Three patients were not included in the 6-month analyses because their follow-up took place after 27 days but before 30 days, and so by protocol definitions could not be included. Of the 56 patients without 1-year data (37 unfractionated heparin–assigned and 19 enoxaparin-assigned), 11 were excluded due to withdrawal of consent and 45 were lost to follow-up.
A, Death or nonfatal MI through 180-day follow-up by intention-to-treat. B, Death or nonfatal MI through 180-day follow-up in patients treated with consistent therapy, defined as patients not receiving prerandomization therapy or assigned same therapy as those receiving prerandomization therapy. C, All-cause death in the overall patient population. D, All-cause death in the consistent-therapy groups. CI indicates confidence interval; HR, hazard ratio.
CABG indicates coronary artery bypass graft surgery; ECG, electrocardiographic; PCI, percutaneous coronary intervention.
Mahaffey KW, Cohen M, Garg J, Antman E, Kleiman NS, Goodman SG, Berdan LG, Reist CJ, Langer A, White HD, Aylward PE, Col JJ, Ferguson JJ, Califf RM, SYNERGY Trial Investigators FT. High-Risk Patients With Acute Coronary Syndromes Treated With Low-Molecular-Weight or Unfractionated HeparinOutcomes at 6 Months and 1 Year in the SYNERGY Trial. JAMA. 2005;294(20):2594-2600. doi:10.1001/jama.294.20.2594
Author Affiliations: Duke Clinical Research Institute, Durham, NC (Drs Mahaffey, Reist, and Califf and Mss Garg and Berdan); Heart Hospital of New Jersey, Newark Beth Israel Medical Center, Newark (Dr Cohen); Cardiovascular Division, Department of Medicine, Brigham and Women’s Hospital, Boston, Mass (Dr Antman); Baylor College of Medicine, Houston, Tex (Dr Kleiman); Canadian Heart Research Centre, Division of Cardiology, St Michael’s Hospital, University of Toronto, Toronto, Ontario (Drs Goodman and Langer); Green Lane Cardiovascular Service, Auckland City Hospital, Auckland, New Zealand (Dr White); Flinders Medical Centre, Bedford Park, South Australia (Dr Aylward); Clinique Universitaire St Luc, Brussels, Belgium (Dr Col); and Texas Heart Institute, Houston (Dr Ferguson).
Context The SYNERGY trial comparing enoxaparin and unfractionated heparin in high-risk patients with acute coronary syndromes (ACS) showed that enoxaparin was not inferior to unfractionated heparin in reducing death or nonfatal myocardial infarction (MI) at 30 days.
Objective To evaluate continued risk in this patient cohort through 6-month and 1-year follow-up.
Design, Setting, and Patients Overall, 9978 patients were randomized from August 2001 through December 2003 in 487 hospitals in 12 countries. Patients were followed up for 6 months and for 1 year.
Main Outcome Measures Six-month outcomes were death, nonfatal MI, revascularization procedures, stroke, and site-investigator–reported need for rehospitalization; 1-year outcome was all-cause death.
Results Six-month and 1-year follow-up data were available for 9957 (99.8%) and 9608 (96.3%) of 9978 patients, respectively; 541 patients (5.4%) had died at 6 months and 739 (7.4%) at 1 year. Death or nonfatal MI at 6 months occurred in 872 patients receiving enoxaparin (17.6%) vs 884 receiving unfractionated heparin (17.8%) (hazard ratio [HR], 0.98; 95% confidence interval [CI], 0.89-1.07; P = .65). In the subgroup of patients receiving consistent therapy, ie, only enoxaparin or unfractionated heparin during the index hospitalization (n = 6138), a reduction in death or nonfatal MI with enoxaparin was maintained at 180 days (HR, 0.85; 95% CI, 0.75-0.95; P = .006). Rehospitalization within 180 days occurred in 858 patients receiving enoxaparin (17.9%) and 911 receiving unfractionated heparin (19.0%) (HR, 0.94; 95% CI, 0.85-1.03; P = .17). One-year all-cause death rates were similar in the 2 treatment groups (380/4974 [7.6%] for enoxaparin vs 359/4948 [7.3%] for unfractionated heparin; HR, 1.06; 95% CI, 0.92-1.22; P = .44). One-year death rates in patients receiving consistent therapy were also similar (251/3386 [7.4%] for enoxaparin vs 213/2720 [7.8%] for unfractionated heparin; HR, 0.95; 95% CI, 0.79-1.14; P = .55).
