- Email: [email protected]
Culprit vessel versus immediate complete revascularization in patients with ST-segment myocardial infarction—a systematic review
Culprit vessel versus immediate complete revascularization in patients with ST-segment myocardial infarction—a systematic review Frederick A. Spencer, MD, FACC, FRCP(c), a Nigar Sekercioglu, MD, MSc, b Manya Prasad, MBBS, c Luciane Cruz Lopes, PhD, d and Gordon H. Guyatt, MD, MSc, FRCP(c) a,b ON, Canada; Haryana, India; and Sorocaba, Brazil
Background Guidelines suggest percutaneous intervention (PCI) of only the culprit artery in patients presenting with ST-segment elevation myocardial infarction (STEMI) and multivessel coronary artery disease. However, recent randomized controlled trials (RCTs) suggest benefit to performing PCI of other stenotic vessels at the same time as culprit vessel PCI. Methods We conducted a systematic review with complete case meta-analysis and sensitivity analyses. Data sources included MEDLINE, EMBASE, Cochrane Register of Controlled Trials, and CINAHL from 1946 to March 2014; MEDLINE and EMBASE from March 2014 to March 2015; and scanning of literature for new studies until August 2015. All RCTs comparing multivessel versus culprit-only PCI in patients with STEMI were eligible. The primary outcomes of interest were recurrent myocardial infarction (MI), recurrent revascularization, and mortality. We combined data from trials to estimate the pooled risk ratio (RR) and associated 95% CIs using random-effects models. Results Five RCTs including 1,606 patients of whom 1,568 had complete data proved eligible. Multivessel revascularization was associated with decreased risk of repeat revascularization (RR 0.36, 95% CI 0.27-0.49, risk difference 9.7% over 2 years) and recurrent nonfatal MI (RR 0.58, 95% CI 0.36-0.93, risk difference 1.8% over 2 years), without increase in mortality (RR 0.82, 95% CI 0.53-1.26) or other adverse events. Conclusions
Pooled data provide moderate-certainty evidence that performance of multivessel PCI will provide an appreciable reduction in nonfatal MI and high-certainty evidence that it will reduce need for repeat revascularization. Patients are likely to place a high value on these benefits. (Am Heart J 2015;170:1133-9.)
Based on high-certainty evidence from randomized controlled trials (RCTs), primary percutaneous coronary intervention (PCI) of the occluded artery is the standard of care in patients presenting with ST-segment elevation myocardial infarction (STEMI). 1-3 However, multivessel coronary artery disease (CAD) (additional coronary stenoses N50%-70%) is present in 40% to 60% of patients with STEMI. 4-6 In comparison to patients with single-vessel disease, patients with multivessel disease have worse
From the
a
Department of Medicine, McMaster University, Hamilton, ON, Canada,
b
Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, Canada, cDepartment of Community Medicine, Pt. B. D. Sharma Postgraduate Institute of Medical Sciences, Rohtak, Haryana, India, and dUniversity of Sorocaba, UNISO, Sorocaba, Brazil. Submitted June 28, 2015; accepted September 1, 2015. Reprint requests: Dr Frederick A. Spencer, MD, FACC, FRCP(c), Department of Medicine, McMaster University–Faculty of Health Sciences, 50 Charlton Avenue East, Hamilton, ON L8N 4A6, Canada. E-mail: [email protected] 0002-8703 © 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ahj.2015.09.002
outcomes after primary PCI of the culprit vessel, including an increased risk of 1-year mortality. A number of studies have evaluated the use of preventive PCI in nonculprit arteries (either at the same time as primary PCI of the culprit artery or shortly thereafter) in patients with STEMI. 7 Most of these studies have been observational and have offered conflicting results. Current guidelines from the European Society of Cardiology and the American College of Cardiology/American Heart Association suggest culprit vessel only PCI in patients with STEMI and multivessel CAD.8,9 Recently, we conducted a meta-analysis of the 3 RCTs including 683 patients examining culprit vessel versus immediate multivessel PCI performed at index catheterization in patients with STEMI.10 Results suggested a decrease in nonfatal myocardial infarction (MI) and subsequent revascularization and a trend toward decreased cardiac mortality, but sample size was limited and certainty of evidence remained low. Since this analysis, 2 additional RCTs have been completed and randomized an additional 923 STEMI patients to culprit-only versus multivessel PCI.11,12 We now provide an updated systematic review including these latest data.
American Heart Journal December 2015
1134 Spencer et al
Methods Eligibility criteria We included studies that (1) enrolled patients with acute STEMI and multivessel CAD; (2) randomized patients to PCI for culprit and nonculprit coronary arteries (before hospital discharge) versus culprit-only PCI; (3) reported ≥1 of the following outcomes: nonfatal MI, revascularization (PCI or coronary artery bypass graft [CABG] surgery), cardiovascular mortality, all-cause mortality; and (4) minimum follow-up of 1 year. We excluded studies in which ≥20% of patients has cardiogenic shock, significant stenosis in a prior coronary bypass graft, or significant (≥50%) left main coronary artery stenosis in which the outcomes of these patients were not reported separately. We also excluded studies in which patients underwent nonculprit artery PCI after discharge from hospital after their acute MI and culprit artery PCI. Data sources and search strategy For our initial systematic review, we searched MEDLINE, EMBASE, the Cochrane Register of Controlled Trials, and the CINAHL from 1946 to March 2014 without any restrictions. (See Supporting Information, Appendix for the search strategy of the MEDLINE database.) For the updated review, we repeated this search in MEDLINE and EMBASE for additional studies from March 2014 to March 2015. We also scanned the literature for new studies until August 2015. For every eligible study we identified and for pertinent review articles, 1 reviewer examined the reference list for additional potential citations. Study selection Teams of 2 investigators (N.S. and L.C.L. for 1946 to March 2014; F.A.S., L.C.L., or N.S. for March 2014 to March 2015) independently screened each title and abstract. If either reviewer identified a citation as relevant, we obtained the full-text manuscript. Three reviewers working in pairs (N.S., L.C.L., and F.A.S.) determined eligibility of full-text studies. Disagreement between reviewers was resolved through discussion. Data abstraction Two reviewers abstracted the following information from each study using a custom-made data collection form: study characteristics, study population, follow-up time, loss to follow-up, intervention, and outcomes. Risk of bias and certainty of evidence To assess risk of bias, 2 reviewers (N.S. and F.A.S.) used a modified version of the Cochrane tool (http:// distillercer.com/resources/). Assessment included adequacy of sequence generation, allocation sequence concealment, blinding, incomplete outcome data, selective outcome reporting, incomplete reporting for loss to follow-up, and stopping early for benefit.
