Targeted Temperature Management for Cardiac Arrest

Authors: Josef Dankiewicz, Tobias Cronberg, Gisela Lilja, Janus C. Jakobsen, Helena Levin

Journal: The New England Journal of Medicine, 2021;384(24)

Type of Study: Randomized Clinical Trial (TTM2 Trial)

DOI: DOI: 10.1056/NEJMoa2100591

Quick Reference Summary

Key Findings: In a randomized trial of 1900 adults with coma after out-of-hospital cardiac arrest, targeted hypothermia at 33°C did not reduce the incidence of death at 6 months compared to targeted normothermia with early fever treatment (P=0.37).

Main Results: Death occurred in 50% of the hypothermia group versus 48% of the normothermia group (relative risk 1.04; 95% CI, 0.94–1.14). Arrhythmias causing hemodynamic compromise were more frequent in the hypothermia group (24% vs. 17%, P<0.001).

Core Clinical Question

Primary Research Question: Does targeted hypothermia at 33°C reduce 6-month mortality compared to targeted normothermia in adults comatose after out-of-hospital cardiac arrest?

Background

Disease Overview: Out-of-hospital cardiac arrest leads to high mortality and often results in hypoxic–ischemic brain injury.

Prior Data:

• Earlier trials suggested improved survival and neurologic outcomes with hypothermia at 33°C in specific patient populations (shockable rhythms).
• Recent studies indicated no dose effect when varying temperature levels (33°C vs. 36°C) or duration (24 vs. 48 hours).

Current Standard of Care: Targeted temperature management (TTM) is recommended for comatose patients post-cardiac arrest, typically involving hypothermia.

Knowledge Gaps: Low certainty in existing evidence supporting TTM; unclear benefits in broader patient populations and nonshockable rhythms.

Study Rationale: To evaluate the efficacy and safety of hypothermia versus normothermia across a diverse population, addressing uncertainties in previous research.

Methods Summary

Study Design: Open-label, randomized controlled trial with blinded outcome assessment.

Setting and Time Period: International trial conducted between November 2017 and January 2020.

Population Characteristics: 1900 adults (≥18 years) comatose after out-of-hospital cardiac arrest of presumed cardiac or unknown cause.

Inclusion Criteria: Coma (FOUR score <4), >20 minutes of spontaneous circulation post-resuscitation.

Exclusion Criteria: Interval >180 minutes from ROSC to screening, unwitnessed asystolic arrest, limitations in care.

Intervention:

• Hypothermia Group: Cooled to 33°C for 28 hours, then rewarmed to 37°C.
• Normothermia Group: Maintained ≤37.5°C with active fever management if ≥37.8°C.

Control/Comparison Group: Normothermia with early fever treatment.

Outcomes:

• Primary: Death from any cause at 6 months.
• Secondary: Functional outcome (modified Rankin scale), health-related quality of life.

Statistical Analysis Approach: Intention-to-treat, mixed-effects generalized linear models, Cox regression for survival.

Sample Size Calculations: 1900 patients to detect a 15% relative risk reduction with 90% power.

Ethics and Funding: Approved by ethics committees; funded by the Swedish Research Council and others. No commercial funding.

Detailed Results

Participant Flow and Demographics:

• Total Evaluated: 1850 patients.
• Groups: 925 hypothermia, 925 normothermia.
• Age: ~64 years; 80% male in hypothermia vs. 79% in normothermia.

Primary Outcome: Death at 6 Months

Group	Death Count	Percentage
Hypothermia	465	50%
Normothermia	446	48%

Relative Risk: 1.04 (95% CI, 0.94–1.14; P=0.37)

Secondary Outcome: Functional Outcome (Modified Rankin Scale ≥4)

Group	Count	Percentage
Hypothermia	488	55%
Normothermia	479	55%

Relative Risk: 1.00 (95% CI, 0.92–1.09)

Adverse Events:

• Arrhythmias Causing Hemodynamic Compromise: 24% hypothermia vs. 17% normothermia (P<0.001)
• Other Adverse Events: No significant differences (pneumonia, sepsis, bleeding).

