This study was designed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standard14 (Supplementary Material, Table 1). The study was also registered in the International Prospective Register of Systematic Reviews database (CRD42023433839).

The search strategy was conducted by ALS and GC. In July 2023, PubMed, Embase, Cochrane Central Register of Controlled Trials, and Web of Science databases were systematically searched, and the reference lists of eligible studies were scanned to capture additional articles by cross-referencing. The search strategy included “Coronary occlusion” or “chronic coronary total occlusion” or “chronic total occlusion of coronary artery” or “CTO”, and “coronary artery bypass grafting” or “surgical revascularization” or “CABG”, and “medical treatment” or “medical management”. The specific search strategy is presented in Supplementary Material, Table 2. Databases were searched without time restrictions, from their inception until July 2023. Literature searches were updated during peer review to include the most up-to-date literature. The search was restricted to involve human subjects only, and no language limits were imposed.

We identified studies investigating adult patients (≥18 years old) and reporting the comparison of surgical revascularization versus MT of the CTO lesion. Additionally, studies comparing bypassed to non-bypassed CTO lesions in patients undergoing CABG were also included. Reports where the comparison received another intervention (e.g., PCI) were excluded from the research.

Observational (cohort and case-control) and interventional (randomized controlled trials) were systematically searched in databases. Reviews, dissertations, theses, editorials, study protocols, clinical guidelines, commentaries, and letters were excluded. Studies without MT as a comparison group were banned.

The primary outcome was all-cause mortality. Secondary outcomes were cardiovascular mortality, acute myocardial infarction, stroke, MACE, and complete revascularization. The definition of MACE differed across studies, although the majority included the composite endpoint of all-cause or cardiovascular mortality, stroke, or myocardial infarction. Further details are provided in Table 2.

Study selection and data extraction were performed in duplicate by two reviewers. Mendeley software was used to store, organize and manage references, and allow a transparent and reproducible systematic search. Both reviewers independently scanned study titles and abstracts against eligibility criteria. The full text of the included studies was assessed independently by two co-authors to comply with the inclusion criteria. Reviewers’ results were compared, and critical points were discussed when disagreements occurred until a consensus was reached. If necessary, other study authors were asked to resolve eligibility issues. The data extracted per study includes the author and year of publication, study design and population, baseline demographic and clinical characteristics, the intervention, and the outcomes mentioned above.

Two co-authors performed the risk of bias assessment independently, using the Newcastle–Ottawa Scale for observational studies. The quality assessment for each study is presented in the Newcastle–Ottawa Scale (Table 1). Publication bias was assessed using funnel plots (Supplementary Material, Figures 1–9).

Statistical analysis was conducted using Review Manager 5.4 from the Cochrane Collaboration to perform meta-analyses for the defined endpoints. The measure of effect was the odds ratio, comparing CABG versus MT or bypassed-CTO versus non-bypassed CTO, with 95% confidence intervals estimated using the Mantel–Haenszel method under a random effects model. Given the clinical and methodological heterogeneity among the included studies, a random effects model was selected to account for between-study variability arising from differences in populations, follow-up durations, and study designs, ensuring a more conservative and generalizable effect size estimate.

The statistical significance of the overall effect was assessed using the Z-statistic and its corresponding p-value, with a significance threshold of 5%. Heterogeneity was visually represented through forest plots and statistically assessed using Cochran's Q test (significance level=0.05) and the I2 index, categorized as low (<25%), moderate (25–50%), or high (>50%). Despite the presence of high heterogeneity in some cases, meta-analyses were performed, with results interpreted in the context of this variability. The certainty of the evidence was evaluated using the GRADE approach (Supplementary Material, Table 3).

