Each year, more than 650,000 adult patients experience a cardiac arrest in the United States (US) alone. 1 Although survival rates for both in-hospital (IHCA) and out-of-hospital cardiac arrest (OHCA) have improved in recent years, overall survival rates remain low. 2,3 Among survivors, there is a substantial risk of neurological injury and disability which is due to a combination of whole body anoxia, ischemia/reperfusion injury, inflammation, and excitotoxicity. 4 We provide a review of the key randomized controlled trials (RCT) of targeted temperature management (TTM) (Table 1).
Table 1: Randomized Control Trials Evaluating Targeted Temperature Management for Cardiac Arrest
| Authors | N | Design | Year | Population | Intervention vs. control | Primary Outcome | Major Findings |
| The Hypothermia after Cardiac Arrest Study Group | 275 | RCT | 2002 | OHCA due to a shockable rhythm | TTM 33°C vs. normothermia (37°C) | Favorable neurologic outcome at 90 days | Significantly higher survival and survival with favorable neurological outcome in hypothermia group |
| Bernard et al. | 77 | RCT | 2002 | OHCA due to ventricular fibrillation | TTM at 33°C vs. 37°C | Survival to hospital discharge with sufficiently good neurologic function to be discharged to home or to a rehabilitation facility | Significantly higher survival with favorable neurological outcome in hypothermia group |
| Bernard et al. [RICH trial] | 234 | RCT | 2010 | OHCA due to ventricular fibrillation | Cool IV fluids en-route to hospital vs. standard of care | Survival to discharge | No significant difference in survival to discharge |
| Nielsen et al. [TTM trial] | 939 | RCT | 2013 | OHCA with a presumed cardiac cause | TTM at 33°C vs. 36°C | All-cause mortality through the end of the trial (mean period of 256 days) | No significant differences between the groups |
| Kim et al. | 1359 | RCT | 2014 | OHCA with either VF or non-shockable rhythms | Pre-hospital cooling with 2 L of 4°C normal saline vs. standard care | Survival to hospital discharge and neurological status at discharge | No improvement in survival or neurological outcomes |
| Lopez et al. [FROST-I trial] | 150 | RCT | 2018 | Witnessed OHCA with shockable rhythms | TTM at 32°C, 33°C, and 34°C | Favorable neurologic outcome at 90 days | No significant difference in primary outcome between study groups |
| Lascarrou et al. [HYPERION trial] | 581 | RCT | 2019 | OHCA and IHCA with non-shockable rhythms | TTM at 33°C vs. normothermia 37.5°C | Survival with a CPC of 1 or 2 at 90 days | Significantly improved survival with a CPC of 1 or 2 in hypothermia group but no difference in overall mortality |
| Dankiewicz et al. [TTM2 trial] | 1861 | RCT | 2021 | OHCA with a presumed cardiac cause | TTM at 33°C vs. targeted normothermia 37.5°C | Death from any cause at 6 months | No significant differences between groups |
Abbreviations: CPC: Cerebral Performance Category; OHCA: Out-of-hospital cardiac arrest; TTM: Targeted Temperature Management
In 2002, two landmark RCTs were published simultaneously that found therapeutic hypothermia (TH) to be effective in reducing the risk of neurological disability in patients with OHCA due to an initial shockable rhythm who were comatose post-arrest. 5,6 These trials led to a rapid adoption of TH into clinical practice for prevention of neurological injury in post-arrest patients. TH also received a class I recommendation in resuscitation guidelines 7 which has since expanded to include patients with non-shockable rhythms and patients with IHCA. 8 However, there are key limitations of the earlier RCTs that merit discussion. First, both trials were small and included a total of 352 patients, raising the possibility of chance findings. Second, temperature was not actively managed in the control arm such that a substantial number of patients randomized to normothermia became febrile. Due to this, it is difficult to discern whether the positive findings were due to a true benefit of TH or due to the deleterious effects of fever on neurological recovery in the normothermia group. Importantly, treatment was not blinded and protocols for neuro-prognostication were not standardized between the treatment groups. Therefore, it is possible that worse survival in the normothermia group may have been driven, at least in part, by premature withdrawal of care. Finally, post-resuscitation care over the years has evolved with development of comprehensive treatment protocols across centers. Whether the large effect of TH seen in the earlier trials is consistent in contemporary practice remains unclear.
Recent trials have attempted to address the above limitations. The Targeted Temperature Management (TTM) trial randomized 950 subjects with an OHCA to two different temperature targets: 33°C versus 36°C, with active temperature management in both groups. 9 Unlike the earlier RCTs, no significant difference in mortality (50% vs. 48%; P = 0.51), or a composite of mortality and poor neurological outcome at 6 months (54% vs. 52%; P = 0.78) between the two treatment arms was observed. Other trials, although smaller than the TTM trial, corroborated the findings of the TTM trial. 10 Collectively, these trials suggest that a strategy of intense cooling may not provide any additional benefit when compared to mild TH with avoidance of fever. However, neither of these studies included a normothermia group. Moreover, non-shockable rhythms were underrepresented in these trials; in the TTM trial non-shockable rhythms comprised only 20%. This is despite >75% of all cardiac arrests are due to asystole or pulseless electrical activity.
