Author + information
- Received February 14, 2008
- Accepted March 19, 2008
- Published online August 1, 2008.
- Mark Y. Chan, MD, MHS⁎,
- Kenneth W. Mahaffey, MD, FACC⁎,
- Lena J. Sun, MS⁎,
- Karen S. Pieper, MS⁎,
- Harvey D. White, MD, FACC†,
- Philip E. Aylward, MD, FACC‡,
- James J. Ferguson, MD, FACC§,
- Robert M. Califf, MD, FACC⁎ and
- Matthew T. Roe, MD, MHS, FACC⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Matthew T. Roe, 2400 Pratt Street, Terrace Level, Room 0311, Durham, North Carolina 27705.
Objectives We sought to characterize the utilization and impact of a conservative medical management strategy for patients with non–ST-segment elevation acute coronary syndromes (NSTE ACS) and significant coronary artery disease on early angiography.
Background Practice guidelines recommend an early invasive management strategy for NSTE ACS, but revascularization procedures may not always be performed after early angiography, even when significant coronary artery disease is present.
Methods We evaluated 8,225 intermediate- to high-risk NSTE ACS patients with at least 1 coronary lesion >50% stenosis on early angiography from the SYNERGY (Superior Yield of the New Strategy of Enoxaparin, Revascularization, and Glycoprotein IIb/IIIa Inhibitors) trial (2001 to 2003), comparing patients treated with conservative medical management with those who underwent in-hospital percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG) within 7 days of randomization.
Results A total of 2,633 patients (32%) were medically managed, 4,294 (52%) underwent PCI, and 1,298 (16%) underwent CABG. The strongest independent predictors of conservative medical management versus any intervention were prior CABG, lower body weight, lack of a reinfarction between randomization and catheterization, and 3-vessel disease. With conservative medical management, the cumulative risk of 1-year mortality after discharge increased rapidly during the first 90 days and thereafter remained higher at 7.7% compared with that seen in patients treated with PCI (3.6%) or CABG (6.2%).
Conclusions One-third of patients with NSTE ACS and significant coronary disease on early angiography were managed without in-hospital revascularization in the SYNERGY trial, and these patients had an increased risk of late mortality. These findings highlight the need for novel treatment approaches for NSTE ACS patients who are not candidates for revascularization. (SYNERGY trial; NCT00043784)
- conservative medical management
- coronary artery disease
- non-ST-segment elevation acute coronary syndrome
Clinical practice guidelines recommend an early invasive management treatment strategy for intermediate- to high-risk patients with non–ST-segment elevation acute coronary syndromes (NSTE ACS) (1,2). However, angiography is a diagnostic procedure that does not confer treatment benefit if patients do not undergo a subsequent revascularization procedure (3). In 4 randomized trials comparing conservative medical management with early invasive management for NSTE ACS, the majority of patients had significant coronary disease identified on angiography in the early invasive treatment arms (75% to 87%), but the use of in-hospital percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG) varied from 59% to 82% (4–7). Data from a contemporary registry showed that approximately 25% of NSTE ACS patients with 3-vessel disease did not undergo revascularization (8). These data suggest that many NSTE ACS patients in contemporary practice with significant coronary artery disease (CAD) on early angiography do not undergo early revascularization, but the long-term consequences of a conservative medical management approach have not been fully explored.
We, therefore, sought to describe the prevalence and predictors of a conservative medical management treatment strategy for patients with NSTE ACS and significant CAD on early angiography, and to investigate the long-term risk of adverse outcomes in these patients compared with that seen in patients who underwent revascularization.
Study design and subjects
We performed a subgroup analysis of patients with significant disease on coronary angiography enrolled in the SYNERGY (Superior Yield of the New Strategy of Enoxaparin, Revascularization, and Glycoprotein IIb/IIIa Inhibitors) trial, which randomized 9,978 patients from 12 countries between August 2001 and December 2003 (9,10). The SYNERGY trial was an open-label trial designed to evaluate the efficacy and safety of enoxaparin versus unfractionated heparin when administered with established guidelines-recommended therapy in patients with NSTE ACS planned for an early invasive strategy. Patients were eligible for study participation if they had ≥2 of the following high-risk features: age ≥60 years, ST-segment changes, and a troponin or creatine kinase-MB level exceeding the upper limit of normal. The majority of subjects underwent cardiac catheterization during the initial hospitalization (92%).
