"Time is brain" in acute stroke treatments, and the complexity of events triggered by ischemic insult makes treatment difficult. Stroke represents an acute heterogeneous syndrome in which areas of ischemia and brain cell death lie adjacent to potentially viable tissue, and reperfusion strategies alone may not be sufficient for maximal improvement of neurologic outcomes. Combined treatment strategies aimed at decreasing multiple ischemic processes may assist in achieving improved neurologic outcomes and functionality. Many neuroprotective therapies aimed at salvaging brain tissue at risk are currently being investigated in combination with acute treatments. Currently, therapeutic hypothermia (TH) appears to be one of the most promising neuroprotective therapies. The purposes of this article were to (1) discuss the effects of ischemia and reperfusion in acute stroke, (2) discuss how TH can potentially limit neurological injury, and (3) review current literature on the use of hypothermia as a treatment for acute stroke.
Ischemia and Reperfusion
Ischemia of the brain initiates a cascade of destructive and often irreversible processes that could destroy brain cells and tissue (Figure 1). A pathway of apoptosis, or energy-dependent programmed cell death, is initiated in which intracellular metabolism converts to anaerobic metabolism (Polderman, 2009). This produces intracellular and extracellular acidosis, an influx of calcium into the cell, and the release of large amounts of the excitatory neurotransmitter glutamate, which further stimulates calcium influx, and eventually causes cell death. Calcium influx triggers apoptosis through mitochondrial dysfunction and initiation of the caspase-enzyme cascade.
Significant disruptions of the blood-brain barrier commence the development of vasogenic edema and intracranial hypertension. In acute neurologic injury, cerebral edema typically occurs several hours after injury, peaking at 24 to 72 hours and persisting for up to 1 week (Bardutzky & Schwab, 2007). Another potential complication of blood-brain barrier disruption is hemorrhagic transformation of the ischemic stroke, which can be devastating (Aviv et al., 2009). A process called cerebro thermopooling can occur, in which heat becomes trapped in injured areas, furthering the extent of neurological injury (Polderman, 2009). Coagulation activation and the formation of microthrombi promote further ischemic insult (Yenari, Kitagawa, Leyden, & Perez-Pinzon, 2008).
When blood flow is restored to ischemic tissue by recanalization, a second wave of excitotoxic substances is released, which can exacerbate the initial ischemic insult. This phenomenon is referred to as reperfusion injury and can increase secondary brain cell injury (Ritter, 2008). An inflammatory response begins about 1 hour after ischemia reperfusion, in which large amounts of proinflammatory cytokines stimulate apoptosis and promote chemotaxis of activated leukocytes across the blood-brain barrier. Persistent cytokine and leukocyte infiltration, free-radical production, and the release of toxic substances increase secondary brain cell injury and contribute to increased infarction.
Hypothermia: How It Works
Hypothermia has been suggested as a potential adjunctive treatment in acute stroke (Figure 2). It simultaneously activates many neuroprotective mechanisms, affects a wide range of destructive mechanisms occurring in ischemic brain tissue, and potentially limits neurological injury (Hemmen & Lyden, 2009; Polderman, 2009). Hypothermia in acute stroke aims to minimize secondary injury, decrease cell death, and salvage neurons within the penumbra (Linares & Mayer, 2009). Hypothermia does little to affect permanently damaged brain tissue.
Cerebral metabolic rate and oxygen consumption decrease by 5% to 10% for each degree Celsius of body temperature reduction with cooling (Polderman, 2009; Yenari et al., 2008). Reduction in cerebral metabolic rate decreases the consumption of glucose and oxygen and slows the breakdown of ATP. The acid-base status of the brain is normalized. As a result, massive depolarization of brain tissue is inhibited, reducing glutamate release, calcium influx, and free-radical production. Infiltration of cytokines and interleukins and subsequent inflammation are attenuated (Axelrod & Diringer, 2008). Hypothermia is thought to diminish apoptosis through inhibition of mitochondrial dysfunction, cytochrome-C release, and caspase activation (Broughton, Reutens, & Sobey, 2009; Yenari et al., 2008). Cytotoxic edema is reduced through preservation of cerebral vascular autoregulation (Linares & Mayer, 2009). Diminished late opening of the blood-brain barrier (22 to 46 hours) in rats treated with hypothermia limited vasogenic edema formation and reduced infarct volume (Huang, Xue, Preston, Karbalai, & Buchen, 1999). Guluma et al. (2008) found that hypothermia decreased postischemic edema in an analysis of 18 human patients cooled intravascularly for 12 or 24 hours to 33[degrees]C, with a 12-hour controlled rewarming period (p < .05). Hypothermia has proven effective in reducing intracranial pressure (ICP) and mortality by 38%-47% in humans when applied in the setting of malignant middle cerebral artery infarct (Bardutzky & Schwab, 2007). Effects on thrombolysis are difficult to predict because thrombolysis activity, platelet function, and coagulation are all altered with hypothermia (Hemmen & Lyden, 2009). Physiologically, it would appear that use of TH in ischemic-reperfusion injuries would limit damage and potentially improve outcomes.
