Early coronary angioplasty fails to lower all‐cause mortality in patients with out‐of‐hospital cardiac arrest without ST‐segment elevation: A systematic review and meta‐analysis

Abstract Introduction Out‐of‐hospital cardiac arrest (OHCA) is defined as the loss of functional mechanical activity of the heart in association with an absence of systemic circulation, occurring outside of a hospital. Immediate coronary angiography (CAG) with percutaneous coronary intervention is recommended for OHCA with ST‐elevation. We aimed to evaluate the effect of early CAG on mortality and neurological outcomes in OHCA patients without ST‐elevation. Methods This meta‐analysis and systemic review was conducted as per principles of Preferred Reporting Items for Systematic Reviews and Meta‐analysis (PRISMA) group. A protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO, Ref No. = CRD42022327833). A total of 674 studies were retrieved after scanning several databases (PubMed Central, EMBASE, Medline, and Cochrane Central Register of Controlled Trials). Results A total of 18 studies were selected for the final analysis, including 6 randomized control trials and 12 observational studies. Statistically, there was no significant difference in primary outcome, i.e., mortality, between early and delayed CAG. In terms of the grade of neurological recovery as a secondary outcome, early and delayed CAG groups also showed no statistically significant difference. Conclusion Early CAG has no survival benefits in patients with no ST elevations on ECG after OHCA.


| INTRODUCTION
Out-of-hospital cardiac arrest (OHCA) is defined as the loss of functional mechanical activity in association with an absence of systemic circulation, occurring outside of a hospital. 1While the incidence and outcome of OHCA vary by country, 2 it affects more than 500,000 patients worldwide each year. 3Despite improvements made within the last two decades, mortality and morbidity remain high, with only 1 in 10 surviving to hospital discharge. 4With such high mortality rates and poor neurological outcomes, 5 OHCA remains a major public health burden that requires immediate identification and appropriate management.
The current American and European guidelines recommend immediate coronary angiography (CAG) with percutaneous coronary intervention (PCI) in patients with OHCA who present with STsegment elevation. 6,7However, in patients with no ST elevation, the role of immediate PCI remains unclear.In the recent TOMAHAWK trial, 554 patients with OHCA without ST-elevation were randomly assigned to either immediate CAG or delayed CAG.No benefit was demonstrated over immediate angiography compared to the delayed strategy with respect to the 30-day risk of death from any cause. 8ving said that, other studies have found a benefit in early revascularization, [9][10][11] indicating mixed evidence.Moreover, a correspondence by field experts highlighted that those trials investigating this issue had stringent selection criteria, limiting the external validity of their results. 12Therefore, collecting and analyzing the data from the comparative literature, including observational studies with less strict inclusion criteria, can provide more clarity as to what the appropriate intervention should be.Therefore, in this systematic review and meta-analysis, we aimed at gathering all the evidence from the available literature to assess the benefit of early revascularization in patients with OHCA without ST elevation.Providing reliable and discrete evidence on this subject matter can help formulate guidelines to improve the outcomes of OHCA.

| Protocol registration
Our team conducted this meta-analysis and systemic review as per the principles of the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement and the checklist (Table S1).Initially, a protocol was formulated and registered with the International Prospective Register of Systematic Reviews (PROSPERO CRD42022327833) before starting this study.

| Search strategy
An extensive search of PubMed, Web of Science, EMBASE, and Cochrane Central Register of Controlled Trials was conducted by a team of two authors (FA, SK).We searched these sources from inception until June 10, 2022.These databases were systemically screened through the construction of a search strategy using the medical subject headings (Mesh).The MeSH terms of "OHCA" OR "Out of hospital cardiac arrest" OR "Out-of-hospital cardiac arrest" AND "Percutaneous Coronary Intervention" OR "PCI" OR "Percutaneous Coronary Angioplasty" OR "Percutaneous Coronary revascularization" were used.Our team also reviewed the references and citations of the published studies, meta-analyses, and systemic reviews to ensure that no relevant literature was left out of the final analysis (manual search).
This study was designed and conducted around the following research question: "Does performing early CAG improve mortality and/or neurological outcomes in patients with OHCA and no STsegment elevations on postarrest electrocardiogram (ECG)?"Therefore, we included randomized control trials (RCTs), observational studies, and cohort studies that reported adults (over 18 years of age) who suffered from cardiac arrest outside of the hospital, diagnosed to have non-ST segment elevation myocardial infarction (NSTEMI), and with reported mortality and/or neurological outcomes, namely the cerebral performance category scores (CPC), which is a 5-point standard scale that ranges from good cerebral performance (1) to brain death (5).Letters to the editors, conference abstracts, case reports and case series, pediatric studies, commentaries, and duplicates were excluded from the finalized pool of studies.

