Acute MI (AMI) is a major cause of significant morbidity and mortality. Optimising treatment strategies for AMI, particularly ST-elevation MI (STEMI), has been a main research concern. Percutaneous coronary intervention (PCI) has proven transformative in STEMI management; however, the challenge of reperfusion injury remains a significant obstacle.1 This meta-analysis aims to assess the efficacy of hypothermia as an adjunctive therapy to PCI.
Timely reperfusion is essential for rescuing at-risk cardiomyocytes. Nevertheless, despite the availability of modern treatments, complications such as congestive heart failure and mortality continue to be serious concerns. Dixon et al. demonstrated that infarct size is a key predictor of both short- and long-term outcomes following AMI, emphasising the priority of reducing its extent.2 Research suggests that myocardial temperature plays a critical role in determining tissue viability during AMI, with even modest temperature reductions potentially leading to significant decreases in infarct size.3
Preclinical studies support the notion that moderate hypothermia, induced prior to reperfusion, may reduce infarct size. However, clinical trials such as COOL AMI, ICE-IT, CHILL MI, and VELOCITY have produced mixed results.4–7 Dallan et al. and Noc et al. reported that while randomised trials did not conclusively show substantial infarct size reduction, endovascular cooling was safe and well-tolerated.4,8 Importantly, patients cooled below 35°C prior to reperfusion showed significant infarct size reductions, highlighting the importance of both timing and degree of cooling.
Götberg et al. conducted a study that specifically examined the safety and practicality of quickly inducing hypothermia in patients with AMI using cold saline infusion and an endovascular cooling catheter prior to reperfusion.9 The objective of these trials was to evaluate not only the safety and practicality of this method but also its effectiveness in lowering the infarct size.
This meta-analysis synthesises these findings to guide future research and clinical practice, with an emphasis on optimising therapeutic hypothermia as an adjunctive treatment to PCI for STEMI management.
Methods
Search Criteria and Study Selection:
A comprehensive search strategy was conducted to identify pertinent studies published until September 2023. The search was executed across electronic databases including PubMed, Embase, and the Cochrane Library. A combination of relevant keywords and Medical Subject Headings (MeSH) terms was employed to ensure the thorough capture of studies related to “adjunctive hypothermia”, “ST-segment elevation MI (STEMI)” and “percutaneous coronary intervention (PCI)”. Boolean operators (AND, OR) were employed to optimise the effectiveness of the search. We used a full search MeSH code as follows:
(“Myocardial infarction”[MeSH terms] OR “ST Elevation Myocardial Infarction”[MeSH terms] OR “STEMI”[text word]) AND (“Hypothermia, Induced”[MeSH terms] OR “Therapeutic Hypothermia”[MeSH terms] OR “cooling”[text word]) AND (“Percutaneous Coronary Intervention”[MeSH terms] OR “Angioplasty, Balloon, Coronary”[MeSH terms] OR “Primary PCI”[text word] OR “PCI”[text word]) AND (“Infarct Size”[MeSH terms] OR “Myocardial Reperfusion Injury”[MeSH terms] OR “Ejection Fraction”[MeSH terms] OR “Myocardium at risk”[text word] OR “Microvascular Obstruction”[text word] OR “Adverse Cardiac Event”[text word] OR “reinfarction”[text word] OR “Heart Failure”[MeSH terms] OR “mortality”[MeSH terms]).
Studies were eligible for inclusion if they fulfilled the following PICO criteria: Population – studies encompassing patients diagnosed with STEMI undergoing PCI; Intervention – studies assessing the effectiveness of adjunctive hypothermia; Comparison – studies comparing adjunctive hypothermia with standard treatment or control interventions; and Outcome – Studies reporting relevant clinical outcomes such as infarct size, myocardium at risk, left ventricular ejection fraction (LVEF), microvascular obstruction, major adverse cardiac events (MACE), reinfarction, all-cause mortality and heart failure.
Studies were excluded if they did not meet the predefined PICO criteria. Specifically, studies were excluded if they: involved patient populations other than those diagnosed with STEMI undergoing PCI; did not evaluate the effectiveness of adjunctive hypothermia; lacked a comparator group involving standard treatment or control interventions; or did not report on key clinical outcomes such as infarct size, myocardium at risk, LVEF, microvascular obstruction, MACE, reinfarction, all-cause mortality or heart failure. Additionally, studies with incomplete or unclear outcome data were also excluded.
Data Extraction:
Two independent reviewers extracted data from the selected studies. A standardised data extraction form was employed, focusing on study characteristics, patient demographics, details of adjunctive hypothermia intervention and comprehensive outcome measures aligned with the PICO framework. Any discrepancies encountered during data extraction were resolved through thorough discussion and mutual consensus (Table 1 ).