Review of Recent Studies
Animal studies have indicated positive effects of hypothermia in ischemic stroke. van der Worp et al. (2007) performed a systematic review and meta-analysis on the efficacy of hypothermia in rat models of ischemic stroke. The review identified 101 publications, with 3,353 animals, reporting the effect of hypothermia on infarct size and functional outcome. Results indicated a 44% reduction in infarct size (95% confidence interval [CI] = 40%-47%) at a target temperature of 33[degrees]C and a 30% reduction in infarct size (95% CI = 21%-39%) at a target temperature of 35[degrees]C when initiation of treatment occurred between 90 and 180 minutes after artery occlusion. The effects of hypothermia not only reduced infarct size but also improved functional outcomes. In addition, hypothermia was more effective with hypertensive animals as compared with normotensive animals. Potentially, these results may be achievable for large numbers of acute ischemic stroke human patients. Kollmar, Blank, Jan, Georgiadis, and Schwab (2007) investigated the optimal depth of hypothermia in a rat model in 2007. In this study, groups of rats were cooled to temperatures of 33[degrees]C, 34[degrees]C, 35[degrees]C, 36[degrees]C, or 37[degrees]C for a period of 4 hours after 90 minutes of middle cerebral artery occlusion. Physiological variables, survival, infarct size, brain edema, neuroscore, and neutrophil invasion were measured at 24 hours and at 5 days. Temperatures of 33[degrees]C and 34[degrees]C showed more positive outcomes in reduction of infarct size and neuroscore (p < .05). Furthermore, only 34[degrees]C produced a significant lowering of neutrophil invasion (p < .05) as compared with 37[degrees]C.
The transition to human patient studies, however, has not been smooth due to small numbers of participants and varying methods and measurements, making comparisons difficult. A recent Cochrane Collaboration review analyzed eight cooling therapy trials (Table 1) for acute ischemic stroke involving a total of 423 patients (Den Hertog, Van der Worp, Tseng, & Dippel, 2009). A meta-analysis of pharmacological and physical temperature reduction trials indicated no significant difference between the active treatment group and the control group with regard to primary outcomes of death or dependency (modified Rankin scale >=3) at 1 and 3 months after stroke (odds ratio [OR] = 0.9, 95% CI = 0.6-1.4). High levels of heterogeneity were found in the physical temperature reduction studies (I2 = 76%). This was not found in the pharmacological temperature reduction studies (I2 = 0%). Secondary outcomes measures included death from all causes, mean body temperature 24 hours after initiation of treatment, intracranial hemorrhages, infections, and other complications. Temperature reduction had no statistical significance on the risk of death (OR = 0.9, 95% CI = 0.5-1.5). The numbers of hemorrhagic transformation of infarcts or intracranial hemorrhage were too small to draw any conclusions. One pharmacological temperature reduction study reported hemorrhagic transformation of infarct in the active treatment group (<1% of randomized individuals in pharmacological reduction studies). In the physical temperature reduction studies, two reports of hemorrhagic transformation of infarct and one intracerebral hematoma were reported during treatment (<1% of randomized individuals in physical reduction studies). More infections were reported in the active treatment groups; however, this did not achieve statistical significance (OR = 1.5, 95% CI = 0.8-2.6). In summary, there was not sufficient evidence to suggest routine use hypothermia with acute stroke; however, positive effects could not conclusively be ruled out. The reviewers suggested further large randomized control trials with emphasis on methodological quality to investigate safety, optimal temperature duration, and effectiveness of both physical and pharmacological temperature reduction in acute stroke.