Results Tables

Outcome	Hypothermia Group	Normothermia Group	Difference (95% CI)	P-value
Death at 6 Months	465/925 (50%)	446/925 (48%)	1.04 (0.94–1.14)	0.37
Modified Rankin ≥4	488/881 (55%)	479/866 (55%)	1.00 (0.92–1.09)	-
Arrhythmias Compromise	222/927 (24%)	152/921 (16%)	1.45 (1.21–1.75)	<0.001
Bleeding	44/927 (5%)	46/922 (5%)	0.95 (0.63–1.42)	0.81
Pneumonia	330/927 (36%)	322/921 (35%)	1.02 (0.90–1.15)	0.75
Sepsis	99/926 (11%)	83/922 (9%)	1.19 (0.90–1.57)	0.23
Skin Complications	10/927 (1%)	5/922 (<1%)	1.99 (0.71–6.37)	0.21

Authors' Conclusions

Primary Conclusions: Targeted hypothermia did not reduce 6-month mortality or improve functional outcomes compared to targeted normothermia in comatose patients post-cardiac arrest.

Interpretation: Hypothermia at 33°C offers no significant survival or neurological benefit over normothermia with early fever management.

Clinical Implications: Current guidelines recommending hypothermia should be re-evaluated in light of these findings.

Future Research: Investigate the benefits of temperature management strategies compared to no temperature control and explore specific patient populations that might benefit.

Critical Analysis

A. Strengths

• Large Sample Size: Included 1900 patients, enhancing statistical power and reliability.
• Randomization and Blinding: Robust randomization with blinded outcome assessment minimizes bias.
• Broad Eligibility Criteria: Increased generalizability across diverse patient populations.
• Consistent Outcome Measures: Use of standardized scales (modified Rankin, EQ-5D-5L) ensures comparability.

B. Limitations

• Open-Label Design: Awareness of group assignments by healthcare providers could introduce performance bias.
• Protocol Adherence: Similar treatment protocols might not reflect real-world variability in ICU practices.
• Exclusion of Certain Populations: Results limited to out-of-hospital cardiac arrest of presumed cardiac or unknown cause; not generalizable to all cardiac arrest scenarios.
• Fever Management in Normothermia Group: Active cooling in the normothermia group may have diluted differences between groups.
• Withdrawal of Life Support: Protocol-guided withdrawal could influence mortality outcomes.

C. Literature Context

• Previous Studies and Meta-Analyses:
  • Hypothermia after Cardiac Arrest Study Group (2002): Reported improved neurologic outcomes with hypothermia (N Engl J Med. 2002;346:549-56).
  • Bernard SA et al. (2002): Showed benefits of induced hypothermia (N Engl J Med. 2002;346:557-63).
  • Lascarrou J-B et al. (2019): Similar findings with nonshockable rhythms (N Engl J Med. 2019;381:2327-37).
• Contrasting Methodological Quality: Current trial larger and with lower risk of bias compared to earlier studies (N Engl J Med. 2021;384:2283-94 vs. N Engl J Med. 2002).
• Comparisons with Guidelines:
  • European Resuscitation Council (2015): Recommends TTM but acknowledged low certainty (Resuscitation 2015;95:202-22).
  • American Heart Association (2020): Supports temperature management post-resuscitation (Circulation 2020;142(Suppl 2): S366-S468).
• This Trial's Contribution:
  • Provides high-certainty evidence against the mortality and functional benefits of hypothermia in a broad population.
  • Confirms findings from recent studies challenging earlier hypothermia benefits (N Engl J Med. 2019;381:2327-37).

Clinical Application

Change in Practice: Routine use of targeted hypothermia at 33°C for comatose post-cardiac arrest patients should be reconsidered.

Applicable Populations: Findings are most relevant to adults with coma after out-of-hospital cardiac arrest of presumed cardiac or unknown cause.

Implementation Considerations: Shift focus to normothermia with proactive fever management to reduce arrhythmia risks associated with hypothermia.

Integration with Existing Evidence: Aligns with recent studies indicating limited benefits of hypothermia, supporting a move towards individualized temperature strategies.

How To Use This Info In Practice

Practical Recommendation: Clinicians should prioritize maintaining normothermia and managing fever in comatose patients post-cardiac arrest, as hypothermia at 33°C does not confer additional survival or functional benefits and may increase arrhythmia risk.