THERAPEUTIC HYPOTHERMIA AND TEMPERATURE MANAGEMENT Volume 00, Number 00, 2014 ª Mary Ann Liebert, Inc. DOI: 10.1089/ther.2014.0019
Is Cooling Still Cool? Ashwin Subramaniam, MMed, FRACP, FCICM, Ravindranath Tiruvoipati, MS, MCh, FRCSEd, FCICM and John Botha, MMed FCP(SA), FRACP, FCICM
Therapeutic hypothermia (TH), where patients are cooled to between 32C and 36C for a period of 12–24 hours and then gradually rewarmed, may reduce the risk of ischemic injury to cerebral tissue following a period of insufficient blood flow. This strategy of TH could improve mortality and neurological function in patients who have experienced out-of-hospital cardiac arrest (OOHCA). The necessity of TH in OOHCA was challenged in late 2013 by a fascinating and potentially practice changing publication, which found that targeting a temperature of 36C had similar outcomes to cooling patients to 33C. This article reviews the current literature and summarizes the uncertainties and questions raised when considering cooling of patients at risk of hypoxic brain injury. Irrespective of whether TH or targeted temperature management is deployed in patients at risk of hypoxic brain injury, it would seem that avoiding hyperpyrexia is important and that a more rigorous approach to neurological evaluation is mandated. supported these guidelines and conclusions (Arrich et al., 2012). In 2003, the recommendations for TH received further endorsement. The International Liaison Committee on Resuscitation (ILCOR) and the American Heart Association (AHA) gave a Class IIA endorsement for all shockable rhythm (ventricular fibrillation and ventricular tachycardia) arrests and a Class IIB endorsement for all nonshockable rhythm (pulseless electrical activity and asystole) arrests that were cardiac in origin (Nolan et al., 2003). TH as therapy in OOHCA with a shockable rhythm OOHCA was upgraded to a Class I endorsement in 2010 (Peberdy et al., 2010). More recently, the value of TH has received further scrutiny. A recent meta-analysis questioned if cooling a person after cardiac arrest with return of spontaneous circulation (ROSC) but without return of consciousness improves neurological outcomes (Xiao et al., 2013). The role of TH in cardiac arrest was further challenged in late 2013 by a fascinating and potentially practice changing publication, which found that a temperature of 36C results in the same outcomes as 33C (Nielsen et al., 2013). Targeted temperature management (TTM) refers to strict temperature control following OOHCA. Current evidence suggests that TTM after cardiac arrest improves neurologically intact survival, although the mechanism is uncertain. Despite this new evidence, there remain questions as to whether TTM should replace TH as therapy in patients after OOHCA. The evidence of randomized trials has revealed conflicting results. The Bernard trial (Bernard et al., 2002) was a small (n = 77) pseudorandomized (alternate days) trial without allocation concealment. The patients in the treatment arm were cooled to 33C for 12 hours when compared to standard care. There was no record of baseline neurological status before the event and no record of Glasgow Coma Scale (GCS) on
ortality rates following out-of-hospital cardiac arrest (OOHCA) are high (Deasy et al., 2011a), and elevated body temperature after cardiac arrest is associated with worse outcomes. Therapeutic hypothermia (TH), where patients are cooled to between 32C and 34C for a period of 12–24 hours and then slowly rewarmed over 12–16 hours, reduces the risk of ischemic tissue injury following a period of insufficient blood flow. Experimental studies (Berntman et al., 1981; Safar et al., 1996; Felberg et al., 2001) and previous clinical trials suggest that TH provides effective neuroprotection and increases tissue tolerance to ischemia by reducing the development of the inflammatory cascade and radical oxygen species production after reperfusion. TH is therefore often introduced as part of the medical treatment bundle in the postarrest management of unconscious patients following cardiac arrest. While associated with some complications (Xiao et al., 2013), death rates of patients in the TH group were lower in a Cochrane Review (Arrich et al., 2012). The NEJM published two studies on TH in 2002, one from Australia and the other from Europe (Bernard et al., 2002; Hypothermia after Cardiac Arrest Study Group, 2002). Even though both studies had inherent flaws, they demonstrated the positive effects of TH following cardiac arrest. The therapeutic role of TH postcardiac arrest has subsequently been extensively investigated and in 2003, Physicians at Rocky Mountain Critical Care Conference agreed that TH should be the standard of care (Nolan et al., 2003). More than 40 nonrandomized trials have reported improved outcomes with TH (Polderman, 2008). A meta-analysis concluded that the number needed to treat (NNT) was 6 with TH (Holzer et al., 2005). The Cochrane Database’s systematic review in 2012
Department of Intensive Care, Frankston Hospital, Frankston, Victoria, Australia.
