Author + information
- Received July 13, 2011
- Revision received September 19, 2011
- Accepted October 28, 2011
- Published online February 1, 2012.
- Peter Damman, MD⁎,
- Nan van Geloven, MSc⁎,
- Lars Wallentin, MD, PhD†,
- Bo Lagerqvist, MD, PhD†,
- Keith A.A. Fox, BSc, MB, ChB‡,
- Tim Clayton, BSc, MSc§,
- Stuart J. Pocock, BSc, MSc, PhD§,
- Alexander Hirsch, MD, PhD⁎,
- Fons Windhausen, MD, PhD⁎,
- Jan G.P. Tijssen, PhD⁎ and
- Robbert J. de Winter, MD, PhD⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Robbert J. de Winter, Department of Cardiology, Cardiac Catheterization Laboratory B2-137, Academic Medical Center, University of Amsterdam, Meibergdreef 9, 1105 AZ, Amsterdam, the Netherlands
Objectives This study sought to investigate long-term outcomes after early or delayed angiography in patients with non–ST-segment elevation acute coronary syndrome (nSTE-ACS) undergoing a routine invasive management.
Background The optimal timing of angiography in patients with nSTE-ACS is currently a topic for debate.
Methods Long-term follow-up after early (within 2 days) angiography versus delayed (within 3 to 5 days) angiography was investigated in the FRISC-II (Fragmin and Fast Revascularization During Instability in Coronary Artery Disease), ICTUS (Invasive Versus Conservative Treatment in Unstable Coronary Syndromes), and RITA-3 (Intervention Versus Conservative Treatment Strategy in Patients With Unstable Angina or Non-ST Elevation Myocardial Infarction) (FIR) nSTE-ACS patient-pooled database. The main outcome was cardiovascular death or myocardial infarction up to 5-year follow-up. Hazard ratios (HR) were calculated with Cox regression models. Adjustments were made for the FIR risk score, study, and the propensity of receiving early angiography using inverse probability weighting.
Results Of 2,721 patients originally randomized to the routine invasive arm, consisting of routine angiography and subsequent revascularization if suitable, 975 underwent early angiography and 1,141 delayed angiography. No difference was observed in 5-year cardiovascular death or myocardial infarction in unadjusted (HR: 1.06, 95% confidence interval [CI]: 0.79 to 1.42, p = 0.61) and adjusted (HR: 0.93, 95% CI: 0.75 to 1.16, p = 0.54) Cox regression models.
Conclusions In the FIR database of patients presenting with nSTE-ACS, the timing of angiography was not related to long-term cardiovascular mortality or myocardial infarction. (Invasive Versus Conservative Treatment in Unstable Coronary Syndromes [ICTUS]; ISRCTN82153174. Intervention Versus Conservative Treatment Strategy in Patients With Unstable Angina or Non-ST Elevation Myocardial Infarction [the Third Randomised Intervention Treatment of Angina Trials (RITA-3)]; ISRCTN07752711)
The routine invasive strategy and selective invasive strategy are alternative treatment strategies for patients presenting with non–ST-segment elevation acute coronary syndrome (nSTE-ACS). With the exception of indications for emergency angiography and revascularization, controversy remains about the optimal timing of angiography of those selected to undergo a routine invasive strategy (1). The American College of Cardiology/American Heart Association guidelines propose 2 approaches within the routine invasive strategy: early (including immediate) angiography or deferred angiography. Early angiography might timely identify those with left main or multivessel disease, expediting coronary artery bypass graft (CABG), or lesions suitable for percutaneous coronary intervention (PCI) (2). In addition, expediting intervention may reduce antithrombotic treatment use and the accompanying increased risk of bleeding (3). By contrast, deferred angiography might prevent procedure-related events after intensive antithrombotic and anti-ischemic therapy (2).
