Outcomes of 6906 Patients Undergoing Percutaneous Coronary Intervention in the Era of Drug-Eluting Stents
Report of the DEScover Registry
Background— The DEScover Registry was designed to characterize patients selected for drug-eluting stents (DES) in routine clinical practice and their outcomes in the United States.
Methods and Results— From January to June 2005, data were collected on 6906 patients who underwent percutaneous coronary intervention at 140 medical centers. Baseline characteristics and outcomes were compared on the basis of treatment with ≥1 bare-metal (BMS; n=397), sirolimus-eluting (SES; n=3873), or paclitaxel-eluting (PES; n=2636) stent. Clinical characteristics and the types of lesion treated for BMS patients differed substantially from those treated with DES, but minimal differences were noted between DES patients receiving SES or PES. At 1 year, the unadjusted cumulative incidence of death/myocardial infarction was higher in BMS than in DES patients (9.0% versus 5.2%; P=0.002) but similar in SES and PES patients (5.2% versus 5.3%; P=0.64). After adjustment, risk of death/MI was not significantly lower in DES- compared with BMS-treated patients (adjusted hazard ratio, 0.74; 95% confidence interval, 0.52 to 1.07). Although target vessel revascularization occurred less often in DES patients (9.5% versus 6.0%; P=0.007), rates were similar between SES and PES patients (6.3% versus 5.5%; P=0.20). Rates of stent thrombosis were similar among BMS (0.8%), SES (0.5%), and PES (0.8%) patients.
Conclusions— In DEScover, differences in patient selection were observed between BMS and DES patients but not between SES and PES patients. DES use resulted in lower rates of clinically driven repeat revascularization with similar rates of stent thrombosis. These observations confirm the effectiveness and safety of both SES and PES in unselected patients.
Received October 2, 2006; revision received October 12, 2006; accepted October 14, 2006.
Several randomized clinical trials have demonstrated the ability of drug-eluting stents (DES) to substantially reduce restenosis and the need for repeat revascularization compared with bare-metal stents (BMS).1–4 Additionally, these studies did not demonstrate any significant short-term safety concerns for DES. Of note, these comparisons were limited to highly selected patient cohorts with stable coronary artery disease and relatively simple coronary lesions. A substantial proportion of patients commonly treated by percutaneous coronary intervention (PCI), however, have more complex disease features, including unstable clinical presentations that may limit the ability to generalize results from the initial DES randomized trials. There have been several relatively small registries of unselected patients treated with sirolimus-eluting stents (SES) or paclitaxel-eluting stents (PES) and a large postmarketing surveillance registry for SES, all outside the United States, showing the relative safety of DES.5–8 In addition, registry studies and randomized trials have sought to compare outcomes for SES- versus PES-treated patients in unselected populations and specific patient subsets, respectively.5,8–11 Few data are available, however, regarding the selection and outcomes of BMS- and of SES- and PES-treated patients in the DES era in the United States. The DEScover Registry was designed to address these issues. Therefore, the purpose of the present report is to describe the findings of DEScover, with particular attention directed at comparing 1-year outcomes of BMS- and DES-treated patients and of SES- and PES-treated patients.
Study Design and Patient Population
DEScover is a prospective, multicenter, observational study designed to characterize PCI patients from a broad sampling of hospitals and practice. A complete list of clinical sites and investigators is available in the online-only Data Supplement. At participating sites (n=140), consecutive patients undergoing PCI were asked to participate in the study. The only exclusion criteria were patient refusal or inability to provide written informed consent and/or HIPAA (Health Insurance Portability and Accountability Act of 1996) authorization. Patients who died during a procedure before informed consent was obtained were not enrolled. A screening log of all potentially eligible patients was maintained at each site to ensure that no purposeful selection bias influenced patient enrollment. All interventional strategies, including the use of stents, choice of stent type, and periprocedural antithrombin and antiplatelet therapy, were at the operator’s discretion. The protocol was approved by a central and/or individual institutional review board.