Conclusions In the SYNERGY trial, patients continued to experience adverse cardiac events through long-term follow-up. The effect of enoxaparin on death or MI compared with that of unfractionated heparin at 6 months was similar to that observed at 30 days in the overall trial and in the consistent-therapy group. One-year death rates were also similar in both groups. High-risk patients with ACS remain susceptible to continued cardiac events despite aggressive therapies.
ClinicalTrials.gov Identifier: NCT00043784.
Patients with non–ST-segment elevation (NSTE) acute coronary syndromes (ACS) comprise a spectrum of risk for adverse cardiac events. In the Superior Yield of the New Strategy of Enoxaparin, Revascularization, and Glycoprotein IIb/IIIa Inhibitors (SYNERGY) trial, patients at high risk for recurrent ischemic cardiac events were randomly assigned to receive low-molecular-weight heparin (enoxaparin) or unfractionated heparin. These patients were treated with an early invasive management strategy. The primary results of the trial end point at 30 days showed that enoxaparin was at least as effective as unfractionated heparin.1
Six-month and 1-year follow-up were preplanned as part of the SYNERGY investigation. We believe this to be valuable, given the high-risk clinical characteristics of the patient population in SYNERGY and the need to understand the long-term outcomes in patients managed with an early aggressive invasive treatment strategy.
The rationale and design as well as the primary results of SYNERGY have been reported previously.1,2 The protocol was approved by the institutional review board or ethics committee at each participating center; all patients provided written informed consent. Patients were randomly assigned from August 2001 through December 2003 to receive enoxaparin or unfractionated heparin and remained eligible if they had started antithrombin therapy at the time of enrollment. The study drug was to be given immediately after randomization and continued until any cardiac procedures were performed or until no further antithrombin was needed.
All patients not allergic to aspirin were to receive aspirin at a dose of 162 to 325 mg/d. Patients were also to be treated with clopidogrel. All other medications through 6 months were decided by the treating clinician with recommendation to follow practice guidelines.3,4
Patients were contacted by telephone or seen in the clinic at 30 days after enrollment (minimum, ≥27 days) and again at 180 days (minimum, ≥120 days). Clinical summaries were collected and renewed for all patients with death or myocardial infarction (MI) between 30- and 180-day follow-up. Only survival status was assessed by telephone at 1 year after enrollment (minimum, ≥10 months). If the patient could not be contacted by telephone, a review of medical records or national death indexes was performed or, for some patients in the United States, a private locator service was used.
Key secondary analyses per protocol include the combined incidence of death or nonfatal MI at 6 months and the incidence of death at 6 months and at 1 year. Overall, 10 027 patients were enrolled in the study, but 9978 were included in the primary efficacy analyses because the first 49 patients enrolled in one country were not randomly assigned due to an error with the interactive voice-activated randomization system. All analyses are based on the intention-to-treat principle; ie, all patients are analyzed as randomized.
The Kaplan-Meier method was used to estimate the probability of death or nonfatal MI through 180 days. The Cox proportional hazards model was used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs). Clinical events committee–adjudicated data were used for MI events occurring during the first 30 days. Events that occurred between 30- and 180-day follow-up were reported by the investigators using standard data collection tools. Rates of death and nonfatal MI were also calculated using the investigator-reported events from enrollment through 180 days. Efficacy comparisons were the time to first event in the treatment groups based on an intention-to-treat strategy using the log-rank test.
Exploratory analyses that were prespecified in the protocol or statistical analysis plan were also performed based on the actual treatment assignment received. Patient groups based on the use of prerandomization antithrombin therapy were defined. Long-term outcomes were evaluated in patients who received no prerandomization therapy (n = 2440) and in patients with investigator-defined consistent therapy (n = 6138), defined as an investigator indicating on the case report form that a patient had received no pretreatment at all or pretreatment in the emergency department using the same drug as the randomized therapy. This patient group, the consistent-therapy group, has been evaluated in detail in a previous report.5
All analyses were performed using SAS version 8.2 (SAS Institute Inc, Cary, NC); P<.05 without adjustment for multiple comparisons was used to determine statistical significance.