We used the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) methodology to rate certainty of evidence for each outcome as high, moderate, low, or very low. 13 Guided by GRADE methodology, we assessed overall risk of bias, 14 imprecision, 15 inconsistency, 16 indirectness, 17 and publication bias 18 and summarized results in an evidence profile. Reviewers resolved any disagreements through discussion.
Data synthesis and statistical analysis We report descriptive statistics as proportions for categorical variables and mean or medians for continuous variables. Our primary analysis for nonfatal MI, revascularization, cardiovascular mortality, and all-cause mortality included only patients with complete follow-up. We conducted a plausible worse case sensitivity analysis for nonfatal MI, revascularization, and cardiovascular mortality in which patients lost to follow-up in the complete revascularization arm were estimated to have a 5-fold increase in the event rates of those not lost to follow-up and patients lost to follow-up in the culprit-only PCI arm had only 1/5 the event rate of patients not lost to followup. We calculated pooled risk ratios (RRs) and associated 95% CIs for each outcome using random-effects models applying the Mantel-Haenszel method. Absolute effects and 95% CI were calculated by multiplying pooled RRs and 95% CI by the pooled control rate of outcomes. Statistical heterogeneity was assessed by the χ 2 test and I 2 statistics. When event rates were very low (nephropathy and stroke), we calculated risk difference (RD) and 95% CI directly by pooling across trials. Funding No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the manuscript, and its final contents. None of the authors have any conflict of interest.
Results Trial identification Our initial search from 1946 to March 2014 yielded 2,837 abstracts of which 83 were duplicates. Sixteen proved potentially eligible and underwent full review. Our search from March 2014 until March 2015 revealed an additional 51 abstracts of which none proved eligible for full review. Of these studies, 13 were excluded—5 had an ineligible population, 3 did not have nonculprit artery revascularization as a comparator, 1 was a prospective cohort study, 3 had follow-up b1 year, and 1 was an editorial letter (Appendix Supplementary Figure 1). We also became aware of a randomized controlled trial reported at the European Society of Cardiology in September 2014. This study was reviewed after it was
American Heart Journal Volume 170, Number 6
Spencer et al 1135
Table I. Study and patient characteristics First author, year published No. of randomized to complete revascularization Number randomized to culprit only PCI Exclusion criteria
Description of multivessel disease Maximum possible follow-up Lost to follow-up, n Study funding
Age (y), mean (±SD) Female (%) Diabetes (%) Prior MI (%) Anterior MI (%) 3-vessel disease (%) Drug-eluting stent (% of patients)
Di Mario, 2004
Politi, 2010
Wald, 2013
Gershlick, 2014
Engstrom, 2015
52
65
234
150
314
17
84
23
146
313
Graft or prior PCI/stent stenosis, cardiogenic shock, left main stenosis
Cardiogenic shock, prior CABG, severe valvular heart disease, unsuccessful procedures
Cardiogenic shock, prior CABG, left main stenosis, chronic total occlusion in nonculprit artery
Cardiogenic shock, stent thrombosis, indication for CABG, increased bleeding risk
≥1 nonculprit artery lesions amenable to stenting 1y
N70% stenosis in ≥1 nonculprit arteries
N50% stenosis is ≥1 nonculprit arteries
5y
4y8m
Cardiogenic shock, prior CABG, previous Q-wave MI, VSD/severe MR, chronic kidney disease, clear indication/ contraindication for multivessel PCI (as per operator) N70% stenosis (or N50% in 2 planes) in ≥1 nonculprit arteries 1y
Not reported
0
Not reported
Not reported
18 (10 complete; 8 culprit only) Private, not for profit
19 (11 complete; 8 culprit only) Not for profit, peer reviewed
Complete PCI
Culprit-only PCI
Complete PCI
Culprit-only PCI
Complete PCI
Culprit-only PCI
63.5 (12.4) 11.8 11.5 NR 51.9 30.8 0⁎
65.3 (7.4) 15.4 41.2 NR 58.8 47.1 0⁎
64.5 (11.7) 23.1 13.8 NR 47.7 29.2 7.7
66.5 (13.2) 23.8 23.8 NR 41.7 25.0 11.9
62 24 15 8 29 39 63
62 19 21 7 39 33 58
N50% stenosis and FFR ≤0.80 or ≥90% stenosis 44 m
1 (complete)
Not for profit, peer reviewed
Complete PCI
Culprit-only PCI
Complete PCI
Culprit PCI
64.6 (11.2) 14.7 12.9 4.8 36 20.7 95.9
65.3 (11.9) 23.3 14.3 3.6 35.6 24.7 90.7
64 20 9 5 33 31 95
63 19 13 9 36 32 93
Abbreviations: VSD, Ventricular septal defect; MR, mitral regurgitation; FFR, fractional flow reserve; NR, not reported. ⁎ All heparin-coated stents.
published in March 2015 and met our criteria for inclusion. An additional eligible RCT was published in August 2015 (after our initial manuscript submission). In all, we included 5 randomized trials enrolling 1,606 patients. 11,12,19-21
Trial and population characteristics Table I presents characteristics of the 5 included studies. All studies enrolled patients with STEMI with b12 hours of symptoms undergoing primary PCI of culprit artery and having at least 1 additional significant nonculprit artery stenosis. Of 1,606 randomized, investi-
gators achieved complete follow-up in 1,568 patients (complete cases). In 1 study, patients with N50% angiographic stenosis in a nonculprit vessel were randomized to consideration for complete revascularization, but multivessel PCI was only performed in those vessels with fractional flow reserve of ≤0.8. 12 In 1 of the 5 studies, PCI was performed using heparin-coated stents 19; in the others, a mix of balloon angioplasty alone, bare-metal stents, and drug-eluting stents were used. One study reported complete follow-up of all patients, 20 and 1 study failed to report loss to follow-up. 19 In the other 3 studies, loss
American Heart Journal December 2015
1136 Spencer et al
Figure 1
Pooled risk of repeat revascularization with complete revascularization versus culprit-only revascularization.