Quality Assessment:
The evaluation of the quality of the obtained clinical trials was executed in accordance with the Cochrane Handbook for Systematic Reviews of Interventions 5.1.0 (revised in March 2011).10 The Cochrane risk of bias assessment tool encompasses distinct domains, namely random sequence generation (for selection bias), allocation concealment (for selection bias), blinding of participants and personnel (for performance bias), blinding of outcome assessment (for detection bias), incomplete outcome data (for attrition bias), selective outcome reporting (for reporting bias) and other potential sources of bias. The authors’ evaluative judgement assigned categorisations of ‘low risk’, ‘high risk’ or ‘unclear risk’ of bias to the identified trials.
Assessment of Heterogeneity:
Visual inspection of the forest plots served as the initial approach to quantify the presence of heterogeneity. Subsequently, an evaluation of heterogeneity was undertaken using both the I2 and χ2 tests. The χ2 test was applied to ascertain the presence of substantial heterogeneity, while the I2 test provided a quantification of the extent of heterogeneity within the effect size. Our assessment and interpretation of heterogeneity closely adhered to the guidelines outlined in the Cochrane Handbook of Systematic Reviews and Meta-analysis (chapter 9).10 In accordance with this handbook, a significance level (α) <0.1 for the χ2 test indicated the presence of significant heterogeneity, whereas the I2 test was interpreted as follows: (0–40%: potentially lacking heterogeneity; 30–60%: suggestive of moderate heterogeneity; 50–90%: indicative of substantial heterogeneity). The fixed effect model was employed due to the absence of significant heterogeneity across the included studies.
Data Synthesis
Synthesis of the collected data was performed using a fixed-effects model. Pooled estimates of effect sizes were pooled, including mean differences (MD) for continuous outcomes and ORs for dichotomous outcomes along with their corresponding 95% CIs.
Publication Bias
As per the insights of Egger et al., the evaluation of publication bias becomes less dependable when applied to fewer than 10 combined studies.11 Consequently, the application of Egger’s test for detecting asymmetry in funnel plots was not feasible within our study, preventing the assessment of potential publication bias.
Statistical Analysis
Continuous variables were pooled using the MD and 95% CIs; categorical variables were synthesised using OR and 95% CIs. Review Manager software (RevMan 5.3) was used. For pooled effects, statistical significance was inferred for values of p<0.05.
Results
The search produced an initial collection of 471 articles. After removing any duplicates and screening the titles and abstracts, 25 articles were selected for detailed examination during the full-text review stage. Ultimately, eight papers were deemed eligible based on strict and predetermined criteria and were included in the comprehensive meta-analysis (Figure 1). Figure 2 shows the results of the quality analysis conducted on the included studies. Baseline characteristics of the participants are presented in Table 2.
Efficacy Analysis Outcomes
Infarct Size
The analysis yielded a MD of −1.03 (95% CI [−3.37, 1.31]; p=0.39), indicating no statistically significant variance. The pooled studies demonstrated moderate heterogeneity (p=0.16; I2=39%; Figure 3A).
Myocardium at Risk
MD was −0.13 (95% CI [−3.19, 2.93]; p=0.94), further reinforcing the absence of significant disparity. Homogeneity persisted within the pooled studies (p=0.94; I2=0%; Figure 3B ).
Left Ventricular Ejection Fraction
An MD of −0.72 (95% CI [−2.56, 1.12]; p=0.44) emerged, reaffirming the absence of a statistically significant difference. Moderate heterogeneity characterised the pooled studies (p=0.12; I2=45%; Figure 3C ).
Microvascular Obstruction
The MD of −0.01 (95% CI [−0.32, 0.29]; p=0.95) reinforced the absence of a statistically significant difference. The pooled studies retained homogeneity (p=0.83; I2=0%; Figure 3D ).
Major Adverse Cardiac Events
The pooled OR was 1.59 (95% CI [0.63–4.01]; p=0.33), indicating the absence of a significant effect. Moderate heterogeneity was present in the pooled studies (p=0.14; I2=45%; Figure 3E ).
Reinfarction
The pooled OR was 1.43 (95% CI [0.50–4.07]; p=0.51), indicating a non-significant difference. Homogeneity was present in the pooled studies (p=0.94; I2=0%; Figure 3F ).
All-cause Mortality
The pooled OR was 0.60 (95% CI [0.24–1.53]; p=0.29), reaffirming the absence of notable differentiation. Homogeneity persisted across the pooled studies (p=0.82; I2=0%; Figure 3G ).
Heart Failure
An OR of 0.46 (95% CI [0.18–1.18]; p=0.11) was obtained, suggestive of no statistically significant difference between the two groups. The pooled studies continued to display low heterogeneity (p=0.2; I2=35%) (Figure 3I ).
Within the context of this meta-analysis, the constant pattern of non-significant differences in several aspects of clinical response is highlighted by the observed outcomes. The studies maintained a decent level of homogeneity (Figures 3A–3I ).
Safety Analysis Outcomes
We assessed bleeding and infection events as primary safety endpoints in the safety analysis. The results of these analyses are presented in Figure 4A (infection) and Figure 4B (bleeding).
Infection Events
The forest plot for infection events revealed an overall OR of 7.22 (95% CI [2.47–21.10]), indicating a significantly higher occurrence of infections in the hypothermia group compared with the control group. The heterogeneity test (I²=0%) suggests that the studies are homogenous, with consistent results across studies. The strong statistical significance (p=0.0003) and the narrow 95% CIs indicate that hypothermia is associated with an increased risk of infection compared with standard PCI (Figure 4A).