Figure 1 shows patient follow-up through 1 year after enrollment. Table 1 shows the proportion of patients with death and death or MI at 30-day and 6-month follow-up and the proportion of patients dead at 30 days, 6 months, and 1 year, by inclusion criteria. Overall, 314 patients died by 30-day follow-up, 227 patients died between 31 and 180 days after enrollment, and 198 patients died between 180-day and 1-year follow-up.
Adverse cardiovascular events continued to accrue in this high-risk patient population between 31 and 180 days after enrollment. Overall, 1769 of 9583 patients (18.5%) were readmitted to the hospital between discharge from the index hospitalization and 6 months, with similar readmission rates in the treatment groups (858/4796 (17.9%) for enoxaparin vs 911/4787 (19.0%) for unfractionated heparin (hazard ratio [HR], 0.94; 95% confidence interval [CI], 0.85-1.03; P = .17). Overall, 61% of the readmissions were for primary cardiovascular events as identified by the site investigator.
Table 2 shows 6-month death or MI outcomes according to treatment assignment. This combined end point occurred in 872 of 4993 patients (17.6%) assigned to receive enoxaparin and 884 of 4985 patients (17.8%) assigned to receive unfractionated heparin through 180 days after enrollment. Stroke was infrequent during 6-month follow-up and occurred in only 148 (1.5%) of all patients, without difference between treatment groups (73 [1.5%] for enoxaparin vs 75 [1.6%] for unfractionated heparin; P = .86). Figure 2 shows the Kaplan-Meier curves for death or nonfatal MI through 180-day follow-up in the overall study population and in the consistent-therapy group, as well as curves for all-cause mortality through 1 year. Six-month hazards for death or MI outcomes for baseline characteristics are shown in Figure 3.
In the group receiving no prerandomization therapy, death or MI occurred at 6 months in 16.0% of patients assigned to receive enoxaparin and 18.1% of patients assigned to receive unfractionated heparin (HR, 0.87; 95% CI, 0.72-1.06).
Percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) surgery was performed in the course of treatment in 73% of patients through 6-month follow-up (Table 2). Of the patients undergoing PCI and CABG surgery between 31 and 180 days, 59 of 78 (75.6%) and 123 of 199 (61.8%) were related to ischemia, respectively.
At 1 year, overall mortality occurred in 739 (7.4%) of 9922 patients and was similar between the 2 treatment groups (Table 2). One-year death rates in patients treated only with enoxaparin or unfractionated heparin (n = 6138) were also similar (251/3386 [7.4%] for enoxaparin vs (213/2720 [7.8%] for unfractionated heparin; HR, 0.95; 95% CI, 0.79-1.14; P = .55).
Overall, 92% of patients were taking aspirin at discharge; 54%, clopidogrel; 85%, a β-blocker; 73%, an angiotensin-converting enzyme inhibitor; and 81%, a statin, without differences between treatment groups. Between 75% and 95% of patients discharged while taking these medications were still taking them at 30-day evaluations. Selected medication was collected through 30-day follow-up only.
The 6-month and 1-year follow-up data from the SYNERGY trial population reveal that this cohort of patients with ACS was at high risk for recurrent cardiovascular ischemic complications and procedures. Overall, nearly 18% died or experienced nonfatal MI through 6 months of follow-up and 7.4% died by 1-year follow-up, despite aggressive coronary revascularization and high use of evidence-based therapies at the time of hospital discharge. More than 70% of patients underwent percutaneous or surgical revascularization during the 6 months after enrollment; of these, 10% had their first revascularization procedure after discharge. Also, of patients who underwent PCI during the index hospitalization, 8.5% underwent repeat PCI and 2.2% underwent CABG surgery between index hospital discharge and 180 days after enrollment.