to follow-up was similar in the complete and culprit-only revascularization arms (4.3% vs 3.5%, 7.3% vs 5.5%, 0.3% vs 0%, respectively) 11,12,21 . In 3 of the 5 studies, complete revascularization was performed at the time of index catheterization. In 1 study, patients underwent complete revascularization 2 days after the procedure, 12 and in the other study, 64% of patients had complete revascularization at time of index catheterization; the remaining patients had complete revascularization before hospital discharge. 11
Assessment of risk of bias All studies met most risk of bias criteria, although none could blind patients, 2 failed to blind outcome assessors, 19,20 and 1 trial stopped early for benefit 21 (Appendix Supplementary Figure 2). We further dealt with loss to follow-up as a risk of bias issue in our sensitivity analysis. Outcomes assessment Revascularization. Complete revascularization was associated with a consistent significant reduction in the risk of repeat revascularization (RR 0.36, 95% CI 0.27-0.49, heterogeneity P = .72, I 2 = 0%) (Figure 1). Our plausible worse case analysis did not significantly change these results (RR 0.41, 95% CI 0.30-0.57), and we judged limitations as minor and, therefore, rated evidence as high certainty (Table II). Based on an estimated baseline risk of repeat revascularization over 2 years in patients undergoing only culprit PCI of 15.1%, complete revascularization will result in 97 fewer repeat revascularizations per 1,000 treated over 2 years (95% CI 110 fewer to 77 fewer) (Table II). Nonfatal MI Complete revascularization was associated with a consistent significant reduction in the risk of recurrent MI (RR 0.58, 95% CI 0.36-0.93, heterogeneity P = .43, I 2 = 0%) (Figure 2). Limitations of the evidence included risk of bias (unblinded adjudication in 2 studies, early stopping in 1 study) and few events; our judgment was that data
provided moderate-certainty evidence (Table II). Based on an estimated baseline risk of recurrent nonfatal MI over 2 years in patients undergoing only culprit PCI of 4.4%, complete revascularization is likely associated with 18 fewer non-fatal MIs per 1,000 treated per 2 years (95% CI 28 fewer to 3 fewer) (Table II). Our plausible worse case analysis did not change these results significantly (RR 0.61, 95% CI 0.39-0.97).
Cardiovascular mortality Complete revascularization was associated with a consistent significant reduction in the risk of cardiovascular mortality (RR 0.46, 95% CI 0.26-0.83, heterogeneity P = .93, I 2 = 0%) (Appendix Supplementary Figure 3). Limitations of the evidence included risk of bias (unblinded adjudication and early stopping) and few events; our judgment was that data provided moderate-certainty evidence (Table II). Based on an estimated baseline risk of cardiovascular mortality over 2 years in patients undergoing culprit PCI of 2.7%, complete revascularization is likely to be associated with 15 fewer cardiovascular deaths per 1,000 treated over 5 years (95% CI 20 fewer to 5 fewer) (Table II). Our plausible worse case analysis did not significantly change these results (RR 0.51, 95% CI 0.29-0.89). Total mortality We could not confirm or exclude a difference in total mortality between patients randomized to complete versus culprit-only PCI (RR 0.82, 95% CI 0.53-1.26, heterogeneity P = .47, I 2 = 0%) (Figure 3). Based on the CI including both benefit and harm, we rated certainty of evidence as moderate (Table II). Major bleeding We could not confirm or exclude a difference in major bleeding between patients randomized to complete versus culprit-only PCI (RR 0.73, 95% CI 0.34-1.55, heterogeneity P = .40, I 2 = 0%). We, therefore, rated confidence as low (decreased due to imprecision [small number of events] and risk of bias).
American Heart Journal Volume 170, Number 6
Spencer et al 1137
Table II. Grade assessment in confidence of estimates of effect
Outcome
No. of participants (no. of studies)
Nonfatal MI
1568 (5)
Repeat 1568 (5) revascularization Cardiovascular 1568 (5) mortality Total mortality 1568 (5)
Risk of bias Serious limitations⁎ No serious limitations Serious limitations⁎ No serious limitations
Consistency Directness No serious limitations No serious limitations No serious limitations No serious limitations
No serious limitations No serious limitations No serious limitations No serious limitations
Precision No serious limitations No serious limitations No serious limitations Serious limitations ||
Publication bias Not detected Not detected Not detected Not detected
Quality
Relative effect (95% CI)
Moderate 0.58 (0.36-0.93) High 0.36 (0.27-0.49) Moderate 0.46 (0.26-0.83) Moderate 0.82 (0.53-1.26)
Absolute effect per 1000 treated per 2 y (95% CI) 18 fewer (28 to 3 fewer)† 97 fewer (110 to 77 fewer)‡ 15 fewer (20 fewer to 5 fewer)§ Not significant
⁎ Serious risk of bias due to lack of blinding for outcome assessment in 2 of the studies and stopping early for benefit in 1 study. † Two-year baseline risk estimate of 44/1,000 derived from control arm of DANAMI-PRIMULTI study12 normalized to 2 years. ‡ Two-year baseline risk estimate of 151/1,000 derived from control arm of DANAMI-PRIMULTI study12 normalized to 2 years. § Two-year baseline risk estimate of 27/1,000 derived from control arm of DANAMI-PRIMULTI study12 normalized to 2 years. || Serious limitations as 95% CI includes possible harm and possible benefit.
Figure 2
Pooled risk of nonfatal MI with complete revascularization versus culprit-only revascularization.
Other adverse events Four studies documented contrast-induced nephropathy. 11,12,20,21 Pooling data from these 4 studies, nephropathy occurred in 10 versus 15 in the complete revascularization versus culprit vessel only revascularization arms (RD −0.01, 95% CI −0.02 to 0.01). Two studies reported stroke (8 vs 3 events, RD 0.01, 95% CI −0.00 to 0.02). 11,12,21
Discussion Our systematic review of 5 RCTs including N1,600 patients demonstrates a decrease in need for repeat revascularization and in subsequent nonfatal MI in patients with STEMI undergoing multivessel PCI compared to those undergoing culprit-only PCI. In absolute terms, our data provide high-certainty evidence that opening nonculprit stenoses at the same time as culprit lesions will provide an absolute reduction of 97 revascularizations over 2 years and moderate-certainty evidence of an absolute reduction of 18 nonfatal MIs over 2 years per 1,000 STEMI patients treated (Table II).
