Highlights
-
•
The present study incorporated data from 9 RCTs and 9,658 patients with time-to-first event data examining CR versus IR strategies in patients with acute MI and MVD.
-
•
CR reduced MACE by 41%, driven by an 18% reduction in CV mortality, a 31% reduction in MI, and a 38% reduction in unplanned repeat revascularizations.
-
•
The present study demonstrates, for the first time, significant reduction in all-cause mortality with CR by 36%.
-
•
Attention should now shift to the optimal approaches to safely achieving CR, including whether proceed with index-procedure CR or staged-procedure CR; whether to guide CR by angiography, physiology or intravascular imaging; and optimizing clinician training.
-
•
Future studies are warranted to determine whether patients with NSTEMI have improved survival with CR.
ABSTRACT
Complete revascularization in patients with ST-segment elevation myocardial infarction (STEMI) and multivessel disease reduces major adverse cardiac events (MACE) compared with incomplete revascularization, although whether survival is improved is uncertain. For this systematic review and meta-analysis, all randomized trials of complete vs incomplete revascularization in patients with acute MI without cardiogenic shock were identified from PubMed, Scopus, Web of Science, and Cochrane Library databases from inception to December 31, 2024. The primary and major secondary endpoints were MACE and all-cause mortality derived from reconstructed time-to-event individual-patient-data from published Kaplan-Meier curves. Additional outcomes included cardiovascular mortality, MI, and unplanned repeat revascularizations. Outcomes were expressed as hazard ratios with 95% confidence intervals. This study was registered with the PROSPERO (number, CRD42023415428). A total of 9 randomized trials with 9,658 patients (86.8% with STEMI) were identified among whom 4,671 (48.4%) patients had complete revascularization. Patients with complete revascularization had a lower 5-year risk of MACE (HR: 0.59, 95% CI: 0.54 to 0.66, P <.001) compared with incomplete revascularization. Complete revascularization was also associated with lower 5-year risks of all-cause mortality (HR: 0.64, 95% CI: 0.56 to 0.72, P <.001), cardiovascular mortality (HR: 0.82, 95% CI: 0.71 to 0.95, P =.008), MI (HR: 0.69, 95% CI: 0.55 to 0.87, P <.001), and unplanned repeat revascularizations (HR: 0.62, 95% CI: 0.54 to 0.71, P <.001). Complete revascularization results in lower risks of all-cause and cardiovascular mortality, MI, unplanned repeat revascularizations and MACE in patients with acute MI and multivessel disease. These results support current guidelines recommending CR in hemodynamically stable patients with STEMI, emphasizing that this approach may improve survival.
Introduction
Patients presenting with ST-segment elevation myocardial infarction (STEMI) have high rates of morbidity and mortality, especially the 40%-50% of patients in whom multivessel disease (MVD) is present. , Following recanalization of the culprit lesion(s) in the infarct artery by percutaneous coronary intervention (PCI), the decision to perform complete versus culprit-only revascularization in patients with STEMI and MVD may influence their prognosis and survival. While numerous randomized controlled trials (RCTs) have explored these strategies, the results have been inconsistent. Proponents of complete revascularization (CR) argue that treating all significant coronary lesions can reduce the risk of future cardiovascular events, including MI and the need for unplanned repeat revascularization procedures. ,, Conversely, supporters of culprit-only incomplete revascularization (IR) emphasize that focusing only on infarct-related lesions reduces procedural risks and costs, shortens recovery time, and avoids many unnecessary procedures. Despite prior RCTs, uncertainties remain, especially whether CR provides a survival benefit.
In this systematic review and meta-analysis, we analyzed the individual-patient-data (IPD) from published Kaplan-Meier (KM) curves of selected studies comparing CR and IR strategies in patients with acute MI to examine whether CR improves survival beyond reducing MACE.
Methods
We followed the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) statement guidelines while conducting this systematic review and meta-analysis. The methodology was carried out per the Cochrane Handbook of systematic reviews and Meta-analysis of interventions, and was registered on PROSPERO (CRD42023415428). Ethics committee approval was not required for this study in which the data from individual patients were not identifiable. This study was performed with no external funding.
Literature search. We performed a comprehensive literature search from PubMed, Scopus, Web of Science, and Cochrane Central from inception until April 2024, using the following search terms: ((myocardial infarction) OR (multivessel disease) OR (multivessel coronary disease)) AND (complete revascularization) AND (culprit-only PCI)) OR (incomplete revascularization)) AND (randomized trial) OR (Controlled trial)). The detailed search strategy for each database is reported in Supplementary Table 1. All duplicates were removed by EndNote and manual backward citation analysis was done for all the references to include all relevant citations.
