Improving neurological outcome after cardiac arrest: Therapeutic hypothermia the best treatment.

Cardiac arrest, irrespective of its etiology, has a high mortality. This event is often associated with brain anoxia which frequently causes severe neurological damage and persistent vegetative state. Only one out of every six patients survives to discharge following in-hospital cardiac arrest, whereas only 2-9% of patients who experience out of hospital cardiac arrest survive to go home. Functional outcomes of survival are variable, but poor quality survival is common, with only 3-7% able to return to their previous level of functioning. Therapeutic hypothermia is an important tool for the treatment of post-anoxic coma after cardiopulmonary resuscitation. It has been shown to reduce mortality and has improved neurological outcomes after cardiac arrest. Nevertheless, hypothermia is underused in critical care units. This manuscript aims to review the mechanism of hypothermia in cardiac arrest survivors and to propose a simple protocol, feasible to be implemented in any critical care unit.

INTRODUCTION

Therapeutic hypothermia (TH) was a widely practiced treatment over a century ago for soldiers to preserve injured limbs and provide analgesia for amputations.[1] Later on, its practice was further propagated by cardiac surgeons and neurosurgeons.[1] By the late 1950s, survival benefit was documented in cardiac arrest, but its routine use fell out of favor due to its complications, particularly infection and coagulopathy.[2] Recent prospective studies have provided credible data showing that mild TH, targeting a temperature of 32°C-34°C, confers significant survival and functional benefits, impelling increasing acceptance and use.[34]

PATHOPHYSIOLOGY OF CENTRAL NERVOUS SYSTEM DAMAGE

Cardiac arrest leads to sudden termination of blood flow, leading to a rapid exhaustion of cerebral oxygen and adenosine triphosphate stores, and depressed cerebral function. Neuronal damage occurs in the CNS during cardiac arrest (Phase I, “No Flow”) and following the return of spontaneous circulation (ROSC) (Phase II, “Low Flow”).[56] The survival of neurons differs, depending on the type, site, and period of global anoxia. In general, neurons in the cerebral cortex, hippocampus, and basal ganglia are the most susceptible.[6]

After the recovery of spontaneous circulation, cerebral blood flow initially attains supernormal levels but falls to below normal over a period of several hours.[67] This reduced cerebral perfusion pressure is a consequence of cerebral vasospasm which occurs due to endothelin release. Leukocyte clumping and microvascular coagulation further reduces the cerebral blood flow.[6] Increased cerebral oxygen requirements during the low-flow phase further result in secondary ischemia.

Inflammatory cascade that begins during resuscitation continues following ROSC.[56789] The main neurotransmitters that are responsible for inflammatory cascade are amino acids (e.g., glutamate) and matrix metalloproteinases, N-methyl-D-asparate (NMDA) which causes receptor activation and increased microvascular permeability which further results in calcium influx that leads to cerebral edema, raised intracranial pressure, and brainstem herniation. Supplementary contributing factors include free radical formation, autodigestion by activated proteases, and apoptosis.[67] Practically, the therapies tried for reducing intracellular edema, including NMDA receptor antagonists, mannitol, albumin, and hypertonic saline, have not proven advantageous.[5]

Following are the benefits of TH which are associated with slowing down of cellular cascade after cardiac arrest by

slowing down the cerebral metabolism (approximately 6-8% per 1°C),

reducing excitatory amino acids (glutamate release),

attenuation and/or reversibility of ischemic depolarization of the CNS, leading to membrane stabilization, electrolyte redistribution, and normalization of intracellular water concentration and intracellular pH (stabilization of the blood–brain barrier),

reduction of oxygen free radical production and lipid peroxidation,

restoration of normal intracellular signalling mechanisms (including calcium modulation) and inhibition of deleterious signalling mechanisms, such as apoptotic signalling,

restoration of protein synthesis and gene expression,

inhibition of deleterious inflammatory products (i.e., cytokines, interleukins, arachidonic acid cascade end products),

attenuation of CSF platelet-activating factor (PAF), and

Inhibition of cytoskeletal breakdown.