Although the 5 studies differed in some respects (proportion of bare-metal to drug-eluting stents, timing of complete revascularization relative to index procedure, etc), heterogeneity across outcomes in our pooled data was minimal (I 2 = 0% for all outcomes, P values for test of heterogeneity from .36 to .72). Strengths of this review include a comprehensive literature search, restriction to RCTs, duplicate assessment of eligibility, risk of bias, and data abstraction, and use of the GRADE approach to rating certainty of evidence. We also conducted a sensitivity analysis (plausible worse case) exploring potential impact of loss to follow-up, which did not appreciably change our findings. Potential limitations are related to limitations in the available evidence. The 5 RCTs that have directly addressed this question enrolled a modest total sample size, the total number of events for all outcomes but revascularization is less than 100, the RCTs had risk of bias limitations, and one of the larger studies stopped early for benefit. Nevertheless, the results warrant moderate confidence for the major morbid outcomes
American Heart Journal December 2015
1138 Spencer et al
Figure 3
Pooled risk of total mortality with complete revascularization versus culprit-only revascularization.
and high confidence for revascularization. It should also be noted that these data do not address whether multivessel PCI performed as a staged procedure within a short time frame may have similar outcomes. The ongoing COMPLETE plans to enroll approximately 3,900 patients to examine this question but will not be completed until approximately 2018. An additional question that merits further study is whether the benefits seen with multivessel PCI in select patients in these randomized trials conducted in high-volume PCI centers are also realized in patients treated in lower volume community-based hospitals. Observational studies examining PCI success and complications in large and small centers may be the best approach to address this issue. Optimization of patient selection based on clinical and angiographic criteria (eg, fractional flow reserve to identify hemodynamically significant lesions) also warrants further study.
Conclusions Pooled data from 5 RCTs provide high-certainty evidence that performance of multivessel PCI will reduce need for repeat revascularization and moderate-certainty evidence that it will provide an important reduction in nonfatal MI. Current guidelines that recommend culprit-only PCI over multivessel PCI in this setting should be reassessed. These guidelines were based on prior observational data suggesting possible harm associated with performing multivessel PCI at time of an acute STEMI and theoretical concerns of an increased risk of contrast-induced nephropathy, PCIrelated complications with multiple interventions (eg, dissection, stent thrombosis), or bleeding with complete revascularization. Our pooled data fail to substantiate these latter concerns: indeed, the absolute rates of bleeding and contrast-induced nephropathy were very low and differ little or not at all between the 2 strategies. Consistent with our results, UpToDate has recently changed their guidelines to suggest multivessel PCI over culprit-only PCI in STEMI patients with additional nonculprit lesions. Our results suggest that other guidelines addressing primary PCI in patients with STEMI also require reconsideration.
Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.ahj.2015.09.002.
References 1. Zijlstra F, Hoorntje JC, de Boer MJ, et al. Long-term benefit of primary angioplasty as compared with thrombolytic therapy for acute myocardial infarction. N Engl J Med 1999;341:1413-9. 2. A clinical trial comparing primary coronary angioplasty with tissue plasminogen activator for acute myocardial infarction. The Global Use of Strategies to Open Occluded Coronary Arteries in Acute Coronary Syndromes (GUSTO IIb) Angioplasty Substudy Investigators. N Engl J Med 1997;336:1621-8. 3. Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361:13-20. 4. Sorajja P, Gersh BJ, Cox DA, et al. Impact of multivessel disease on reperfusion success and clinical outcomes in patients undergoing primary percutaneous coronary intervention for acute myocardial infarction. Eur Heart J 2007;28:1709-16. 5. Jaski BE, Cohen JD, Trausch J, et al. Outcome of urgent percutaneous transluminal coronary angioplasty in acute myocardial infarction: comparison of single-vessel versus multivessel coronary artery disease. Am Heart J 1992;124:1427-33. 6. Claessen BE, Dangas GD, Weisz G, et al. Prognostic impact of a chronic total occlusion in a non–infarct-related artery in patients with ST-segment elevation myocardial infarction: 3-year results from the HORIZONS-AMI trial. Eur Heart J 2012;33:768-75. 7. Bainey KR, Mehta SR, Lai T, et al. Complete vs culprit-only revascularization for patients with multivessel disease undergoing primary percutaneous coronary intervention for ST-segment elevation myocardial infarction: a systematic review and meta-analysis. Am Heart J 2014;167:1-14.e2. 8. Task Force on the management of STseamiotESoC, Steg PG, James SK, et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation. Eur Heart J 2012;33:2569-619. 9. O'Gara PT, Kushner FG, Ascheim DD, et al. 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2013;127:e362-425. 10. Sekercioglu N, Spencer FA, Cruz Lopes L, et al. Culprit vessel only vs immediate complete revascularization in patients with acute
American Heart Journal Volume 170, Number 6
11.
12.
13.
14.
15.
ST-segment elevation myocardial infarction: systematic review and meta-analysis. Clin Cardiol 2014;37:765-72. Gershlick AH, Khan JN, Kelly DJ, et al. Randomized trial of complete versus lesion-only revascularization in patients undergoing primary percutaneous coronary intervention for STEMI and multivessel disease: the CvLPRIT trial. J Am Coll Cardiol 2015;65:963-72. Engstrøm T, Kelbæk H, Helqvist S, et al. Complete revascularisation versus treatment of the culprit lesion only in patients with ST-segment elevation myocardial infarction and multivessel disease (DANAMI-3—PRIMULTI): an open-label, randomised controlled trial. Lancet 2015;386:665-71. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008;336:924-6. Guyatt GH, Oxman AD, Vist G, et al. GRADE guidelines: 4. Rating the quality of evidence—study limitations (risk of bias). J Clin Epidemiol 2011;64:407-15. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines 6. Rating the quality of evidence—imprecision. J Clin Epidemiol 2011;64: 1283-93.
Spencer et al 1139
16. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 7. Rating the quality of evidence—inconsistency. J Clin Epidemiol 2011;64: 1294-302. 17. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 8. Rating the quality of evidence—indirectness. J Clin Epidemiol 2011;64: 1303-10. 18. Guyatt GH, Oxman AD, Montori V, et al. GRADE guidelines: 5. Rating the quality of evidence—publication bias. J Clin Epidemiol 2011;64:1277-82. 19. Di Mario C, Mara S, Flavio A, et al. Single vs multivessel treatment during primary angioplasty: results of the multicentre randomised HEpacoat for cuLPrit or multivessel stenting for Acute Myocardial Infarction (HELP AMI) Study. Int J Cardiovasc Interv 2004;6:128-33. 20. Politi L, Sgura F, Rossi R, et al. A randomised trial of target-vessel versus multi-vessel revascularisation in ST-elevation myocardial infarction: major adverse cardiac events during long-term follow-up. Heart 2010;96:662-7. 21. Wald DS, Morris JK, Wald NJ, et al. Randomized trial of preventive angioplasty in myocardial infarction. N Engl J Med 2013;369: 1115-23.