In the overall intention-to-treat cohort, the rates of death and nonfatal MI at 180-day follow-up were similar in the 2 treatment groups. The Kaplan-Meier curves remain parallel over time, which is consistent with other long-term data on enoxaparin, other antithrombin agents, and glycoprotein IIb/IIIa inhibitors.6- 8 The benefit with enoxaparin observed at 30 days in the prespecified subgroups defined by prerandomization therapy and consistent therapy1 was maintained through 6-month follow-up. A statistically significant 15% relative risk reduction in death or nonfatal MI through 180 days was observed in patients who received consistent therapy as part of the initial drug assignment in the SYNERGY trial. There was no significant difference in 1-year death rates between treatments in this subset. These results are similar to the treatment effect seen with enoxaparin compared with unfractionated heparin in prior trials of patients with NSTE ACS.7 A comprehensive analysis of the no-prerandomization therapy group has shown similar results.5
In SYNERGY, death or nonfatal MI occurred in 14% to 15% of patients through 30-day evaluation, despite a large proportion of patients managed with an early invasive treatment strategy; more than 70% of patients underwent percutaneous or surgical revascularization and were treated with relatively high use of evidence-based therapies at discharge and through the first 30 days after enrollment. By 6 months, nearly 18% of all patients had died or had experienced a nonfatal MI. In addition, of those patients in SYNERGY with highest risk at baseline, defined by age 60 years or older, elevated levels of cardiac biomarkers, and electrocardiographic changes, 7.7% had died and 21% had either died or experienced nonfatal MI by 6-month follow-up, and 10.2% had died by 1 year. Therefore, despite contemporary revascularization techniques, glycoprotein IIb/IIIa inhibition, and other proven pharmacotherapies, these rates of morbidity and mortality are higher than those reported in virtually all contemporary NSTE ACS trials.9
In SYNERGY, MI within 30 days was adjudicated by a clinical events committee using an MI definition that required elevation of creatine kinase-MB levels at least 2 times the upper limit of normal for non–procedural-related MI events. This definition is more stringent than that recommended by the American College of Cardiology/European Society of Cardiology MI consensus document10 but was chosen to provide a robust assessment of treatment differences and allow more clarity in the identification of early recurrent events in the trial, since the majority of patients already had elevated levels of biomarkers at the time of enrollment from the index event. Nonfatal infarctions between 30- and 180-day follow-up were to be reported by the site investigators and were not adjudicated by a clinical events committee. Prior studies have shown that investigators underreport MIs in this patient population, so the risk for recurrent infarction may be underestimated.11 Methods to ascertain MI events over long-term follow-up are needed to better characterize the risk in this patient population during long-term care.
It is clear that patients with an ACS event with high-risk baseline features are at substantial risk for adverse cardiac events over time. The results from secondary prevention trials have shown benefits with long-term therapy with lipid-lowering agents, angiotensin-converting enzyme inhibitors, or antiplatelet therapies.12- 14 The impact of these treatments in this patient population needs to be specifically addressed in future studies, since we did not collect medication compliance and success rates with lifestyle modifications during follow-up. Others have shown that adherence to pharmacological and lifestyle interventions can be challenging but are associated with improved outcomes when used.15,16 These data and the SYNERGY results suggest that continued careful monitoring and reassessment of adequate secondary prevention strategies may be helpful.
The rate of revascularization procedures in patients with percutaneous procedures during the index hospitalization was approximately 10% between 31 days and 180 days. This is lower than has been reported in prior studies but is similar to that recently reported by the Randomized Evaluation in PCI Linking Angiomax to Reduced Clinical Events 2 (REPLACE-2) investigators, although the patient populations are not directly comparable.8 More aggressive use of antiplatelet, β-blocker, angiotensin-converting enzyme inhibitor, and statin therapies is a potential contributor to the low rate of repeat procedures. During the course of the SYNERGY trial, drug-eluting coronary stents became commercially available in both Europe and the United States, but information about the type of coronary stents used was not collected.