Eligibility criteria. The literature search results were screened in a 2-step process. Initially, the titles and abstracts of all articles were assessed for eligibility. Subsequently, full-text screening was conducted to identify RCTs that met the eligibility criteria.
We included RCTs that included patients presenting with an acute MI (STEMI or non-STEMI [NSTEMI]) and MVD that reported outcomes in patients who underwent a CR vs an IR strategy in an intention-to treat analysis. The studies had to assess the primary outcome of interest, MACE (the composite of all-cause or cardiovascular death, MI or unplanned revascularization) and the major secondary outcome, all-cause mortality, using KM curves. Additional secondary outcomes were cardiovascular mortality, MI, and unplanned repeat revascularization. The risks of stroke, major bleeding, acute kidney injury and stent thrombosis were also assessed.
CR was defined as planned PCI of the culprit lesion(s) responsible for the MI and all other significant nonculprit lesions either immediately during the index procedure, at a later stage during the index hospitalization, or during an elective re-admission. IR was defined as planned PCI of the culprit lesion only. We further excluded studies that were not in English, were conference abstracts, had nonrandom or uncertain treatment allocation, and enrolled patients predominantly with cardiogenic shock.
Quality assessment. Three authors (A.H, M.M, R.A) independently evaluated the quality of the included clinical trials using the Cochrane Risk of Bias 2 tool (ROB-2) for RCTs, which involves assessing 5 domains: randomization process (selection bias), deviation from intended interventions (performance bias), outcome measurement (detection bias), missing outcome data (attrition bias), selection of reported results (reporting bias), and other potential sources of bias. The authors’ assessment decisions were categorized as “low risk of bias”, “high risk of bias”, or “some concerns”. Any discrepancies among the 3 authors were resolved through discussion with a fourth author.
Data extraction and analysis. Data from the included studies was extracted and recorded in a standardized data extraction sheet. The extracted data encompassed 4 main categories: (1) characteristics of the included studies; (2) characteristics of the study population; (3) risk of bias domains; and (4) outcome measures.
The principal outcomes analysis was conducted from reconstructed IPD from published KM curves, analyzed using the 2-step approach as described by Liu et al., based on “IPDfromKM” R package version 0.1. In the first step, we extracted the raw data of time and survival probability from each intervention arm in each KM curve using WebPlotDigitizer. Then, the data were analyzed in conjunction with the number at risk at each given time point, and finally IPD were reconstructed. This technique has been shown to accurately replicate original source IPD with high accuracy and reliability. ,, Two investigators independently performed the digitization of the KM curves, and discrepancies in any given values were re-evaluated by a third investigators.
The reconstructed IPD were pooled to create the dataset and pooled time-to-first event curves. The cumulative incidence of MACE and other clinical outcomes at follow-up after both interventions were assessed using KM estimates from the R packages “survival” and “survminer”. Hazard ratios (HRs) with 95% CIs for the pooled effect estimates were generated from a Cox regression model using the R package “coxphw”. To account for clustering of patients within trials and intra-study correlation, we used the coxph () function from the “survival” package in R with the “cluster (study)” option which provides robust standard errors for KM and Cox model analyses. Additionally, to account for the heterogeneity in the HRs across the trials, we specified a trial-level random slope for the treatment effect using a mixed-effects Cox model. We also assessed the restricted median survival time (RMST) for each event from the KM curves, representing the mean time-weighted difference in survival between groups during follow-up.
A standard 2-stage meta-analysis was then performed as a secondary analysis. Dichotomous data derived from the reconstructed IPD were expressed as the frequency of events and total number of patients at the latest reported follow-up duration from each study. For each categorical outcome, the pooled risk ratio (RR) with its 95% confidence interval (CI) was calculated using the Der Simonian-Laird random-effect model. Statistical heterogeneity among studies was evaluated by the Chi-square test (Cochrane Q test). I 2 test for heterogeneity was then determined according to the equation: I 2 = (Q-df/ Q) x100%. A Chi-square P value less than.05 and I-square values ≥50% were considered to represent significant heterogeneity.
For the primary MACE outcome, several sensitivity analyses were performed. We excluded 1 study at a time (“leave-one-out” analysis) to ensure that the overall effect size was not heavily influenced by any single study. Galbraith plots were also constructed to identify outlier studies in case of significant heterogeneity. To investigate publication bias, we created funnel plots that present the relationship between effect size and standard error. We used Egger’s regression test method to assess for small-study effects. Additionally, we visualized the risk of each study using L’Abbe plot. STATA 18MP and R-Studio software were used for all statistical analyses.