Background Guidelines suggest percutaneous intervention (PCI) of only the culprit artery in patients presenting with ST-segment elevation myocardial infarction (STEMI) and multivessel coronary artery disease. However, recent randomized controlled trials (RCTs) suggest benefit to performing PCI of other stenotic vessels at the same time as culprit vessel PCI. Methods We conducted a systematic review with complete case meta-analysis and sensitivity analyses. Data sources included MEDLINE, EMBASE, Cochrane Register of Controlled Trials, and CINAHL from 1946 to March 2014; MEDLINE and EMBASE from March 2014 to March 2015; and scanning of literature for new studies until August 2015. All RCTs comparing multivessel versus culprit-only PCI in patients with STEMI were eligible. The primary outcomes of interest were recurrent myocardial infarction (MI), recurrent revascularization, and mortality. We combined data from trials to estimate the pooled risk ratio (RR) and associated 95% CIs using random-effects models. Results Five RCTs including 1,606 patients of whom 1,568 had complete data proved eligible. Multivessel revascularization was associated with decreased risk of repeat revascularization (RR 0.36, 95% CI 0.27-0.49, risk difference 9.7% over 2 years) and recurrent nonfatal MI (RR 0.58, 95% CI 0.36-0.93, risk difference 1.8% over 2 years), without increase in mortality (RR 0.82, 95% CI 0.53-1.26) or other adverse events. Conclusions
Pooled data provide moderate-certainty evidence that performance of multivessel PCI will provide an appreciable reduction in nonfatal MI and high-certainty evidence that it will reduce need for repeat revascularization. Patients are likely to place a high value on these benefits. (Am Heart J 2015;170:1133-9.)
Based on high-certainty evidence from randomized controlled trials (RCTs), primary percutaneous coronary intervention (PCI) of the occluded artery is the standard of care in patients presenting with ST-segment elevation myocardial infarction (STEMI). 1-3 However, multivessel coronary artery disease (CAD) (additional coronary stenoses N50%-70%) is present in 40% to 60% of patients with STEMI. 4-6 In comparison to patients with single-vessel disease, patients with multivessel disease have worse
From the
a
Department of Medicine, McMaster University, Hamilton, ON, Canada,
b
Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, Canada, cDepartment of Community Medicine, Pt. B. D. Sharma Postgraduate Institute of Medical Sciences, Rohtak, Haryana, India, and dUniversity of Sorocaba, UNISO, Sorocaba, Brazil. Submitted June 28, 2015; accepted September 1, 2015. Reprint requests: Dr Frederick A. Spencer, MD, FACC, FRCP(c), Department of Medicine, McMaster University–Faculty of Health Sciences, 50 Charlton Avenue East, Hamilton, ON L8N 4A6, Canada. E-mail: [email protected] 0002-8703 © 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ahj.2015.09.002
outcomes after primary PCI of the culprit vessel, including an increased risk of 1-year mortality. A number of studies have evaluated the use of preventive PCI in nonculprit arteries (either at the same time as primary PCI of the culprit artery or shortly thereafter) in patients with STEMI. 7 Most of these studies have been observational and have offered conflicting results. Current guidelines from the European Society of Cardiology and the American College of Cardiology/American Heart Association suggest culprit vessel only PCI in patients with STEMI and multivessel CAD.8,9 Recently, we conducted a meta-analysis of the 3 RCTs including 683 patients examining culprit vessel versus immediate multivessel PCI performed at index catheterization in patients with STEMI.10 Results suggested a decrease in nonfatal myocardial infarction (MI) and subsequent revascularization and a trend toward decreased cardiac mortality, but sample size was limited and certainty of evidence remained low. Since this analysis, 2 additional RCTs have been completed and randomized an additional 923 STEMI patients to culprit-only versus multivessel PCI.11,12 We now provide an updated systematic review including these latest data.
American Heart Journal December 2015
1134 Spencer et al
Methods Eligibility criteria We included studies that (1) enrolled patients with acute STEMI and multivessel CAD; (2) randomized patients to PCI for culprit and nonculprit coronary arteries (before hospital discharge) versus culprit-only PCI; (3) reported ≥1 of the following outcomes: nonfatal MI, revascularization (PCI or coronary artery bypass graft [CABG] surgery), cardiovascular mortality, all-cause mortality; and (4) minimum follow-up of 1 year. We excluded studies in which ≥20% of patients has cardiogenic shock, significant stenosis in a prior coronary bypass graft, or significant (≥50%) left main coronary artery stenosis in which the outcomes of these patients were not reported separately. We also excluded studies in which patients underwent nonculprit artery PCI after discharge from hospital after their acute MI and culprit artery PCI. Data sources and search strategy For our initial systematic review, we searched MEDLINE, EMBASE, the Cochrane Register of Controlled Trials, and the CINAHL from 1946 to March 2014 without any restrictions. (See Supporting Information, Appendix for the search strategy of the MEDLINE database.) For the updated review, we repeated this search in MEDLINE and EMBASE for additional studies from March 2014 to March 2015. We also scanned the literature for new studies until August 2015. For every eligible study we identified and for pertinent review articles, 1 reviewer examined the reference list for additional potential citations. Study selection Teams of 2 investigators (N.S. and L.C.L. for 1946 to March 2014; F.A.S., L.C.L., or N.S. for March 2014 to March 2015) independently screened each title and abstract. If either reviewer identified a citation as relevant, we obtained the full-text manuscript. Three reviewers working in pairs (N.S., L.C.L., and F.A.S.) determined eligibility of full-text studies. Disagreement between reviewers was resolved through discussion. Data abstraction Two reviewers abstracted the following information from each study using a custom-made data collection form: study characteristics, study population, follow-up time, loss to follow-up, intervention, and outcomes. Risk of bias and certainty of evidence To assess risk of bias, 2 reviewers (N.S. and F.A.S.) used a modified version of the Cochrane tool (http:// distillercer.com/resources/). Assessment included adequacy of sequence generation, allocation sequence concealment, blinding, incomplete outcome data, selective outcome reporting, incomplete reporting for loss to follow-up, and stopping early for benefit.