In 2 recent large randomized clinical trials, an early invasive strategy was compared with a deferred invasive strategy. The TIMACS (Timing of Intervention in Acute Coronary Syndromes) trial (4) showed that overall, a routine early intervention strategy (coronary angiography within 24 h) was not superior to a delayed-intervention strategy (angiography after 36 h) for the prevention of the composite of death, myocardial infarction (MI), or stroke at 6 months. In the ABOARD (Angioplasty to Blunt the Rise of Troponin in Acute Coronary Syndromes Randomized for an Immediate or Delayed Intervention) trial (5), where an immediate angiography (“primary PCI approach”) was compared with delayed angiography, no difference was observed in the composite of death, MI, or urgent revascularization at 1 month. However, both trials showed a benefit in high-risk patients. Several smaller randomized trials and post hoc analyses of randomized trials have shown inconsistent results regarding the timing of angiography or intervention (6–10). Furthermore, data on long-term follow-up is lacking.
We performed a collaborative analysis of data from the FRISC-II (Fragmin and Fast Revascularization During Instability in Coronary Artery Disease), ICTUS (Invasive Versus Conservative Treatment in Unstable Coronary Syndromes), and RITA-3 (Intervention Versus Conservative Treatment Strategy in Patients With Unstable Angina or Non-ST Elevation Myocardial Infarction) (FIR) trials of patients with nSTE-ACS to assess the effect of timing of angiography in patients originally randomized to the routine invasive strategy on the occurrence of long-term clinical outcomes. By pooling individual patient data, the effect of timing of angiography can adjusted for variables associated with timing of angiography and outcomes.
Setting and data collection
The principal investigators (L.W., R.J.d.W., K.A.F.) initiated this collaborative analysis, and a protocol was written summarizing the main pre-specified analyses and a common set of variables. Investigators from the 3 trials provided individual patient data to form a patient-pooled database. The database included core variables on demographics, clinical history, risk factors for coronary artery disease, baseline electrocardiographic characteristics, laboratory results, and 5-year clinical outcomes. Datasets from each trial were sent for merging to the coordinating Academic Medical Center in Amsterdam, the Netherlands. The merged database was checked for completeness and consistency by all 3 participating sites.
Study population and procedures
The details of the design of the FRISC-II, ICTUS, and RITA-3 trials have been published previously (11–13). These trials compared a routine invasive strategy with a selective invasive strategy in patients with nSTE-ACS. Patients in the routine invasive group were scheduled to undergo early coronary angiography (within 24 to 48 h in ICTUS, within 72 h in RITA-3), with subsequent revascularization when appropriate. In the FRISC-II trial, the aim was to perform angiography and revascularization, if appropriate, within 7 days. CABG was recommended with severe left main stem or 3-vessel disease. The selective invasive strategy consisted of initial medical treatment with coronary angiography and revascularization only in the case of refractory angina despite optimal medical treatment (or in the case of hemodynamic or rhythmic instability in ICTUS). In the FRISC-II and ICTUS trials, a pre-discharge ischemia detection test was performed.
Timing of angiography
For our current analyses, we included data from all patients originally randomized to the routine invasive strategy who underwent angiography. Patients undergoing a selective invasive management were excluded. Two study groups were formed: the early angiography group consisted of patients receiving coronary angiography on the day of randomization (day 1) or the next day and the delayed angiography group consisted of patients receiving coronary angiography within days 3 to 5. A second analysis consisted of the time to angiography divided into 1-day intervals (from 1 day up to 5 days and a remaining group, including angiography after 5 days or no angiography).
The main outcomes for the current analysis were the 5-year composite outcome of cardiovascular (CV) death or MI and the individual 5-year outcomes CV death and MI. CV death was defined as all-cause death, unless an unequivocal noncardiovascular cause could be established. The original definition of MI per trial was used; readjudication of individual events to accommodate common definitions was not possible. In the FRISC-II trial, MI was defined by the occurrence of 2 of the following conventional criteria: typical chest pain, diagnostic electrocardiographic recording (new Q waves), and elevation in 1 cardiac biomarker above the upper limit of normal (ULN) with spontaneous MIs, or elevation in 1 cardiac biomarker up to 1.5 the ULN with procedure-related MIs. In the RITA-3 trial, MI was defined as diagnostic electrocardiographic recording (new Q waves) or by the combination of a typical clinical event, electrocardiographic evidence of acute infarction, and an elevation in 1 cardiac biomarker up to twice the ULN. In the ICTUS trial, MI was defined as myocardial necrosis in the setting of myocardial ischemia. Myocardial necrosis was defined as an elevation in 1 cardiac biomarker above the ULN with spontaneous MIs, or 3 times the ULN in case of a procedure-related MI (14).
