Recommendation for Application of Therapeutic Hypothermia to Cardiac Arrest Victims
Nicholas Rockwell, MD
Albert T. Cheung, MD
University of Pennsylvania
In the June 2003 issue of Resuscitation and the July 8, 2003 issue of Circulation, the Advanced Life Support Task Force of the International Liaison Committee on Resuscitation (ILCOR) concurrently published an advisory statement on the use of therapeutic mild hypothermia following cardiac arrest. Specifically the advisory states: "Unconscious adult patients with spontaneous circulation after out-of-hospital cardiac arrest should be cooled to 32øC to 34øC for 12 to 24 hours when the initial rhythm was ventricular fibrillation (VF)."1 This advisory came as a result of the review of the basic science, animal models, case reports, and recently published randomized controlled trials of mild therapeutic hypothermia for the prevention of post arrest encephalopathy. The two randomized controlled trials that provided the Level 1 evidence for the above recommendation were published in the February 21, 2002 issue of The New England Journal of Medicine.
The first of the two studies was conducted by The Hypothermia After Cardiac Arrest Study Group, a European multi-center initiative (9 centers in 5 countries), with data collected between March 1996 and January 2001.2 The second study was a coordinated effort among four hospitals in Melbourne, Australia with data collected between September 1996 and June 1999.3 These studies prospectively examined the effect of standard ICU care versus standard ICU care with the induction of mild hypothermia for comatose patients following cardiac arrest. Strict inclusion and exclusion criteria were defined for both studies. Enrollment in the European study required a witnessed cardiac arrest of presumed cardiogenic origin, VF or pulseless VT as initial rhythm, an interval less than 15 minutes from collapse to initial resuscitation attempts by EMS personnel, return of spontaneous circulation within 60 minutes from time of arrest, and comatose state defined as failure to respond to verbal commands prior to randomization. Exclusion criteria were age greater than 75 and less than 18, pregnancy, hypotension (MAP less than 60 mmHg) or hypoxemia (SpO2 less than 85%) following return of spontaneous circulation, or drug induced CNS depression preceding arrest. Inclusion and exclusion criteria in the Australian study were similar. In the European study only 275 of the screened 3,551 patients (7.8%) met criteria with 137 patients randomized to the hypothermia group and 138 patients randomized to the normothermia group. In the Australian study, 77 patients were enrolled with 43 randomized to the hypothermia group and 34 to the normothermia group.
The therapeutic protocols differed between the two studies. In the European study, patients in the hypothermia group received standard ICU care (including sedation, neuromuscular blockade, and mechanical ventilation) as well as deliberate hypothermia using a forced cool air bed-blanket system to a target bladder temperature of 32øC to 34øC. Cooling was begun on admission to the ICU with a goal of reaching the target temperature within 4 hours of return of spontaneous circulation and maintaining this target temperature for 24 hours. Passive rewarming to a target temperature of 36øC was permitted after 24 hours. Patients in the normothermia group received standard ICU care with temperatures maintained between 36øC to 38øC. In the Australian study, hypothermia was delivered with cold packs applied to the head and torso by EMS providers in the field. Topical cooling was continued in the emergency department and ICU to a target temperature of 33øC for 12 hours followed by passive rewarming for 6 hours, then active rewarming with forced air warming to normothermia. Patients randomized to the normothermia group were maintained at a target temperature of 37øC with otherwise similar ICU care protocols.
Both studies employed blinded assessment of outcome. In the European study, the primary endpoint was a favorable neurological outcome at six months with secondary endpoints of mortality at six months and complications within the initial seven days. Neurological outcome at six months was assessed using the Pittsburgh cerebral performance categories (1=good recovery, 2=moderate disability, 3=severe disability, 4=vegetative state, 5=death) with a score of 1 or 2 considered as a favorable outcome. A favorable outcome was observed in 55% (75/136) of the hypothermia group compared to 39% (54/137) of the normothermia group (RR=1.40, 95% CI=1.08 to 1.81; NNT= 6, 95% CI 4-25). Six-month mortality was also significantly lower in the hypothermia group (41% v. 55%, RR 0.74, 95% CI 0.58 to 0.95, NNT=7 95% CI 4-33). In the Australian study, the primary endpoint was favorable neurological recovery defined as the ability to be discharged to home or a rehabilitation facility versus death or discharge to a chronic nursing facility assessed by a blinded rehabilitation medicine specialist. In the hypothermia group, 49% (21/43) had a favorable outcome compared to 26% (9/34) in the normothermia group (P=0.046). The death rate of 51% (22/43) in the hypothermia group and 68% (23/34) in the normothermia group was not statistically different.
The European study also analyzed the rate of complications that may have been related to hypothermia such as bleeding, pneumonia, sepsis, pancreatitis, renal failure, pulmonary edema, seizures, arrhythmias, or pressure sores. Although there was a trend towards more bleeding complications, pneumonia, and sepsis in the hypothermia group, these differences did not reach statistical significance. In the Australian trial, the hypothermia group exhibited statistically significant decreases in cardiac index, increases in systemic vascular resistance, and increases in serum glucose levels.
The ILCOR recommendations on the clinical application of mild hypothermia may be considered somewhat premature because it was based on only two randomized controlled trials that involved only highly selected patients. Despite limited evidence, the ILCOR advisory included wording that "such cooling may also be beneficial for other rhythms or in-hospital cardiac arrest."1 However, ILCOR acknowledged that the clinical use of deliberate hypothermia to protect the brain in the setting of global ischemia for heart operations has been used widely and successfully since the 1950's. Whether therapeutic hypothermia is effective in other neurologic conditions or patient populations remains to be established. The National Acute Brain Injury Study: Hypothermia failed to demonstrate the efficacy of hypothermia for treatment of head trauma.4 In the editorial that accompanied the publication of the cardiac arrest trials5, Safar and Kochanek called for additional clinical trials of therapeutic mild hypothermia in stroke, traumatic brain injury, spinal cord injury, and hemorrhagic shock. In addition to defining what patient populations may benefit most from hypothermic therapy, future research needs to address also the method of cooling, duration of therapy, target temperature, monitoring techniques, complications, and the effectiveness of adjunctive neuroprotective therapies such as hypertensive reperfusion.4,6 Finally, ILCOR recognized that current methodology for the delivery of mild hypothermia have not been standardized and remain cumbersome, unreliable, and inefficient. Widespread application of therapeutic hypothermia will require the development of delivery devices that are safe, portable, inexpensive, easy to use, and are efficient yet predictable.