Outcomes of Unselected Recipients of Sirolimus-Eluting Stents: The Cypher® Stent U.S. Post-Marketing Surveillance Registry

ABSTRACT: Objective. To examine the 1-year safety and clinical outcomes associated with the post-marketing early unselected use of sirolimus-eluting stents (SES) in the United States. Background. The safety and effectiveness of SES has been assessed in selected patients enrolled in pivotal randomized trials. This PMS registry was initiated to examine the safety and effectiveness of SES in an unselected population. Methods. Consecutive patients who underwent implantation of ≥1 SES at 38 participating U.S. centers were enrolled in this registry. Results were compared according to “off-” versus “on-label” use of SES. Multivariate regression analyses were carried out in search of predictors of 1-year MACE and stent thrombosis. Results. The mean age of the 2,067 patients (3,367 treated lesions) was 63.7 years. The 12-month follow up was completed by 1,964 patients (95%). SES were implanted for “off-label” indications in 1,173 patients (57%). The 12-month rates of MACE and TLR in that subgroup were 9.2% and 6.2% (p

J INVASIVE CARDIOL 2010;22:48–55

Key words: coronary artery disease; coronary stenosis; drug-eluting stent; percutaneous coronary intervention

___________________________________________________________
ABBREVIATIONS LIST ARC = Academic Research Consortium BMS = bare-metal stent CEC = Clinical Events Committee CI = confidence interval DES = drug-eluting stent MACE = major adverse cardiac events MI = myocardial infarction OR = odds ratio PCI = percutaneous coronary intervention QCA = quantitative coronary angiography SD = standard deviation SES = sirolimus-eluting stent SVG = saphenous vein graft TLR = target lesion revascularization TVR = target vessel revascularization
The approval of the sirolimus-eluting stent (SES) for clinical use was based on safety and effectiveness outcomes observed in pivotal randomized trials, which compared their performance with that of bare-metal stents for the treatment of de novo native coronary artery stenosis.^1,2 These initial randomized trials enrolled low to moderate-risk patients, who do not represent the general population undergoing percutaneous coronary intervention (PCI) in the “real world.” Therefore, the U.S. Food and Drug administration required a post-marketing surveillance study after approval of the Cypher^® SES (Cordis Corp., a Johnson and Johnson Company, Miami Lakes, Florida) in the United States to monitor its unrestricted utilization in a large population. This was a fully monitored and adjudicated protocol to examine the safety and clinical outcomes associated with the broad, unselected clinical use of SES in a large sample of U.S. hospitals to identify uncommon device-related adverse events that might not have been noted during the conduct of the initial randomized trials.