Baseline clinical and angiographic characteristics and certain procedural and clinical in-hospital events were recorded for the enrolled patients. Follow-up was obtained from patients at 1, 6, and 12 months by a central telephone facility. For those patients reporting an event, a specially trained research coordinator then obtained additional information.
Data Management and Definitions
Data were collected and submitted from participating sites using password-protected Internet-based electronic case report forms and study codes to protect subject identity. A private independent contract research organization was retained to serve as the administrator for the study, which is sponsored by Cordis, a Johnson & Johnson Company (Miami Lakes, Fla), the manufacturer of the Cypher stent (an SES). The University of Pittsburgh serves as the statistical analysis core. Follow-up was 93.2% complete at 6 months and 89.6% complete at 1 year.
Angiographic success was defined as the achievement of a minimum stenosis diameter reduction to <50% in the presence of Thrombosis in Myocardial Infarction grade 3 flow. Procedural success was defined as complete if angiographic success was achieved in all attempted lesions without the occurrence of a major complication (myocardial infarction [MI], repeat revascularization during the hospitalization) or death. Partial success was angiographic success in at least 1 lesion without an in-hospital major complication. Procedural failure was the absence of angiographic success in all lesions or the occurrence of an in-hospital major complication. All angiographic variables were reported on the basis of visual assessment by the individual sites.
Major adverse coronary events were defined as death, MI, repeat PCI, or coronary artery bypass graft surgery (CABG). Deaths were classified as cardiac, noncardiac, or unknown. For this report, all-cause mortality is presented. MI was diagnosed by evolutionary ST-segment elevation, development of new Q waves or left bundle-branch block on the ECG, or biochemical evidence of necrosis, including a total creatine kinase ≥2 times the normal upper limit with an elevated creatine kinase-MB or a positive troponin. Repeat PCI was categorized according to whether the index lesion or artery was attempted. Stent thrombosis was considered acute (0 to 24 hours), subacute (>24 hours to 30 days), or late (>31 days). Stent thrombosis was defined as either definite or probable stent thrombosis. Definite stent thrombosis was defined as the presence of angiographic thrombus in a stent that previously had been successfully deployed accompanied by an acute coronary syndrome. Angiographic thrombus was defined as complete occlusion (Thrombosis in Myocardial Infarction grade 0 or 1 flow) with a stent diameter stenosis <30% or evidence of flow-limiting thrombus (ovoid or linear filling defect) within or immediately adjacent to the stent. Probable stent thrombosis is defined as unexplained sudden cardiac death or Q-wave MI in the distribution of the stented artery. In each instance, the stented segment was not known to be patent. An independent committee adjudicated episodes possibly considered to be stent thrombosis. In this report, rates of stent thrombosis include those classified as definite or probable stent thrombosis. Death and MI were not adjudicated.
Differences in baseline, procedural, and angiographic characteristics and in-hospital adverse events were compared between BMS and DES patients and between SES and PES patients by use of χ2 tests for categorical variables and Student t and Wilcoxon rank sum tests for continuous data. Cumulative 30-day and 1-year adverse event rates were estimated by the Kaplan-Meier approach and compared by the log-rank statistic.12,13 Patients lost to follow-up were censored at the last known date of contact. Multivariable Cox proportional hazards regression was used to estimate adjusted hazard ratios (HRs) of adverse clinical outcomes by DES versus BMS and SES versus PES treatment strategy.14 Variables selected for adjustment were those significantly associated with the clinical outcome of interest and those considered biologically relevant. The primary study outcomes were the composite of death and MI (safety outcome) and target vessel revascularization (TVR) (effectiveness outcome). Values of P<0.05 were considered significant.
The authors had full access to the data and take responsibility for their integrity. All authors have read and agree to the manuscript as written.
Baseline and Procedural Characteristics
Between January and June 2005, 7752 patients were enrolled; 7745 had complete data on PCI treatment strategy. Of these, 7420 (96%) received at least 1 stent, and the remaining 325 had attempted angioplasty alone. This analysis includes 6906 patients who received only 1 type of stent, including those treated with at least 1 BMS (n=397), SES (n=3873), or PES (n=2636), during the index procedure. The remaining patients had combinations of BMS, SES, and PES stents. This cohort, selected to permit comparisons between BMS and DES and between SES and PES, made up 89% of the entire study cohort.