Hypothermia is helpful in cardiac injury as it may restrict the area of injury, improve epicardial reflow, decrease myocardial metabolic demand, and conserve intracellular high-energy phosphate stores.[101112]

METHODOLOGY

Who to cool

Two milestone studies by Bernard et al. and Hypothermia after Cardiac Arrest (HACA) group, which were published in the New England Journal of Medicine in February 2002, demonstrated improved survival and neurological outcomes with induction of mild TH for comatose survivors of out-of-hospital cardiac arrest. These studies now have been incorporated in 2010 American Heart Association guidelines in cardiopulmonary resuscitation (CPR) in post-cardiac arrest care. The HACA study group showed that when applied to unconscious out-of-hospital cardiac arrest patients with ROSC, mild hypothermia (cooling to 32°C–34°C) provided significant improvement in functional recovery at hospital discharge (55% vs. 39%) and lower 6-month mortality rate, when compared with patients who were not cooled (41% vs. 55%).[13] The number of patients requiring treatment was very low and comparable to other important emergent treatments such as cardiac catheterization for acute coronary syndrome.[14] Bernard et al. examined the endpoint of survival to hospital discharge to home or a rehabilitation facility (good outcome) in 77 patients and demonstrated 49% in the hypothermia group compared with 26% in the normothermic group.[3] A 2011 meta-analysis of randomized controlled trials found that TH with conventional cooling methods improves both survival and neurological outcomes at hospital discharge for patients who experienced cardiac arrest.[15] Additional studies with historical control groups show improved neurological outcome after TH for comatose survivors of ventricular fibrillation (VF) cardiac arrest.[1617]

No randomized controlled trials have compared outcome between hypothermia and normothermia for non-VF arrest. However, six studies with historical control groups reported a beneficial effect on outcome from use of TH in comatose survivors of out-of-hospital cardiac arrest associated with any arrest rhythm.[161718] Only one study with historical controls reported better neurological outcome after VF cardiac arrest but no difference in outcome after cardiac arrest associated with other rhythms.[19] Two nonrandomized studies with concurrent controls indicate a possible benefit of hypothermia after in- and out-of-hospital cardiac arrest associated with non-VF initial rhythms.[1920]

Case series have reported the feasibility of using TH after ROSC in the setting of cardiogenic shock and TH in combination with emergent percutaneous coronary intervention (PCI).[2122232425262728] Case series also report successful use of fibrinolytic therapy for acute myocardial infarction (AMI) after ROSC, but data are lacking about interactions between fibrinolytics and hypothermia in this population.[293031] The eligibility criteria for therapeutic hypothermia are given in Table 1.

Table 1

The eligibility criteria for therapeutic hypothermia

When to cool

The best time to initiate hypothermia after cardiac arrest is not completely understood till date. It is logical to conclude that the benefits may be maximized if hypothermia is initiated as soon as possible after successful resuscitation. Experimental studies of cardiac arrest in animal models have shown that the beneficial effect of hypothermia was lost when hypothermia was delayed for more than 1 h after ROSC.[313233] Beyond the initial minutes of ROSC and when hypothermia is prolonged (>12 h), the relationship between the onset of hypothermia and the resulting neuroprotection is less clear.[3435] HACA group has shown that the interval between successful resuscitation and achieving the required temperatures has an interquartile range (IQR) 4–16 h with a mean of 8 h and has demonstrated better outcomes in hypothermia-treated patients than in normothermia-treated patients.[13] A case series consisting of 49 consecutive comatose cardiac arrest patients cooled intravascularly after pre-hospital cardiac arrest has also shown that time of target temperature (mean 6.8 h) (IQR 4.5–9.2 h) was not an independent predictor of neurological outcome as previously thought. Another registry based case series of 986 comatose post-cardiac arrest patients suggested that time to initiation of cooling (IQR 1–1.8 h) and time to achieving target temperature (IQR 3–6.7 h) were not associated with improved neurological outcome after discharge.[27] Five studies have indicated that the combination of TH and PCI is feasible and safe after cardiac arrest caused by acute myocardial infarction.

The current recommendation of International Liaison Committee on Resuscitation (ILCOR) is to start cooling within 4–6 h.[36] The most favorable period of induced hypothermia is at least 12 h and may be > 24 h. Hypothermia was maintained for 12 or 24 h in the studies of out-of-hospital patients presenting in VF. Most case series of adult patients have reported 24 h of hypothermia. The effect of a longer duration of cooling on outcome has not been studied in adults, but hypothermia for up to 72 h was used safely in newborns.[37]

How to cool

The hypothermia comprises of three phases-induction, maintenance, and rewarming.

