- Email: [email protected]
Role of routine early angiography post-fibrinolysis for ST elevation myocardial infarction — A meta-regression analysis using angiography rate in the non-routine arm
International Journal of Cardiology 167 (2012) 1888–1891
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Role of routine early angiography post-fibrinolysis for ST elevation myocardial infarction — A meta-regression analysis using angiography rate in the non-routine arm Cheuk-Kit Wong a,⁎, Sophia Leon de la Barra b, Peter Herbison b a b
Department of Cardiology, Dunedin School of Medicine, University of Otago, Dunedin Public Hospital, Dunedin, New Zealand Department of Preventive and Social Medicine, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand
a r t i c l e
i n f o
Article history: Received 2 February 2012 Received in revised form 24 April 2012 Accepted 28 April 2012 Available online 21 May 2012 Keywords: Routine angiography Fibrinolysis Meta-regression
a b s t r a c t Background: The current European and American Guidelines differ with regard to the recommended level for the use of routine early angiography after fibrinolysis for STEMI. Previous meta-analyses on randomized controlled trials have supported the routine early approach, but its advantage may be because of an excessively low angiography rate among patients in the non-routine strategy arm of the trials. Methods: We update the meta-analysis and apply meta-regression to evaluate whether the difference in outcome between the 2 randomized arms could be explained by the angiography rates in the non-routine strategy arm. Because reinfarction and recurrent ischemia are often the reported indication for angiography, we only use mortality endpoint in our meta-regression analysis. Results: Among the eight trials included with 3195 patients, the angiography rate in the non-routine strategy arms ranges from 15% to 100%. The overall odds ratio for 30-day mortality comparing the routine early angiography arm vs the non-routine arm is 0.86 (95% confidence interval 0.60–1.24). On the plot listing the eight trials according to angiography rates, there is no visual trend in the odds ratio estimates for mortality when comparing the 2 treatment strategies as angiography rate decreases. In meta-regression analysis, angiography rate does not predict 30-day mortality (p = 0.461). Conclusion: For STEMI, mortality endpoint trumps the softer endpoints of recurrent infarction and ischemia. The current study shows that the equipoise between the routine early invasive versus the non-routine strategy on 30-day mortality cannot be explained by the variable performance of angiography in the non-routine strategy arm. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The European recommendations give a class I recommendation (level of evidence A) for routine early percutaneous coronary intervention (PCI) within 24 h following successful fibrinolysis [1], but the American recommendations take a softer stance. The ACC/AHA 2009 update to the STEMI guidelines [2] gives a class IIb recommendation (level of evidence C) stating that “Patients who are not at high risk who receive fibrinolytic therapy as primary reperfusion therapy at a non‐PCI-capable facility may be considered for transfer as soon as possible to a PCI-capable facility where PCI can be performed either when needed or as a pharmacoinvasive strategy”, and a class IIa (level of evidence B) recommendation for this strategy for high-risk patients. The 2011 ACCF/AHA/SCAI recommendations for
⁎ Corresponding author. Tel.: + 64 3 4747980; fax: + 64 3 4747655. E-mail address: [email protected] (C-K. Wong). 0167-5273/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2012.04.151
Percutaneous Coronary Intervention [3] for STEMI patients initially treated by fibrinolysis are similar. In the latest meta-analysis [4] on clinical trials that randomly assigned patients following fibrinolysis to routine early coronary angiography and PCI vs the “non-routine” strategy, there were respectively 38% and 79% significant reductions in recurrent MI and recurrent ischemia in the routine early strategy arm with no substantial reduction in mortality. While almost all patients in the routine early strategy arm had angiography, the proportion undergoing angiography in the non-routine strategy arm of the included studies varied widely. Angiography, by providing unique information for revascularization, might have been the crucial step altering outcome. However, the decision on angiography in the non-routine arm would have been affected by factors including recurrent MI and ischemia, which are also the softer endpoints of the randomized trials. Mortality is the hard endpoint in STEMI. The impact on mortality from a routine early invasive strategy over a non-routine strategy may
C-K. Wong et al. / International Journal of Cardiology 167 (2012) 1888–1891
vary with the proportion of patients that underwent angiography with the latter strategy even when the overall effect between the 2 strategies is neutral (i.e. studies with high angiography rates producing an effect which is offset by studies with low angiography rates). We address this question using the established statistical technique of metaregression. 2. Methods MEDLINE was searched using the search terms “coronary, thrombolysis, early or immediate stenting or angioplasty, and acute ST elevation myocardial infarction”, in order to obtain randomized controlled trials comparing the routine early angiography strategy vs the non-routine strategy. This method was identical to the most recently published meta-analysis [4] of the period 1977 to March 2010 except that the search was extended for a further year to March 2011. The search on this extended year yielded a further 124 articles but none was randomized controlled trial suitable for inclusion. The current study therefore included the same eight trials [5–12] as in the previous meta-analysis [4], but data were analyzed by the new method of meta-regression.
