Sawyer KN, Humbert A, Leroux BG, et al. Relationship between duration of targeted temperature management, ischemic interval, and good functional outcome from out-of-hospital cardiac arrest. Crit Care Med 2020; 48(3):370-377.
Reviewed by Alexis Steinberg, MD
Read the article.*
*You will need to log in to US National Library of Medicine National Institutes of Health to read this article.
Targeted temperature management (TTM), or induced hypothermia, likely offers neuroprotection in patients after cardiac arrest. However, the optimal treatment duration of TTM is unclear. Guidelines recommend a minimum of 24 hours, with no clear upper limit. One possibility is that the duration of TTM may need to be tailored to individualized patients and the extent of injury, but the majority of hospitals have a protocol with a single approach.
Ischemic time (time pulseless) could guide decisions about duration of TTM, as longer ischemic time could signify worse brain injury. Hypothermia/ischemia (HI) ratio (calculated as the total TTM time [hours] divided by ischemic time [minutes]) has been a predictor of survival after out-of-hospital cardiac arrest (OHCA), making it a potential modifier of duration of TTM. Using existing data from the large North American clinical trial Resuscitation Outcome Consortium (ROC) Amiodarone, Lidocaine, and Placebo Study (ALPS), a retrospective cohort study was conducted to investigate whether prolonged hypothermia for patients with a longer ischemic time (HI ratio) was associated with improved functional outcome.
Nontraumatic EMS-treated OHCA patients presenting with any initial rhythm who were screened for eligibility for ROC-ALPS, regardless of randomization in the parent trial, were included in the study, which spanned May 2012 – October 2015. Patients included in the analysis had to survive to hospital admission with a recorded time of ROSC and have documentation of TTM. Patients were excluded if they were not intubated (awake), not treated with TTM or died within 4 hours of hospital arrival.
Local hospital protocols dictated criteria for initiating TTM, method, timing and duration, and all other critical care interventions. The outcome of interest was survival to hospital discharge with good functional status, defined as a modified Rankin Scale (mRS) of 0-3, which was obtained from chart review by trained hospital personnel. The exposure was HI ratio. Ischemia time or time to ROSC was calculated as the difference in minutes from time of call to 9-1-1 dispatch or EMS-witnessed collapse to time to ROSC. Total hypothermia treatment time was measured from initiation of cooling, through TTM time to rewarming.
Patients that received TTM were included in the study, but a comparison of patients who underwent TTM were compared to those who did not to assess for any biases. The proportion of patients with good functional outcome were stratified by terciles of hypothermia length and ischemia time. Primary analysis used logistic regression to assess the association between HI ratio and outcome, adjusting for presence of lay-rescuer cardiopulmonary resuscitation (CPR), witnessed arrest status, initial rhythm, age, gender and ROC site. Patients with missing variables were excluded from analysis. Sensitivity analyses was performed for the same model with four subsets of witnessed arrests, nonwitnessed arrest, shockable rhythm and nonshockable first rhythm. A multivariate model where HI ratio was treated as an interaction, including main effects was done as another sensitivity analysis that investigated the association of HI ratio with outcome.
A total of 37,898 were screened for the original ROC-ALPS trial, with 3,429 surviving to hospital discharge and only 2,616 included in this study based on completed data for the intended primary analysis.
Those receiving TTM had a higher proportion of patients with VF/VT as initial rhythm, public location of arrest and a lay-rescuer-witnessed arrest compared to those without TTM but a lower proportion of EMS-witnessed arrest. Other demographic characteristics including age, gender, time from 9-1-1 call to first EMS arrival and performance of lay CPR did not differ between the two groups. About 28% (28.9%) of patients with OHCA receiving TTM had a good functional discharge. Median interval ischemia time was 21 minutes, and median duration of hypothermia was 32.9 hours.
After adjusting for age, gender and lay-site, a larger HI ratio (odds ratio, 2.01; 95% CI, 1.82-2.23) was associated with good functional recovery; however, the data suggests a nonlinear relationship. Based on sensitivity analysis, there is no evidence for effect modification between hypothermia and ischemia duration on functional outcome in this patient cohort. To note, HI ratio had an OR of 1.12 (95% CI, 0.75 – 1.66), for 1/ischemia time OR was 4.82 (95% CI, 2.59 – 8.97) and for hypothermia time OR 1.0 (0.99 – 1.00).
The authors conclude that in this retrospective, cohort study for OHCA that a larger HI ratio was associated with good functional outcome, but this is primarily driven by time to ROSC. Overall, this study suggests the need for prospective investigation as to whether patients with longer duration of pulselessness, a surrogate for more severe brain injury, requires longer duration of TTM. A planned clinical trial called Influence of Cooling Duration on Efficacy in Cardiac Arrest Patients (ICECAP), NCT04217551, will use an adaptive trial design to identify the optimal duration of cooling. Enrollment should start shortly.
The authors address the most apparent limitation is that this is a retrospective secondary analysis of a larger clinical trial that was not intended or powered for the specific research question. Both the retrospective methodology and missing data can introduce a variety of biases. Recurrent arrest periods were not included as part of the definition of time to ROSC, which could underestimate the ischemic injury. Changes to temperature over time was not considered and could impact the different durations of TTM. Adjustments for individual hospitals was not accounted for and does lead to large variations. However, this improves the generalizability of results. Similarly, adjustments based on the degree of cooling (32°C to 36°C) was not done. Inherent to all observation studies of post-cardiac arrest patients, the results are influenced by survivor bias, as patients included in the study need to survive long enough to get TTM. Patients who woke up quickly had TTM stopped early. Lastly, other confounders related to inpatient treatment, such as coronary angiography and withdrawal of life-sustaining therapy, were not measured.
To note, the distribution of hypothermia duration is more variable than most TTM cohort studies. Hypothermia durations of more than 28 hours seemed to the proportion of patients with good functional outcome across terciles of ischemia duration. Most interestingly, the relationship does not seem linear, which indicate that possibly there is a threshold of TTM duration that after which there is no further benefit of cooling. Currently, there is only one small, underpowered, prospective trial that assessed 24 hours compared to 48 hours of TTM (34°C). Hopefully, the future trial (ICECAP) will give insight into clearer clinical parameters regarding TTM.