Hypothermia induction phase

During the induction phase, the goal is to achieve the target temperature of 32°C–34°C as quickly as possible.[38] Specialists from various medical fields (neurology, pulmonary and critical care, and cardiology) and ancillary medical staff should be alerted. A baseline neurological exam should be conducted. Preliminary patient monitoring should include electrocardiogram, fluid balance, invasive blood pressure measurement, and core temperature measurement by either vesicle catheter, esophageal thermometer, or pulmonary artery catheter. As hypotension is commonly observed during TH, intra-arterial blood pressure monitoring is important. Profuse diuresis during TH leads to hypovolemia.[35]

All baseline investigations should include complete blood count, platelets, coagulation, electrolytes, and arterial blood gas, at induction and every 6 or 12 h. Mild coagulation changes are seen in hypothermia; however, massive bleedings are not associated with mild hypothermia.[39] Cutaneous vasoconstriction due to the hypothermia makes pulse oximetry less reliable; therefore, blood oxygenation and ventilatory settings are better evaluated using arterial blood gas. Hypothermia may lead to severe arrhythmias which occur due to potassium, magnesium, calcium, and phosphorus cell inflow. Electrolyte replacement should take place during the induction phase, but should be discontinued during rewarming.[35] Sedatives are initiated during this phase to reduce cerebral metabolism in patients with anoxic brain injury.[40] In the advent of brain injury, sedation acts along with TH to reduce CNS oxygen consumption. Normally, benzodiazepines, propofol, and short-acting narcotics are used.[41] These include midazolam, 0.125 mg/kg/h; fentanyl, 2 μg/kg/h; and propofol, 5 μg/kg/min.[4142] Doses are adjusted to facilitate mechanical ventilation and keep the patient deeply sedated, i.e., unarousable and unresponsive to verbal/tactile stimuli. Richmond or Ramsay Scale for monitoring sedation may be helpful.[943] Sedation may also aid cooling by preventing shivering.[44]

Shivering is common between 36°C and 34°C.[1541] Boluses of fentanyl and other short-acting narcotics may also be effective.[41] The rate of narcotic/benzodiazepine infusions may be increased to prevent shivering. Pethidine (10–25 mg IV) is often used to treat shivering during TH when other measures are ineffective.[41] Magnesium at doses of 2–5 g infused over 5 h may also control shivering.[43] In addition, it has vasodilator effects and may speed cooling.[43] Magnesium may also prevent and treat TH-related hypomagnesemia and provide some neuroprotection.[43]

If shivering is unmanageable and impairs cooling, neuromuscular blockade may be required. Often, pancuronium (0.1 mg/kg q 2 h) or cisatracurium (0.15 mg/kg bolus followed by a 3 μg/kg/min infusion) may be used, although other agents are acceptable.[4142] Paralytics can suppress shivering, but also mask seizure activity and may increase the risk of critical illness myoneuropathy.[4243]

The ideal cooling method is one that is able to induce rapid hypothermia, without overcooling risk; to keep the desired temperature without wide variations, to provide a controlled and slow rewarming; to be minimally invasive and affordable.[45] Cooling may be invasive or non-invasively induced. The non-invasive or conventional methods include use of ice packs, thermal blankets, surface cooling devices, and infusion of cold solutions. These methods are quite effective to induce hypothermia; however, the rate of temperature change is less accurate and there is a higher overcooling risk, in addition to a more difficult rewarming. Combinations of these methods have been used in several TH studies, showing good results. Infusion of cold fluids at a 30-40 mL/kg dose, either peripheral or centrally, is capable of inducing a temperature drop by 2°C-4°C, with the advantage of practical use even in pre-hospital setting.[4647] Large cold saline volumes do not appear to be associated with severe adverse effects in cardiac arrest survivors, with no pulmonary edema seen in these patients.[48] Along with the use of cold saline, using ice packs over the neck, axilla, and groin surfaces is a simple and easy way to keep cooling. The external ice packs should be changed when melted, and attention should be paid to cold-induced skin injuries. No association is obvious between the body surface and the time to hypothermia induction.[49] Ice packs together with thermal blankets are the most inexpensive way to induce TH; however, studies have shown that overcooling is almost a rule, with potential severe complications if very deep or prolonged.[50] Overcooling is uncommon with surface cooling devices. These devices are made of pads covered with thermal gel, connected to a thermoregulator unit. The system either increases or reduces the circulating water temperature in response to both the target temperature and the patient's temperature. The mean elapsed time to reach the target temperature is about 1.4°C/h. This is a safe and effective method, as the temperature range is better controlled both during induction and rewarming.[51] Cooling methods are given in Table 2 (2a- Internal cooling and 2b-external cooling).

Table 2a

Cooling methods: Internal cooling

Presently, the most effective method to induce hypothermia is the use of endovascular catheters that give optimized temperature control for induction, maintenance, or rewarming. It has a very fast induction, reducing the temperature by 2°C-2.5°C/h. This system uses a special coated metal central venous catheter where water circulates from an external cooler system. The catheter may be installed either via femoral, subclavian, or jugular accesses. The experience with these devices is still inadequate as they are more expensive but, on the other hand, less troublesome than the other conventional methods.[45]

Hypothermia maintenance phase

The temperature should be regularly measured, aiming to be kept between 32°C and 34°C over 24 h.[52] An important point for these patients’ care involves hemodynamic parameters. Mean blood pressure levels above 80 mmHg are suggested for cardiac arrest survivors, and volume replacement and vasopressors may be required to keep these values. The most commonly used vasopressor during TH is norepinephrine.