Overall, the mean age of the patients was 64.1 years; 67.9% were male; and 31.8% had diabetes mellitus (Table 1). There were substantial baseline differences between patients receiving BMS compared with those treated with DES. BMS patients were older, less often had prior PCI, but more often had prior CABG, presence of severe noncardiac disease, acute MI as the indication for PCI, and emergent procedural circumstances. Left ventricular ejection fraction was lower among BMS patients. Conversely, other than SES patients being younger and trending toward more diabetes, the baseline features of patients treated with SES and PES were remarkably similar. There were no differences in intraprocedural medications among stented patients, although use of aspirin and use of clopidogrel were more common at discharge among DES-treated patients (data not shown).
There were 515 lesions treated (n=471 stented) in BMS patients and 9368 lesions treated (n=8814 stented) in DES patients. Nonstented lesions were treated with balloon angioplasty, atherectomy, or another nonstent device. Lesions treated with BMS more often were de novo and total occlusions, were located in venous bypass grafts, and had reference vessel diameters <2.5 or >3.5 mm (Table 2). Multilesion and left anterior descending PCI was performed less often in BMS-treated patients. The treatment of bifurcation and ostial or calcified lesions was similar in the 2 groups. Procedural success was high for both groups. Aside from a reduction in flow being more common among BMS-treated patients, no significant differences were noted in rates of lesion complications.
Among patients treated with DES, SES patients had 5655 lesions treated (n=5315 stented), and PES patients had 3713 lesions treated (n=3499 stented). Small, statistically significant differences were observed in angiographic and procedural characteristics between SES and PES patients. The SES patients more often had multiple lesions attempted (35% versus 32%), the left anterior descending artery as the site of attempted lesion (45% versus 40%), and a bifurcation side branch treated (6% versus 5%) and less often had vessel calcification (24% versus 29%). In addition, SES patients were nominally more likely to receive >1 stent (29% versus 27%) and have stent overlap (17% versus 15%), longer mean stent length (20.2 versus 18.6 mm), and larger-diameter stents (3.0 versus 2.9 mm). The angiographic success rate was similar and exceeded 98% in both groups, but PES patients had more bailout use of a glycoprotein IIb/IIIa inhibitor (2.7% versus 1.8%), abrupt vessel closure (0.6% versus 0.2%), persistent flow reduction (0.7% versus 0.3%), and a trend for more side-branch occlusion (1.5% versus 1.0%) than SES patients.
BMS Versus DES
There were significant differences in outcomes in the BMS and DES groups at 1 year (Table 3). The rate of death was higher in BMS patients (5.9% for BMS versus 3.1% for DES; P=0.005), whereas the rates of MI only trended to be higher in BMS patients. Repeat TVR was less common among DES patients (9.5% for BMS versus 6.0% for DES; P=0.007). In particular, CABG was needed less often among DES patients (3.5% for BMS versus 1.4% for DES; P=0.0007). Rates for stent thrombosis were low (<1%) for both groups and not significantly different.
Adjusted 1-year HRs comparing BMS with DES for death or MI and for TVR among all patients and in various subsets are shown in Figure 1. For the overall cohort, death or MI was not significantly different according to DES or BMS. A benefit of DES was suggested among patients without a history of advanced coronary disease, namely those with no prior MI, PCI, or CABG. For TVR, a substantial benefit was observed with DES (adjusted HR, 0.58; 95% confidence interval, 0.40 to 0.83). This benefit was observed in most subgroups but not among women and patients with a history of MI or PCI.
SES Versus PES
In-hospital and 30-day outcomes were similar between SES and PES patients (Table 4). At 1 year, the cumulative incidence of death and MI remained similar. Rates of stent thrombosis were low and not statistically different between SES and PES patients (0.5% versus 0.8%; P=0.06). In terms of additional revascularization, the rates for overall TVR via PCI or CABG were similar (6.3% for SES versus 5.5% for PES; P=0.20). Figure 2 displays the adjusted HRs of death/MI and TVR overall and among subgroups for SES versus PES patients. No significant differences were observed between SES and PES for death or MI overall or among subgroups. Similarly, for TVR, no difference was noted for the entire DES group between SES and PES patients. A benefit for PES was suggested among patients with a history of prior CABG.