arrival to the emergency department. There was no structured assessment of outcome—they considered a patient discharged home or rehabilitation facility as a good outcome. The positive outcome of the trial would have been lost if one patient in the good outcome group had a bad outcome. The study reported a 23% absolute improvement in outcome in patients with OOHCA presenting with a shockable rhythm. The trial had an absolute risk reduction (ARR) for death or severe disability of 23%, and the NNT was 4.5. The Hypothermia After Cardiac Arrest (HACA) trial (Hypothermia after Cardiac Arrest Study Group, 2002) was a larger (n = 273) multicenter randomized control trial. The patients in the treatment arm were cooled to T 33C for 24 hours, and patients in the control arm received standard care and were not cooled postarrest. The primary outcome was a favorable neurologic outcome as determined by a structured grading system within 6 months after cardiac arrest. The patients in the usual care group had no active temperature control and tended to be hyperthermic (average T 37.6C) rather than normothermic. The limitation of this study was that due to very slow recruitment and strict inclusion criteria with only 8% of screened emergency department (ED) patients included, the trial was stopped early and therefore underpowered. The study reported a 16% absolute improvement in outcome in patients with OOHCA presenting with a shockable rhythm. The ARR for unfavorable outcome and the NNT were similar to the Bernard trial (24% and 4%, respectively). The more recent publication by Nielsen et al. (2013) was an international multicenter randomized control trial that recruited 950 unconscious adults patients after OOHCA of presumed cardiac cause, where patients were randomly assigned to TTM at either 33C or 36C. The primary outcome measured was all-cause mortality at the end of the trial. The secondary outcomes included a composite of poor neurologic function or death at 180 days, as evaluated with the Cerebral Performance Category (CPC) scale and the modified Rankin scale. In this study, there were no differences in outcomes between cooling to 33C or 36C. In total, 939 patients were included in the primary analysis. The primary endpoint was death from any cause through the conclusion of the trial, specified as 180 days after the last patient’s enrollment. Survival at the end of the trial was 50% in the low-temperature group and 48% in the higher temperature group (hazard ratio 1.06, 95% confidence interval [CI] 0.89–1.28; p = 0.51). At 180 days after randomization, 54% of the patients in the 33C group had died or had poor neurologic function according to the CPC, as compared with 52% of patients in the 36C group (risk ratio, 1.02; 95% CI 0.88–1.16; p-value = 0.78). There were no differences in the duration of mechanical ventilation, in median days affecting the neurological assessment or the length of ICU or hospital stay. Despite a recent meta-analysis that found an increased incidence of pneumonia in the patients treated with TH (Geurts et al., 2014), the Nielson trial reported no statistical differences in the rate of pneumonia ( p = 0.089) between the two groups. There were no significant differences in other serious infections ( p = 0.52), severe sepsis (0.92), and bleeding ( p = 0.076). There were no significant differences in the distribution of CPCs or modified Rankin scale scores between the two groups. Furthermore, there were no differences across prespecified subgroups regarding age, sex, time from arrest to ROSC, shockable versus nonshockable initial rhythms, whether patients were in shock at admission, and the
SUBRAMANIAM ET AL.
