Strategies to cool out-of-hospital cardiac arrest patients more or faster had little extra impact on outcomes, two trials showed.
In one, maintaining hypothermia at 33°C (91°F) didn't reduce mortality or its composite with poor neurologic outcome compared with a more modest 36°C (97°F) to largely prevent fever (P=0.51 and P=0.78), Niklas Nielsen, MD, PhD, of Sweden's Helsingborg Hospital, and colleagues found.
In the other, prehospital cooling got core temperatures down to 34°C (93°F) faster but didn't affect survival to hospital discharge or neurologic status significantly either, Francis Kim, MD, of Harborview Medical Center in Seattle, and colleagues found.
The trials were reported online in the New England Journal of Medicine and Journal of the American Medical Association, respectively.
Altogether, "the results might temper and introduce some healthy skepticism about [hypothermia] interventions," Nielsen told MedPage Today.
Clinical Message
Whereas the prehospital warming study affirmed the lack of benefit seen in a prior smaller study, prior clinical trials had suggested better outcomes with hypothermia in-hospital, leading to guidelines for maintenance at 32°C to 34°C after out-of-hospital cardiac arrest.
However, Neilsen's study wasn't a negative trial for the concept of therapeutic hypothermia, argued Robert Neumar, MD, PhD, chair of emergency medicine at the University of Michigan Health System in Ann Arbor.
Most patients actually develop fever after cardiac arrest, so keeping patients cooled slightly below the normal body temperature of 37°C (99°F) wasn't the same as no treatment at all, as done in prior trials' control groups, he said in an interview.
"We should not go back to the pre-2002 standards and not have temperature managing in our patients," Nielsen agreed in speaking to MedPage Today, highlighting the "quite good possibility of good outcome compared to other intensive care diagnoses."
And there aren't a lot of other good options to improve outcome in out-of-hospital arrest, Kim said.
"Hypothermia has been one of the first treatments in the early 2000s that's actually shown a benefit. A lot of other trials and therapies we have had in the past have just not been that successful."
Before that time point, only about a third of patients who regained a pulse survived hospitalization and overall survival was only about 8%, Jon C. Rittenberger, MD, and Clifton W. Callaway, MD, PhD, both of the University of Pittsburgh, pointed out in an editorial accompanying the NEJM editorial.
"Perhaps the most important message to take from this trial is that modern, aggressive care that includes attention to temperature works, making survival more likely than death when a patient is hospitalized after CPR," they concluded. "Future studies can continue to refine protocols, define subgroups that benefit from individual therapies, and clarify how to best adjust temperature or other interventions to each patient's illness."
However, Myron Weisfeldt, MD, chair of medicine at Johns Hopkins Hospital in Baltimore, lamented that such studies hadn't been done already, while there was still clinical equipoise to do a no-hypothermia control arm.
"We have two studies from 10 years ago that the FDA does not accept as definitive, and what we're lacking is a large-scale study with lots of data so we can really show what works and what doesn't," he said in an interview. "We don't know how much, we don't know when."
In-Hospital Hypothermia
Nielsen's international trial included 950 adults unconscious on admission after out-of-hospital arrest of presumed cardiac cause, irrespective of initial rhythm, who regained spontaneous circulation for at least 20 minutes.
They were randomized to active maintenance of a body temperature of 33°C or 36°C for 28 hours under sedation with use of ice packs, intravascular, or surface temperature management devices, or other tools at the sites' discretion and then rewarming through 36 hours to maintain normal body temperature at 37°C until 72 hours post-arrest.
Initial body temperature at presentation was 35°C in both groups.
While the healthcare team wasn't blinded to the treatment group, those performing neurologic and other outcome assessment were.
For the primary endpoint, the rate of mortality at the end of the trial was similar between groups at 50% in the 33°C group and 48% in the 36°C group (hazard ratio 1.06, 95% CI 0.89 to 1.28).
Likewise, the rate of death or poor neurologic outcome didn't differ between 33° and 36°C as evaluated with either the Cerebral Performance Category scale (54% versus 52%, P=0.78) or the modified Rankin scale (both 52%, P=0.87).
Results adjusted for known prognostic factors likewise didn't favor the lower body temperature target. No patient subgroup showed a benefit either.
The only impact that reached statistical significance was more frequent hypokalemia in the 33°C group (19% versus 13%, P=0.02).
But "you don't really know why it's not different," Weisfeldt argued. "Is it really that there's an ideal temperature that's in between the two? Or is it that at the time they gave the 33 or 36°C wasn't the right time? We're very confused."
Pre-Hospital Hypothermia
Kim's trial included 1,359 patients with out-of-hospital arrest resuscitated by paramedics in King County, Wash., randomized to standard care or to get up to 2 L of 4°C (39°F) normal saline on the way to the hospital.
Among the 583 with an initially shockable ventricular fibrillation rhythm, that strategy brought body temperature from an average 36°C at baseline down by an average 1.2°C by time of hospital arrival compared with a 0.2°C decrease in the control group.
Of those who survived to hospital admission, 77% had therapeutic hypothermia in the hospital.
Prehospital cold saline sped up the time to reach the 34°C target by about an hour (from 5 hours to 4 hours).
The prehospital intervention had a similar impact on temperature and time to target in the patients without ventricular fibrillation.
But it had no impact on survival to hospital discharge versus standard care for either patients with ventricular fibrillation (62.7% versus 64.3%, P=0.69) or those without that initial rhythm (19.2% versus 16.3%, P=0.30).
Nor was there a benefit for proportion of patients achieving full neurological recovery or mild impairment at discharge either with ventricular fibrillation (57.5% versus 61.9% of controls, P=0.69) or without ventricular fibrillation (14.4% versus 13.4% of controls, P=0.30).
And there was actually some evidence of harm.
Re-arrest during transport occurred more often with the intervention (26% versus 21%,P=0.008).
Prehospital cold saline also was associated with significantly lower oxygenation, increased pulmonary edema on first chest x-ray, and greater use of diuretics during the first 12 hours of hospitalization.
Weisfeldt chalked those risks up to the large volume of fluid administered.
Kim countered that his group's pilot study suggested no safety concern with the volume of fluid they used, but acknowledged that animal studies have suggested reduced coronary artery perfusion with IV volume loading in cardiac arrest.
He suggested that there's no future for routine use of the methods used in the trial.
Other methods of cooling are available, such as external skin cooling devices and intravenous cooling catheters, Christopher B. Granger, MD, of the Duke Clinical Research Institute in Durham, N.C., and Lance B. Becker, MD, of the University of Pennsylvania Health System in Philadelphia, wrote in an editorial accompanying the JAMA paper.
But "the use of intravenous saline for cooling after cardiac arrest is common in the United States, and this study should provide a note of caution for the use of rapid infusions for hypothermia by all clinicians who use this method," they warned.
Both studies were published in conjunction with presentation at the American Heart Association meeting in Dallas.