The Intravascular Cooling in the Treatment of Stroke-Longer Window (ICTuS-L) trial was recently completed, and results are not yet published (Stroke Trials Registry, 2010). The ICTuS-L trial investigated the combination of intravenous tissue plasminogen activator/alteplase (tPA) and intravascular hypothermia within 6 hours after ischemic stroke (Hemmen, et al., 2006; Hemmen & Lyden, 2007). Enrollment was stratified to time windows of 0 to 3 hours and 3 to 6 hours. A total of 58 patients were enrolled, with 30 patients randomized to normothermia and 28 patients randomized to hypothermia treatment, using intravascular cooling to 33[degrees]C. All of the 44 patients who presented within 3 hours of stroke onset received intravenous tPA, whereas only 4 of 14 patients within the 3- to 6-hour window received intravenous tPA. Hypothermia treatment did not have a statistically significant effect on death or disability at 3 months. Symptomatic intracranial hemorrhage occurred in four patients, all of whom received tPA less than 3 hours, one of whom additionally received hypothermia treatment. Pneumonia was reported in 14 patients who received hypothermia treatment versus 3 in the normothermia group (p = .001); however, this did not adversely affect disability outcomes at 3 months (p = .32). This study was a step to establish safety in the use of hypothermia as a neuroprotective therapy in stroke. A phase 2 study is planned in which researchers plan to enroll more than 400 participants to better determine outcomes data.
A study combining early (15 +/- 6 hours) decompressive hemicraniectomy with mild hypothermia (35[degrees]C) for 48 hours showed a trend toward better outcomes with the combined therapies (Els et al., 2006). Twenty-five consecutive patients with malignant ischemic stroke (>2/3 infarct of one hemisphere) were randomized to receive either hemicraniectomy alone (13 patients) or a combination of hemicraniectomy and mild hypothermia (12 patients). Hypothermia was applied via intravascular catheter with 10 patients, whereas surface cooling was applied to two patients. The study did not show additional risk of side effects with hypothermia and did show a tendency toward improved functional status at 6 months, with National Institutes of Health Stroke Scale scores of 10 +/- 1 for the hypothermia group versus 11 +/- 3 for the normothermia group (p < .08).
Optimizing TH in Acute Stroke
Key factors that determine the outcomes of TH are timing of cooling, core temperature measurement, speed of induction and rewarming, duration of cooling, and management or prevention of side effects. Specifically in relation to acute stroke, the delivery of hypothermia within a neuroprotective window of time would be most beneficial.
Animal Studies
Numerous animal studies have investigated the application of hypothermia. In the previously discussed meta-analysis of animal studies by van der Worp et al. (2007), target temperature was achieved in 20 minutes and duration of cooling was 180 minutes. The most robust effects of cooling were obtained in rats when hypothermia was initiated before or at the time of vessel occlusion, which showed a decrease in infarct size by 44% (95% CI = 40%-47%). Maier, Sun, Kunis, Yenari, and Sterlin (2001) discovered that significant neuroprotection was still achieved when hypothermia (33[degrees]C) was delayed for up to 120 minutes and maintained for 2 hours after transient middle cerebral artery occlusion in rats when compared with the normothermic group (p < .05). In addition, the neuroprotective effect of mild hypothermia induced during the ischemic period was sustained over 2 months. Kollmar et al. (2002) demonstrated the effectiveness of neuroprotection in rats with moderate hypothermia (33[degrees]C) induced 180 minutes after middle cerebral artery occlusion and 1 hour after reperfusion in rats with an extended duration of hypothermia for 5 hours (p < .05). Thus, animal studies have substantiated early induction of hypothermia in ischemic stroke.
Temperature Measurement
Consideration must also be given to the method of temperature measurement because this can impact accuracy of true cooling measurement and influence cooling outcomes. The generally accepted gold standard for core temperature measurement is a pulmonary artery catheter, offering highly precise and rapid temperature measurement (Polderman & Herold, 2009). Pulmonary artery catheters, however, do require a complex insertion procedure. Alternate temperature measurement sites, such as esophageal, bladder, or rectal, have lag times that differ significantly and can affect potential overshoot of the target temperature, thereby dropping the actual core temperature below the target. Esophageal temperature offers an accurate reflection of the gold standard, with a lag time of 5 minutes. Esophageal probes can become displaced downward, with resultant changes in temperature and lag time that may not be noticed. Precise insertion to a depth of 32-38 cm and proper securement can alleviate these issues. Bladder temperature offers a fair level of accuracy with a lag time of 20 minutes. Bladder temperature accuracy can also be affected by the rate of urine flow, seen in diuresis or oliguria. Rectal temperature measurement offers a fair level of accuracy with a lag time of 15 minutes. Insertion of rectal probes is quick and easy; however, identification of dislocation of the probe can be missed. Tympanic temperature measurement is quick and easy, although readings can be inaccurate and have lag times of 10 minutes. The correlation between brain temperature and other temperature sources is also variable, and evidence supports that brain temperatures are higher than core, bladder, and rectal temperatures (Mcilvoy, 2004). Temperatures within the brain also vary by site and depth. Temperatures deeper within the brain have higher temperatures, with the lateral ventricle recording the highest brain temperatures. Initial temperatures in ischemic areas are also greater than in nonischemic areas (Juttler et al., 2007).
