Cognitive function and quality of life are similar among cardiac arrest patients receiving a targeted temperature management of 33°C or 36°C, a new analysis shows.
The previously reported primary analysis of the Targeted Temperature Management (TTM) study showed that a lower temperature target did not reduce all-cause mortality, nor did it show benefit on a secondary composite endpoint of poor neurologic function or death at 180 days.
A new exploratory analysis looking at cognitive function and quality of life in the TTM trial using performance, observer-reported, and patient-reported measures still shows no difference between groups on the basis of target temperature.
The new results could be interpreted to mean that the way that hypothermia is typically administered is not effective, said lead author Tobias Cronberg, MD, PhD, associate professor, Department of Neurology and Rehabilitation Medicine, Skåne University Hospital, Lund, Sweden.
“But maybe we’re starting too late, maybe you have to cool for a longer period of time; or maybe you have to target specific categories of patients.”
The new analysis, the largest assessment to date of cognitive outcomes after cardiac arrest, was published online April 6 in JAMA Neurology.
The TTM trial compared two temperature management regimens with a target of 33°C or 36°C after an out-of-hospital cardiac arrest in unconscious patients at 36 centers in Europe and Australia. Using outcome measures, including the Cerebral Performance Category (CPC) and modified Rankin scale (mRS), the study showed no difference in the two intervention groups.
However, these measures are “crude” and not designed to pick up mild cognitive impairment, said Dr Cronberg. “We wanted to try to do something more, try to investigate more, the real outcome of the patients, so we decided to use quite a new approach.”
That approach in the new analysis used four different perspectives: clinician-reported measures (CPC and mRS); patient-reported outcomes (Two Simple Questions) and quality of life (Medical Outcomes Study 36-Item Short Form Health Survey version 2 [SF-36v2]); performance measures (Mini-Mental State Examination [MMSE]); and relative-reported cognitive outcome (Informant Questionnaire of Cognitive Decline in the Elderly [IQCODE]).
In the IQCODE, a relative or close acquaintance is typically asked to compare the patient’s current status with that of 10 years before on a scale that can range from 26 to 130, with lower scores indicating better function and a score of 78 suggesting no change. In this study, however, informants were asked to compare post–cardiac arrest function to that before the arrest.
The study included 939 unconscious adults admitted to the hospital after an out-of-hospital cardiac arrest of presumed cardiac cause. Eligible patients were randomly assigned to management with 33°C or 36°C as the target temperature of cooling. In the 33°C group, 51.8% survived to follow-up compared with 52.8% of the 36°C group.
The analysis showed that at 6 months, more than 90% of survivors returned home and about 18% needed help with daily activities. Less than half returned to their previous state of employment.
No Group Differences
The median MMSE score for survivors was 28 for both groups (P = .61). From the Two Simple Questions assessments, there was no difference in the percentage of patients with an increased need for help in activities of daily living (18.8% for 33°C and 17.5% for the 36°C; P = .71), or those who thought they had made a complete recovery (66.9% vs 61.8%; P = .32).
There was also no between-group difference (P = .77) for survivors in the mean mental component summary score of the SF-36v2 (49.1 in the 33°C group vs 49.0 in the 36°C group). The mean physical component summary scores were 46.8 and 47.5 in the 33° and 36°C groups respectively (P = .44). Both the mental and physical summary scores were similar to population norms.
Although about 50% of patients or their relatives reported that the patient had cognitive problems, “still their quality of life is good,” commented Dr Cronberg.
But the study did pick up a difference in one measurement. While in the primary analysis, including nonsurvivors, the ICODE was identical in the intervention groups, there was a statistically significant difference of 1.2 IQCODE points favoring the 33°C group when analyzing survivors (P = .04). However, Dr Cronberg was very cautious about this finding.
Sensitive to Bias
“We found that it is very, very sensitive to bias,” he said. He noted that there was a higher mortality among patients in the 33°C group, “so it could be a survivor bias affecting this outcome.”
There were also other potentially confounding issues; for example, more patients in the lower temperature target group had less than 12 years of education, so they may not have recognized decline as easily as members of the other group.
“I think one should be very, very careful about overinterpreting that result,” said Dr Cronberg.
Another study from the same research group published online February 13 in Circulation carried out more extensive tests, including those for memory (Rivermead Behavioural Memory Test), executive function (Frontal Assessment Battery), and attention/mental speed (Symbol Digit Modalities Test), and found no difference between groups cooled to the two target temperatures.
But that’s not to say that hypothermia doesn’t benefit some patients, he cautioned. “We haven’t seen it yet in our analysis, but there may be more refined ways to select patients who need hypothermia treatment, and perhaps they need to be treated for several days,” said Dr Cronberg, adding that more research is needed to find such subgroups.
Earlier and Faster?
It is possible that initiating hypothermia earlier and faster is the way to go, he speculated. However, in other studies that started hypothermia earlier there was at least a trend toward negative effects that may be due to the cooling method.
“If you give a lot of cold fluids prehospital, patients tend to go into cardiogenic shock and it could be that those kinds of effects counterbalance the effects of hypothermia,” he said. “You need to cool quite rapidly to see a strong effect.”
Ideally, hypothermia management should be initiated before a cardiac arrest. “There’s no question about it from experimental evidence that if you cool somebody before a cardiac arrest then it’s really a very forceful treatment.”
One previously published study showed trans-nasal evaporative cooling during resuscitation is safe and feasible in humans and is associated with a significant improvement in the time intervals required to cool patients.
In light of the new results indicating that outcomes for 33° and 36°C temperatures are similar, Dr. Cronberg anticipates that new guidelines from the International Liaison Committee on Resuscitation (ILCOR) later this year will recommend targeting either temperature.
However, in his own group, “we have decided to target 36° because it requires less cooling energy,” and is less invasive, he said. As well, cooling has been linked to risk for infection.
Dr Cronberg stressed that more than 90% of the survivors participated in the follow-up assessments at 6 months and more than 90% of the assessments were done in person rather than by telephone, as is often the case. Median time from the cardiac arrest to follow-up was 186 days.
“It’s amazing to see how willing the patients and their relatives have been,” said Dr. Cronberg. “I think that we have shown that such interviews are possible.”
These rates are “noteworthy” and the investigators should be “commended for their rigor and persistence” while conducting this study, said Venkatesh Aiyagari, MBBS, DM, Departments of Neurological Surgery and Neurology and Neuro-therapeutics, The University of Texas Southwestern Medical Center, Dallas, and Michael Diringer, MD, Departments of Neurology, Neurosurgery, Anesthesiology, and Occupational Therapy, Washington University School of Medicine, St Louis, Missouri, in an accompanying editorial.
Although none of the individual measures used in the study is particularly good at assessing cognitive outcome, the authors took the “unique” approach of combining the tests to highlight differences between what is reported by patients and observers, Dr Aiyagari and Dr Diringer write.
“As we move forward and focus on cognitive outcome, we must identify, refine, and validate more sensitive measures and apply them in a standardized format,” they said. “Currently, we know little about long-term cognitive outcomes and changes over time in patients after CA [cardiac arrest], and a longitudinal study of a cohort such as this would advance the field of resuscitation research.”
An important take-home message for neurologists who are often called upon to render an opinion on the prognosis of unconscious patients after cardiac arrest is that although cognitive changes are common, the overall long-term outcome of patients who survive to hospital discharge is quite good, said the editorial writers.
“Most of these patients are discharged home and report no problem with self-care, and a significant number are gainfully employed,” they note.