We used the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) methodology to rate certainty of evidence for each outcome as high, moderate, low, or very low. 13 Guided by GRADE methodology, we assessed overall risk of bias, 14 imprecision, 15 inconsistency, 16 indirectness, 17 and publication bias 18 and summarized results in an evidence profile. Reviewers resolved any disagreements through discussion.
Data synthesis and statistical analysis We report descriptive statistics as proportions for categorical variables and mean or medians for continuous variables. Our primary analysis for nonfatal MI, revascularization, cardiovascular mortality, and all-cause mortality included only patients with complete follow-up. We conducted a plausible worse case sensitivity analysis for nonfatal MI, revascularization, and cardiovascular mortality in which patients lost to follow-up in the complete revascularization arm were estimated to have a 5-fold increase in the event rates of those not lost to follow-up and patients lost to follow-up in the culprit-only PCI arm had only 1/5 the event rate of patients not lost to followup. We calculated pooled risk ratios (RRs) and associated 95% CIs for each outcome using random-effects models applying the Mantel-Haenszel method. Absolute effects and 95% CI were calculated by multiplying pooled RRs and 95% CI by the pooled control rate of outcomes. Statistical heterogeneity was assessed by the χ 2 test and I 2 statistics. When event rates were very low (nephropathy and stroke), we calculated risk difference (RD) and 95% CI directly by pooling across trials. Funding No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the manuscript, and its final contents. None of the authors have any conflict of interest.
Results Trial identification Our initial search from 1946 to March 2014 yielded 2,837 abstracts of which 83 were duplicates. Sixteen proved potentially eligible and underwent full review. Our search from March 2014 until March 2015 revealed an additional 51 abstracts of which none proved eligible for full review. Of these studies, 13 were excluded—5 had an ineligible population, 3 did not have nonculprit artery revascularization as a comparator, 1 was a prospective cohort study, 3 had follow-up b1 year, and 1 was an editorial letter (Appendix Supplementary Figure 1). We also became aware of a randomized controlled trial reported at the European Society of Cardiology in September 2014. This study was reviewed after it was
American Heart Journal Volume 170, Number 6
Spencer et al 1135
Table I. Study and patient characteristics First author, year published No. of randomized to complete revascularization Number randomized to culprit only PCI Exclusion criteria
Description of multivessel disease Maximum possible follow-up Lost to follow-up, n Study funding
Age (y), mean (±SD) Female (%) Diabetes (%) Prior MI (%) Anterior MI (%) 3-vessel disease (%) Drug-eluting stent (% of patients)
Di Mario, 2004
Politi, 2010
Wald, 2013
Gershlick, 2014
Engstrom, 2015
52
65
234
150
314
17
84
23
146
313
Graft or prior PCI/stent stenosis, cardiogenic shock, left main stenosis
Cardiogenic shock, prior CABG, severe valvular heart disease, unsuccessful procedures
Cardiogenic shock, prior CABG, left main stenosis, chronic total occlusion in nonculprit artery
Cardiogenic shock, stent thrombosis, indication for CABG, increased bleeding risk
≥1 nonculprit artery lesions amenable to stenting 1y
N70% stenosis in ≥1 nonculprit arteries
N50% stenosis is ≥1 nonculprit arteries
5y
4y8m
Cardiogenic shock, prior CABG, previous Q-wave MI, VSD/severe MR, chronic kidney disease, clear indication/ contraindication for multivessel PCI (as per operator) N70% stenosis (or N50% in 2 planes) in ≥1 nonculprit arteries 1y
Not reported
0
Not reported
Not reported
18 (10 complete; 8 culprit only) Private, not for profit
19 (11 complete; 8 culprit only) Not for profit, peer reviewed
Complete PCI
Culprit-only PCI
Complete PCI
Culprit-only PCI
Complete PCI
Culprit-only PCI
63.5 (12.4) 11.8 11.5 NR 51.9 30.8 0⁎
65.3 (7.4) 15.4 41.2 NR 58.8 47.1 0⁎
64.5 (11.7) 23.1 13.8 NR 47.7 29.2 7.7
66.5 (13.2) 23.8 23.8 NR 41.7 25.0 11.9
62 24 15 8 29 39 63
62 19 21 7 39 33 58
N50% stenosis and FFR ≤0.80 or ≥90% stenosis 44 m
1 (complete)
Not for profit, peer reviewed
Complete PCI
Culprit-only PCI
Complete PCI
Culprit PCI
64.6 (11.2) 14.7 12.9 4.8 36 20.7 95.9
65.3 (11.9) 23.3 14.3 3.6 35.6 24.7 90.7
64 20 9 5 33 31 95
63 19 13 9 36 32 93
Abbreviations: VSD, Ventricular septal defect; MR, mitral regurgitation; FFR, fractional flow reserve; NR, not reported. ⁎ All heparin-coated stents.
published in March 2015 and met our criteria for inclusion. An additional eligible RCT was published in August 2015 (after our initial manuscript submission). In all, we included 5 randomized trials enrolling 1,606 patients. 11,12,19-21
Trial and population characteristics Table I presents characteristics of the 5 included studies. All studies enrolled patients with STEMI with b12 hours of symptoms undergoing primary PCI of culprit artery and having at least 1 additional significant nonculprit artery stenosis. Of 1,606 randomized, investi-
gators achieved complete follow-up in 1,568 patients (complete cases). In 1 study, patients with N50% angiographic stenosis in a nonculprit vessel were randomized to consideration for complete revascularization, but multivessel PCI was only performed in those vessels with fractional flow reserve of ≤0.8. 12 In 1 of the 5 studies, PCI was performed using heparin-coated stents 19; in the others, a mix of balloon angioplasty alone, bare-metal stents, and drug-eluting stents were used. One study reported complete follow-up of all patients, 20 and 1 study failed to report loss to follow-up. 19 In the other 3 studies, loss
American Heart Journal December 2015
1136 Spencer et al
Figure 1
Pooled risk of repeat revascularization with complete revascularization versus culprit-only revascularization.