Hypothermia Timing
The greatest periods of instability occur with induction of cooling and also with rewarming (Polderman & Herold, 2009). Attention must directed toward the speed of cooling because it is a key factor in maximizing the protective effects of hypothermia by limiting secondary injury (Yenari et al., 2008). Studies in a variety of populations have investigated the use of ice-cold saline (ICS) infusions to rapidly induce hypothermia, followed by different types of surface or intravascular cooling (Kim, Olsufka, Nichol, Copass, & Cobb, 2009; Polderman & Herold, 2009; Polderman, Rijnsburger, Peerderman, & Girbes, 2005). The early use of ICS in acute stroke patients was piloted in a recent study (Kollmar, Schellinger, Steigleder, Kohrmann, & Schwab, 2009). With the use of ICS infusions, TH was induced in 10 acute ischemic stroke patients, 9 of whom were treated with thrombolysis. Within 3 hours of onset of stroke symptoms, 25 ml/kg body weight of ICS (4[degrees]C) was infused over 17 minutes. Tympanic temperature decreased significantly by a maximum of 1.6[degrees]C +/- 0.3[degrees]C (p < .005) at 52 +/- 16 minutes after ICS was started. The National Institutes of Health Stroke Scale score also significantly improved from an admission median score of 5.5 (range = 4-12) to a median score of 1 (range = 1-15) at discharge (p < .02). Hypothermia was induced without major side effects of shivering, fluid overload, changes in vital signs, ejection fraction, or blood sample results. Limitations of this study include tympanic temperature measurement as a less reliable temperature measurement source and a small number of participants. This pilot study, however, is the first report of infusion of ICS in acute ischemic stroke patients within the neuroprotective window, offering a potential initial treatment that could be administered in an ambulance.
The rewarming phase is also associated with instability. Rapid rates of rewarming lead to impaired cerebrovascular reactivity, brain hypoxia, and severe brain hyperthermia, even when core temperatures measured at other sites appear normal (Polderman & Herold, 2009). Catecholamine levels increase significantly during rewarming, affecting the brain, cardiac, and other organ systems. The rewarming process appears to be crucial because rebound intracranial hypertension can occur and contribute to mortality, especially with rewarming periods less than 16 hours (Bardutzky & Schwab, 2007). Very gradual rewarming can be started after ICP is under control for a minimum of 24 hours. Faster rewarming rates are associated with increased ICP and deaths, with mortality up to 38% (Linares & Mayer, 2009). In comparison with rewarming rates of 0.25[degrees]C per hour in cardiac arrest victims, slow and controlled rewarming of 0.1[degrees]C per hour produces less alterations in ICP and is essential in stroke patients with elevated ICP. Rewarming rates of 0.25[degrees]C to 0.33[degrees]C can be tolerated by patients without elevated ICP (Linares & Mayer, 2009; Polderman, 2009). Slow rewarming and maintenance of normothermia after TH help to preserve the neuroprotective effects of hypothermia.
The research on hypothermia in cardiac arrest and brain injury populations is relevant and can provide a foundation for TH protocols for acute stroke. In cardiac arrest, cooling to 33[degrees]C and a duration of 12 to 24 hours are traditionally applied. In comparison, the therapeutic window of brain injury suggests that earlier mild hypothermia (33[degrees]C-35[degrees]C) and a longer duration of cooling (more than 48 hours) may be beneficial (Jiang et. al., 2006; Stocchetti et al., 2007). In a previously discussed systematic review and meta-analysis of animal studies by van der Worp et al. (2007), mild hypothermia at 35[degrees]C produced a 30% reduction in cerebral infarct volume (95% CI = 21%-39%). Milder cooling (35[degrees]C) would necessitate less need for sedatives, paralytics, and mechanical intubation and potentially permit neurological assessment of an awake stroke patient. A longer duration of cooling may be more effective when being used to control refractory intracranial hypertension and cerebral edema in ischemic stroke (Bardutzky & Schwab, 2007).