Results
Literature search. Our search yielded 364 citations. After duplicates were removed, 14 studies were assessed for their eligibility. Four RCTs ,,, were excluded as they did not report survival and 1 RCT was excluded because it was not a RCT of CR vs ICR. Finally, 9 RCTs ,,,,,,,, were included in the quantitative analysis. The PRISMA flow diagram is shown in Supplementary Figure 1.
Characteristics and risk of bias assessment of the included studies. Among the 9 RCTs 9,658 patients were enrolled, 4,671 patients (48.4%) of whom had CR and 4,987 (51.6%) had IR. The baseline and summary characteristics are presented in Table . Most patients (8,382 [86.8%]) presented with STEMI, whereas the remainder had NSTEMI. All studies had a low risk of bias using the ROB-2 tool, except for Politi et al., which showed some concerns from missing outcome data (Supplementary Figure 2 ) . The follow-up duration ranged from 12 months to 60 months. The median duration of follow-up was 30 months (25th percentile 17.5, 75th percentile 59).
Table
Characteristics of the included studies
| Study ID | Complete 2019 | Politi et al. | Compare-acute 2017 | Prami 2013 | Fire 2023 | Full revasc 2024 | Danami-3—Primulti 2015 | Correct II 2019 | CvLPRIT 2015 | |
|---|---|---|---|---|---|---|---|---|---|---|
| Total n randomized | 4,041 | 149 | 885 | 465 | 1,445 | 1,542 | 627 | 208 | 296 | |
| STEMI/NSTEMI/UA, n (%) | 4,041 (100)/0 (0)/0 (0) | 149 (100)/0 (0)/0 (0) | 885 (100)/0 (0)/0 (0) | 465 (100)/0 (0)/0 (0) | 509 (35.2)/936 (64.8)/0 (0) | 1,410 (91.5)/132 (8.5)/0 (0) | 627 (100)/0 (0)/0 (0) | 0 (0)/76 (36.5)/129 (62) | 295 (100)/0 (0)/0 (0) | |
| Follow-up (months) | 36 | 30 | 12 | 23 | 12 | 58 | 27 | 60 | 12 | |
| Groups | CR/IR | CR/IR | CR/IR | CR/IR | CR/IR | CR/IR | CR/IR | CR/IR | CR/IR | |
| Patients (n) with CR/IR | 2016/2025 | 65/84 | 295/590 | 234/231 | 720/725 | 764/778 | 314/313 | 105/103 | 150/146 | |
| Age, years (mean) | 61.6/62.4 | 64.5/66.5 | 62/61 | 62/62 | 81/80 | 65.0/65.7 | 64/63 | 66.4/64.4 | 64.6/65.3 | |
| Male, n (%) | 1623 (80.5)/1602 (79.1) | 50 (76.9)/64 (76.2) | 233 (79)/450 (76.3) | 177 (76)/186 (81) | 457 (63.5)/460 (63.4) | 601 (78.7)675 (/74) | 251 (80)/255 (81) | 72 (69)/78 (76) | 128 (85.3)/112 (76.7) | |
| Killip class II–IV, n (%) | 212 (10.6)/218 (10.9) | NA | 15 (5.1)/30 (5.1) | NA | 204 (28.3)/208 (28.7) | 34 (4.6)/37 (4.8) | 22 (7)/20 (6) | NA | 10 (6.8)/13 (9.4) | |
| 3-vessel disease, n (%) | 459 (23.9)/442 (22.9) | 19 (29.2)/21 (25) | 91 (30.8)/194 (32.9) | 91 (39)/76 (33) | NA | NA | 97 (31)/100 (32) | 6 (6)/6 (6) | 31 (20.7)/36 (24.7) | |
| Timing of staged procedure, days | 45 | Immediately | Immediately | Immediately | 3 | 2 | 2 | 14 | NA | |
| CV risk factors n, (%) | Hypertension | 982 (48.7)/1027 (50.7) | 32 (49.2%)/50 (59.5%) | 136 (46.1)/282 (47.8) | 94 (40)/93 (40) | 593 (82.4)/592 (81.7) | 384 (50.3)/405 (52.2) | 130 (41)/146 (47) | 52 (50)/55 (53) | 54 (36.6)/51 (36.4) |