We included only patients who underwent cardiac catheterization during the initial hospitalization with 1 or more lesions >50% severity in a native epicardial coronary artery (left main, left anterior descending, circumflex artery, or right coronary artery) or a major side branch of a native artery. Angiographic lesion severity was recorded by site investigators without utilization of an angiographic core laboratory. We excluded patients who did not undergo catheterization during the initial hospitalization (n = 880) and patients with normal coronaries on angiography or all lesions on angiography ≤50% (n = 873), as these patients would not be candidates for revascularization (Fig. 1).
Data on in-hospital outcomes were obtained by reviewing clinical records for all enrolled patients. Outcomes after discharge were captured by contacting patients by telephone or through scheduled patient visits through 1 year, and 1-year mortality was ascertained through medical records or national death indexes. Adjudication was performed for all nonfatal myocardial events occurring through 6 months. The total number of subjects who were lost to follow-up for the mortality end point at 6 and 12 months were 16 and 35, respectively.
We divided the study population into 3 groups: 1) those who underwent in-hospital PCI; 2) those who underwent in-hospital CABG; and 3) those who were medically managed without in-hospital revascularization performed within 7 days of randomization. Data were displayed as percentages for categorical variables and medians (25th percentile, 75th percentile) for continuous variables. Comparison of baseline characteristics, angiographic results, in-hospital medication use, and outcomes was performed using the Pearson chi-square test for categorical variables and the Kruskal-Wallis test for continuous variables. The Global Registry of Acute Coronary Events (GRACE) score has been established as a robust tool for predicting 6-month post-discharge mortality in patients with ACS (11). In order to provide a single numerical estimate of risk for each of the 3 treatment groups, we calculated the GRACE risk scores for all subjects studied. Unadjusted in-hospital events rates in the 3 groups were compared using the chi-square test, while long-term event rates were estimated using the Kaplan-Meier method with pairwise comparisons between groups made using the log-rank test.
Predictors of in-hospital revascularization
We developed a logistic regression model to identify significant predictors of in-hospital conservative medical management versus in-hospital revascularization (Model 1) using both stepwise and backward selection techniques on 37 candidate predictors derived from a large observational registry of NSTE ACS patients (12). Nonfatal clinical outcomes captured before catheterization were entered as predictors in the model. The spline transformation method was used to determine the functional form for continuous variables that did not meet linearity assumptions. Whenever appropriate, we applied piecewise linear splines as the preferred method of transformation. A p value of 0.05 was used as the criterion for variable entry and retention in the model.
To evaluate the association between conservative medical management and long-term outcomes, we developed a proportional hazards model comparing the conservative medical management group with the PCI and CABG groups individually (conservative medical management vs. PCI and conservative medical management vs. CABG), as patients undergoing PCI and CABG often have very different clinical profiles (13). We examined the association between conservative medical management as an independent variable and death (all-cause mortality) or myocardial infarction (MI) at 6 months as a dependent variable occurring after a discharge landmark time point, defined as time of discharge, or 7 days from randomization if hospitalization was prolonged (Model 2). Modeling outcomes using a landmark time point enabled accounting for the ascertainment bias expected when comparing early MI rates between conservative medical management patients (in whom markers were drawn only when a suspected recurrent ischemic event occurred) and PCI or CABG patients (in whom serial markers were recommended to be routinely drawn post-procedure); as such a comparison artificially inflates the relative rate of MI in the PCI and CABG groups (14). The use of a discharge landmark further permitted the modeling of nonfatal events occurring before the landmark time point, including recurrent ischemia, recurrent MI, acute heart failure, stroke, and major bleeding, as time-updated covariates reflecting changes in dynamic risk status (15). We then adjusted Model 2 for the propensity for in-hospital conservative medical management by including propensity scores derived from data captured before catheterization (Model 1) as a covariate in Model 2.
Time-dependent covariate analysis
We developed a Cox proportional hazards model that included conservative medical management versus PCI and conservative medical management versus CABG as time-dependent covariates and death from catheterization through 1 year as the outcome variable (Model 3). Time-dependent covariates account for the timing of PCI and CABG relative to the time of catheterization and the outcome of interest, therefore serving as a means to reduce survivor bias (16).
All p values were two-tailed and alpha set at 0.05 with confidence intervals calculated to the 95th percentile. Statistical analyses were performed using SAS version 8.2 (SAS Institute Inc., Cary, North Carolina).