Cooling Methods
Current cooling methods can be divided into noninvasive/surface cooling devices and intravascular catheters (Table 2). Consensus does not currently exist on which method is superior for TH; however, there is consensus that cooling of the head alone is not sufficient (Juttler et al., 2007). Prospective studies have not compared surface and intravascular cooling devices in a standardized way to establish any differences in efficacy of cooling, maintenance of hypothermia, rewarming, and side effects (Polderman & Herold, 2009). Intravascular cooling may prove to be more efficient in cooling the brain-injured patient due to rapid cooling rates, maintenance of a stable temperature, and control of rewarming (Polderman, 2009; Yenari et al., 2008). The use of ICS was covered in detail earlier and is best applied in conjunction with other cooling methods. Ice packs have been used to induce and maintain hypothermia and represent an inexpensive, readily available, and nurse-driven method. Maintenance of target temperatures within a narrow range using ice packs is complicated and labor intensive, precise and slow rewarming can be difficult, and skin issues can arise with direct contact with a cold source. Water-circulating blankets, pads, and wrapping garments are inexpensive, are reusable or disposable, and can be initiated quickly by nurses at the bedside compared with most other devices. Conversely, these water-circulating devices do not cool as rapidly as other devices, maintenance of target temperature is not as precise, and they carry a risk of skin lesions. Hydrogel-coated water-circulating pads have more reliable maintenance of target temperature; allow for slow, controlled rewarming; can be initiated quickly by nurses at the bedside; and are less labor intensive for controlling hypothermia (Mayer et al., 2004). The hydrogel-coated pads do carry a risk of skin lesions with prolonged use but have a decreased incidence in comparison with higher rubber content pads found in blankets, pads, and wrapping garments. Intravascular catheters have rapid cooling rates and are highly reliable in terms of maintenance of target temperature and slow, controlled rewarming (Hoedemaekers, Ezzahti, Gerritsen, & van der Hoeven, 2007). An invasive insertion procedure is required and can delay cooling if a physician to perform the procedure is unavailable. Another potential problem is the risk of catheter-related thrombosis. Data detailing the risk of cooling catheter-related thrombosis are scarce and require future studies. Anecdotally, the range of asymptomatic thrombosis is from 33% to 75%, depending on the dwell time of the catheter. Currently, the risk of symptomatic thrombus formation with intravascular cooling catheters is very low (Polderman & Herold, 2009).
Risks of Hypothermia
Hypothermia is not without risks and possible complications. Initiating a more rapid induction of hypothermia partially reduces the impact of these adverse effects. Anticipating the side effects of hypothermia permits prevention and control, highlighting the importance of vigilant monitoring and prompt management (Yenari et al., 2008).
Shivering is a major adverse effect of lowering body temperature, and 35.5[degrees]C appears to be the shivering threshold in humans (van Zanten & Polderman, 2009). Once below this temperature, a shivering response occurs. Shivering can cause discomfort, increases catecholamine release and oxygen consumption, and possibly increases ICP. By employing sedation and neuromuscular blockade, the adverse effects of shivering can be diminished. Protocols, such as the Bedside Shivering Assessment Scale, have been developed to assess shivering and offer guidelines to treat or prevent shivering (Mayer, 2008). Most hypothermia protocols incorporate sedation and neuromuscular blockade, necessitating intubation and mechanical ventilation in an intensive care unit setting. Neuromuscular blockade implementation may eliminate the muscular response to shivering but does nothing to inhibit the neural response; thus, the brain may still be shivering (van Zanten & Polderman, 2009).
Skin counterwarming (CW) has assisted in the suppression of shivering in patients undergoing cooling (van Zanten & Polderman, 2009; Mayer, 2008). Surface warming appears to affect the feedback loop from the skin to the thermoregulatory centers in the hypothalamus. A study published by Badjatia et al. (2009) investigated the use of surface CW as an adjunctive therapy to reduce shivering in both TH and controlled normothermia. The prospective study included 50 mechanically ventilated patients with brain injury treated with hypothermia (9 with intracerebral hemorrhage, 32 with subarachnoid hemorrhage, 3 with cardiac arrest, 3 with traumatic brain injury, and 3 with ischemic stroke). The method of cooling, either intravascular cooling (5 participants) or surface cooling using hydrogel-coated water-circulating pads (45 participants), and duration of cooling were based on the neurointensivist decision. In addition to the administration of acetaminophen 650 mg orally every 4 hours and buspirone 30 mg orally every 8 hours, the entire anterior surface area of the body was covered and warmed with an air circulating blanket, revealing a simple, nonsedative method to oppose shivering and its metabolic effects. Although surface CW produced significant reductions in resting energy expenditure (p < .001), oxygen consumption (p < .001), and carbon dioxide production (p < .001), it did not prove beneficial for all patients, and implementation of multiple methods to reduce shivering may be required. Subarachnoid hemorrhage patients exhibited a higher incidence of shivering (Bedside Shivering Assessment Scale score >=1 = 81% vs. 56%; p = .05). The presence of shivering was significantly associated with hypomagnesemia (1.8 +/- 0.3 vs. 2.4 +/- 0.3 mg/dL; p = .001), lower body mass index (26.3 +/- 4.6 vs. 29.3 +/- 1.9 kg/m2; p = .02), and lower body surface area (1.8 +/- 0.2 vs. 2.0 +/- 0.1 m2; p = .001). The use of intravascular cooling does allow for more exposed skin and potentially may increase the beneficial use of skin CW. However, limitations in generalizability of this study include that most patients (n = 44) underwent normothermia using surface cooling, only five patients were cooled intravascularly, direct skin temperature measurement was not assessed, and the effect of CW on core temperature was not discussed.