| Study selection and data extraction
First, the titles and abstracts of the articles obtained from the databases were checked against the eligibility criteria.Then, the full texts of the shortlisted abstracts were retrieved and evaluated.The data from the shortlisted studies was extracted using a standardized form.All these steps were done by two independent reviewers with complete blinding.
Then, their work was compared, and a consensus through a formal discussion was reached if differences were detected.In case of disagreements between these authors, a third author (MW) provided input.
The data recorded included the study authors, design, location, population, basic demographics, cardiovascular comorbidities, and nature of cardiac arrest were obtained in addition to data on mortality and neurological outcomes.

| Outcomes
The definition of early versus delayed CAG was variable among the included studies and ranged from 2 to 24 h of OHCA.4][15][16][17][18] One study defined early as less than 3 h. 19One study also defined early as less than 6 h 20 and four studies defined early as less than 24 h. 21,22For the purpose of this study, the main analysis was conducted using the studies that defined early versus delayed CAG with cutoff of 2 h.In this analysis, the intervention group was defined as those who underwent early CAG (within 2 h of OHCA) and the control group was defined as those who underwent delayed CAG (more than 2 h of OHCA) or no CAG.Supplementary analyses for studies were conducted for studies that defined early vs delayed CAG with cutoff other than 2 h, with a special focus to the cutoff of 24 h.The primary outcome of this study was all-cause mortality, and the secondary outcome was defined as the neurological status of these patients.A CPC score of 1-2 was considered a good neurological outcome.Thirteen studies utilized CPC scoring as the neurological recovery parameter.Crude propensity scoring was used in one study, and in four of the studies there was no available information with regards to neurological status.When CAG indicated a substantial coronary artery lesion, PCI was defined as revascularization with stent insertion.

| Statistical analysis
The statistical analysis was carried out using R programming language and the "meta" package. 19Data were entered in the format of a comparison between two categories to perform the analysis, and analysis was then carried out for dichotomous data.The categories were named in accordance with the data, which was recorded as the number of incidents in each category.As indicated in the tables, the precise calculations were reported in the odds ratio and 95% confidence interval (CI).A random effect model was employed to perform this meta-analysis as the analysis included a combination of observational studies and clinical trials.Calculating the Q values, df values, Q-df values, T, and I 2 statistics were the techniques utilized to evaluate heterogeneity and variability.

| Risk of bias assessment
The risk of bias was assessed using the Cochrane risk-of-bias tool for randomized trials (RoB 2) 23 or the risk of bias in non-randomized studies of interventions (ROBINS-I) 24 in the respective study designs using the same methodology adopted in the study selection and data extraction.Publication bias was neither assessed using Egger's regression test nor visualized using Begg's funnel plot as the number of studies in each forest plot did not exceed 10 studies. 25,26| RESULTS

| Searching databases
A total of 674 studies were retrieved after searching the various databases.Following the removal of duplicated articles (n = 129), 545 studies were further screened by title and abstract.Of these, 491 studies were omitted as they did not match the research question or were reviews, case reports, conference abstracts, letters to editors, or pediatric research.A total of 54 studies were selected, and their full texts were obtained.Finally, 18 articles were included in the final review and the analysis after being reviewed for PICO (Population, Intervention, Comparison, and Outcome) using the retrieved full-text articles (Figure 1).

| Study characteristics
Of the 18 articles, 6 articles were RCTs and 12 were observational studies.Several studies were conducted (n = 12) in multiple centers, whereas six studies were carried out at a single center (Table 1).Across all the studies, a total of 335,737 patients were included in the final analysis.The number of patients in each study varied from 33 to 325,563, with the median number of patients in a single study being 541 patients.The average age in each study varied from 57 to 71 years, with the median being 65.Males represented most of the patients in all studies.Some studies were more comprehensive than others in detailing the characteristics of the included participants (Table S1).