The literature search identified 749 articles. After removing duplicates, 690 publications were screened. Based on titles, abstracts, study type, and study population assessments, 669 studies were excluded from the medical therapy versus CABG comparison (Figure 1), and 680 studies were excluded from the CTO-bypass versus no CTO-bypass analysis (Figure 2). For the medical therapy versus CABG comparison, 21 full-text articles were evaluated for eligibility (Figure 1). Of these, 11 were excluded due to the absence of a direct CTO-CABG versus CTO-MT comparison (n=7), inappropriate study design (n=2), or lack of relevant CTO population data (n=2). In the CTO-bypass versus no CTO-bypass analysis, 10 full-text articles were assessed for eligibility (Figure 2). Five articles were excluded due to the absence of a CTO-bypass versus no CTO-bypass comparison (n=3) or insufficient data on the CTO population (n=2).

Figure 1.

Literature search flow diagram of medical therapy versus CABG analysis.

(0.37MB).

Figure 2.

Literature search flow diagram of CTO-bypass versus no CTO-bypass analysis.

(0.36MB).

Following this procedure, ten studies met the eligibility criteria for the CTO-CABG versus CTO-MT comparison and five for the CTO-bypass versus no CTO-bypass analysis. One study fulfilled the criteria for both comparisons and was included in both analyses. Additionally, one study was excluded from the quantitative analysis of medical therapy versus CABG comparison due to the absence of a comparable outcome.

The characteristics of the fourteen studies that met the eligibility criteria are summarized in Table 2, with one study included in both analyses. Ten studies conducted a comparative analysis of clinical outcomes between CABG and MT of CTO lesions, enrolling 6458 patients (34.1% underwent CABG, while 65.9% received MT). Additionally, five studies compared groups undergoing CABG with and without CTO bypass, involving a total of 1949 patients (73.7% with CTO bypass compared to 26.3% without). Demographic and baseline characteristics of included patients are represented in Table 3. Follow-up time ranged between 1 and 11.2 years, with a median of 4 (IQR 1.6) years.

Some studies narrow their scope to specific CTO populations. For instance, Flores-Umanzor et al. exclusively examined diabetic patients in one study,15 and older patients (≥75 years) in another.16 Yan et al. focused on diabetic patients with a CTO of the right coronary artery (RCA).17 Ishida's study included only patients with reduced ejection fraction (FEVE <40%), and Kim et al. specifically targeted patients with two or more CTOs.18,19

First, we assessed the clinical outcomes of patients in which CTO was treated with CABG versus MT alone, focusing on all-cause mortality as primary endpoint, and cardiovascular mortality, acute MI, stroke, and MACE as secondary endpoints.

All-cause mortality rates between the above-mentioned groups were evaluated in eight studies. Of the 3815 patients included in the analysis, 1107 died during follow-up. Our study showed significantly lower all-cause mortality in the CABG group compared to the MT group, with a moderate heterogeneity in the magnitude of effect between studies (OR 0.31, 95% CI 0.24–0.40, p<0.001, I2=36%) (Figure 3).

Figure 3.

Forest plots of (A) all-cause mortality, (B) cardiovascular mortality, (C) acute myocardial infarction, (D) stroke, and (E) major adverse cardiovascular events (MACE) comparing coronary artery bypass grafting (CABG) versus medical therapy. CI: confidence interval; M–H: Mantel–Haenszel.

(1.03MB).

This divergence persisted in cardiovascular mortality, where CABG patients showed fewer cardiac deaths compared to those under medical management, although with high heterogeneity between studies (OR 0.37, 95% CI: 0.24–0.57, p<0.001, I2=59%) (Figure 3). When examining acute MI, CABG showed a lower event rate when compared with MT, with no heterogeneity identified between the included studies (OR 0.41, 95% CI: 0.30–0.56, p<0.001, I2=0%) (Figure 3). The risk of stroke was assessed in two studies, with no statistically significant difference between the MT and the CABG groups (OR 3.54, 95% CI 0.73–17.17, p=0.12, I2=71%) (Figure 3). Regarding MACE, only three studies were included in the meta-analysis. Despite the high heterogeneity identified between studies, the analysis revealed lower event rates in the surgical revascularization group compared with the MT group (OR 0.37, 95% CI 0.15–0.92, p=0.03, I2=80%) (Figure 3). To ensure the robustness of our findings, a sensitivity analysis was conducted using a fixed effects model (Supplementary Material, Table S10).