3. Statistical analysis
1889
Table 1 Criteria/indications for urgent angiography in the “non-routine” arm and overall angiography rate in the “non-routine” arm. Trials (years of study)
Criteria/indications for urgent angiography
Overall % angiography
SIAM III [5] (1998–2001) TRANSFER [6] (2004–2007) NORDISTEMI [7] (2005–2008) CAPITAL AMI [8] (2001–2004)
Ongoing ischemia
100%a
CARESS-AMI [9] (2002–2007) GRACIA-1 [10] (2000–2005) PRAGUE-1 [11] (1997–1999) WEST [12] (2005)
Persistent chest pain and b 50% reduction of ST elevation 60–90 min post‐lysis Persistent chest pain and b 50% reduction of ST elevation 60–90 min post-lysis Persistent chest pain and of ST elevation ≥ 90 min after initiation of thrombolysis or deteriorating hemodynamic state Rescue PCI
88.7% 86% 67.5%
35.7%
Rest angina with ECG changes or CCS class III/IV effort angina or positive stress Rescue PCI after failed lysis
21% 15%b
Rescue PCI
15%b
a
Meta-regression is a random effects regression weighted by the inverse of the variance of the effect size. In contrast to meta-analysis, meta-regression aims to relate the size of effect to one or more characteristics of the individual studies involved. Meta-regression analysis is a technique that allows evaluation of outcomes to include some baseline variables that are different in the studies being analyzed [13,14]. In this analysis, the angiography rate among patients not randomized to routine early invasive strategy (“non-routine” group) was used as the predictor variable of the endpoint 30-day mortality. Meta-regression was performed using the metareg function of Stata, which performs random-effects meta-regression using aggregate-level data weighted by the inverse of the variance of the effect size. This function uses an iterative method to produce estimates of regression parameters, their asymptotic variances, and the residual heterogeneity variance. While the impact of any measurable characteristic may theoretically be investigated using meta-regression, results are easier to interpret when the characteristic has high variability across studies. Angiography rates vary widely across studies included in this meta-regression, so it is likely that statistical non-significance indicates the absence of a true relationship [13,14]. Secondary analysis was performed to discern other potential independent predictors in the meta-regression analysis. Parameters included the PCI procedural success rate in the routine early invasive arm; and the rates of PCI and CABG before hospital discharge in the non-routine arm. Exploratory analysis was also performed in an attempt to compare the performance of urgent versus non-urgent angiography in the non-routine arm. 4. Results Eight trials were included for meta-regression with a total of 3195 patients. Table 1 reports these trials, the years of each trial, the reported indications for urgent angiography in the non-routine arm, and the overall angiography rates (including both urgent and non-urgent angiography) in the non-routine arm for each trial. Trials are listed in descending order of the overall angiography rates in the nonroutine group. Apart from SIAM-III where all patients eventually had angiography by protocol, the overall angiography rates would be according to the standard of care during the trial period and was higher in the more recent trials of TRANSFER-AMI (88.7%) and NORDISTEMI (86%). Fig. 1 lists the eight trials according to the descending order in angiography rates within the non-routine strategy (remainder) arm which ranges from 100% to 15%. The overall odds ratio for 30-day mortality
By protocol all patients had undergone angiography 2 weeks after lysis. Angiography rates were not reported in the original publications and were estimated from the reported revascularization rates. b
comparing the routine early angiography arm vs the remainder nonroutine arm is 0.86 (95% confidence interval 0.60–1.24). In Fig. 1, there is no visual trend in the odds ratio estimates for mortality comparing the 2 treatment strategies as angiography rate decreases. Meta-regression analysis indicates that angiography rate is not a predictor of 30-day mortality (p = 0.461). In the secondary analysis, all parameters tested are not significant: procedural success rate in the routine early invasive arm (p = 0.311), rate of CABG in the nonroutine arm (p = 0.342), and rate of PCI in the non-routine arm (p = 0.64). Table 2 reports the indications for angiography in the non-routine group and timing of the angiography for both groups. None of the original reports of the 8 included trials [5–12] provides direct information comparing outcomes of patients undergoing urgent angiography versus patients undergoing non-urgent angiography in the nonroutine arm. Further hand search for subsequent publications from these trials reveals no further information.
5. Discussion This study demonstrates that the neutral effect on mortality comparing routine early invasive strategy versus the non-routine strategy cannot be explained by the variable performance of angiography in the latter. The trials with low angiography rates could have correctly selected the patients who benefited from angiography including those with failed reperfusion by fibrinolysis, as depicted in Table 1. This cannot be further assessed as information on the outcome of patients according to the indications for angiography was not available from the included trials. Alternatively, the routine early strategy could have submitted some sick patients to high risks during lengthy inter-hospital transfers that overrode any potential benefits. This latter explanation has some support from the 1059 patient TRANSFER-AMI trial [6]. Despite having an overall significant reduction of the composite cardiac endpoints and a non-significant reduction of combined death and reinfarction within 30 days with the early routine invasive strategy following fibrinolysis, 16% of their patients deemed at “high risk” by the GRACE risk score actually had harm from the invasive strategy whereas the lower risk patients had benefit, with a significant interaction between baseline risk and treatment effect regarding the outcome of death or reinfarction (p = 0.001) [15]. In TRANSFER AMI
1890
C-K. Wong et al. / International Journal of Cardiology 167 (2012) 1888–1891
Fig. 1. Odds ratio for 30-day mortality comparing routine early invasive strategy versus remainder (i.e. “non-routine” group). Results were pooled using random effects model and inverse variance (IV) weights of the effect size. The first column is angiography rates in the remainder group of the included studies [5–12].