| Outcomes
Meta-analysis was performed using a random effect model, effect size was calculated for both primary and secondary outcomes in terms of risk ratios and log risk ratios.Heterogeneity in the analysis was assessed by calculating the τ 2 and measuring the prediction interval from the statistics, moreover, I 2 was also used as the measure of heterogeneity.

| Primary outcome
In our meta-analysis, we have found that those who underwent early CAG (≤2 h) had an increase in the 30-day mortality, compared to the those who underwent delayed CAG (>2 h) (relative risk [RR]: 1.57, 95% CI: 0.84-2.93),yet not statistically significant.This effect was mainly derived from the findings of the observational studies (RR: 2.98, 95% CI: 1.47-6.06)as randomized trials have found equivocal effect between both groups in terms of 30-day mortality (RR: 0.90, 95% CI: 0.71-1.13)(Figure 2).
On the other hand, when comparing those who underwent CAG within 24 h of OHCA to those who underwent CAG after 24 h, we observed a trivial reduction in risk of 30-day mortality (RR: 0.86, 95% CI: 0.62-1.19), yet statistically insignificant.No significant was observed between findings from observational and randomized trials (Figure 3).
In two observational studies which used different cutoffs, we have found that: (1) in the study conducted by Irene et al., they defined the cutoff as 3 h from the OHCA and have found that those who underwent CAG within 3 h had lower mortality than others (RR: 0.72, 95% CI: 0.57-0.91)(Figure S1); (2) Dankiewicz et al. 27 have defined early versus delayed CAG using a cutoff of 6 h and they observed lower 30-day mortality for patients who underwent CAG within 6 h (RR: 0.39, 95% CI: 0.32-0.48)(Figure S1).

| Secondary outcome
Compared to those who underwent CAG >2 h, we have found that undergoing CAG within 2 h of OHCA had insignificant reduction in probability of good CPC score (RR: 0.82, 95% CI: 0.64-1.06).Yet, it is to be noted that results from observational studies have  4).
Again, when using the cutoff of 24 h, we observed a significant increase in probability of good CPC score when CAG was performed within 24 h (RR: 1.68, 95% CI: 1.05-2.70).This effect was mostly because of the observational studies results (RR: 1.91, 95% CI: 1.15-3.18)(Figure 5).

| DISCUSSION
The benefit of early CAG is still debatable in patients' achieving ROSC after OHCA without ST elevation on ECG.9][30][31]   Our study has added discreet evidence to the already available literature, as we have included two recently reported RCTs on this topic in our final analysis (TOMAWAHAK and EMERGE). 8,14Our results showed that there was no significant reduction in mortality when early CAG was performed in a subgroup of patients who had achieved ROSC and had no ST elevation after OHCA.Moreover, neurological recovery was similar in patients who underwent early or late CAG.These results are well in line with the multiple RCTs, including TOMAHAWK and EMERGE. 8,14Preliminary results from other similar ongoing trials, DISCO (Direct or Subacute Coronary Angiography in Out-of-Hospital Cardiac Arrest) and ARREST (Randomized Trial of Expedited Transfer to a Cardiac Arrest Center for Non-ST Elevation OHCA), also suggest that early CAG showed no benefit in short-term survival rate when compared to delayed CAG. 16,20though our results are consistent with most studies included, there are some aspects that need to be considered in the included papers.For example, while the TOMAHAWK trial had a good randomization process, baseline characteristics were balanced, clearly defined and focused end point and was conducted in multiple centers across Europe increasing, 8 the trial had several limitations such as the unblinding nature of the investigation and the higher rates of cross-over from delayed/selective cohort to CAG within 24 h due to cardiogenic shock, large myocardial injury, new ST elevation, electrical instability, and violating study protocol.In addition, the CI for secondary outcomes was not adjusted for multiplicity leading to multiplicity bias, and as the authors rightly mentioned secondary outcomes are hypothesis-generating only, limiting external applicability of the study. 12Similarly, the French multicentric EMERGE trial 14 had some limitations mainly being underpowered due to the nonachievement of the pre-planned sample size with difficulties obtaining informed consent and overestimation of cardiac arrests meeting the inclusion and exclusion criteria.Physicians not being blinded was another limitation of the study.Furthermore, the angiograms, echocardiograms, and follow-up visits were not assessed by a core laboratory. 33is data supports the hypothesis that clinical factors other than early CAG may have an influence on survival in patients with non-ST elevated OHCA, emphasizing the importance of ICU management and careful selection of patients for interventional revascularization strategies.Researchers hypothesize that the paucity of benefits with early CAG might be attributable to the absence of ACS.As evident from a study, coronary artery vasospasm was observed in only 11.7% of on-ST elevated OHCA patients. 34Similarly, the COACT trial found that only 14.8% of the sample population had coronary lesions that met the criteria for ACS. 16,35Hence, it is essential to focus patient selection on identifying individuals who would benefit most from early interventional revascularization techniques.