Lenhardt et al. (2009) investigated the use of medications to reduce the shivering threshold in young, healthy volunteers to assess the possibility of applying hypothermia in awake patients with minimal sedation. The study investigated the combination use of buspirone, which reduced the shivering threshold by 0.7[degrees]C, from 36.6[degrees]C to 35.9 [degrees]C +/- 0.4[degrees]C, and dexmedetomidine, which reduced the shivering threshold by 2.0[degrees]C to 34.7[degrees]C +/- 0.4[degrees]C (both p < .001). A total of 60 mg of Buspirone was administered orally in three equal doses at 90 minutes before active cooling, 45 minutes before cooling, and at the initiation of cooling. Dexmedetomidine was administered via a computer-controlled intravenous infusion to a target plasma concentration of 0.6 ng/ml. The combination of the medications, with different mechanisms of action, produced a synergistic effect lowering the shivering threshold by 2.5[degrees]C to 34.1[degrees]C +/- 0.4[degrees]C and also lacked the side effects of respiratory depression and drug-related seizures, as was found in the combination of buspirone and meperidine. There were a number of cited limitations to the study, such as restricted use of male participants, no direct application to patients with stroke/brain injury, and the use of dexmedetomidine, which is restricted by the Food and Drug Administration to the postoperative care of critically ill patients for a period of 24 hours.
Hypothermia also inhibits systemic inflammatory responses through inhibition of proinflammatory cytokines and leukocytes, thus increasing the risk of infection (Polderman, 2009). One of the most frequent complications of hypothermia is pneumonia. Furthermore, increased risk of infection is associated with longer periods of cooling (>24 hours).
Cardiac adverse effects include changes in heart rate (bradycardia), prolonged PR and QT intervals, widening of the QRS complex, and occasional appearance of Osborne waves. Cardiac output decreases 25%-40% mainly due to a decrease in heart rate. Hypotension can develop related to a decreased cardiac output and hypovolemia due to hypothermia-induced "cold diuresis." All of these factors can decrease brain tissue perfusion and promote further ischemic injury.
Hypothermia is also associated with electrolyte abnormalities such as decreases in potassium, magnesium, phosphorus, and calcium. During the rewarming phase, patients are at risk for hyperkalemia due to extracellular shifts and other changes in laboratory values (increase in liver enzymes, amylase, and lactate; decrease in white blood cells and platelets). Hyperglycemia and insulin resistance, potential for skin pressure ulcers, changes in drug clearance, and delayed gastric emptying are also possible side effects of hypothermia. Vigilant monitoring and prompt management are key in diminishing these adverse effects of TH.
Discussion
Hypothermia represents an evolution in the future of neuroprotective therapy, as it affects a broad range of cellular death mechanisms and offers an exciting frontier for further study. The preponderance of evidence suggests the positive effects of TH if it is appropriately applied in early phases of neurologic injury; however, the ideal protocol for the use of TH in stroke is yet to be developed. Key elements still need to be determined, such as optimal initiation time, cooling method, temperature monitoring method, target temperature, and duration of cooling and rewarming. Currently, intravascular cooling, as compared with traditional surface cooling methods, appears to be a more efficient method of cooling in brain-injured patients, and suppression of shivering may be a key in allowing cooling in the awake stroke patient. Further development of the application, implementation, and evaluation of TH shows promise and can potentially offer new hope for preventing secondary brain injury and improving the neurologic outcomes of acute stroke patients.
References