30-day mortality was 4.5% in the routine early PCI group and 3.4% in the non-routine group (p = 0.39). “High risk” can be due to many factors including age and renal dysfunction which contribute to the GRACE score but these are generally irreversible risk factors despite a successful PCI. On the contrary, high risk can also be due to a tight discrete residual stenosis of the infarct-artery serving a large territory of viable myocardium and a successful PCI should provide dramatic benefit. In the American guidelines which take a softer stance on the routine early angiography [2], only high risk “cardiac” characteristics are included for decisions on transferring patients to PCI centers. These characteristics include hemodynamic disturbances, large ECG infarct size including right ventricular infarction, low ejection fractions including prior MIs, and new onset LBBB [2]. Indeed, clinicians almost always use the amount of ST changes on the 12-lead baseline ECG as hints of the potential size of the infarct territory. However, information from ST changes in aVR [16] or V1
Table 2 Time to angiography or PCI from lysis or randomization and indications for angiography in the non-routine arm. Trials (years of study)
Time to angiography/ PCI in the routine arm
Time to angiography/ Indications for angiography in the PCI in the nonnon-routine arm (% of routine arm patients in the arm)
SIAM III [5] (1998–2001)
Mean of 3.5 h
Mean of 11.7 days; 23% unplanned procedures within a mean of 3.1 days
TRANSFER [6] (2004–2007)
Median of 2.8 h
Median of 32.5 h
NORDISTEMI [7] (2005–2008)
Median of 2.7 h; 83% within 3 h, 99% within 12 h CAPITAL AMI [8] Within 3 h; (2001–2004) median of 1.6 h CARESS-AMI [9] Median of (2002–2007) 2.3 h GRACIA-1 [10] Mean of 4.8 h (2000–2005)
PRAGUE-1 [11] Mean of 1.8 h (1997–1999) WEST [12] Median of (2005) 5.4 h
Median of 3 days; 12% within 3 h, 33% within 12 h
Unplanned procedures in 23% including ECG signs of ischemia 11%, angina 11%, and hemodynamic instability 1% 35% urgent angiography within 12 h — reinfarction 4%, cardiogenic shock or worsening failure 2%, recurrent ischemia 25%, unknown 4% Rescue in 28%
Median of 2.5 days
Rescue in 9.5%; recurrent ischemia 39%
Median of 3.5 h
Rescue in 30%
Not reported
Not reported
Spontaneous ischemia in 12% and stress induced ischemia in 8% Not reported
Median of 5.9 h
Rescue in 14%
elevation during inferior STEMI [17], and prolonged QRS duration (particularly during RBBB) [18] are also helpful. The issue of LBBB is complex [18]. If found on the presenting ECG of patients with ischemic symptoms, fibrinolytic therapy is often administered presuming the LBBB as new, as it is impossible to have ECG recordings right before the onset of symptoms. The Sgarbossa criteria of ST changes during LBBB are specific but not sensitive for diagnosing MI [18]. At the same time, “STEMI” can be over-diagnosed in many patients not fulfilling the Sgarbossa criteria. For example, enzymatic MI was not confirmed in 27% of the LBBB patients without those ST changes during LBBB in the HERO-2 cohort although all received fibrinolysis [19]. Even high “cardiac” risks from the index STEMI itself do not necessarily imply benefit from PCI of the infarct-related lesion because of uncertainty on the viability of the subtended myocardium. Q waves in the infarct leads may reflect the “transmural extent” of the infarct. Classified dichotomously, the presence of Q waves portends a worse outcome regardless of the reperfusion approach [20]. Q waves should be further studied as a continuous variable (including depth and lead distribution) to confirm its utility and if so the cut-points used for clinical decisions. Despite the above considerations, routine early PCI clearly reduces recurrent MI and ischemia among the same patients included in this analysis, as previously demonstrated [4]. We did not perform metaregression analysis using these 2 endpoints versus the rates of angiography in the non-routine arm because these non-mortality endpoints would be inter-related with the performance of angiography. Indeed, they often were the reported indications for urgent angiography in the non-routine arm (Table 1). Detailed patient level data on the timing of angiography versus timing of these endpoints are needed to dissect their complex relationships. Notwithstanding the complexities of such patient level analysis requiring the timing of events, recurrent MI and ischemia are softendpoints dependent on ascertainment and adjudication. These events are expected after medical treatment of STEMI. The benefit of a routine invasive strategy in reducing these endpoints [4] reflects the efficacy of PCI in preventing spontaneous events, but there can also be underdiagnosis of ischemic events related to PCI. Diagnostic difficulties for peri-PCI infarctions have been acknowledged in detail by the joint ESC/ACCF/AHA/WHF Task force in their document “Universal Definition of Myocardial Infarction” [21]. Thus, only 30-day mortality endpoint was used in this meta-regression analysis. For STEMI, mortality endpoint clearly trumps the softer endpoints of recurrent MI and ischemia. The current study shows that the equipoise between the routine early invasive versus the non-routine strategy on 30-day mortality cannot be explained by the variable performance of angiography in the non-routine strategy arm. Angiography rates after fibrinolysis has been increasing in recent years and likely has contributed to improved patient care. In the more recent TRANSFER-AMI and NORDISTEMI trials, almost up to 90% of patients underwent angiography despite being randomized
C-K. Wong et al. / International Journal of Cardiology 167 (2012) 1888–1891
into the non-routine arm. Outcome according to the indications for angiography (urgent versus non-urgent) in these patients will provide useful information and may be available in future reports from the two groups. Meanwhile, routine urgent transferral within 24 h for angiography after fibrinolysis [1–3] should be applied according to the local situation. The assessment of risk–benefit ratio for individual clinical patient can be inherently complex but careful assessment of the always available ECG will help. Other factors such as the safety of urgent transferral, extent of available support and resources will also influence the final outcome. Acknowledgment
[8]
[9]
[10]
[11]
The authors of this manuscript have certified that they comply with the principles of ethical publishing in the International Journal of Cardiology.
