Therapeutic hypothermia did not improve survival or functional outcomes in unconscious patients after out-of-hospital cardiac arrest with initial nonshockable rhythm, a meta-analysis of patient-level data found.
The primary outcome of all-cause mortality at 3 months was 80.1% for patients in the hypothermia group and 82.1% for patients in the normothermia group (RR 1.04, 95% CI 0.89-1.20, P=0.63), Fabio Silvio Taccone, PhD, of Hôpital Universitaire de Bruxelles in Belgium, and colleagues reported in JAMA Neurology.
On the last day of follow-up at 3 to 6 months, 90% of those in the hypothermia group and 89.2% in the normothermia group had an unfavorable functional outcome (RR 0.99, 95% CI 0.87-1.15, P=0.97), they found.
“Our study, along with recent systematic reviews that used traditional and Bayesian meta-analyses, suggests that the current type of temperature control used to induce and maintain hypothermia (i.e., target of 33°C [about 91°F]; duration of the intervention of 24 to 40 hours; associated sedation), which has been used over the past 2 decades, does not provide the intended benefit, as shown in landmark trials,” Taccone and colleagues wrote.
“Our subgroup analyses showed no association of hypothermia with improved outcomes in some populations of patients, such as those with a noncardiac cause of arrest (i.e., mostly hypoxic/respiratory), prolonged resuscitation, absence of bystander CPR, and higher CAHP [cardiac arrest hospital prognosis] score, in whom previous studies suggested potential benefits,” they added.
Recent European guidelines recommend fever prevention for at least 72 hours in unconscious patients resuscitated after cardiac arrest, the researchers noted, but “these guidelines also highlighted the absence of evidence to support or discourage temperature control at lower body temperatures in specific patient populations.”
The study represents the largest analysis to date “of individual patient-level data from two large randomized clinical trials seeking to understand the effects of targeted temperature management following cardiac arrest presenting with nonshockable rhythm,” observed William Spears, MD, MS, and David Greer, MD, MA, of Boston University in Massachusetts, in an accompanying editorial.
“This is an important population in that nonshockable rhythm represents the most common rhythm following cardiac arrest worldwide,” the editorialists pointed out.
“Given these results, future studies should focus on the role of fever prevention alone in neurological outcomes following cardiac arrest,” they suggested. “This will be the focus of the STEPCARE study, which will analyze, among other outcomes, fever management with or without a feedback-controlled device: essentially, fever prevention versus fever reaction.”
In their meta-analysis, Taccone and colleagues analyzed two trials — TTM2 and HYPERION — using individual patient data. The analysis included only those participants who had experienced out-of-hospital-cardiac arrest and had a nonshockable rhythm with all causes of cardiac arrest.
Participants were followed for 6 months in TTM2 and 90 days in HYPERION. They had a mean age of about 65 and about two-thirds were male. In both trials, patients had been randomly assigned to hypothermia and normothermia groups at 90 hospitals in Australia, New Zealand, Europe, and the U.S. Of these, 490 people from TTM2 and 422 from HYPERION were included in the meta-analysis.
The trials used different measures for functional outcomes: the Glasgow Outcome Scale-Extended (GOSE) in TTM2 and the Cerebral Performance Category (CPC) for HYPERION. To facilitate the analysis, CPC was derived from GOSE for all participants.
The meta-analysis revealed no significant differences in the occurrence of arrhythmias, bleeding, or pneumonia between the hypothermia and normothermia groups. The trial sequential analysis showed hypothermia was associated with futility — “i.e., the inability of these clinical trials to achieve a statistically significant and/or clinically relevant difference from hypothermia,” Taccone and colleagues noted.
The study was limited by the potential influence of other interventions like sedation, paralysis, and mechanical ventilation on patient outcomes, and differences in underlying trial design, baseline characteristics, and interventions, Taccone and co-authors said.
The trials used different protocols to induce hypothermia and normothermia, and did not have a third group without temperature management. The researchers also acknowledged a lack of data about cognitive function and longer-follow up information and noted that CPC is no longer recommended in assessing functional outcomes in similar trials.
Taccone disclosed relationships with Becton Dickinson and Zoll. Co-authors had multiple relationships with industry and non-profit organizations.
Greer disclosed relationships with Becton Dickinson. Spears had no conflicts of interest.
Source Reference: Taccone FS, et al “Hypothermia vs normothermia in patients with cardiac arrest and nonshockable rhythm: a meta-analysis” JAMA Neurol 2023; DOI: 10.1001/jamaneurol.2023.4820.
Source Reference: Spears WE, Greer DM “Hypothermia to 33°C following cardiac arrest: time to close the freezer door for good?” JAMA Neurol 2023; DOI: 10.1001/jamaneurol.2023.4831.