Additionally, we focused on studies involving patients with multivessel disease undergoing CABG, and compared clinical outcomes between those who received a bypass to CTO and those with a non-bypassed CTO. This assessment allowed us to evaluate the impact of complete revascularization. All-cause mortality, MACE, acute MI, and stroke were evaluated. Cardiovascular mortality was not assessed due to insufficient data in the individual studies.

In terms of all-cause mortality, our study demonstrated no statistically significant difference between the bypassed and non-bypassed CTO groups, although a trend was observed favoring CTO bypass (OR 0.64, 95% CI 0.40–1.02, p=0.06, I2=57%) (Figure 4). Conversely, there was a statistically significant difference in MACE between the two groups, demonstrating a lower event rate in the CTO-bypass group, with a moderate heterogeneity identified between the two included studies (OR 0.49, 95% CI 0.30–0.81, p=0.005, I2=44%) (Figure 4). Regarding acute MI and stroke, no differences were found between the bypass-CTO and no bypass-CTO groups (acute MI: OR 0.86, 95% CI 0.39–1.86, p=0.70, I2=47%, Figure 4; stroke OR 0.95, 95% CI 0.53–1.68, p=0.85, I2=0%, Figure 4).

Figure 4.

Forest plots of (A) all-cause mortality, (B) major adverse cardiovascular events (MACE), (C) acute myocardial infarction, and (D) stroke comparing CTO-bypass versus no CTO-bypass in patients undergoing coronary artery bypass grafting (CABG). CI: confidence interval; M–H: Mantel–Haenszel.

(0.76MB).

This meta-analysis has several limitations. First, the included studies have different study populations (including baseline and demographic clinical characteristics), entry criteria, clinical outcomes, and follow-up times. Despite a multivariate analysis with adjustments for baseline differences being performed in most studies, it may not fully compensate for the differences that influence the outcomes. A key limitation of this meta-analysis is the heterogeneity of the included studies and the broad categorization of CTOs. Factors such as vessel size, myocardial mass at risk, myocardial viability, and collateral circulation were not consistently reported, despite their significant impact on outcomes. For example, an occluded but well-collateralized proximal LAD differs substantially from a non-dominant posterior descending artery without viable myocardium.

Another important limitation of our study is the lack of consistent data on the completeness of revascularization (CR) in studies comparing MT versus CABG. Although some studies included information on CR during CABG, many lacked such data, and those that reported it demonstrated substantial variability in CR rates, potentially influencing the outcomes.

Additionally, the lack of detailed information on why occluded vessels were not bypassed in the non-bypassed CTO group introduces bias. Decisions not to bypass are often due to diffuse disease, small vessel size, or poor myocardial viability – markers of higher cardiovascular risk that likely influenced outcomes independently of revascularization strategy. These findings should therefore be interpreted cautiously, as they may reflect the inherent risk profiles of these patients.

On the other hand, our meta-analysis only included observational studies, which, as noted in the literature, are subject to inherent limitations that statistical adjustments cannot fully correct. This is consistent with findings from Simsek et al., which demonstrated better outcomes in 19 observational studies comparing CTO-PCI versus no PCI, but not in the 6 RCTs included.8 This is an example of how observational designs may overestimate the benefits of revascularization. Despite our findings providing valuable insights into real-world outcomes, they should be interpreted cautiously, as only prospective RCTs can definitively establish causality. Thus, due to the non-randomized nature of our study, selection bias cannot be completely ruled out, which limits the power and generalizability of our analysis.

Furthermore, the selection of the treatment group in the included studies was likely influenced by patients’ characteristics and Heart Team preferences, which means that the MT group might have a poorer baseline prognosis.

Lastly, given that there was no difference in all-cause mortality between CABG with and without CTO revascularization, the secondary outcomes are exploratory and should be interpreted with caution.