Utilization of revascularization procedures
A total of 8,225 patients were found to have significant CAD on early angiography, and 4,752 of these patients (57.8%) were enrolled from U.S. sites. In-hospital revascularization before the discharge landmark time point was performed in 5,592 patients (68%), with 4,294 patients undergoing PCI (52.2%) and 1,298 patients undergoing CABG (15.8%). The median time from randomization to cardiac catheterization was 21 (6, 42) h, while the median times from catheterization to PCI and CABG were 0.5 (0.2, 1.2) h and 64.8 (23.5, 136.1) h, respectively. A total of 2,633 patients (32%) were medically managed without in-hospital revascularization before the discharge landmark time point (Fig. 1).
Compared with patients undergoing in-hospital revascularization (PCI/CABG), patients in the conservative medical management group were older, more likely to be women, and more likely to have peripheral vascular disease, prior stroke, hypertension, diabetes mellitus, hypercholesterolemia, prior CABG, and lower creatinine clearance values (Table 1). The median GRACE risk score was similar for the CABG group at 124 (110, 141) and conservative medical management group at 121 (106, 136), but was lower for the PCI group at 103 (90, 119).
Patients in the CABG group had the greatest frequency of left main and 3-vessel disease, while patients in the PCI group had the greatest frequency of 1- or 2-vessel disease (Table 2). More than half of the patients in the conservative medical management group had 3-vessel disease. Median left ventricular ejection fraction values were lowest in the conservative medical management and CABG groups.
Concomitant medication utilization
The use of aspirin, beta-blockers, angiotensin-converting enzyme inhibitors, and statins was high in all groups during the initial hospitalization, at discharge, and at 30 days (Table 3). The use of evidence-based medications was highest in the PCI group followed by the conservative medical management group.
Predictors of conservative medical management
Several variables were independently associated with conservative medical management (Table 4). The strongest predictors of conservative medical management were prior CABG, lower body weight, lack of a recurrent MI from randomization until the time of catheterization, and 3-vessel disease demonstrated on angiography.
Revascularization procedures during follow-up
Overall, rates of late or repeat revascularization occurring after discharge were low (Fig. 2A). Patients who underwent in-hospital CABG before 7 days were the least likely to undergo repeat revascularization procedures, while the rate of late revascularization was similar in the conservative medical management and PCI groups. The median time to late revascularization (PCI) after discharge for the conservative medical management group was substantially shorter than the median time to repeat revascularization for those patients who had received PCI/CABG during the index hospitalization (Fig. 2B).
Unadjusted in-hospital events occurring between catheterization and the discharge landmark time point occurred most frequently in the CABG group (Table 5). However, the in-hospital mortality curves appeared to be similar for the 3 comparison groups (Fig. 3A).
Unadjusted post-discharge events occurred most frequently in the conservative medical management group (Table 6). The post-discharge mortality curves demonstrated an early hazard for both the conservative medical management and CABG groups, but after 90 days, the curves diverged and the mortality hazard appeared higher in the conservative medical management group (Fig. 3B). Because the conservative medical management group had both an increased rate of post-discharge revascularization procedures and an earlier median time to revascularization compared with the PCI and CABG groups, we repeated the landmark analysis excluding all medically managed patients with revascularization performed after the landmark time point (discharge or 7 days) through 30 days (Table 6, bottom). We observed that the unadjusted and adjusted risk of MI declined in the conservative medical management group when events through 30 days were censored, but the adjusted risks of 6-month and 1-year mortality remained similar.
We analyzed overall mortality at 1 year over a period of observation that included both the in-hospital and post-discharge time periods (Fig. 4). The early risk of death from CABG was seen in the first 90 days, but, after that period, the risk in the medical cohort increased at a greater rate and the 2 curves cross. Thus, for those who survived the initial surgery, the risk became less than that of the medically treated patients with time. Overall mortality was highest in the medically managed group and lowest in the PCI group, with the difference persisting even after adjusting for conservative medical management as a time-dependent covariate (Table 7).
Our study showed that, in a fairly contemporary cohort of high-risk NSTE ACS managed with an early invasive management strategy, the majority of which were from U.S. sites, a substantial proportion of patients did not receive in-hospital revascularization despite documentation of significant coronary disease on angiography. Medically managed patients had a higher risk of long-term adverse outcomes despite widespread utilization of evidence-based medications beyond the initial hospitalization.