| Diabetes | 385 (19.1)/402 (19.9) | 9 (13.8%)/20 (23.8%) | 43 (14.6)/94 (15.9) | 35 (15)/48 (21) | 230 (31.9)/233 (32.1) | 122 (16.0)/127(16.3) | 29 (9)/42 (13) | 25 (24)/17 (17) | 19 (12.9)/20 (14.3) | |
| Dyslipidemia | 764 (37.9)/797 (39.4) | NA | 95 (32.2)/176 (29.8) | NA | 384 (53.3)/375 (51.7) | 177 (23.3)/171 (22.2) | NA | 31 (30)/33 (32) | 41 (27.9)/34 (24.3) | |
| Smoking | 819 (40.6)/787 (38.9) | NA | 120 (40.8)/287 (48.7) | 118 (50)/103 (45) | 61 (8.5)/62 (8.6) | 266 (35.8)/251 (33.6) | 160 (51)/151 (48) | 25 (25)/18 (19) | 50 (34.3)/37 (26.8) | |
| Comorbidities (%) | Prior PCI | 7/7 | NA | 8.5/7.5 | NA | 16.8/18.8 | 9.3/8.1 | NA | 20/14 | 4.1/2.1 |
| Prior MI | 7.3/7.6 | NA | 7.5/8.1 | 8/7 | 14.4/16.0 | 9.5/6.8 | 5/9 | 34/34 | 4.8/3.6 | |
| Prior stroke | 3.2/3.1 | NA | 3.4/4.4 | 4/4 | 7.8/8.7 | NA | NA | NA | 4.1/2.1 | |
| Renal failure | 2/2.3 | 26.6/29.3 | 1/1.2 | NA | NA | NA | NA | NA | NA | |
| Culprit lesion n, (%) | Posterior | NA | NA | 53 (18.0)/96 (16.3) | NA | NA | NA | 10 (3)/20 (6) | NA | NA |
| Inferior | NA | NA | 149 (50.5)/307 (52.0) | 154 (66)/128 (55) | NA | NA | 195 (62)/179 (57) | NA | NA | |
| Anterior | NA | NA | 105 (35.6)/206 (34.9) | 67 (29)/89 (39) | NA | NA | 105 (33)/112 (36) | NA | NA | |
| Lateral | NA | NA | 41 (13.9)/86 (14.6) | 10 (4)/14 (6) | NA | NA | NA | NA | NA | |
| Culprit lesion n, (%) | LM | 3 (0.2)/4 (0.2) | NA | NA | NA | 35 (4.9)/41 (5.7) | 6 (0.8)/1 (0.1) | NA | NA | NA |
| LAD | 660 (34.4)/657 (33.9) | 31 (47.7)/35 (41.7) | NA | NA | 329 (45.7)/330 (45.5) | 260 (34.0)/264 (33.9) | NA | 48 (46)/44 (43) | 29 (19.3)/31 (21.2) | |
| LCx | 346 (18.0)/307 (15.8) | NA | NA | NA | 136 (18.9)/133 (18.3) | 154 (20.2)/166 (21.3) | NA | NA | 9 (6)/13 (8.9) | |
| RCA | 909 (47.4)/972 (50.1) | NA | NA | NA | 204 (28.3)/209 (28.8) | 351 (45.9)/356 (45.8) | NA | 35 (33)/30 (29) | 29 (19.3)/30 (20.5) | |
| Medical therapy n, (%) | Aspirin | 2011 (99.8)/2015 (99.5) | 62 (98.4)/74 (96.1) | NA | 233 (100)/229 (100) | 692 (96.1)/683 (94.2) | 742 (97.5)/758 (97.6) | 303 (96)/308 (96) | NA | 141 (99.3)/131 (97) |
| P2Y12 inhibitor | 2003 (99.4)/2018 (99.7) | 61 (96.8)/71 (92.2) | NA | 234 (100)/229 (100) | 713 (50)/711 (50) | 746 (98.4)/764 (98.3) | 310 (99)/309 (99) | NA | 136 (94.5)/136 (98.8) | |
| Beta-blocker | 1776 (88.1)/1804 (89.1) | 52 (82.5)/62 (80.5) | NA | 207 (88)/210 (92) | 556 (77.2)/541 (74.6) | 621 (81.6)/628 (80.8) | 290 (92)/285 (91) | 90 (86)/90 (88) | 137 (93.2)/126 (93.3) | |
| ACEi/ARB | 1723 (85.5)/1714 (84.6) | 35 (55.6)/48 (62.3) | NA | 218 (93)/209 (91) | 556 (77.2)/552 (76.1) | 607 (79.8)/610 (78.5) | 142 (45)/139 (44) | 46 (44)/43 (42) | 142 (96.6)/129 (95.6) | |
| Statin | 1980 (98.2)/1968 (97.2) | 57 (90.5)/68 (88.3) | NA | 222 (95)/223 (97) | 680 (94.4)/661 (91.2) | 742 (97.6)/752 (96.9) | 310 (99)/308 (98) | 93 (89)/96 (93) | 146 (100)/ 133 (98.5) | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