Patient Population and Methods

The U.S. Cypher Post Marketing Surveillance Registry prospectively enrolled consecutive patients who underwent implantation of ≥ 1 SES at 38 participating medical centers, representing all geographical regions of the continental United States (Appendix). All eligible patients were included, whether the indication for SES implantation was “on-” or “off-label.” An indication was considered “off-label” if: a) the treated lesion was in the left main coronary artery, at a bifurcation or ostial region, in a previously implanted stent or in a saphenous vein graft; b) the treated vessel was totally occluded; c) the procedure was performed within 24 hours of an acute myocardial infarction (MI); d) the vessel reference diameter was 3.5 mm; e) the lesion length was > 30 mm; or f) multiple vessels were treated. All patients had granted their informed consent to be included in the registry, and its protocol was reviewed and approved by the Research Ethics Committee of each participating institution. Objectives. The purpose of this registry was to collect post-marketing surveillance data of the early use of SES in “real world” U.S. practice to 1) ascertain the safety of SES and examine the clinical outcomes associated with their use; 2) identify uncommon device-related adverse events that may not have been noted during controlled clinical trials; and 3) collect data on procedural, epidemiological and demographic characteristics, and assess their impact on the outcome of the interventions. Patient Management and Follow-up. The stenting procedure was performed, and peri-procedural medications were prescribed according to the usual practice of each implanting physician, who; however, were advised to adopt the practices described in the Instructions for Use of the stent issued by the study sponsor. Procedural success was defined as the achievement of a Antithrombotic therapy. The prescription of antithrombotic medications was left to the discretion of each institution or cardiologist participating in the registry. However, as in the case of stent use, prescribing physicians were recommended to follow the guidelines prepared by the sponsor, that is, at least 3 months of clopidogrel and indefinite use of aspirin. Data were collected with respect to the type of medications administered at the time of discharge from the hospital, and at each subsequent follow up. Data collection, quality control and study monitoring. The registry and post-marketing surveillance electronic data capture database were developed and implemented by Clinsights, an independent contract research organization (PPD, GlobalView Web Services, New Hope, Minnesota). This database was tested and fully validated by both Clinsights and the study sponsor in multiple audits. The Registry was under the supervision of an Advisory Board (Appendix), who could request a safety overview and review procedural data, treatment techniques, epidemiological observations and patient demographic information, to independently oversee the proper conduct of the registry and critically analyze the outcomes data. Data were collected without a pre-specified endpoint, with a view to enable multiple analyses. Baseline demographic, clinical and lesion information, procedural data and results of measurements of cardiac enzymes and blood lipids were recorded, as well as short- and long-term adverse events and follow-up information during clinic visits or by telephone at 1, 6 and 12 months. On-site monitoring visits were performed at ≤ 3 month intervals, and the database was remotely monitored on an ongoing basis. Ultimately, 100% of the case report forms were verified by review of the source documents. All clinical events were adjudicated by an independent Clinical Events Committee (CEC; Appendix) composed of 3 interventional cardiologists considered to be experts in the diagnosis and management of coronary artery disease, and experienced in the use of interventional cardiovascular and endovascular devices. The committee was supported by a project manager, a research associate, a medical safety officer, a safety data manager and additional support staff, as necessary, whose efforts were dedicated to the facilitation of timely review of cases identified as having potential study endpoints, by providing supporting information, including laboratory results, procedure notes, and angiograms sent to the core laboratory. The adjudications, which included 100% of clinical events, were based on the information contained in the electronic database and on the responses to queries addressed to the participating study sites. This information was gathered by questionnaires sent to the sites by the sponsor, who coordinated the data acquisition. Definitions of clinical events. Myocardial infarction was classified as Q-wave, when new, pathological Q waves in 2 or more contiguous leads of the surface electrocardiogram were accompanied by a rise in creatine kinase (CK) or CK-MB concentrations above normal; or non-Q-wave when the CK concentrations were greater than twice the upper normal limit with elevated CK-MB, in absence of new pathological Q waves (Cardiac enzymes quantification post percutaneous intervention was not mandatory per protocol, and it was left to the discretion of the physicians). Stent thrombosis was defined by the protocol as any death not attributed to a non-cardiac cause, MI (Q-wave or non-Q wave) in the territory of the stented vessel, or target vessel occlusion requiring TLR, occurring within 0–24 hours post index procedure (acute) or >24 hours to 30 days after stent deployment (subacute). Late stent thrombosis was defined as any MI attributable to the target vessel with angiographic documentation of thrombus or total occlusion at the target site > 30 days after the index procedure in the absence of an intervening revascularization of the target vessel. Stent thrombosis was also classified