| LIMITATIONS
The present study acknowledges certain limitations in the interpretation of its findings.Including observational studies, which are considered weaker evidence compared to randomized controlled trials, may restrict the generalizability and robustness of the results.
Observational studies are prone to biases and confounding, making it challenging to establish causal relationships.However, the study aimed to provide a comprehensive overview by incorporating both RCTs and observational studies.Subgroup analyses were conducted to address this limitation, examining variables and potential confounders to gain a nuanced understanding of the associations.
Random effects models were employed to account for heterogeneity across studies and enhance the reliability of the findings.Despite these adjustments, caution should be exercised when applying the results to broader populations or contexts.Further research using rigorous experimental designs and larger sample sizes is needed to strengthen the evidence base and improve generalizability.However, we sub-grouped the analyses to account for this issue and used random effects models.Second, there was variation between the definitions reported in different studies.Few studies used the term "early CAG" to refer to operations conducted within 2 h of an accident, whereas others used 3, 6, or 24 h.To overcome this issue, we performed three analyses for different definitions as follows: 2 h, 24 h, and all definitions other than 2 h.Third, various trials studied the 30-, 90-, and 180-day mortality rates that might limit the generalizability of a single-window period.Aside from that, we only evaluated the mortality and neurological status of the patients.Other factors such as procedure-associated complications, time of hospital stay, and quality of life, were not assessed due to the paucity of the data.Finally, a previous systematic review and meta-analysis addressed the same research questions as ours 36 ; however, they have not included observational studies, which may add another perspective to the problem.

| CONCLUSIONS
After analyzing the relevant literature, we conclude that performing emergency CAG fails to reduce mortality and improve neurological outcomes in patients with OHCA without ST elevations on post-ROSC ECG.Therefore, in this cohort of patients, early CAG should not be a preferred approach while evaluating and managing the cause of OHCA, and nonemergent delayed CAG should be performed to look for a cardiac cause of OHCA.
s 2015 updated recommendations state emergency CAG is reasonable for select adult patients with electrical or hemodynamic instability and who are comatose after an OHCA with no ST elevation (class IIa recommendations). 29According to the 2020 European Society of Cardiology (ESC) guidelines outlined by the task force for the management of acute coronary syndrome (ACS) in patients presenting without persistent ST-elevation, if the post-resuscitation ECG does not show ST-elevation, patients should undergo urgent angiography within 2 h after a quick evaluation to exclude noncoronary cause.The guideline also recommends considering factors related to poor neurological outcomes when considering urgent CAG.This is because severe neurologic impairment may offset any benefit from immediate coronary reperfusion.The presence of an unwitnessed cardiac arrest, initial non-shockable rhythm, prehospital F I G U R E 2 Early versus delayed CAG outcome on 30-day mortality.CAG, coronary angiography.F I G U R E 3 Early versus delayed CAG outcome on 30-day mortality.CAG, coronary angiography.F I G U R E 4 Early versus delayed CAG outcome on neurological outcomes.CAG, coronary angiography.team's late arrival without basic life support (>10 min) or more than 20 min of advanced life support without return to spontaneous circulation should all be taken strongly into account to argue against invasive coronary intervention.32

F I G U R E 5
Early versus delayed CAG outcome on 30-day mortality.CAG, coronary angiography.
Flowchart of the search and screening process.T A B L E 1 Baseline characteristics of included studies.