Data concerning major bleeding was collected in the ICTUS and RITA-3 trials. Major bleeding was defined in ICTUS as fatal bleeding, intracranial bleeding, a need for blood transfusion, a decrease of 3 mmol/l or more in hemoglobin levels, and bleeding resulting in hemodynamic compromise. Major bleeding in RITA-3 was defined as fatal, intracerebral hemorrhage, or transfusion of 2 or more units.
Data with a normal distribution are described as the mean ± SD, data with a non-normal distribution as the median (with interquartile range). The Student t test or a 1-way analysis of variance was used to compare data with a normal distribution, whereas a nonparametric Kruskal-Wallis test was used to compare data with a non-normal distribution. Categorical data are presented as frequencies (%) and compared with a chi-square test. The primary analysis was the comparison between patients who received angiography within 2 days of randomization and those for which angiography was performed 3 to 5 days after randomization regarding the composite outcome of CV death or MI. Cumulative event rates for unadjusted analyses were estimated using the Kaplan-Meier method and compared with the log-rank test. Follow-up for the composite outcome was censored at the actual date of last contact or at 5 years, whichever came first. To adjust for the nonrandomized allocation of the timing of angiography, the relation of timing to CV death or MI was investigated using Cox proportional hazards regression using inverse probability weighting (IPW) (15). IPW aims to reweigh the observations such that the timing of angiography becomes independent of measured confounders. We considered the following baseline and angiographic characteristics as potential confounders for the effect of timing on CV death or MI: study, age, sex, body mass index, current smoking, hypertension, hyperlipidemia, diabetes mellitus, history of MI or revascularization, and the presence of ST-segment depression. For the construction of IPW weights, we selected those variables that showed a relevant association with the timing in logistic regression models (variables with a p value >0.1 by the likelihood-ratio test were excluded from the model using backward selection). On top of the IPW correction, we adjusted the Cox models for relevant predictors of CV death or MI summarized in the FIR risk score and for original study (16). Patients who did not receive angiography within any of the time frames used in the primary analysis were not part of that analysis. We also performed landmark analyses, in which patients who received angiography within a 1-day landmark were compared with all patients who received angiography after the landmark or had no angiography at all. Patients with events or angiography before the landmark were excluded. The relation of timing of angiography in 1-day intervals to CV death or MI was investigated using Cox proportional hazards regression in 2 sets of landmark models: models adjusted for relevant predictors of CV death or MI and study, and models with adjustments for predictors of the timing angiography using IPW. Regarding IPW, we identified predictors for angiography within each landmark compared with later or no angiography.
Pre-specified secondary analyses included comparisons of outcomes in the early (2 days) versus delayed (3 to 5 days) angiography groups: 1) interaction between timing of angiography and baseline risk as indicated by the FIR low-risk, intermediate-risk, and high-risk categories (16); 2) stratified by angiographic left main or 3-vessel disease; and 3) restricted to patients that underwent intervention (PCI or CABG) within 30 days. The proportional hazards assumption for all analyses were verified graphically by checking parallelism of log-log survival curves and with Schoenfeld tests, no relevant violations were observed.
A total of 2,721 patients were originally randomized to the routine invasive arm, of which 244 had missing angiography data. Regarding the early and delayed study groups, 975 patients underwent early angiography (randomization days 1 or 2) and 1,141 patients underwent delayed angiography (days 3 to 5). Of the remaining patients, 317 underwent angiography later than 5 days and 44 underwent no angiography at all. The latter 2 groups could only be used in the landmark analyses. The baseline characteristics of patients in the early and delayed angiography groups and in 1-day groups are shown in Tables 1 and 2,⇓ respectively. Patients undergoing early angiography more often had a history of PCI or CABG, more often were smokers or had hypercholesterolemia, and were younger. Most patients in the early angiography group were originally from the ICTUS trial.