according to the academic research consortium (ARC) definition; as definitive = acute coronary syndrome and angiographic or pathologic evidence of stent thrombosis; probable = unexplained death within 30 days or Q-wave MI in the distribution of the stented artery; and possible = unexplained death after 30 days of stent placement. TLR was defined as any clinically-driven repeat treatment of the target lesion or bypass graft surgery on the target vessel. Clinically-driven was defined as prompted by a) results of a non-invasive study indicative of stress-induced myocardial ischemia; or b) symptoms or ECG changes in the distribution of the target vessel consistent with myocardial ischemia, and the presence of a ≥ 50 in-lesion %DS by quantitative coronary angiography (QCA); or c) TLR for > 70%DS by QCA, in absence of other manifestations of myocardial ischemia. Non-emergent repeat TLR for a Statistical analysis. The statistical design of the registry was prepared, and the data were analyzed, by an independent organization (Clinsight). Descriptive statistics are presented. Continuous variables are expressed as means ± standard deviation (SD) and categorical variables as percentages. The chi-square test was used for comparisons between proportions and Student’s t-test was used for comparisons of mean values. When appropriate, odds ratio (OR) and 95% confidence intervals (CI), are presented. Single and multiple variable logistic regression analysis, including demographic, clinical, lesion and procedural observations, was performed in search of predictors of MACE and stent thrombosis. Multiple variable predictors were chosen by a stepwise procedure using an entry criterion of 0.10 with a stay criterion 0.05. P-values Results Overall population Characteristics of the patient sample and follow-up compliance. The registry included 2,067 patients who received ≥ 1 SES for the treatment of 3,367 lesions. The 12-month follow-up was completed by 1,964 patients (95%). All clinical events were adjudicated by the CEC. The patient demographic, clinical and baseline characteristics are presented in Table 1. The characteristics of the 2,067 index procedures and 3,367 treated lesions are shown in Table 2. The vast majority of lesions were type B2/C (76%) and in 40% of the cases more than one lesion was treated. Outcomes of the overall registry sample up to 1 year In-hospital. One patient died from a Q-wave MI during the index hospitalization. Other MACE that occurred before discharge from the hospital included 11 (2 Q-wave and 9 non-Q-wave) MI (0.5%), and 5 (4 percutaneous and 1 surgical) TLR (0.2%). In addition, in-hospital acute and subacute protocol defined stent thromboses occurred in 3 (0.1%) and 2 (0.1%) patients, respectively. Post discharge. The cumulative 12-month rates of MACE, TLR and stent thrombosis were 7.3%, 4.6% and 1.2%, respectively (Table 3). Antithrombotic therapy. At 1 month, 97.7% of patients were being treated with a combined aspirin and clopidogrel or ticlopidine antithrombotic regimen. At 6 and 12 months, 75.9% and 61.6% remained on combined antithrombotic therapy, while 24.1% and 38.4% of patients, respectively, were being treated with aspirin alone. Clinical characteristics and outcomes of patients treated for “off-label” indications versus “on-label” indications. Baseline clinical characteristics of 1,173 patients who underwent SES implantation for at least 1 “off-label” indication, versus those of 894 patients who were treated for approved indications are compared in Table 1. Patients in the “off-label” indications group were more likely to be men with histories of coronary artery bypass graft surgery or disease of multiple coronary arteries and more frequently presented as unstable angina, though less likely to undergo stent implantation for silent ischemia. Baseline lesions, procedural and stents characteristics of the patients treated for “off-label” indications, versus patients treated for approved indications are compared in Table 2. Patients with “off-label” indications were more likely to have more than one lesion treated. The baseline medication was similar in both groups, except for the higher percent of anti-anginal medication, aspirin and IIb/IIIa utilization in the “off-label” population, (62% vs. 55.6%, p = 0.003; 99.1% vs. 97.9%, p = 0.025 and 45% vs. 37.9%, p = 0.001). It is noteworthy that, after the procedure, TIMI flow 3 was similar in both groups but the rate of procedural success favored the “on-label” group (89.1% vs. 95.6%, p 3.5 mm and lesions related with acute myocardial infarction. Mortality rates were higher in patients treated because of in-stent restenosis, bifurcation lesions and SVG lesions. Protocol defined stent thrombosis to 12 months was identical between the “off-label” and approved indications group (0.6%). The incidence of any ARC type of stent thrombosis was higher in the “off-label” group (2.4% vs. 0.8%, p = 0.005), definitive or probable stent thrombosis was higher in the “off-label” population (1.6% vs. 0.6%, p = 0.026). Late ST occurred more frequently in the “off label” group (0.9% vs. 0.1%, p = 0.014). Subgroups with higher incidence of late stent thrombosis were: total occlusion (1.8% vs 0.1%, p = 0.001), lesion length > 30 mm (2.8% vs. 0.6%, p = 0.014), ostial (1.7% vs. 0.1%, p = 0.001), bifurcation lesions (2.1% vs. 0.1%, p Diabetic sample. A total of 640/2012 (32%) included in this registry were diabetic. These patients had a higher incidence of prior bypass surgery (40.3% vs. 28.7%, p Predictors of MACE and stent thrombosis. A total of 33 demographic, clinical, angiographic and procedural variables were entered in the logistic regression analysis, in search of predictors of 360-day MACE (Table 4) and stent thrombosis (definitive or probable, ARC criteria at 360 days) (Table 5). None of the individual “off-label” indications was an independent predictor of MACE or stent thrombosis.