[12]
References
[13]
[1] Wijns W, Kohl P, Danchin N, et al. Guidelines on myocardial revascularization: the Task Force on Myocardial Revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J 2010;31:2501–55. [2] Kushner FG, Hand M, Smith SC, et al. Focused updates: ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction (updating the 2004 guideline and 2007 focused update) and ACC/AHA/SCAI guidelines on percutaneous coronary intervention (updating the 2005 guideline and 2007 focused update): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2009;54:2205–41. [3] Levine GN, Bates ER, Blankenship JC, et al. 2011 ACCF/AHA/SCAI guideline for percutaneous coronary intervention: executive summary: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines and the Society for Cardiovascular Angiography and Interventions. J Am Coll Cardiol 2011;58:2550–83. [4] D'Souza SP, Mamas MA, Fraser DG, Fath-Ordoubadi F. Routine early coronary angioplasty versus ischaemia-guided angioplasty after thrombolysis in acute ST-elevation myocardial infarction: a meta-analysis. Eur Heart J 2011;32:972–82. [5] Scheller B, Hennen B, Hammer B, et al. Beneficial effects of immediate stenting after thrombolysis in acute myocardial infarction. J Am Coll Cardiol 2003;42: 634–41. [6] Cantor WJ, Fitchett D, Borgundvaag B, et al. Routine early angioplasty after fibrinolysis for acute myocardial infarction. N Engl J Med 2009;360:2705–18. [7] Bohmer E, Hoffmann P, Abdelnoor M, Arnesen H, Halvorsen S. Efficacy and safety of immediate angioplasty versus ischemia-guided management after thromboly-
[14] [15]
[16] [17]
[18]
[19]
[20]
[21]
1891
sis in acute myocardial infarction in areas with very long transfer distances results of the NORDISTEMI (NORwegian study on DIstrict treatment of ST-Elevation Myocardial Infarction). J Am Coll Cardiol 2010;55:102–10. Le May MR, Wells GA, Labinaz M, et al. Combined angioplasty and pharmacological intervention versus thrombolysis alone in acute myocardial infarction (CAPITAL AMI study). J Am Coll Cardiol 2005;46:417–24. Di Mario C, Dudek D, Piscione F, et al. Immediate angioplasty versus standard therapy with rescue angioplasty after thrombolysis in the Combined Abciximab REteplase Stent Study in Acute Myocardial Infarction (CARESS-in-AMI): an open, prospective, randomised, multicentre trial. Lancet 2008;371:559–68. Fernandez-Aviles F, Alonso JJ, Castro-Beiras A, et al. Routine invasive strategy within 24 hours of thrombolysis versus ischaemia-guided conservative approach for acute myocardial infarction with ST-segment elevation (GRACIA-1): a randomised controlled trial. Lancet 2004;364:1045–53. Widimsky P, Groch L, Zelizko M, Aschermann M, Bednar F, Suryapranata H. Multicentre randomized trial comparing transport to primary angioplasty vs immediate thrombolysis vs combined strategy for patients with acute myocardial infarction presenting to a community hospital without a catheterization laboratory. The PRAGUE study. Eur Heart J 2000;21:823–31. Armstrong PW. A comparison of pharmacologic therapy with/without timely coronary intervention vs. primary percutaneous intervention early after ST-elevation myocardial infarction: the WEST (Which Early ST-elevation myocardial infarction Therapy) study. Eur Heart J 2006;27:1530–8. Thompson S, Higgin JPT. How should meta-regression analyses be undertaken and interpreted? Stat Med 2002;21:1559–73. van Houwelingen HC, Arends LR, Stijnen T. Advanced methods in meta-analysis: multivariate approach and meta-regression. Stat Med 2002;21(4):589–624. Yan AT, Yan RT, Cantor WJ, et al. Relationship between risk stratification at admission and treatment effects of early invasive management following fibrinolysis: insights from the Trial of Routine ANgioplasty and Stenting After Fibrinolysis to Enhance Reperfusion in Acute Myocardial Infarction (TRANSFER-AMI). Eur Heart J 2011;32:1994–2002. Wong CK. Detecting ST deviations: does body surface mapping help or have we missed information from lead aVR? Int J Cardiol 2011;152:403–7. Wong CK. Isolated right ventricular infarction as a cause for anterior (V1–V3) ST elevation — review of publications since 2006 and re-appraisal of direct vs reciprocal ST changes. Int J Cardiol 2011;151:239–41. Wong CK. The evolution of reperfusion management in patients with suspected myocardial infarction and bundle branch block: how ECG language will intertwine with angiographic findings. Int J Cardiol 2011;153:327–8. Wong CK, French JK, Aylward PE, et al. Patients with prolonged ischemic chest pain and presumed-new left bundle branch block have heterogeneous outcomes depending on the presence of ST-segment changes. J Am Coll Cardiol 2005;46: 29–38. Wong CK, Herbison P. Initial Q waves and outcome after reperfusion therapy in patients with ST elevation acute myocardial infarction: a systematic review. Int J Cardiol 2011;148:305–8. Thygesen K, Alpert JS, White HD. Universal definition of myocardial infarction. J Am Coll Cardiol 2007;50:2173–87.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Role of routine early angiography post-fibrinolysis for ST elevation myocardial infarction — A meta-regression analysis using angiography rate in the non-routine arm Cheuk-Kit Wong a,⁎, Sophia Leon de la Barra b, Peter Herbison b a b
Department of Cardiology, Dunedin School of Medicine, University of Otago, Dunedin Public Hospital, Dunedin, New Zealand Department of Preventive and Social Medicine, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand
a r t i c l e
i n f o
Article history: Received 2 February 2012 Received in revised form 24 April 2012 Accepted 28 April 2012 Available online 21 May 2012 Keywords: Routine angiography Fibrinolysis Meta-regression
a b s t r a c t Background: The current European and American Guidelines differ with regard to the recommended level for the use of routine early angiography after fibrinolysis for STEMI. Previous meta-analyses on randomized controlled trials have supported the routine early approach, but its advantage may be because of an excessively low angiography rate among patients in the non-routine strategy arm of the trials. Methods: We update the meta-analysis and apply meta-regression to evaluate whether the difference in outcome between the 2 randomized arms could be explained by the angiography rates in the non-routine strategy arm. Because reinfarction and recurrent ischemia are often the reported indication for angiography, we only use mortality endpoint in our meta-regression analysis. Results: Among the eight trials included with 3195 patients, the angiography rate in the non-routine strategy arms ranges from 15% to 100%. The overall odds ratio for 30-day mortality comparing the routine early angiography arm vs the non-routine arm is 0.86 (95% confidence interval 0.60–1.24). On the plot listing the eight trials according to angiography rates, there is no visual trend in the odds ratio estimates for mortality when comparing the 2 treatment strategies as angiography rate decreases. In meta-regression analysis, angiography rate does not predict 30-day mortality (p = 0.461). Conclusion: For STEMI, mortality endpoint trumps the softer endpoints of recurrent infarction and ischemia. The current study shows that the equipoise between the routine early invasive versus the non-routine strategy on 30-day mortality cannot be explained by the variable performance of angiography in the non-routine strategy arm. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The European recommendations give a class I recommendation (level of evidence A) for routine early percutaneous coronary intervention (PCI) within 24 h following successful fibrinolysis [1], but the American recommendations take a softer stance. The ACC/AHA 2009 update to the STEMI guidelines [2] gives a class IIb recommendation (level of evidence C) stating that “Patients who are not at high risk who receive fibrinolytic therapy as primary reperfusion therapy at a non‐PCI-capable facility may be considered for transfer as soon as possible to a PCI-capable facility where PCI can be performed either when needed or as a pharmacoinvasive strategy”, and a class IIa (level of evidence B) recommendation for this strategy for high-risk patients. The 2011 ACCF/AHA/SCAI recommendations for
⁎ Corresponding author. Tel.: + 64 3 4747980; fax: + 64 3 4747655. E-mail address: [email protected] (C-K. Wong). 0167-5273/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2012.04.151
Percutaneous Coronary Intervention [3] for STEMI patients initially treated by fibrinolysis are similar. In the latest meta-analysis [4] on clinical trials that randomly assigned patients following fibrinolysis to routine early coronary angiography and PCI vs the “non-routine” strategy, there were respectively 38% and 79% significant reductions in recurrent MI and recurrent ischemia in the routine early strategy arm with no substantial reduction in mortality. While almost all patients in the routine early strategy arm had angiography, the proportion undergoing angiography in the non-routine strategy arm of the included studies varied widely. Angiography, by providing unique information for revascularization, might have been the crucial step altering outcome. However, the decision on angiography in the non-routine arm would have been affected by factors including recurrent MI and ischemia, which are also the softer endpoints of the randomized trials. Mortality is the hard endpoint in STEMI. The impact on mortality from a routine early invasive strategy over a non-routine strategy may
C-K. Wong et al. / International Journal of Cardiology 167 (2012) 1888–1891
vary with the proportion of patients that underwent angiography with the latter strategy even when the overall effect between the 2 strategies is neutral (i.e. studies with high angiography rates producing an effect which is offset by studies with low angiography rates). We address this question using the established statistical technique of metaregression. 2. Methods MEDLINE was searched using the search terms “coronary, thrombolysis, early or immediate stenting or angioplasty, and acute ST elevation myocardial infarction”, in order to obtain randomized controlled trials comparing the routine early angiography strategy vs the non-routine strategy. This method was identical to the most recently published meta-analysis [4] of the period 1977 to March 2010 except that the search was extended for a further year to March 2011. The search on this extended year yielded a further 124 articles but none was randomized controlled trial suitable for inclusion. The current study therefore included the same eight trials [5–12] as in the previous meta-analysis [4], but data were analyzed by the new method of meta-regression.