These analyses are limited by several issues. The open-label trial design could potentially bias the reporting of events over time due to knowledge of the treatment assignment, although this is less likely to affect reports of deaths and revascularization procedures than the more subjective end points of MI or recurrent ischemia. Follow-up was pursued aggressively in all patients. Overall, only 19 did not have any 6-month follow-up due to withdrawal of consent or inability to contact patients. Of the 56 patients without 1-year data (37 assigned to receive unfractionated heparin and 19 to receive enoxaparin), 11 were excluded due to withdrawal of consent and 45 were lost to follow-up. Implementation of Health Insurance Portability and Accountability Act (HIPAA) regulations in April 2003 during the trial required careful planning to achieve this level of long-term evaluation and data collection. Trial investigators should consider multiple strategies to perform long-term follow-up so the federal regulations for informed consent and HIPAA authorization can be followed.
The SYNERGY trial studied a high-risk cohort of patients with NSTE ACS. The 30-day, 6-month, and 1-year data show that this cohort of patients remains at substantial risk for recurrent cardiovascular events and coronary revascularization procedures: nearly 20% of patients died or experienced reinfarction by 6 months. Overall, the rates of death and nonfatal MI were similar at 6 months between treatment groups. The reduction in death or nonfatal MI at 30 days seen in the subgroup of patients treated with consistent therapy during the initial study drug assignment was sustained through 6 months, but mortality at 1 year was similar. Despite aggressive revascularization strategies and high use of evidence-based therapies, patients with high-risk ACS features remain at risk for continued adverse cardiac morbidity and mortality.
Corresponding Author: Kenneth W. Mahaffey, MD, Duke Clinical Research Institute, PO Box 17969, Durham, NC 27715 (email@example.com).
Author Contributions: Drs Mahaffey, Ferguson, and Califf had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Mahaffey, Cohen, Goodman, Berdan, Reist, White, Ferguson, Califf.
Acquisition of data: Mahaffey, Cohen, Kleiman, Berdan, Reist, Langer, White, Aylward, Col, Califf.
Analysis and interpretation of data: Mahaffey, Cohen, Garg, Antman, Kleiman, Goodman, Reist, Langer, White, Aylward, Ferguson, Califf.
Drafting of the manuscript: Mahaffey, Cohen, Aylward.
Critical revision of the manuscript for important intellectual content: Mahaffey, Cohen, Garg, Antman, Kleiman, Goodman, Berdan, Reist, Langer, White, Aylward, Col, Ferguson, Califf.
Statistical analysis: Mahaffey, Cohen, Garg, Antman.
Obtained funding: Berdan, Califf.
Administrative, technical, or material support: Mahaffey, Goodman, Berdan, Reist, Aylward, Califf.
Study supervision: Col, Ferguson.
Financial Disclosures: Dr Mahaffey has received funding from Sanofi-Aventis (research grants, consulting/honoraria, speakers bureau, advisory board), as have Drs Cohen (research grants, speakers bureau), Antman (research grants), Kleiman (research grants, consulting/honoraria, speakers bureau), Goodman (research grants, consulting/honoraria, speakers bureau), Langer (research grants, consulting/honoraria), White (research grants), Aylward (research grants, consulting/honoraria, speakers bureau), Col (consulting/honoraria), Ferguson (research grants, consulting-honoraria, speakers bureau), and Califf (research grants, consulting/honoraria).
Funding/Support: This study was funded by Aventis Pharmaceuticals, a member of the Sanofi-Aventis Group, Bridgewater, NJ.
Role of the Sponsor: Aventis Pharmaceuticals collaborated with the academic steering committee and principal investigators on all aspects of the trial, but trial operations were run through the Duke Clinical Research Institute. The manuscript was reviewed by the entire academic steering committee. Aventis Pharmaceuticals reviewed and provided comments on the manuscript but did not have veto power on submission or publication.
Acknowledgment: We thank all the SYNERGY investigators for their hard work to achieve exemplary follow-up in this clinical trial. We would like to especially thank Kimberly Schwabe and Dinah Ballon, Sanofi-Aventis, for their tireless efforts and Erin Allingham, BA, Kim Ferrer, BS, and Karen S. Pieper, MS, Duke Clinical Research Institute, for their expert editorial and design assistance. Finally, we would like to thank Louise Traylor, PhD, and Min Chen, MS, Sanofi-Aventis, for contributing their statistical expertise.