size of the treating centers. The results of analyses adjusted for known prognostic factors were also similar. This study was more than twice the size of the original trials combined (which enrolled a total of 352 patients) and was conducted with meticulous attention to modern intensive care. Protocolized withdrawal of life-sustaining treatment was the first of its kind in such a large study. Before this trial, treatment was withdrawn based on the consensus of futility of treatment. Methods of treatment limitation or withdrawal were not specified in the HACA trial. In the Bernard trial, active life support was withdrawn from patients who remained deeply comatosed at 72 hours. Patients with uncertain prognoses were tracheotomized and discharged from intensive care. The TTM trial used a standardized protocol for neurological prognostication to guide decisions regarding treatment withdrawal. All patients in the trial were actively treated for a minimum of 72 hours after the intervention period, that is, after rewarming, when neurological evaluation was done on patients not regaining consciousness. They used a multimodal approach of clinical examination and the use of EEG and somatosensory evoked potentials. Biomarkers for brain damage were not used for operational prognostication (collected for further analyses). There were even plans for patients whose treatments had to be withdrawn before 72 hours. Patients with a Glasgow Motor Score of 1–2 at 72 hours or later who had retained N20-peak on the somatosensory evoked potential (SSEP), or patients in hospitals where SSEP was not available were reexamined daily and treatments withdrawn if GCS-Motor did not improve and metabolic and pharmacological affections were ruled out. Active treatment was withdrawn before 72 hours after the intervention period for ethical reasons in patients with previously unknown information about disseminated end-stage cancer or refractory shock with end-stage multiorgan failure. There are numerous explanations as to why the recent Nielson trial results differed from previous studies where TH was initiated. In the HACA trial, despite TH, hyperthermia (*37.6C) was noted in the control arm (Hypothermia after Cardiac Arrest Study Group, 2002). In the Nielson study, hyperthermia was not permitted and it may well be that the attention to normothermia significantly affected patient outcomes. In the Nielson publication, the study population was less selective than in the previous studies and there was no benefit to cooling to 33C in the shockable rhythm group in the predetermined subgroup analyses. Encouragingly, the mortality rate in both groups of this trial was lower than the control group of the HACA trial (Hypothermia after Cardiac Arrest Study Group, 2002), suggesting that the ICU care of such patients has improved over the last two decades. The mechanisms of benefit are controversial with many potential nonmutually exclusive possibilities. They include avoidance of hyperthermia (decreased metabolic demand and fever-related tissue injury); reduction in ischemic–reperfusion injury; and improved overall care. There are some possible advantages of TTM when compared to TH. Even though boluses of cold fluids are unlikely to cause acute pulmonary edema, positive fluid balances are associated with negative sequelae in certain patient populations. In patients treated with TTM, cumulative fluid balances are likely to be less in comparison to patients treated with large volumes of cold intravenous fluid where TH is achieved. Furthermore, the side effects of TH can be avoided; lesser sedative and/or
IS COOLING STILL COOL?
paralytic medications can be used so that patients can be neurologically evaluated with the opportunity for prognostication. Despite the potential advantages of TTM, there are aspects of the Nielson study that warrant further discussion. The study found that the average time to ROSC was 25 minutes with a wide interquartile range (18–40 minutes in the hypothermic group and 16–40 minutes in the normothermic group). In prolonged cardiac arrest, reducing the brain metabolism by hypothermia may not have a real benefit on the already damaged structures (Kuboyama et al., 1993). Even though the authors confirmed that sites consecutively screened all patients meeting the inclusion criteria and randomly assigned every patient not meeting the exclusion criteria, a recent editorial questioned the extremely low rate of recruitment of patients in the study (Varon et al., 2014). The study was conducted in 36 intensive care units in just over 2 years, with 939 of the 1431 enrolled. This meant that 18 patients were screened and 12 enrolled per center per year or 1 patient per center per month, which seems exceptionally low. Taccone and colleagues (2014) raised concerns that there was a higher percentage of bystander-initiated cardiopulmonary resuscitation (CPR) (73%) than in previous clinical trials (49–58%) (Bernard et al., 2002; Hypothermia after Cardiac Arrest Study Group, 2002) and whether such results could be widely applied to communities with a longer time to resuscitation. Nielsen and team in response clarified that there has been a continuous rise in bystander-initiated CPR during the past decade, with positive consequences on the overall outcome (Adielsson et al., 2011). The patients studied had a delay of several hours from resuscitation until the target temperature had been reached. The authors discuss the clinical relevance of this finding