Hypothermia causes insulin resistance. Blood glucose monitoring should be performed with venous blood samples as skin vasoconstriction may alter the results. Laboratory tests may be planned every 6 or 12 h, depending on previous results, and include the same tests as during the induction phase. Pulse oximetry is not a reliable parameter during hypothermia, and the mechanical ventilation parameters should be based on arterial blood gas values.

Feeding is not indicated during TH, as stomach emptying is tardy in these patients. Moreover, there is an increased mechanical ventilation associated pneumonia (VAP) risk for probable aspiration during cardiac arrest. Strict VAP prevention measures are thus recommended.

Another primary aspect for this phase management is sedation and analgesia. In addition to continued midazolam and fentanyl infusions, additional doses may be needed to maintain proper sedation. A better sedation control is using sedation scores or the Bispectral Index (BIS). Seizure activity may be masked by sedation and muscular blockade, so continued electroencephalogram (EEG) is indicated, if available. Use of BIS or EEG is a protocol refinement, and is not obligatory. Convulsive seizures and shivering require aggressive therapy in any phase, as they increase metabolic oxygen demands. Neuromuscular blockers should be reassessed after 12 h and stopped if there is no evidence of shivering.[53]

Rewarming phase

This phase takes place 24 h after cooling is started, and should be slow, by 0.2°C–0.4°C/h for 12 h, until a temperature between 35°C and 37°C is reached. Rewarming may be either passive or active. Passive rewarming up to a 35°C central temperature usually takes about 8 h.[21] If a thermal blanket is used, it should be removed when the temperature reaches 35°C. If commercial external cooling devices or endovascular catheters are used, the rewarming speed is set. One of the main advantages of these devices is a better control of the rate of temperature changes.[45]

Hemodynamic instability, with peripheral vasodilation and hypotension is part of post-reperfusion syndrome, and is very common as the temperature increases. It may require higher vasopressor doses. Another concern during rewarming is hyperkalemia, as the potassium cell inflow during hypothermia now flows out. This may cause arrhythmias. All potassium- or magnesium-containing solutions should be discontinued. Insulin infusion is also discontinued due to hypoglycemia risk.

When 35°C is reached, sedation is discontinued. After the TH protocol ends, aggressive fever (if occurs) therapy is recommended, as it is associated with unfavorable outcomes after cardiac arrest.[54]

Based on the above information, we propose a simple protocol that is easy to perform and can be immediately implemented in low-, medium-, or high-complexity ICUs. The medication doses are given in Appendix 1 and the suggested ICU protocol is shown in Appendix 2.

Appendix 1

Sedative and paralysis options

Appendix 2

Summary of ICU protocol

Adverse effects are described in Table 3

Table 3

Adverse effects

CHANGES IN PROGNOSTICATION WITH HYPOTHERMIA

There is not enough data about the utility of physical examination, EEG, and evoked potentials in patients who have been treated with induced hypothermia. Physical examination (motor response, pupillary light and corneal reflexes), EEG, somatosensory evoked potential (SSEP), and imaging studies are less reliable for predicting poor outcome in patients treated with hypothermia. Durations of observation greater than 72 h after ROSC should be considered before predicting poor outcome in patients treated with hypothermia (Class I, Level C).[36] With the advent of hypothermia, the current indicators of poor prognosis of neurological outcome need revision.

The prognostic ability of median nerve short latency SSEPs does not seem to be affected by TH. Decreasing levels of serum neuron specific enolase (NSE) after TH may indicate selective attenuation of delayed neuronal death in victims of cardiac arrest.

CONCLUSION

Hypothermia is proving to be an extremely robust and important therapy for cardiac arrest survivors and so far the only therapy consistently shown to reduce mortality and improve neurological outcomes in cardiac arrest survivors. Although more research is needed to define optimal timing and duration of therapy, we believe that TH should be initiated rapidly once the indication is clear, particularly among survivors of cardiac arrest. Hospitals must prioritize establishing hypothermia protocols and systems to improve compliance with treatment recommendations which are acceptable, available, and affordable. TH can also be started by paramedics in pre-hospital setting by storing cold IV fluids and ice packs in ambulances, thereby hastening the process of cooling. Individual institutions should choose methods most likely to succeed in their local environment, pending further data comparing the available techniques. Assembly of a TH team and development of treatment protocols are likely to optimize implementation.

Table 2b

Cooling methods: External cooling