Use of revascularization after early catheterization
Most observational studies and randomized controlled trials studying early invasive management strategies for NSTE ACS have focused on early catheterization rather than early revascularization as a management strategy (4–7). The potential consequences of this approach were highlighted by a recent meta-analysis of clinical trials of early invasive management in NSTE ACS in which the survival benefit with an early invasive strategy was most evident when there was proportionally greater use of revascularization in the invasive versus conservative treatment arms (17). While a recent analysis from the SYNERGY trial examined outcomes in NSTE ACS patients undergoing CABG for intermediate- to high-risk NSTE ACS, this analysis did not focus upon patients with angiographically defined significant coronary disease, so the medical management population was diluted with patients with insignificant coronary disease (18). Our study demonstrates that many patients with NSTE ACS and significant coronary disease are determined to not be candidates for revascularization procedures, and these patients have a significantly increased risk of adverse events, despite widespread use of evidence-based medications.
Reasons for nonuse of revascularization procedures
Possible reasons for patients not undergoing in-hospital revascularization despite significant coronary disease include death before planned revascularization, unfavorable coronary anatomy, serious comorbidities, resource availability, socioeconomic disparities, and patient refusal. Practice guidelines recommend consideration of early invasive management for patients with NSTE ACS who have had prior revascularization (1), but our data show that patients with prior CABG were least likely to undergo in-hospital revascularization after early catheterization. Patients with prior CABG comprise a subgroup in which repeat revascularization may be technically challenging due to diffuse native vessel and graft disease (19). Despite the availability of 1 subgroup analysis of a randomized trial to guide selection of revascularization strategies in these patients (20), the risk of periprocedural adverse events with any revascularization strategy remains high for patients with prior CABG. Additionally, many unmeasured variables likely contributed to decision-making regarding revascularization procedures, as reflected in the relatively poor c-index (0.64) of the model we developed to determine predictors of conservative medical management. This matter requires prospective study to understand whether there are opportunities to expand the utilization of PCI and CABG for these patients.
Data from the CRUSADE (Can Rapid risk stratification of Unstable angina patients Suppress ADverse outcomes with Early implementation of the American College of Cardiology/American Heart Association guidelines) trial showed that the widespread availability of drug-eluting stents (DES) from 2003 to 2005 was associated with a sharp increase in PCI use in patients with NSTE ACS and significant CAD, while the rate of CABG procedures remained relatively stable (8). This was accompanied by a concordant decline in the proportion of medically managed patients, from 49% in 2002 to 36% in 2005, suggesting that physicians had become more willing to offer percutaneous revascularization options to patients who might have been considered poor candidates for PCI before the advent of DES. The high rate of off-label DES usage since its introduction, primarily for complex lesion subsets, supports this conclusion (21). Given recent safety concerns about DES and uncertainty over the actual benefit of an early invasive strategy (22), the proportion of medically managed patients may remain stable, or actually even increase over time.
The risk-treatment paradox refers to progressive diminution in the use of evidence-based therapies as baseline risk and future probability of death increases (23). However, treatment intensification in high-risk patients remains a major challenge because these patients have an increased susceptibility to drug toxicity and treatment complications (24). The greatest difference in medication use among the comparison groups was with clopidogrel, which was targeted for patients who underwent PCI. A subgroup analysis of the CURE (Clopidogrel in Unstable Angina to Prevent Recurrent Ischemic Events) trial showed a similar benefit of clopidogrel in patients treated with conservative medical management compared with the benefit seen in patients treated with revascularization (25). However, in the recent TRITON (Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel) study, older patients and those with lower body weight were also more likely to have serious bleeding complications and less likely to derive a net treatment benefit with more potent thienopyridine treatment (26).
Similar challenges face the widespread implementation of intensive lipid-lowering in the medically managed population because of their greater prevalence of comorbid conditions. In a meta-analysis of 4 large clinical trials that included the PROVE-IT–TIMI 22 (Pravastatin or Atorvastatin Evaluation and Infection Therapy–Thrombolysis In Myocardial Infarction 22) trial (27), investigators confirmed the safety and efficacy of high-dose statin therapy compared with standard-dose statin therapy (28). However, patients with serum creatinine >2.5 mg/dl were excluded from the aforementioned clinical trials, limiting generalizability of the findings to this population. Impaired renal function is a key concern, and the majority of NSTE ACS patients with renal impairment are managed medically (29). Hence, questions remain about the efficacy, safety, and correct dosing of guidelines-recommended medications in medically managed patients excluded from landmark randomized controlled trials because of excess comorbidity. Dedicated trials with alternative dosing strategies may be a step toward resolving concerns regarding drug metabolism and the increased likelihood of medication side effects in this population.