and cite data from the Hypothermia Network Registry, which showed no association between the time to the initiation of temperature management and the 6-month neurologic outcome (Nielsen et al., 2009). The data on when to cool have revealed conflicting results. Studies have shown that the severity of neuronal damage is dependent on the delay in initiation of cooling after reperfusion (Kuboyama et al., 1993); yet, data from the recent randomized trials showed that early initiation of temperature management does not improve outcomes (Bernard et al., 2010; Kim et al., 2014). Animal data suggest that intraarrest cooling is clearly superior to postresuscitation cooling (Abella et al., 2004). There are a few clinical trials underway to test the effect of inducing hypothermia during CPR using icecold intravenous saline (Deasy et al., 2011b) or nasopharyngeal cooling (Castren et al., 2010; Nordberg et al., 2013). In the Nielson trial, there was a rapid rate of rewarming from 33C to 36C that occurred over 6–8 hours (0.5C/hour). This process was quicker than in the previous studies and may also have impacted on patient outcomes (Polderman et al., 2009a; Polderman et al., 2009b). Even though not statistically significant, there were many favorable factors such as bystanderwitnessed arrest and appropriate resuscitation and shockable rhythms prevalent in the 36C group, when compared with unfavorable factors such as circulatory shock and absence of pupillary and corneal reflexes in the 33C group (Varon et al., 2014). There were other factors indicating that patients cooled to 33C degrees were sicker. Many patients had spontaneous hypothermia (before start of active cooling), potentially indicating greater severity of brain injury with diminished shivering response (Polderman et al., 2009a; Polderman et al.,
2009b). Despite the well-recognized antiseizure effects of hypothermia (Polderman, 2008), more patients in the 33C group had seizures. The patients in the 33C group also had worse cardiovascular impairment (cardiogenic shock and circulatory failure) (75% vs. 70% on day 2 and 67% vs. 54% on day 3) and probably met the criteria for earlier redirection of care, indicating greater severity. This probably suggests inadequate organ perfusion and potentially harmful effects at a lower target temperature (Taccone et al., 2014). The differences are small, but may be cumulative. It is possible that some readers will take away a message from the Nielson study that hypothermia therapy, and therefore cooling, is not important. It is vital to point out that the Nielson study did manage temperature in a controlled way and avoided hyperthermia. Such management of temperature is vital even if conventional TH is not practiced and close monitoring and management of temperature should not be ignored. Despite recent results suggesting that patient outcomes from OOHCA are improving, many important questions remain unanswered. These include the optimum target temperature, how soon cooling should be commenced, the duration of TTM, the importance of avoiding hyperthermia, and the rate of rewarming in those patients who are cooled. The management of in hospital arrest has also been less well studied and it is unclear whether temperature targets should be individualized. With more recent recommendation that extracorporeal membrane oxygenation (ECMO) should be considered during CPR, the relevance of temperature control in the group of patients could be further debated. Even though Bernard, the author of the original 2002 NEJM article, has recommended that targeting 36C is appropriate, and has changed guidelines in his institution (Bernard, 2014). The Australian Resuscitation Council (ARC), in their statement in December 2013 ( Jacobs, 2013), clearly identified the reasons why they did not want to endorse the change. They have reasoned that the Nielson study did not show a difference in mortality between the two target temperature groups. Both groups were undergoing active target temperature management, which prevented fever, and were part of intensive postresuscitation care. The mean temperature of patients in both groups at the time of recruitment was 35C. No differences in complications between the two groups were observed. The study included both shockable and nonshockable arrest rhythms. For the above reasons, the ARC still recommended 32–34C. However, they also state that anyone choosing target temperature of 36C should take precautions to avoid fever. Conclusion
Irrespective of whether TH or TTM is deployed in patients at risk of hypoxic brain injury, it would seem that avoiding hyperpyrexia is important and that a more rigorous approach to neurological evaluation is mandated. Despite these recent data suggesting a role for TTM, there remain many unanswered questions, including patient selection, initiation, and duration of a cooling strategy. Clinicians caring for adults following cardiac arrest should pay more not less attention to temperature management. The need for robust guidelines and rigorous adherence to departmental standards of care are vital during periods of practice change. We strongly believe that protocolized withdrawal of life-sustaining treatment should be adopted in
managing these patient cohorts. As has been observed with the evolution of tight glucose control (van den Berghe et al., 2001; NICE-SUGAR Study Investigators et al., 2009), steroid administration (Sprung et al., 2008), and goal-directed therapy in the critically ill (Rivers et al., 2001; Pro et al., 2014), a more circumspect approach to cooling may be prudent at present. Disclosure Statement