During initial hospitalization, 654 (67%) patients were revascularized in the early angiography and 732 (64%) patients in the delayed angiography group (p = 0.16). In the early angiography group, PCI was the first procedure in 477 (73%) patients and CABG was first in 175 (27%) patients. Two patients received both on 1 day. In the delayed angiography group, these numbers were respectively 438 (60%), 292 (40%), and 2 (0%). The median time to first revascularization procedure during initial hospitalization was 1 day (interquartile range [IQR]: 1 to 5 days) in the early angiography and 4 days (IQR: 3 to 6 days) in the delayed angiography group (p < 0.001).
At a median of 5-year follow-up, early angiography did not result in a lower cumulative (unadjusted) CV death or MI rate compared with delayed angiography (15.4% vs. 14.8%, p = 0.61). Moreover, no differences in event rates were observed in the individual outcomes CV death (p = 0.94) or MI (p = 0.37). Divided into 1-day intervals or angiography after 5 days/no angiography at all, no difference was observed in the cumulative CV death or MI rate (log rank, p = 0.61), CV death rate (log rank, p = 0.95), or MI rate (log rank, p = 0.21). The unadjusted Kaplan-Meier event rates are presented in Table 3. While awaiting angiography, 6 patients endured a spontaneous MI. Of these 6 MIs, 3 occurred on day 0, 1 on day 2, and 2 on day 3.
In unadjusted Cox proportional hazards models, early angiography showed a similar CV death or MI hazard (hazard ratios [HR]: 1.06, 95% confidence interval [CI]: 0.79 to 1.42, p = 0.61), when compared with the delayed angiography group. After adjustment for the FIR risk score, study, and IPW, no difference was observed when comparing the early angiography group with the delayed angiography group (HR: 0.93, 95% CI: 0.75 to 1.16, p = 0.54). Moreover, no difference was observed regarding the individual outcomes. These results were consistent when assessed per individual trial. Regarding timing of angiography and FIR risk score, no interaction was observed (p value for interaction in unadjusted and adjusted analyses, respectively: 0.98 and 0.89). The unadjusted and adjusted Cox models are shown in Table 4. Kaplan-Meier survival curves and adjusted hazard ratios are presented in Figure 1.
In the 1-day interval, Cox proportional-hazards models adjusted for the FIR score and study or predictors for timing of angiography using IPW, angiography within none of the time intervals was associated with an increased CV death or MI hazard when compared with later or no angiography. The landmark models are shown in Figure 2.
Early versus delayed angiography in revascularized patients
In a separate analysis, we compared long-term outcomes between patients in the early and delayed angiography groups who underwent intervention within 30 days. In the early angiography and delayed angiography groups, respectively, 666 and 763 patients were revascularized. No difference was observed between 30 days and the end of follow-up regarding CV death (p = 0.59). Kaplan-Meier curves for CV death are shown in Figure 3.
Data concerning major bleeding during initial hospitalization was collected in the ICTUS and RITA-3 trials. In the early angiography arm, 28 of 857 patients (3.3%) endured a major bleeding. This was not significantly different when compared with 15 of 486 patients (3.1%) who endured a major bleeding in the delayed angiography arm (p = 0.86).
Left main or 3-vessel disease
In the ICTUS and RITA-3 trials, data were collected during coronary angiography. Of 1,461, patients with known angiography data, 343 were diagnosed as having a ≥50% stenosis in the left main or in 3 major epicardial arteries (≥50% in other arteries than the left main). Within this subgroup, comparable CV death or MI rates were observed between early or delayed angiography (19.6% vs. 12.0%, p = 0.09). These event rates were, respectively, 13.6% and 12.8% in patients without left main or 3-vessel disease (p = 0.62).
Several implications can be drawn from the current FIR patient-pooled analysis. First, early angiography within 2 days was not associated with lower 5-year CV death or MI when compared with delayed angiography within 3 to 5 days. There were no differences in individual long-term CV death or MI rates. These results were consistent after adjustment for predictors for long-term outcomes, study, and predictors for timing of angiography. In 1-day interval landmark analyses, angiography within any of the time intervals was not associated with an increased CV death or MI hazard compared with later angiography or no angiography at all.