Discussion

The Cypher Post-Market Surveillance is the first large U.S. multicenter registry conducted after approval of SES. The rigorous data monitoring and meticulous adjudication of clinical events is unique among DES registries. This study reveals that > 55% of the early U.S. patients treated with SES were “off-label,” including multivessel and left main coronary disease, acute MI, saphenous vein graft disease, chronic total occlusions, and restenotic lesions. The high proportion of diabetic patients and the frequent implantation of SES without predilatation are also noteworthy. Similar practices have been observed in European registries.^3,4 This study also confirms the safety and efficacy of SES implanted for “on-label” indications, and reports clinical outcomes for “off-label” indications similar to that observed in the pre-approval controlled pivotal trials. Clopidogrel, aspirin and unfractionated heparin were the prevailing antithrombotic medications administered during SES implantation, with 40% of patients also treated with glycoprotein IIb/IIIa inhibitors. Despite a high proportion of high-risk patients treated with aspirin alone at 12 months, the cumulative incidence of stent thrombosis after the index procedure was low and similar to that observed in randomized clinical trials including mostly low-risk recipients of bare metal⁵ or drug-eluting stents.⁶ The present results compare favorably with those reported by U.S. PCI registries conducted before the availability of drug-eluting stents.^7-8 While their early results were similar to those observed in this registry, the 12-month incidence of MACE in the Dynamic registry was nearly 4 times higher.⁹ Likewise, the 1-year rate of MACE observed in this sample was similar to that observed in selected patients treated with SES in the SIRIUS trial,¹⁰ despite the higher-risk characteristics of the patients entered in this registry. These observations are concordant with those made in the worldwide e-Cypher registry,¹¹ as well as in the single-center European RESEARCH registry,³ both contemporary studies which included unselected patients treated with SES. Diabetes population. Diabetes is historically a strong predictor of adverse outcomes in patients undergoing percutaneous coronary intervention. DES has been shown to be superior to BMS in the treatment of both insulin-dependent and orally treated diabetic patients,^12,13 but diabetes remains a strong predictor of worse outcomes even in the DES era. The relatively low one-year rates of definite/probable stent thrombosis (1.3%) and cardiac mortality (2.0%) observed in the general diabetic population is encouraging. In spite of its efficacy in decreasing the need for repeat revascularization compared with BMS in insulin-dependent patients,¹³ the observed 4% rate of cardiac death among those taking insulin in the present study highlights that local vascular therapy alone may not be enough for this high-risk subgroup of patients. “On-label” vs. “off –label” SES indications. In a simplistic view, “off-label” indications refer to treatment of conditions not present in the “Instructions for Use” in the device package.¹⁴ Likely a result of the selective nature of initial pivotal trials, the “off-label” category represents a large proportion of patients treated in everyday practice, and includes a very heterogeneous group of clinical and anatomic scenarios. These off-label situations have been implicated with worse clinical outcomes after DES, but these results may be associated with comorbidities and not related to a specific off-label indication. The off-label indications with higher incidence of 1-year MACE in the univariate analysis were: lesion length > 30 mm and restenotic lesions. However, these indications were not independently associated with MACE or stent thrombosis. Previous reported predictors of stent thrombosis, such as total stent length,^5,6 were not found to influence thrombosis outcomes in this large population. Indeed, outcomes have been shown to be similar for on-label and off-label indications when adjusted for other confounding factors.¹⁵ Outcomes were influenced by systemic conditions such as diabetes mellitus. Our data support the notion that the term “label” and “clinically appropriate” are not interchangeable. The latter represents a more comprehensive assessment that includes symptoms, anatomical and clinical characteristics that impact clinical outcomes, beyond device labeling. Further investigations on the best therapeutic approach for patients with diabetes mellitus, particularly those taking insulin, are warranted given the high-risk profiles of these patients. The analysis of this registry, which is representative of the broad use of drug-eluting stents in the United States, revealed a high “off-label” implantation rate of SES. Overall, the outcomes observed in the present study are somewhat reassuring, as off-label treatment remains an important component of medical practice in the U.S. and worldwide. Study limitations. Most of the current practice adopts longer duration of dual antiplatelet therapy than used in the present study, which could have impacted the observed incidence of late thrombosis. The average maximum inflation pressure for stent deployment of 10 atm accompanied by only 38% of post-dilatation had dramatically changed in the last years. Unfortunately, the effect of these changes is not measurable. A relatively small number of patients presenting with acute myocardial infarction were included in this registry, and this limits conclusions on the safety of the device on this sub-population.