Several limitations were present with this analysis. First, detailed lesion-level angiographic findings were not captured, so we were not able to characterize angiographic suitability for revascularization. Second, significant lesions in the major coronary arteries were recorded in a dichotomized approach (>50% vs. ≤50%) on the SYNERGY case report form, so we could not assess different classification schemes for significant coronary disease such as stenoses >70%. Third, details about physician decision-making regarding revascularization were not captured. However, patients with clinical contraindications to catheterization and revascularization were excluded from the SYNERGY trial, so the study population represented an ideal group of patients for early invasive management at the time of randomization (30). Fourth, despite our best efforts to adjust for confounding, the conservative medical management group represents a particularly high-risk cohort in whom survival is unlikely to approximate that of the lower-risk PCI group because of unmeasured confounders making it difficult to draw definitive conclusions about the benefits of early revascularization procedures. Nonetheless, our intent was not to show that early revascularization was associated with improved outcomes, but rather to fully characterize the short- and long-term risks associated with a conservative medical management strategy. Last, the 7-day landmark analysis, although necessary for the investigation of MI outcomes in this particular study, may introduce biases associated with left censoring of longitudinal data.
Almost one-third of patients with NSTE ACS documented to have significant coronary disease on angiography did not undergo revascularization procedures in the SYNERGY trial. These patients have a higher risk of long-term adverse outcomes compared with those who undergo PCI or CABG despite widespread utilization of guidelines-recommended medications. These findings highlight the need for novel strategies to mitigate the increased risk of adverse outcomes in this population and for additional prospective studies to examine patient and physician decision-making regarding revascularization decisions.
The authors would like to thank the following people for their invaluable contributions: Amanda McMillan for editorial assistance and Kimberly Ferrer for graphics assistance.
Dr. Chan receives salary support from the National Medical Research Council, Singapore. The SYNERGY trial was funded by Sanofi-Aventis.
- Abbreviations and Acronyms
- coronary artery bypass grafting
- coronary artery disease
- drug-eluting stent(s)
- Global Registry of Acute Coronary Events score
- myocardial infarction
- NSTE ACS
- non–ST-segment elevation acute coronary syndromes
- percutaneous coronary intervention
- Received February 14, 2008.
- Accepted March 19, 2008.
- American College of Cardiology Foundation
- Anderson J.L.,
- Adams C.D.,
- Antman E.M.,
- et al.
- Bassand J.P.,
- Hamm C.W.,
- Ardissino D.,
- et al.
- Fox K.A.,
- Poole-Wilson P.A.,
- Henderson R.A.,
- et al.
- FRagmin and Fast Revascularisation during InStability in Coronary artery disease Investigators. Invasive compared with non-invasive treatment in unstable coronary-artery disease: FRISC II prospective randomised multicentre study. Lancet 1999;354:708-15.
- McCullough P.A.,
- O'Neill W.W.,
- Graham M.,
- et al.
- Effects of tissue plasminogen activator and a comparison of early invasive and conservative strategies in unstable angina and non–Q-wave myocardial infarction
- Ferguson J.J.,
- Califf R.M.,
- Antman E.M.,
- et al.
- Wang T.Y.
- Anderson J.R.,
- Cain K.C.,
- Gelber R.D.
- Zhou Z.,
- Rahme E.,
- Abrahamowicz M.,
- et al.
- Bavry A.A.,
- Kumbhani D.J.,
- Rassi A.N.,
- Bhatt D.L.,
- Askari A.T.
- Eagle K.A.,
- Guyton R.A.,
- Davidoff R.,
- et al.
- Morrison D.A.,
- Sethi G.,
- Sacks J.,
- et al.
- FDA Statement on Coronary Drug-Eluting Stents (September 14, 2006)
- Fox K.A.,
- Mehta S.R.,
- Peters R.,
- et al.
- Cannon C.P.,
- Braunwald E.,
- McCabe C.H.,
- et al.
- Cannon C.P.,
- Steinberg B.A.,
- Murphy S.A.,
- Mega J.L.,
- Braunwald E.