We declare that we have no proprietary, financial, professional, or other personal interest of any nature or kind in any product, service, and/or company that could be construed as influencing the position presented in the submitted manuscript. References
Abella BS, et al. Intra-arrest cooling improves outcomes in a murine cardiac arrest model. Circulation 2004;109:2786–2791. Adielsson A, et al. Increase in survival and bystander CPR in out-of-hospital shockable arrhythmia: bystander CPR and female gender are predictors of improved outcome. Experiences from Sweden in an 18-year perspective. Heart 2011;97: 1391–1396. Arrich J, et al. Hypothermia for neuroprotection in adults after cardiopulmonary resuscitation. Cochrane Database Syst Rev 2012;9:CD004128. Bernard S. Inducing hypothermia after out of hospital cardiac arrest. BMJ 2014;348:g2735. Bernard SA, et al. Induction of therapeutic hypothermia by paramedics after resuscitation from out-of-hospital ventricular fibrillation cardiac arrest: a randomized controlled trial. Circulation 2010;122:737–742. Bernard SA, et al. Treatment of comatose survivors of out-ofhospital cardiac arrest with induced hypothermia. N Engl J Med 2002;346:557–563. Berntman L, Welsh FA, Harp JR. Cerebral protective effect of low-grade hypothermia. Anesthesiology 1981;55:495–498. Castren M, et al. Intra-arrest transnasal evaporative cooling: a randomized, prehospital, multicenter study (PRINCE: PreROSC IntraNasal Cooling Effectiveness). Circulation 2010; 122:729–736. Deasy C, et al. Cardiac arrest outcomes before and after the 2005 resuscitation guidelines implementation: evidence of improvement? Resuscitation 2011a;82:984–988. Deasy C, et al. Design of the RINSE trial: the rapid infusion of cold normal saline by paramedics during CPR. BMC Emerg Med 2011b;11:17. Felberg RA, et al. Hypothermia after cardiac arrest: feasibility and safety of an external cooling protocol. Circulation 2001;104:1799–1804. Geurts M, et al. Therapeutic hypothermia and the risk of infection: a systematic review and meta-analysis. Crit Care Med 2014;42:231–242. Holzer M, et al. Hypothermia for neuroprotection after cardiac arrest: systematic review and individual patient data metaanalysis. Crit Care Med 2005;33:414–418. Hypothermia after Cardiac Arrest Study Group. Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. N Engl J Med 2002;346:549–556. Jacobs I. Therapeutic Hypothermia in Cardiac Arrest: An Information Update. 2013. www.resus.org.au/files/arc_therapeutic_ hypothermia.pdf Kim F, et al. Effect of prehospital induction of mild hypothermia on survival and neurological status among adults with cardiac arrest: a randomized clinical trial. JAMA 2014;311:45–52.
SUBRAMANIAM ET AL.
Kuboyama K, et al. Delay in cooling negates the beneficial effect of mild resuscitative cerebral hypothermia after cardiac arrest in dogs: a prospective, randomized study. Crit Care Med 1993;21:1348–1358. NICE-SUGAR Study Investigators, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:1283–1297. Nielsen N, et al. Outcome, timing and adverse events in therapeutic hypothermia after out-of-hospital cardiac arrest. Acta Anaesthesiol Scand 2009;53:926–934. Nielsen N, et al. Targeted temperature management at 33 degrees C versus 36 degrees C after cardiac arrest. N Engl J Med 2013;369:2197–2206. Nolan JP, et al. Therapeutic hypothermia after cardiac arrest: an advisory statement by the advanced life support task force of the International Liaison Committee on Resuscitation. Circulation 2003;108:118–121. Nordberg P, et al. Design of the PRINCESS trial: pre-hospital resuscitation intra-nasal cooling effectiveness survival study (PRINCESS). BMC Emerg Med 2013;13:21. Peberdy MA, et al. Part 9: post-cardiac arrest care: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2010;122(18 Suppl 3):S768–S786. Polderman KH. Induced hypothermia and fever control for prevention and treatment of neurological injuries. Lancet 2008;371:1955–1969. Polderman KH. Mechanisms of action, physiological effects and complications of hypothermia. Crit Care Med 2009a;37: 186–202. Polderman KH, et al. Therapeutic hypothermia and controlled normothermia in the intensive care unit: practical considerations, side effects and cooling methods Crit Care Med 2009b;37:1101–1120. Pro CI, et al. A randomized trial of protocol-based care for early septic shock. N Engl J Med 2014;370:1683–1693. Rivers E, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345: 1368–1377. Safar P, et al. Improved cerebral resuscitation from cardiac arrest in dogs with mild hypothermia plus blood flow promotion. Stroke 1996;27:105–113. Sprung CL, et al. Hydrocortisone therapy for patients with septic shock. N Engl J Med 2008;358:111–124. Taccone FS, et al. Targeted temperature management after cardiac arrest. N Engl J Med 2014;370:1357–1358. van den Berghe G, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359–1367. Varon J, et al. Targeted temperature management after cardiac arrest. N Engl J Med 2014;370:1358–1359. Xiao G, et al. Safety profile and outcome of mild therapeutic hypothermia in patients following cardiac arrest: systematic review and meta-analysis. Emerg Med J, 2013;30:91–100.
Address correspondence to: Ashwin Subramaniam, MMed, FRACP, FCICM Department of Intensive Care Frankston Hospital Frankston, Victoria 3199 Australia E-mail: [email protected]