The DEScover Registry has several unique features that distinguish it from other published registry reports characterizing the outcomes of patients treated with DES in routine clinical practice. First, prior registry descriptions of DES have been performed outside the United States.5–8 Given that approximately half of the 2 million stent procedures worldwide are performed in the United States and that practice patterns probably differ according to region, the information obtained in DEScover is significant and unique. Second, DEScover included all patients having PCI rather than just those receiving stents. This design permitted assessment of the frequency at which stents are used in general, and we observed that nearly all patients (96%) undergoing PCI received a stent. Acknowledging that a stent may not be deliverable to a lesion because of anatomic and technical considerations, we believe that physicians attempt to implant a stent whenever feasible. Third, DEScover enrolled patients treated with BMS and DES. The acceptance of DES as preferred therapy is reflected in the overwhelming proportion (94%) of stented lesions treated with DES rather than BMS.
Small but significant differences were noted in the manner in which physicians selected patients for DES compared with BMS. Physicians favored BMS for patients with acute MI, possibly because of concerns of stent thrombosis or potential lack of effectiveness. Two recent randomized clinical trials, however, showed no excess thrombosis with DES compared with BMS in patients with acute ST-elevation MI. Additionally, DES showed similar effectiveness in this population, with a significant reduction in target lesion revascularization in the SES trial and a trend in the PES trial that had a low BMS revascularization rate.15,16 BMS also was preferred for patients with prior CABG. On a lesion level, vein graft lesions were attempted substantially more often in BMS patients. These observations may be explained by the lack of DES availability in diameters >3.5 mm, the stent size commonly required for a venous graft. BMS was used less often than DES for patients with a history of prior PCI and for lesions previously treated with a stent. This practice pattern probably was based on the belief that the DES would be effective in the treatment of bare-stent restenosis. The effectiveness of this strategy has now been validated by 2 randomized clinical trials and 1 observational report.17–19 Unadjusted event rates indicated that death and MI occurred more frequently among BMS-treated patients at the 1-year follow-up. This difference was not present, however, after statistical adjustment for confounding factors as might be expected because more BMS patients were treated for MI and as a group had a higher prevalence of high-risk features.
DEScover supports the effectiveness of DES for reducing TVR by either PCI or CABG compared with BMS. Although there was again an imbalance of baseline clinical and procedural factors that could potentially influence the likelihood of repeat revascularization, this benefit was confirmed after statistical adjustment. This is an important observation because the efficacy of DES for reducing the incidence of restenosis has been unequivocally demonstrated only in highly selected patients. Our findings indicate that clinicians and patients can expect a similar, albeit less pronounced, benefit of a relative reduction in TVR of ≈42% from DES in a broad population of patients. Furthermore, from a safety standpoint, there was certainly no excess of death, MI, or stent thrombosis among DES patients over the 1-year follow-up period.
DEScover also compared the use and outcomes of patients treated with both types of commercially available DES, PES and SES. Of interest, baseline clinical features of patients selected for either SES or PES were nearly identical. Thus, despite the fact that the 2 DES have very different antirestenotic drugs, physicians did not demonstrate any selectivity according to clinical or angiographic features. During and since the completion of DEScover, differences in effectiveness of SES and PES for certain clinical subgroups have been demonstrated by randomized clinical trials.8–11
Of particular interest, early, intermediate, and 1-year patient outcomes were quite similar between SES and PES patients, and the overall rates of major adverse cardiac events were low. These findings are reassuring, given concerns regarding the possibility of excess rates of stent thrombosis, death, and MI that might be related to the components of DES.20 Observational registries from Europe have reported 1-year rates of death ranging from 1.4% to 5.3%, with combined death and MI rates ranging from 2.7% to 8.8%.5–7 Differences in completeness of follow-up and possibly in the types of patients treated and interventional techniques used may explain the variation in these rates. Certainly, the results from DEScover compare favorably with those reported earlier. In terms of adjudicated stent thrombosis, we observed that rates between SES and PES were similar and low. DEScover was not designed to detect a difference in thrombosis rates between SES and PES.