Bleeding Events
The forest plot for bleeding events revealed an OR of 2.27 (95% CI [0.76–6.78]), indicating a non-significant difference in bleeding between the hypothermia and control groups. The overall effect test (p=0.14) and the I² value of 0% suggest no significant heterogeneity or difference between the groups. This implies that hypothermia does not significantly increase the risk of bleeding compared with standard treatment (Figure 4B ).
Discussion
This meta-analysis was designed to compare the effectiveness and procedural outcomes of hypothermia as an adjuvant therapy to PCI compared with standard PCI in patients with STEMI. Eight studies involving 488 patients were included in this analysis. The findings indicate that adjunctive hypothermia did not show significant clinical benefits compared with standard PCI alone. The pooled effect estimates revealed no statistically significant differences in the incidence of all-cause mortality and MACE between hypothermia and standard PCI groups. Additionally, the use of hypothermia did not show significant differences in infarct size, left ventricular function or microvascular obstruction compared with the control group. Furthermore, the analysis showed a significantly higher risk of infection with hypothermia, while no significant difference was found in the risk of bleeding. These findings suggest that infection risk should be carefully considered in the clinical application of hypothermia as an adjunctive therapy in STEMI patients undergoing PCI. Therefore, the results of our study reveal a controversy, raising important questions about the utility and safety of adjunctive hypothermia in this patient population.
Multiple factors may contribute to the lack of observed benefits with adjunctive hypothermia. First, the timing and duration of hypothermia induction could be critical determinants of its efficacy. Previous studies have shown that hypothermia may exert different effects depending on when it is initiated relative to the onset of AMI.12,13 Additionally, the extent and duration of hypothermia might influence outcomes, with more profound and prolonged cooling potentially offering greater protection.14.15 Interestingly, some studies show a maximal reduction in the infarction size when the therapeutic hypothermia is initiated as early as possible even before the occlusion occurs; however, this would not be feasible in our patient population.14–16
In animal studies, therapeutic hypothermia has shown a promising cardioprotective effect in limiting the reperfusion injury and reducing the infarction size.17 The infarct size reduction has been estimated to be >40%.18 At the patient level, many RCTs have failed to show positive clinical outcomes with the use of hypothermia as an adjunct therapy to primary coronary intervention in patients with STEMI.2,6,7,8 However, these trials demonstrated the feasibility and safety of the adjuvant therapeutic hypothermia without significant delay in the procedure time.
A pooled analysis of the Rapid MI-ICE and Chill-MI trials revealed that therapeutic hypothermia was effective in the reduction of infarct size and post-infarction heart failure by 1–3 hours with endovascular cooling adjuvant to primary PCI in acute STEMI patients, predominantly with a larger area of myocardium at risk.1 Another pooled analysis of six RCTs that used endovascular therapeutic hypothermia during primary PCI in STEMI patients showed a significant reduction in infarct size when the temperature decreased below 35°C at the time of reperfusion.19 However, it should be noted that they included two conference abstracts, which represents a major percentage of their patient population and raises concerns about the risk of reporting bias. A meta-analysis by Mhanna et al. showed that adjuvant therapeutic hypothermia did not show a statistically significant difference in reducing infarction size, microvascular obstruction and MACE when compared with conventional PCI.20 The results of the latter meta-analysis are consistent with our study.
Limitations and Strengths
The limitations of this meta-analysis are multifaceted. The studies included in this analysis vary significantly in terms of their design and the characteristics of the patients included. This variability can impact the reliability and consistency of the results. Furthermore, the limited sample sizes in several studies restrict the statistical power and the capacity to identify significant disparities. A concern arises from the inconsistency in hypothermia induction techniques observed in the literature, which encompasses variations in the timing, length and degree. This inconsistency adds complexity to the comparison of data. Furthermore, the majority of research only examines immediate results, offering restricted insights into the effects on long-term outcomes.
Despite these limitations, the meta-analysis has several advantages. It provides a thorough analysis of a diverse array of studies with different techniques and patient groups. The emphasis on the safety and practicality of hypothermia as an additional treatment is a vital consideration for its prospective use in clinical settings. In addition, the study offers useful insights into the impact of particular factors, such as the extent of cooling and the time of induction in relation to reperfusion.
Conclusion
Ultimately, this meta-analysis offers a comprehensive perspective on the use of hypothermia as an adjunctive treatment in PCI for the management of STEMI. While the study emphasises the feasibility of the method, it also highlights safety concerns, particularly an increased risk of infection, without a significant impact on bleeding events. Moreover, the analysis does not definitively prove substantial therapeutic advantages compared with traditional PCI. The results are limited by factors such as variations in the studies, a limited number of participants and an emphasis on immediate effects. The findings underscore the need for further investigation, including studies with larger sample sizes, uniform hypothermia methods, extended follow-up periods and a focus on safety outcomes, to refine treatment approaches for STEMI patients.