to follow-up was similar in the complete and culprit-only revascularization arms (4.3% vs 3.5%, 7.3% vs 5.5%, 0.3% vs 0%, respectively) 11,12,21 . In 3 of the 5 studies, complete revascularization was performed at the time of index catheterization. In 1 study, patients underwent complete revascularization 2 days after the procedure, 12 and in the other study, 64% of patients had complete revascularization at time of index catheterization; the remaining patients had complete revascularization before hospital discharge. 11
Assessment of risk of bias All studies met most risk of bias criteria, although none could blind patients, 2 failed to blind outcome assessors, 19,20 and 1 trial stopped early for benefit 21 (Appendix Supplementary Figure 2). We further dealt with loss to follow-up as a risk of bias issue in our sensitivity analysis. Outcomes assessment Revascularization. Complete revascularization was associated with a consistent significant reduction in the risk of repeat revascularization (RR 0.36, 95% CI 0.27-0.49, heterogeneity P = .72, I 2 = 0%) (Figure 1). Our plausible worse case analysis did not significantly change these results (RR 0.41, 95% CI 0.30-0.57), and we judged limitations as minor and, therefore, rated evidence as high certainty (Table II). Based on an estimated baseline risk of repeat revascularization over 2 years in patients undergoing only culprit PCI of 15.1%, complete revascularization will result in 97 fewer repeat revascularizations per 1,000 treated over 2 years (95% CI 110 fewer to 77 fewer) (Table II). Nonfatal MI Complete revascularization was associated with a consistent significant reduction in the risk of recurrent MI (RR 0.58, 95% CI 0.36-0.93, heterogeneity P = .43, I 2 = 0%) (Figure 2). Limitations of the evidence included risk of bias (unblinded adjudication in 2 studies, early stopping in 1 study) and few events; our judgment was that data
provided moderate-certainty evidence (Table II). Based on an estimated baseline risk of recurrent nonfatal MI over 2 years in patients undergoing only culprit PCI of 4.4%, complete revascularization is likely associated with 18 fewer non-fatal MIs per 1,000 treated per 2 years (95% CI 28 fewer to 3 fewer) (Table II). Our plausible worse case analysis did not change these results significantly (RR 0.61, 95% CI 0.39-0.97).
Cardiovascular mortality Complete revascularization was associated with a consistent significant reduction in the risk of cardiovascular mortality (RR 0.46, 95% CI 0.26-0.83, heterogeneity P = .93, I 2 = 0%) (Appendix Supplementary Figure 3). Limitations of the evidence included risk of bias (unblinded adjudication and early stopping) and few events; our judgment was that data provided moderate-certainty evidence (Table II). Based on an estimated baseline risk of cardiovascular mortality over 2 years in patients undergoing culprit PCI of 2.7%, complete revascularization is likely to be associated with 15 fewer cardiovascular deaths per 1,000 treated over 5 years (95% CI 20 fewer to 5 fewer) (Table II). Our plausible worse case analysis did not significantly change these results (RR 0.51, 95% CI 0.29-0.89). Total mortality We could not confirm or exclude a difference in total mortality between patients randomized to complete versus culprit-only PCI (RR 0.82, 95% CI 0.53-1.26, heterogeneity P = .47, I 2 = 0%) (Figure 3). Based on the CI including both benefit and harm, we rated certainty of evidence as moderate (Table II). Major bleeding We could not confirm or exclude a difference in major bleeding between patients randomized to complete versus culprit-only PCI (RR 0.73, 95% CI 0.34-1.55, heterogeneity P = .40, I 2 = 0%). We, therefore, rated confidence as low (decreased due to imprecision [small number of events] and risk of bias).
American Heart Journal Volume 170, Number 6
Spencer et al 1137
Table II. Grade assessment in confidence of estimates of effect
Outcome
No. of participants (no. of studies)
Nonfatal MI
1568 (5)
Repeat 1568 (5) revascularization Cardiovascular 1568 (5) mortality Total mortality 1568 (5)
Risk of bias Serious limitations⁎ No serious limitations Serious limitations⁎ No serious limitations
Consistency Directness No serious limitations No serious limitations No serious limitations No serious limitations
No serious limitations No serious limitations No serious limitations No serious limitations
Precision No serious limitations No serious limitations No serious limitations Serious limitations ||
Publication bias Not detected Not detected Not detected Not detected
Quality
Relative effect (95% CI)
Moderate 0.58 (0.36-0.93) High 0.36 (0.27-0.49) Moderate 0.46 (0.26-0.83) Moderate 0.82 (0.53-1.26)
Absolute effect per 1000 treated per 2 y (95% CI) 18 fewer (28 to 3 fewer)† 97 fewer (110 to 77 fewer)‡ 15 fewer (20 fewer to 5 fewer)§ Not significant
⁎ Serious risk of bias due to lack of blinding for outcome assessment in 2 of the studies and stopping early for benefit in 1 study. † Two-year baseline risk estimate of 44/1,000 derived from control arm of DANAMI-PRIMULTI study12 normalized to 2 years. ‡ Two-year baseline risk estimate of 151/1,000 derived from control arm of DANAMI-PRIMULTI study12 normalized to 2 years. § Two-year baseline risk estimate of 27/1,000 derived from control arm of DANAMI-PRIMULTI study12 normalized to 2 years. || Serious limitations as 95% CI includes possible harm and possible benefit.
Figure 2
Pooled risk of nonfatal MI with complete revascularization versus culprit-only revascularization.
Other adverse events Four studies documented contrast-induced nephropathy. 11,12,20,21 Pooling data from these 4 studies, nephropathy occurred in 10 versus 15 in the complete revascularization versus culprit vessel only revascularization arms (RD −0.01, 95% CI −0.02 to 0.01). Two studies reported stroke (8 vs 3 events, RD 0.01, 95% CI −0.00 to 0.02). 11,12,21
Discussion Our systematic review of 5 RCTs including N1,600 patients demonstrates a decrease in need for repeat revascularization and in subsequent nonfatal MI in patients with STEMI undergoing multivessel PCI compared to those undergoing culprit-only PCI. In absolute terms, our data provide high-certainty evidence that opening nonculprit stenoses at the same time as culprit lesions will provide an absolute reduction of 97 revascularizations over 2 years and moderate-certainty evidence of an absolute reduction of 18 nonfatal MIs over 2 years per 1,000 STEMI patients treated (Table II).