3. Statistical analysis
1889
Table 1 Criteria/indications for urgent angiography in the “non-routine” arm and overall angiography rate in the “non-routine” arm. Trials (years of study)
Criteria/indications for urgent angiography
Overall % angiography
SIAM III [5] (1998–2001) TRANSFER [6] (2004–2007) NORDISTEMI [7] (2005–2008) CAPITAL AMI [8] (2001–2004)
Ongoing ischemia
100%a
CARESS-AMI [9] (2002–2007) GRACIA-1 [10] (2000–2005) PRAGUE-1 [11] (1997–1999) WEST [12] (2005)
Persistent chest pain and b 50% reduction of ST elevation 60–90 min post‐lysis Persistent chest pain and b 50% reduction of ST elevation 60–90 min post-lysis Persistent chest pain and of ST elevation ≥ 90 min after initiation of thrombolysis or deteriorating hemodynamic state Rescue PCI
88.7% 86% 67.5%
35.7%
Rest angina with ECG changes or CCS class III/IV effort angina or positive stress Rescue PCI after failed lysis
21% 15%b
Rescue PCI
15%b
a
Meta-regression is a random effects regression weighted by the inverse of the variance of the effect size. In contrast to meta-analysis, meta-regression aims to relate the size of effect to one or more characteristics of the individual studies involved. Meta-regression analysis is a technique that allows evaluation of outcomes to include some baseline variables that are different in the studies being analyzed [13,14]. In this analysis, the angiography rate among patients not randomized to routine early invasive strategy (“non-routine” group) was used as the predictor variable of the endpoint 30-day mortality. Meta-regression was performed using the metareg function of Stata, which performs random-effects meta-regression using aggregate-level data weighted by the inverse of the variance of the effect size. This function uses an iterative method to produce estimates of regression parameters, their asymptotic variances, and the residual heterogeneity variance. While the impact of any measurable characteristic may theoretically be investigated using meta-regression, results are easier to interpret when the characteristic has high variability across studies. Angiography rates vary widely across studies included in this meta-regression, so it is likely that statistical non-significance indicates the absence of a true relationship [13,14]. Secondary analysis was performed to discern other potential independent predictors in the meta-regression analysis. Parameters included the PCI procedural success rate in the routine early invasive arm; and the rates of PCI and CABG before hospital discharge in the non-routine arm. Exploratory analysis was also performed in an attempt to compare the performance of urgent versus non-urgent angiography in the non-routine arm. 4. Results Eight trials were included for meta-regression with a total of 3195 patients. Table 1 reports these trials, the years of each trial, the reported indications for urgent angiography in the non-routine arm, and the overall angiography rates (including both urgent and non-urgent angiography) in the non-routine arm for each trial. Trials are listed in descending order of the overall angiography rates in the nonroutine group. Apart from SIAM-III where all patients eventually had angiography by protocol, the overall angiography rates would be according to the standard of care during the trial period and was higher in the more recent trials of TRANSFER-AMI (88.7%) and NORDISTEMI (86%). Fig. 1 lists the eight trials according to the descending order in angiography rates within the non-routine strategy (remainder) arm which ranges from 100% to 15%. The overall odds ratio for 30-day mortality
By protocol all patients had undergone angiography 2 weeks after lysis. Angiography rates were not reported in the original publications and were estimated from the reported revascularization rates. b
comparing the routine early angiography arm vs the remainder nonroutine arm is 0.86 (95% confidence interval 0.60–1.24). In Fig. 1, there is no visual trend in the odds ratio estimates for mortality comparing the 2 treatment strategies as angiography rate decreases. Meta-regression analysis indicates that angiography rate is not a predictor of 30-day mortality (p = 0.461). In the secondary analysis, all parameters tested are not significant: procedural success rate in the routine early invasive arm (p = 0.311), rate of CABG in the nonroutine arm (p = 0.342), and rate of PCI in the non-routine arm (p = 0.64). Table 2 reports the indications for angiography in the non-routine group and timing of the angiography for both groups. None of the original reports of the 8 included trials [5–12] provides direct information comparing outcomes of patients undergoing urgent angiography versus patients undergoing non-urgent angiography in the nonroutine arm. Further hand search for subsequent publications from these trials reveals no further information.
5. Discussion This study demonstrates that the neutral effect on mortality comparing routine early invasive strategy versus the non-routine strategy cannot be explained by the variable performance of angiography in the latter. The trials with low angiography rates could have correctly selected the patients who benefited from angiography including those with failed reperfusion by fibrinolysis, as depicted in Table 1. This cannot be further assessed as information on the outcome of patients according to the indications for angiography was not available from the included trials. Alternatively, the routine early strategy could have submitted some sick patients to high risks during lengthy inter-hospital transfers that overrode any potential benefits. This latter explanation has some support from the 1059 patient TRANSFER-AMI trial [6]. Despite having an overall significant reduction of the composite cardiac endpoints and a non-significant reduction of combined death and reinfarction within 30 days with the early routine invasive strategy following fibrinolysis, 16% of their patients deemed at “high risk” by the GRACE risk score actually had harm from the invasive strategy whereas the lower risk patients had benefit, with a significant interaction between baseline risk and treatment effect regarding the outcome of death or reinfarction (p = 0.001) [15]. In TRANSFER AMI
1890
C-K. Wong et al. / International Journal of Cardiology 167 (2012) 1888–1891
Fig. 1. Odds ratio for 30-day mortality comparing routine early invasive strategy versus remainder (i.e. “non-routine” group). Results were pooled using random effects model and inverse variance (IV) weights of the effect size. The first column is angiography rates in the remainder group of the included studies [5–12].