In the earlier mentioned TIMACS and ABOARD trials, no difference was shown in clinical outcomes when comparing an early angiography with a delayed angiography treatment strategy (4,5). However, the TIMACS and ABOARD trials demonstrated no difference at short-term outcomes (6 months and 1 month, respectively) when comparing a very early or immediate angiography, whereas our data concerns angiography within 2 days compared with later angiography. We note that a benefit of early angiography was observed in high-risk patients in both TIMACS and ABOARD. We did not observe an interaction between timing of angiography and baseline risk profile as indicated by the FIR risk score.
In the smaller ISAR-COOL (Intracoronary Stenting With Antithrombotic Regimen Cooling-Off) trial, patients were randomized to expedited angiography (within 6 h) or to a strategy of 3 to 5 days of prolonged antithrombotic treatment followed by angiography (6). An excessive event rate incurred during antithrombotic pre-treatment in the prolonged pre-treatment arm, resulting in a significant benefit in 30-day death or MI with the expedited angiography strategy. We did not observe this excessive hazard with increasing times to angiography in our data. Few patients endured a spontaneous MI while awaiting angiography or subsequent revascularization.
In a subanalysis of the ACUITY (Acute Catheterization and Urgent Intervention Triage Strategy) trial (17), PCI after 24 h was associated with 1-year mortality and adverse ischemic outcomes. The highest hazard was observed in high-risk patients, as identified by the TIMI (Thrombolysis In Myocardial Infarction) risk score. It is not possible to directly compare our current analysis with these results, because we investigated time to angiography instead of time to PCI. Potentially, medically stabilized patients undergoing delayed angiography do not undergo PCI and are thus excluded from time-to-PCI analyses. Moreover, patients referred for CABG are also excluded. We explored the relation between time to PCI and outcomes in our early versus delayed study group and did not observe a significant difference in unadjusted event rates when comparing patients undergoing early versus delayed PCI (data not shown).
In both the TIMACS trial and the ACUITY substudy, the Kaplan-Meier estimates or curves for mortality continue to diverge up to 30 days and thereafter. Apparently, the risk of delaying angiography and subsequent intervention extends to the period after the intervention (18,19). However, when we compared early with delayed angiography in patients who underwent revascularization within 30 days, no difference in CV mortality was observed. Because outcomes after revascularization can be expected to be similar and potential mechanisms for late mortality are not elucidated in the previously mentioned studies, further studies are required to confirm our findings.
Major bleeding has been identified as an independent predictor of mortality (20). Patients in the delayed angiography group receive a prolonged medical treatment with antiplatelet and antithrombin agents, potentially placing them at a higher bleeding risk. However, the duration of antiplatelet and antithrombin administration did not influence major bleeding in the ACUITY trial. This is supported by the results of the TIMACS and ABOARD trials, where no differences in major bleedings were observed. In our current analysis, no difference in major bleeding was observed when comparing early angiography with delayed angiography. Furthermore, we did not observe any difference in mortality between the timing study groups.
Left main or 3-vessel disease
According to the American College of Cardiology/American Heart Association guidelines, patients with left main coronary artery disease or multivessel disease are candidates for expedited CABG to avoid a risky waiting period (2). Our current analyses suggest that CV death or MI rates in these patients were comparable with early angiography or delayed angiography. However, one should be cautious when interpreting subgroup analyses, albeit pre-specified.
Clinical implications and future perspectives
Because we did not observe a relation between timing of angiography and outcomes, a potential implication could be that angiography can be postponed to the next working day in patients presenting outside working hours. On the other side, early angiography might potentially reduce hospitalization time and associated costs. Keeping limitations of a post hoc analysis in mind, these findings should be interpreted as hypothesis-generating. Furthermore, future research should focus on the patients at highest baseline risk and cost-effectiveness analyses. Finally, the current analysis does not address the appropriateness of the very early compared with immediate angiography.