There were no significant differences between SES and PES for rates of any repeat PCI, CABG, or TVR. This finding is distinct from prior registries and recent randomized trials showing equal or greater effectiveness of SES over PES for reducing TVR.5,8,9,11,20,21 In all, this is the largest published registry that includes both SES and PES, and no difference in effectiveness of the 2 stents for reducing TVR was observed.
As an observational study, certain factors limit our report. The number of BMS patients was relatively small compared with the DES groups and thus lacked precision for uncommon clinical events, including stent thrombosis. In addition, we fully expected a selection bias between BMS- and DES-treated patients because our primary goal was to describe usage patterns unconstrained by eligibility criteria. Accordingly, despite our best efforts to adjust for confounding factors, our comparative findings need to be viewed with this consideration in mind. Unfortunately, we did not request information regarding the use of antiplatelet agents during follow-up. Such information would be valuable in augmenting our understanding of the impact of this therapy and for formulating guidelines for its use. Additionally, the potential exists that we did not capture all adverse events, but we achieved high rates of follow-up and had stent thrombosis events adjudicated. Our longest follow-up interval was 1 year; there would be considerable value in more extended follow-up. Outcome beyond 1 year would be quite desirable, given recent awareness of very late DES thrombosis.
DEScover provides considerable information regarding PCI as performed in the era of DES, the usage patterns of DES, and the outcomes of patients treated with DES in the United States. Our findings support the use of DES for the types of patients enrolled in DEScover and the manner in which physicians selected their patients for treatment.
Source of Funding
This study was funded by Cordis Corporation (a Johnson & Johnson Company), the manufacturer of the Cypher stent.
Drs Williams and Kip have received research support from Cordis Corporation (a Johnson & Johnson Company), and Dr Williams has served as a consultant for Cordis Corporation. Dr Abbott reports no conflicts.
The online-only Data Supplement, consisting of a complete list of clinical sites and investigators, is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.106.667915/DC1.
Moses JW, Leon MB, Popma JJ, Fitzgerald PJ, Holmes DR, O’Shaughnessy C, Caputo RP, Kereiakes DJ, Williams DO, Teirstein PS, Jaeger JL, Kuntz RE, 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.
Schofer J, Schluter M, Gershlick AH, Wijns W, Garcia E, Schampaert E, Breithardt G, for the E-SIRIUS Investigators. Sirolimus-eluting stents for treatment of patients with long atherosclerotic lesions in small coronary arteries: double-blind, randomised controlled trial (E-SIRIUS). Lancet. 2003; 362: 1093–1099.
Schampaert E, Cohen EA, Schluter M, Reeves F, Traboulsi M, Title LM, Kuntz RE, Popma JJ, for the C-SIRIUS Investigators. 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). J Am Coll Cardiol. 2004; 43: 1110–1115.
Stone GW, Ellis SG, Cox DA, Hermiller J, O’Shaughnessy C, Mann JT, Turco M, Caputo R, Bergin P, Greenberg J, Popma JJ, Russell ME, for the TAXUS-IV Investigators. One-year clinical results with the slow-release, polymer-based, paclitaxel-eluting TAXUS stent: the TAXUS-IV trial. Circulation. 2004; 109: 1942–1947.
Ong AT, Serruys PW, Aoki J, Hoye A, van Mieghem CA, Rodriguez-Granillo GA, Valgimigli M, Sonnenschein K, Regar E, van der Ent M, de Jaegere PP, McFadden EP, Sianos G, van der Giessen WJ, de Feyter PJ, van Domburg RT. The unrestricted use of paclitaxel- versus sirolimus-eluting stents for coronary artery disease in an unselected population: one-year results of the Taxus-Stent Evaluated at Rotterdam Cardiology Hospital (T-SEARCH) registry. J Am Coll Cardiol. 2005; 45: 1135–1141.