References

1. Morice MC, Serruys PW, Sousa JE, et al. A randomized comparison of a sirolimus-eluting stent with a standard stent for coronary revascularization. N Engl J Med 2002;346:1773–1780. 2. Moses JW, Leon MB, Popma JJ, et al. Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315–1323. 3. Lemos PA, Serruys PW, van Domburg RT, et al. 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. 4. Zahn R, Hamm CW, Zeymer U, et al. [Safety and current indications during “real life” use of sirolimus-eluting coronary stents in Germany. Results from the prospective multicenter German Cypher Registry]. Herz 2004;29(2):181–186. 5. Cutlip DE, Baim DS, Ho KK, et al. Stent thrombosis in the modern era: A pooled analysis of multicenter coronary stent clinical trials. Circulation 2001;103:1967–1971. 6. Moreno R, Fernandez C, Hernandez R, et al. Drug-eluting stent thrombosis: Results from a pooled analysis including 10 randomized studies. J Am Coll Cardiol 2005;45:954–959. 7. Shaw RE, Anderson HV, Brindis RG, et al. Updated risk adjustment mortality model using the complete 1.1 dataset from the American College of Cardiology National Cardiovascular Data Registry (ACC-NCDR). J Invasive Cardiol 2003;15:578–580. 8. Cohen HA, Williams DO, Holmes DR Jr, et al. Impact of age on procedural and 1-year outcome in percutaneous transluminal coronary angioplasty: A report from the NHLBI Dynamic Registry. Am Heart J 2003;146:513–519. 9. Al Suwaidi J, Yeh W, Cohen HA, et al. Immediate and one-year outcome in patients with coronary bifurcation lesions in the modern era (NHLBI dynamic registry). Am J Cardiol 2001;87:1139–1144. 10. Holmes DR Jr, Leon MB, Moses JW, et al. Analysis of 1-year clinical outcomes in the SIRIUS trial: A randomized trial of a sirolimus-eluting stent versus a standard stent in patients at high risk for coronary restenosis. Circulation 2004;109:634–640. 11. Urban P, Gershlick AH, Guagliumi G, et al. Safety of coronary sirolimus-eluting stents in daily clinical practice: One-year follow-up of the e-Cypher registry. Circulation 2006;113:1434–1441. 12. Garg P, Normand SL, Silbaugh TS, et al. Drug-eluting or bare-metal stenting in patients with diabetes mellitus: Results from the Massachusetts Data Analysis Center Registry. Circulation 2008;118:2277-85, 7p following 2285. 13. Mulukutla SR, Vlachos HA, Marroquin OC, et al. Impact of drug-eluting stents among insulin-treated diabetic patients A report from the NHLBI Dynamic Registry. JACC Cardiovasc Interv 2008;1:139–147. 14. Price MJ, Teirstein PS. The off- versus on-label use of medical devices in interventional cardiovascular medicine: Clarifying the ambiguity between regulatory labeling and clinical decision-making, Part 1: PCI. Catheter Cardiovasc Interv 2008;72:500–504. 15. Beohar N, Davidson CJ, Kip KE, et al. Outcomes and complications associated with off-label and untested use of drug-eluting stents. JAMA 2007;297:1992–2000.

__________________________________________________________ From ^aUniversity Hospitals Case Medical Center, Cleveland, Ohio; ^bTexas Heart Institute, Houston, Texas; ^cGeisinger Clinic, Danville, Pennsylvania; ^dEmory University, Atlanta, Georgia; ^eOchsner Clinic, New Orleans, Louisiana; ^fNorth Shore University Hospital, Manhasset, New York; ^gMaine Medical Center, Portland, Maine; ^hCardiovascular Medicine, PC, Davenport, Iowa; ⁱRush University Medical Center, Chicago, Illinois; ^jCordis Corp., Bridgewater, New Jersey. Manuscript submitted August 7, 2009, provisional acceptance given September 15, 2009, final version accepted November 3, 2009. Clinicaltrials.gov identifier NCT00245401 Address for correspondence: Prof. Marco A. Costa, University Hospitals Case Medical Center, Cardiovascular Dept., 11100 Euclid Avenue, Cleveland, OH 44106-6031. E-mail: Marco.Costa@UHhospitals.org