Although the 5 studies differed in some respects (proportion of bare-metal to drug-eluting stents, timing of complete revascularization relative to index procedure, etc), heterogeneity across outcomes in our pooled data was minimal (I 2 = 0% for all outcomes, P values for test of heterogeneity from .36 to .72). Strengths of this review include a comprehensive literature search, restriction to RCTs, duplicate assessment of eligibility, risk of bias, and data abstraction, and use of the GRADE approach to rating certainty of evidence. We also conducted a sensitivity analysis (plausible worse case) exploring potential impact of loss to follow-up, which did not appreciably change our findings. Potential limitations are related to limitations in the available evidence. The 5 RCTs that have directly addressed this question enrolled a modest total sample size, the total number of events for all outcomes but revascularization is less than 100, the RCTs had risk of bias limitations, and one of the larger studies stopped early for benefit. Nevertheless, the results warrant moderate confidence for the major morbid outcomes
American Heart Journal December 2015
1138 Spencer et al
Figure 3
Pooled risk of total mortality with complete revascularization versus culprit-only revascularization.
and high confidence for revascularization. It should also be noted that these data do not address whether multivessel PCI performed as a staged procedure within a short time frame may have similar outcomes. The ongoing COMPLETE plans to enroll approximately 3,900 patients to examine this question but will not be completed until approximately 2018. An additional question that merits further study is whether the benefits seen with multivessel PCI in select patients in these randomized trials conducted in high-volume PCI centers are also realized in patients treated in lower volume community-based hospitals. Observational studies examining PCI success and complications in large and small centers may be the best approach to address this issue. Optimization of patient selection based on clinical and angiographic criteria (eg, fractional flow reserve to identify hemodynamically significant lesions) also warrants further study.
Conclusions Pooled data from 5 RCTs provide high-certainty evidence that performance of multivessel PCI will reduce need for repeat revascularization and moderate-certainty evidence that it will provide an important reduction in nonfatal MI. Current guidelines that recommend culprit-only PCI over multivessel PCI in this setting should be reassessed. These guidelines were based on prior observational data suggesting possible harm associated with performing multivessel PCI at time of an acute STEMI and theoretical concerns of an increased risk of contrast-induced nephropathy, PCIrelated complications with multiple interventions (eg, dissection, stent thrombosis), or bleeding with complete revascularization. Our pooled data fail to substantiate these latter concerns: indeed, the absolute rates of bleeding and contrast-induced nephropathy were very low and differ little or not at all between the 2 strategies. Consistent with our results, UpToDate has recently changed their guidelines to suggest multivessel PCI over culprit-only PCI in STEMI patients with additional nonculprit lesions. Our results suggest that other guidelines addressing primary PCI in patients with STEMI also require reconsideration.
Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.ahj.2015.09.002.
References 1. Zijlstra F, Hoorntje JC, de Boer MJ, et al. Long-term benefit of primary angioplasty as compared with thrombolytic therapy for acute myocardial infarction. N Engl J Med 1999;341:1413-9. 2. A clinical trial comparing primary coronary angioplasty with tissue plasminogen activator for acute myocardial infarction. The Global Use of Strategies to Open Occluded Coronary Arteries in Acute Coronary Syndromes (GUSTO IIb) Angioplasty Substudy Investigators. N Engl J Med 1997;336:1621-8. 3. Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361:13-20. 4. Sorajja P, Gersh BJ, Cox DA, et al. Impact of multivessel disease on reperfusion success and clinical outcomes in patients undergoing primary percutaneous coronary intervention for acute myocardial infarction. Eur Heart J 2007;28:1709-16. 5. Jaski BE, Cohen JD, Trausch J, et al. Outcome of urgent percutaneous transluminal coronary angioplasty in acute myocardial infarction: comparison of single-vessel versus multivessel coronary artery disease. Am Heart J 1992;124:1427-33. 6. Claessen BE, Dangas GD, Weisz G, et al. Prognostic impact of a chronic total occlusion in a non–infarct-related artery in patients with ST-segment elevation myocardial infarction: 3-year results from the HORIZONS-AMI trial. Eur Heart J 2012;33:768-75. 7. Bainey KR, Mehta SR, Lai T, et al. Complete vs culprit-only revascularization for patients with multivessel disease undergoing primary percutaneous coronary intervention for ST-segment elevation myocardial infarction: a systematic review and meta-analysis. Am Heart J 2014;167:1-14.e2. 8. Task Force on the management of STseamiotESoC, Steg PG, James SK, et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation. Eur Heart J 2012;33:2569-619. 9. O'Gara PT, Kushner FG, Ascheim DD, et al. 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2013;127:e362-425. 10. Sekercioglu N, Spencer FA, Cruz Lopes L, et al. Culprit vessel only vs immediate complete revascularization in patients with acute
American Heart Journal Volume 170, Number 6
11.
12.
13.
14.
15.
ST-segment elevation myocardial infarction: systematic review and meta-analysis. Clin Cardiol 2014;37:765-72. Gershlick AH, Khan JN, Kelly DJ, et al. Randomized trial of complete versus lesion-only revascularization in patients undergoing primary percutaneous coronary intervention for STEMI and multivessel disease: the CvLPRIT trial. J Am Coll Cardiol 2015;65:963-72. Engstrøm T, Kelbæk H, Helqvist S, et al. Complete revascularisation versus treatment of the culprit lesion only in patients with ST-segment elevation myocardial infarction and multivessel disease (DANAMI-3—PRIMULTI): an open-label, randomised controlled trial. Lancet 2015;386:665-71. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008;336:924-6. Guyatt GH, Oxman AD, Vist G, et al. GRADE guidelines: 4. Rating the quality of evidence—study limitations (risk of bias). J Clin Epidemiol 2011;64:407-15. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines 6. Rating the quality of evidence—imprecision. J Clin Epidemiol 2011;64: 1283-93.
Spencer et al 1139
16. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 7. Rating the quality of evidence—inconsistency. J Clin Epidemiol 2011;64: 1294-302. 17. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 8. Rating the quality of evidence—indirectness. J Clin Epidemiol 2011;64: 1303-10. 18. Guyatt GH, Oxman AD, Montori V, et al. GRADE guidelines: 5. Rating the quality of evidence—publication bias. J Clin Epidemiol 2011;64:1277-82. 19. Di Mario C, Mara S, Flavio A, et al. Single vs multivessel treatment during primary angioplasty: results of the multicentre randomised HEpacoat for cuLPrit or multivessel stenting for Acute Myocardial Infarction (HELP AMI) Study. Int J Cardiovasc Interv 2004;6:128-33. 20. Politi L, Sgura F, Rossi R, et al. A randomised trial of target-vessel versus multi-vessel revascularisation in ST-elevation myocardial infarction: major adverse cardiac events during long-term follow-up. Heart 2010;96:662-7. 21. Wald DS, Morris JK, Wald NJ, et al. Randomized trial of preventive angioplasty in myocardial infarction. N Engl J Med 2013;369: 1115-23.