Urban P, Gershlick AH, Guagliumi G, Guyon P, Lotan C, Schofer J, Seth A, Sousa JE, Wijns W, Berge C, Deme M, Stoll HP, for the e-Cypher Investigators. Safety of coronary sirolimus-eluting stents in daily clinical practice: one-year follow-up of the e-Cypher registry. Circulation. 2006; 113: 1434–1441.
Lemos PA, Serruys PW, van Domburg RT, Saia F, Arampatzis CA, Hoye A, Degertekin M, Tanabe K, Daemen J, Liu TK, McFadden E, Sianos G, Hofma SH, Smits PC, van der Giessen WJ, de Feyter PJ. Unrestricted utilization of sirolimus-eluting stents compared with conventional bare stent implantation in the “real world”: the Rapamycin-Eluting Stent Evaluated at Rotterdam Cardiology Hospital (RESEARCH) registry. Circulation. 2004; 109: 190–195.
Saia F, Piovaccari G, Manari A, Santarelli A, Benassi A, Aurier E, Sangiorgio P, Tarantino F, Geraci G, Vecchi G, Guastaroba P, Grilli R, Marzocchi A. Clinical outcomes for sirolimus-eluting stents and polymer-coated paclitaxel-eluting stents in daily practice: results from a large multicenter registry. J Am Coll Cardiol. 2006; 48: 1312–1318.
Windecker S, Remondino A, Eberli FR, Juni P, Raber L, Wenaweser P, Togni M, Billinger M, Tuller D, Seiler C, Roffi M, Corti R, Sutsch G, Maier W, Luscher T, Hess OM, Egger M, Meier B. Sirolimus-eluting and paclitaxel-eluting stents for coronary revascularization. N Engl J Med. 2005; 353: 653–662.
Mehilli J, Kastrati A, Wessely R, Dibra A, Hausleiter J, Jaschke B, Dirschinger J, Schomig A, for the Intracoronary Stenting and Angiographic Restenosis–Test Equivalence Between 2 Drug-Eluting Stents (ISAR-TEST) Trial Investigators. Randomized trial of a nonpolymer-based rapamycin-eluting stent versus a polymer-based paclitaxel-eluting stent for the reduction of late lumen loss. Circulation. 2006; 113: 273–279.
Kalbfleisch JD, Prentice RL. Statistical Analysis of Failure Time Data. New York, NY: Wiley; 1980.
Cox DR. Regression models and life tables (with discussion). J Royal Stat Soc B. 1972; 34: 187–220.
Spaulding C, Henry P, Teiger E, Beatt K, Bramucci E, Carrie D, Slama MS, Merkely B, Erglis A, Margheri M, Varenne O, Cebrian A, Stoll HP, Snead DB, Bode C, for the TYPHOON Investigators. Sirolimus-eluting versus uncoated stents in acute myocardial infarction. N Engl J Med. 2006; 355: 1093–1104.
Liistro F, Fineschi M, Angioli P, Sinicropi G, Falsini G, Gori T, Ducci K, Bravi A, Bolognese L. Effectiveness and safety of sirolimus stent implantation for coronary in-stent restenosis: the TRUE (Tuscany Registry of Sirolimus for Unselected In-Stent Restenosis) Registry. J Am Coll Cardiol. 2006; 48: 270–275.
Holmes DR Jr, Teirstein P, Satler L, Sketch M, O’Malley J, Popma JJ, Kuntz RE, Fitzgerald PJ, Wang H, Caramanica E, Cohen SA, for the SISR Investigators. Sirolimus-eluting stents vs vascular brachytherapy for in-stent restenosis within bare-metal stents: the SISR randomized trial. JAMA. 2006; 295: 1264–1273.
Stone GW, Ellis SG, O’Shaughnessy CD, Martin SL, Satler L, McGarry T, Turco MA, Kereiakes DJ, Kelley L, Popma JJ, Russell ME, for the TAXUS V ISR Investigators. Paclitaxel-eluting stents vs vascular brachytherapy for in-stent restenosis within bare-metal stents: the TAXUS V ISR randomized trial. JAMA. 2006; 295: 1253–1263.