30-day mortality was 4.5% in the routine early PCI group and 3.4% in the non-routine group (p = 0.39). “High risk” can be due to many factors including age and renal dysfunction which contribute to the GRACE score but these are generally irreversible risk factors despite a successful PCI. On the contrary, high risk can also be due to a tight discrete residual stenosis of the infarct-artery serving a large territory of viable myocardium and a successful PCI should provide dramatic benefit. In the American guidelines which take a softer stance on the routine early angiography [2], only high risk “cardiac” characteristics are included for decisions on transferring patients to PCI centers. These characteristics include hemodynamic disturbances, large ECG infarct size including right ventricular infarction, low ejection fractions including prior MIs, and new onset LBBB [2]. Indeed, clinicians almost always use the amount of ST changes on the 12-lead baseline ECG as hints of the potential size of the infarct territory. However, information from ST changes in aVR [16] or V1
Table 2 Time to angiography or PCI from lysis or randomization and indications for angiography in the non-routine arm. Trials (years of study)
Time to angiography/ PCI in the routine arm
Time to angiography/ Indications for angiography in the PCI in the nonnon-routine arm (% of routine arm patients in the arm)
SIAM III [5] (1998–2001)
Mean of 3.5 h
Mean of 11.7 days; 23% unplanned procedures within a mean of 3.1 days
TRANSFER [6] (2004–2007)
Median of 2.8 h
Median of 32.5 h
NORDISTEMI [7] (2005–2008)
Median of 2.7 h; 83% within 3 h, 99% within 12 h CAPITAL AMI [8] Within 3 h; (2001–2004) median of 1.6 h CARESS-AMI [9] Median of (2002–2007) 2.3 h GRACIA-1 [10] Mean of 4.8 h (2000–2005)
PRAGUE-1 [11] Mean of 1.8 h (1997–1999) WEST [12] Median of (2005) 5.4 h
Median of 3 days; 12% within 3 h, 33% within 12 h
Unplanned procedures in 23% including ECG signs of ischemia 11%, angina 11%, and hemodynamic instability 1% 35% urgent angiography within 12 h — reinfarction 4%, cardiogenic shock or worsening failure 2%, recurrent ischemia 25%, unknown 4% Rescue in 28%
Median of 2.5 days
Rescue in 9.5%; recurrent ischemia 39%
Median of 3.5 h
Rescue in 30%
Not reported
Not reported
Spontaneous ischemia in 12% and stress induced ischemia in 8% Not reported
Median of 5.9 h
Rescue in 14%
elevation during inferior STEMI [17], and prolonged QRS duration (particularly during RBBB) [18] are also helpful. The issue of LBBB is complex [18]. If found on the presenting ECG of patients with ischemic symptoms, fibrinolytic therapy is often administered presuming the LBBB as new, as it is impossible to have ECG recordings right before the onset of symptoms. The Sgarbossa criteria of ST changes during LBBB are specific but not sensitive for diagnosing MI [18]. At the same time, “STEMI” can be over-diagnosed in many patients not fulfilling the Sgarbossa criteria. For example, enzymatic MI was not confirmed in 27% of the LBBB patients without those ST changes during LBBB in the HERO-2 cohort although all received fibrinolysis [19]. Even high “cardiac” risks from the index STEMI itself do not necessarily imply benefit from PCI of the infarct-related lesion because of uncertainty on the viability of the subtended myocardium. Q waves in the infarct leads may reflect the “transmural extent” of the infarct. Classified dichotomously, the presence of Q waves portends a worse outcome regardless of the reperfusion approach [20]. Q waves should be further studied as a continuous variable (including depth and lead distribution) to confirm its utility and if so the cut-points used for clinical decisions. Despite the above considerations, routine early PCI clearly reduces recurrent MI and ischemia among the same patients included in this analysis, as previously demonstrated [4]. We did not perform metaregression analysis using these 2 endpoints versus the rates of angiography in the non-routine arm because these non-mortality endpoints would be inter-related with the performance of angiography. Indeed, they often were the reported indications for urgent angiography in the non-routine arm (Table 1). Detailed patient level data on the timing of angiography versus timing of these endpoints are needed to dissect their complex relationships. Notwithstanding the complexities of such patient level analysis requiring the timing of events, recurrent MI and ischemia are softendpoints dependent on ascertainment and adjudication. These events are expected after medical treatment of STEMI. The benefit of a routine invasive strategy in reducing these endpoints [4] reflects the efficacy of PCI in preventing spontaneous events, but there can also be underdiagnosis of ischemic events related to PCI. Diagnostic difficulties for peri-PCI infarctions have been acknowledged in detail by the joint ESC/ACCF/AHA/WHF Task force in their document “Universal Definition of Myocardial Infarction” [21]. Thus, only 30-day mortality endpoint was used in this meta-regression analysis. For STEMI, mortality endpoint clearly trumps the softer endpoints of recurrent MI and ischemia. The current study shows that the equipoise between the routine early invasive versus the non-routine strategy on 30-day mortality cannot be explained by the variable performance of angiography in the non-routine strategy arm. Angiography rates after fibrinolysis has been increasing in recent years and likely has contributed to improved patient care. In the more recent TRANSFER-AMI and NORDISTEMI trials, almost up to 90% of patients underwent angiography despite being randomized
C-K. Wong et al. / International Journal of Cardiology 167 (2012) 1888–1891
into the non-routine arm. Outcome according to the indications for angiography (urgent versus non-urgent) in these patients will provide useful information and may be available in future reports from the two groups. Meanwhile, routine urgent transferral within 24 h for angiography after fibrinolysis [1–3] should be applied according to the local situation. The assessment of risk–benefit ratio for individual clinical patient can be inherently complex but careful assessment of the always available ECG will help. Other factors such as the safety of urgent transferral, extent of available support and resources will also influence the final outcome. Acknowledgment
[8]
[9]
[10]
[11]
The authors of this manuscript have certified that they comply with the principles of ethical publishing in the International Journal of Cardiology.
