, median and range, or as a percentage, as appropriate. To maximize the amount of information for comparison, pair-wise analyses were performed for each variable individually. If one or both members of a pair were missing for a particular variable, that variable was set to missing. Continuous or ordinal data were compared using the Student’s t-test for paired data, proportions with two levels were compared using McNemer’s chi-square test, and proportions with three or more levels were compared using Bowker’s test of symmetry. A p-value < 0.05 was considered statistically significant. Sample size was determined beforehand using expected TVR rates from the literature of 25% for BMS,³ a relative reduction of 40–50% and a power of approximately 0.80. A total of 210 pairs of patients were to be enrolled, but due to slow enrollment, this number was decreased to 175 pairs, yielding a power of 0.60–0.82 using the original assumptions.

Results

Baseline patient characteristics are provided in Table 1. The average patient age was 69 years, and 77% of patients were male. Thirty-nine percent of patients were diabetic, 75% presented with unstable angina, the average SVG age was 9.9 ± 6.3 years, reference vessel diameter was 3.35 ± 0.36 mm, and lesion length was 17.4 ± 8.4 mm. Patients were well matched for all measured baseline characteristics. Treatments rendered and in-hospital outcomes are presented in Table 2. Most patients had only 1 lesion treated, and angiographic and clinical outcomes were excellent. Between-group differences were, however, noted with regard to glycoprotein IIb/IIIa use (more common with BMS) and deployment pressure and embolic protection device used (higher or more common with Cypher stents). Long-term outcomes are available for 171/175 (97.7%) Cypher and 173/175 (98.8%) BMS patients, and paired data are provided in Table 3. In follow up, more patients received clopidogrel therapy beyond 6 months in the Cypher group. There was no difference in survival through 12 months. TVR occurred in 6.8% of Cypher-treated patients and 11.8% of BMS-treated patients (p = 0.14). Subgroup analyses in prespecified groups are presented in Figure 1.

Discussion

The principal finding of this case-control study involving more Cypher-treated patients than the combined number in all reported previous randomized and nonrandomized head-to-head comparisons was that: Cypher DES appeared to reduce clinical restenosis (TVR) by a relative 40–45% without compromise of early (in-hospital MACE) or 12-month (mortality) safety. This relative reduction in TVR is perhaps somewhat less than that seen with de novo native vessel lesions,¹ although the impact of mandatory coronary angiography increasing TLR1,¹¹ in the SIRIUS trial must be noted.

To place these results in context, it is useful to review previous reports on the same topic. Vermeersch and colleagues⁷ performed the only randomized, controlled trial addressing the same issue, in which 75 patients (96 SVGs) were randomized using entry criteria similar to those in this study, with the exception that patients who required more than 2 stents or had chronic renal insufficiency were excluded. Follow-up angiography was mandated. This study demonstrated a marked reduction in TVR with Cypher versus BMS (5.3% compared with 21.6%; p = .047). This incidence of BMS-related TVR is as high or higher than other recent reports, which have shown incidences of 6–37%, perhaps due to the requirement for angiographic follow up.¹⁰ Some nonrandomized comparisons have shown an apparent reduction in clinical restenosis with Cypher versus BMS utilized to treat SVG lesions, whereas others have shown no apparent benefit.^8–10 Follow up of the Vermeersch study showed a significant, albeit not fully explained, mortality excess in the DES-randomized patients.¹² This is not apparent in our considerably larger study.

Study Limitations

The present study is limited by the following: (1) its nonrandomized origin, resulting in an imbalances in some variables (although those known to play a key role in the risk for restenosis were well matched);¹³ (2) a lesser actual power than anticipated due to lower-than-expected TVR rates in the control population; (3) a reduced number of subjects due to slow enrollment; (4) a lack of long-term data on TVR and stent thrombosis; (5) the fact that available stent sizes limited the study to small-to-medium diameter bypass grafts; and (6) the fact that quantitative coronary angiography was not routinely performed. Subgroup data, in particular, should be interpreted with considerable caution.

Conclusion

In summary, taken in conjunction with other existing data, this study supports a reduction in restenosis-related events by Cypher stents in diseased de novo SVG stenoses (but it is likely of lesser magnitude than seen in the previous small, randomized trial testing the same hypothesis), but finds no evidence to confirm concerns raised about DES safety in this clinical setting.

References

1. Moses JD, Leon MB, Popma JJ, et al, for the SIRIUS Investigators. Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315–1323.

2. Stone GW, Ellis SG, Cox DA, et al. One-year clinical results with the slow-release, polymer-based, paclit axel-eluting TAXUS stent – The TAXUS-IV trial. Circulation 2004;109:1942–1947.

3. Savage MP, Douglas JS Jr, Fischman DL, et al. Stent placement compared with balloon angioplasty for obstructed bypass grafts. Saphenous Vein De Novo Trial Investigators. N Engl J Med 1997;337:740–747.

4. Sacks H, Chalmers TC, Smith H Jr. Randomized versus historical controls for clinical trials. Am J Med 1982;72:233–240.

5 . Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N Engl J of Med 2000;342:1878–1886.

6 . Concato J, Shah N, Horwitz R. Randomized, controlled trials, observational studies, and the hierarchy of research designs. N Engl J Med 2000;342:1887–1896.

7 . Vermeersch P, Agostoni P, Verheye S, et al. Randomized double-blind comparison of sirolimus-eluting stent versus bare-metal stent implantation in diseased saphenous veingrafts . J Am Coll Cardiol 2006;48:2424–2431.

8 . Ge L, Iakovou I, Sangiorgi GM, et al. Treatment of saphenous vein graft lesions with drug-eluting stents: Immediate and midterm outcome. J Am Coll Cardiol 2005;45:989–994.

9 . Lee MS, Shah AP, Aragon J, et al. Drug-eluting stenting is superior to bare metal stenting in saphenous vein grafts. Catheter Cardiovasc Interv 2005;66:512–513.

10. Chu WW, Rha SW, Kuchulakanti PK, et al. Efficacy of sirolimus-eluting stents compared with bare metal stents for saphenous vein graft intervention. Am J Cardiol 2006;97:34–37.

11. Ruygrok PN, Melkert R, Morel MAM, et al, on behalf of the Benestent II Investigators, Cardialysis, Rotterdam, The Netherlands. Does angiography six months after coronary intervention influence management and outcome? J Am Coll Cardiol 1999;34:1507–1511.

12. Vermeersch P, Agostoni P, Verheye S, et al. Increased late mortality after sirolimus-eluting stents versus bare-metal stents in diseased saphenous vein grafts. J Am Coll Cardiol 2007;50:261–267.

13. Chan AW, Moliterno DJ. Restenosis: The clinical issues. In: Topol EJ (ed). Textbook of Interventional Cardiology, 4th Edition. Saunders Co.: Philadelphia. 2003, pp. 415–453.