The HYPERION trial studied the efficacy of TH (33°C) versus targeted normothermia (37°C) in 584 comatose patients admitted to an intensive care unit (ICU) after a cardiac arrest due to an initial non-shockable rhythm and included both OHCA and IHCA patients. The trial found a significant benefit in favor of TH for the primary end point of 90-day survival with a favorable neurological outcome (10.2% vs. 5.7%; 95% [CI], 0.1 to 8.9, P = 0.04), but no difference in overall mortality. 11 While the trial demonstrated a benefit of TH compared to normothermia and is the only RCT to include IHCA patients, the effect estimate had a large confidence interval such that the occurrence of one additional primary endpoint in the normothermia group would have eliminated the statistical significance (i.e., a fragility index of the trial was 1). 12
Accordingly, the recently published TTM2 trial is highly relevant as it compares TH with targeted normothermia in a large and diverse cohort of OHCA patients. 13 The trial included 1,861 patients with OHCA at 61 sites across Australia, Europe and the US who were comatose post-arrest. Patients were randomized to TH of 33°C versus targeted normothermia (goal temperature: <37.5°C). Overall, there was no difference in survival at 6 months (50% vs. 48%, P = 0.37) and no difference in survival with severe disability on the modified Rankin scale (55% vs. 55%; RR 1.00; 95% CI, 0.92-1.09). There was also no difference in health-related quality of life between the groups, but a higher incidence of arrhythmias with the use of hypothermia. These findings are in direct contrast to the original RCTs published in 2002 and the HYPERION trial, 5,6 but do align with the findings reported in the TTM trial.
There are several strengths of the TTM2 trial that merit discussion. First, it is the largest study of targeted temperature management to date and had 90% power to detect an absolute 7.5% difference in mortality. For context, the observed mortality difference in the Bernard and HACA study was 17% and 14%, respectively. Aside from the randomized treatment, both groups were treated similarly including use of mechanical ventilation, sedation, and neuromuscular blockers. Importantly, neuro-prognostication was performed using a standardized protocol and withdrawal of care before 96 hours was strongly discouraged, unless further treatment was deemed unethical or if there was evidence of brain death. Furthermore, while patients and physicians were not blinded, outcome assessors were blinded thereby reducing bias. Finally, follow up was nearly 100% and there were few protocol deviations.
A few limitations of the TTM2 trial also deserve mention. First, >90% of cardiac arrests were witnessed and approximately 80% received bystander cardiopulmonary resuscitation (CPR). Moreover, a high proportion of patients had an initial shockable rhythm, suggesting that the study findings may not be generalizable to all OHCA patients. Despite this, survival rates in this study were similar to survival reported in the earlier trials. Similarly, only OHCA patients were included and therefore findings may not apply to patients with IHCA. However, data supporting a benefit of TH in IHCA patients is limited given the limitations of the HYPERION trial noted above. 14 Second, active cooling was used in 46% of patients in the normothermia group which highlights the high prevalence of fever and underscores the importance of active temperature management in the post-arrest setting. Therefore, the TTM2 trial should not be interpreted as no benefit of any temperature management. Third, although hypothermia was initiated quickly with a median time to achieve 34°C of 3 hours after randomization, it remains to be seen whether more rapid cooling would have demonstrated a benefit of TH. However, prior RCTs have not shown pre-hospital cooling to improve survival in OHCA. 15,16 Finally, it should be noted that there were on average 136 minutes from patient contact until randomization, which increased the total time from patient contact to temperature goal to over 5 hours. This likely represents the real-world experience of cardiac arrest survivors, but future RCTs should test whether more rapid time-to-cooling will confer neuroprotection.
In summary, based on the findings of TTM2 trial, TH is not superior to targeted normothermia for neuroprotection in OHCA patients. It is possible that the time window during which TH may provide patient benefit is narrow, and unrealistic to achieve using current cooling protocols. However, until future RCTs studies such as those evaluating intra-arrest cooling and selective cerebral hypothermia are completed, we believe that routine use of TH for neuroprotection may no longer be necessary. We recommend active temperature management to maintain normothermia in post-cardiac arrest patients.
References
Keywords: Quality of Life, Out-of-Hospital Cardiac Arrest, Hypothermia, Survival Rate, Coma, Respiration, Artificial, Brain Death, Neuroprotection, Confidence Intervals, Temperature, Follow-Up Studies, Goals, Random Allocation, Cardiopulmonary Resuscitation, Hypothermia, Induced, Arrhythmias, Cardiac, Intensive Care Units, Survivors, Clinical Protocols, Neuromuscular Blocking Agents, Hospitals, Physicians, Hypoxia, Brain, Inflammation, Reperfusion Injury