[12]
References
[13]
[1] Wijns W, Kohl P, Danchin N, et al. Guidelines on myocardial revascularization: the Task Force on Myocardial Revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J 2010;31:2501–55. [2] Kushner FG, Hand M, Smith SC, et al. Focused updates: ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction (updating the 2004 guideline and 2007 focused update) and ACC/AHA/SCAI guidelines on percutaneous coronary intervention (updating the 2005 guideline and 2007 focused update): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2009;54:2205–41. [3] Levine GN, Bates ER, Blankenship JC, et al. 2011 ACCF/AHA/SCAI guideline for percutaneous coronary intervention: executive summary: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines and the Society for Cardiovascular Angiography and Interventions. J Am Coll Cardiol 2011;58:2550–83. [4] D'Souza SP, Mamas MA, Fraser DG, Fath-Ordoubadi F. Routine early coronary angioplasty versus ischaemia-guided angioplasty after thrombolysis in acute ST-elevation myocardial infarction: a meta-analysis. Eur Heart J 2011;32:972–82. [5] Scheller B, Hennen B, Hammer B, et al. Beneficial effects of immediate stenting after thrombolysis in acute myocardial infarction. J Am Coll Cardiol 2003;42: 634–41. [6] Cantor WJ, Fitchett D, Borgundvaag B, et al. Routine early angioplasty after fibrinolysis for acute myocardial infarction. N Engl J Med 2009;360:2705–18. [7] Bohmer E, Hoffmann P, Abdelnoor M, Arnesen H, Halvorsen S. Efficacy and safety of immediate angioplasty versus ischemia-guided management after thromboly-
[14] [15]
[16] [17]
[18]
[19]
[20]
[21]
1891
sis in acute myocardial infarction in areas with very long transfer distances results of the NORDISTEMI (NORwegian study on DIstrict treatment of ST-Elevation Myocardial Infarction). J Am Coll Cardiol 2010;55:102–10. Le May MR, Wells GA, Labinaz M, et al. Combined angioplasty and pharmacological intervention versus thrombolysis alone in acute myocardial infarction (CAPITAL AMI study). J Am Coll Cardiol 2005;46:417–24. Di Mario C, Dudek D, Piscione F, et al. Immediate angioplasty versus standard therapy with rescue angioplasty after thrombolysis in the Combined Abciximab REteplase Stent Study in Acute Myocardial Infarction (CARESS-in-AMI): an open, prospective, randomised, multicentre trial. Lancet 2008;371:559–68. Fernandez-Aviles F, Alonso JJ, Castro-Beiras A, et al. Routine invasive strategy within 24 hours of thrombolysis versus ischaemia-guided conservative approach for acute myocardial infarction with ST-segment elevation (GRACIA-1): a randomised controlled trial. Lancet 2004;364:1045–53. Widimsky P, Groch L, Zelizko M, Aschermann M, Bednar F, Suryapranata H. Multicentre randomized trial comparing transport to primary angioplasty vs immediate thrombolysis vs combined strategy for patients with acute myocardial infarction presenting to a community hospital without a catheterization laboratory. The PRAGUE study. Eur Heart J 2000;21:823–31. Armstrong PW. A comparison of pharmacologic therapy with/without timely coronary intervention vs. primary percutaneous intervention early after ST-elevation myocardial infarction: the WEST (Which Early ST-elevation myocardial infarction Therapy) study. Eur Heart J 2006;27:1530–8. Thompson S, Higgin JPT. How should meta-regression analyses be undertaken and interpreted? Stat Med 2002;21:1559–73. van Houwelingen HC, Arends LR, Stijnen T. Advanced methods in meta-analysis: multivariate approach and meta-regression. Stat Med 2002;21(4):589–624. Yan AT, Yan RT, Cantor WJ, et al. Relationship between risk stratification at admission and treatment effects of early invasive management following fibrinolysis: insights from the Trial of Routine ANgioplasty and Stenting After Fibrinolysis to Enhance Reperfusion in Acute Myocardial Infarction (TRANSFER-AMI). Eur Heart J 2011;32:1994–2002. Wong CK. Detecting ST deviations: does body surface mapping help or have we missed information from lead aVR? Int J Cardiol 2011;152:403–7. Wong CK. Isolated right ventricular infarction as a cause for anterior (V1–V3) ST elevation — review of publications since 2006 and re-appraisal of direct vs reciprocal ST changes. Int J Cardiol 2011;151:239–41. Wong CK. The evolution of reperfusion management in patients with suspected myocardial infarction and bundle branch block: how ECG language will intertwine with angiographic findings. Int J Cardiol 2011;153:327–8. Wong CK, French JK, Aylward PE, et al. Patients with prolonged ischemic chest pain and presumed-new left bundle branch block have heterogeneous outcomes depending on the presence of ST-segment changes. J Am Coll Cardiol 2005;46: 29–38. Wong CK, Herbison P. Initial Q waves and outcome after reperfusion therapy in patients with ST elevation acute myocardial infarction: a systematic review. Int J Cardiol 2011;148:305–8. Thygesen K, Alpert JS, White HD. Universal definition of myocardial infarction. J Am Coll Cardiol 2007;50:2173–87.