Patients were not randomized to have coronary angiography within a particular time from randomization; therefore, differences in outcomes may reflect differences in patient characteristics or practice patterns rather than differences in time to coronary angiography. One of these differences might be the number of patients that required urgent angiography because of severe ongoing angina, profound or dynamic electrocardiogram changes, major arrhythmias, or hemodynamic instability. Although we corrected for important predictors of CV death or MI, study, and predictors for receiving angiography, our results could have been biased by unavailable variables, such as stent placement and cardiac biomarkers. Second, the outcome MI was composed of all spontaneous and procedure-related MIs from the FRISC-II, ICTUS, and RITA-3 trials, and a uniform definition could not be achieved. The use of a uniform definition of MI would allow a more accurate assessment of the effect of timing of angiography on the clinical outcome MI (21). This limitation should be taken into account in the 1-day analysis where there is a varying contribution of MIs to the combined outcome. Third, cardiac biomarkers rise, peak, and decline over time relative to the onset of myocardial ischemia, and an additional biomarker increase due to an early procedure is difficult to distinguish from already elevated biomarker levels due to the index event. In the ICTUS trial, where patients were eligible if they were troponin-positive, recurrent MI during the first 48 h was diagnosed when there was a 50% decrease from a previous peak biomarker value, followed by a subsequent rise to a level exceeding the ULN as measured by routine serial sampling. The detection of procedure-related MI might thus be prone to detection bias, especially in the early angiography group. However, the procedure-related MIs were adjudicated by a clinical endpoint committee.
We conclude that in patients presenting with nSTE-ACS, early angiography within 48 h does not reduce the incidence of 5-year death or MI, when compared with delayed angiography within 48 to 120 h.
The authors would like to thank all investigators from the FRISC-II, ICTUS, and RITA-3 trials. Most important, they would like to thank all patients participating in these trials.
The collaboration and the meta-analysis were conducted using resources from the host institutions for the respective studies (Dr. Wallentin is supported by the Swedish Heart Foundation. Dr. Fox is supported by the British Heart Foundation) and from the London School of Hygiene and Tropical Medicine. Dr. Fox has received grants and honoraria from Sanofi-Aventis/Bristol-Myers Squibb, GlaxoSmithKline, Eli Lilly, and AstraZeneca. Dr. Pocock has consulted for The Medicines Company, and has served on a Boston Scientific-sponsored Data Monitoring Committee. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- coronary artery bypass graft
- confidence interval
- FRISC-II, ICTUS, RITA-3
- hazard ratios
- inverse probability weighting
- interquartile range
- myocardial infarction
- non–ST-segment elevation acute coronary syndrome
- percutaneous coronary intervention
- Thrombolysis In Myocardial Infarction
- upper limit of normal
- Received July 13, 2011.
- Revision received September 19, 2011.
- Accepted October 28, 2011.
- American College of Cardiology Foundation
- Bassand J.P.,
- Hamm C.W.,
- Ardissino D.,
- et al.,
- Task Force for Diagnosis and Treatment of Non–ST-Segment Elevation Acute Coronary Syndromes of the European Society of Cardiology
- Anderson J.L.,
- Adams C.D.,
- Antman E.M.,
- et al.
- van't Hof A.W.,
- de Vries S.T.,
- Dambrink J.H.,
- et al.
- Ronner E.,
- Boersma E.,
- Laarman G.J.,
- et al.
- Tricoci P.,
- Lokhnygina Y.,
- Berdan L.G.,
- et al.
- Wallentin L.,
- Lagerqvist B.,
- Husted S.,
- Kontny F.,
- Ståhle E.,
- Swahn E.
- Fox K.A.,
- Poole-Wilson P.A.,
- Henderson R.A.,
- et al.,
- RITA Investigators
- Damman P.,
- Hirsch A.,
- Windhausen F.,
- Tijssen J.G.,
- de Winter R.J.,
- ICTUS Investigators
- Fox K.A.,
- Clayton T.C.,
- Damman P.,
- et al.,
- FIR Collaboration
- Sorajja P.,
- Gersh B.J.,
- Cox D.A.,
- et al.
- Damman P.,
- van Geloven N.,
- Tijssen J.G.,
- de Winter R.J.
- Manoukian S.V.,
- Feit F.,
- Mehran R.,
- et al.
- Thygesen K.,
- Alpert J.S.,
- White H.D.,
- Joint ESC/ACCF/AHA/WHF Task Force for the Redefinition of Myocardial Infarction