Temperature Management in the Neurological and Neurosurgical ICU

Temperature management in the neurosurgical intensive care units (ICUs) is now an important component of caring for patients. While mild cooling may be initiated to target secondary injury mechanisms, strategies to inhibit periods of reactive hyperthermia are also important to ICU management techniques. In the area of traumatic brain injury and spinal cord injury, for example, early cooling has been shown in some clinical studies to reduce mortality and improve functional outcome in specific patient populations. Experimentally, temperature has been shown to target a variety of pathophysiological mechanisms thought to be important in producing structural damage and long-term functional deficits. While mild cooling reduces many of these pathophysiological events, mild elevations in temperature aggravate some of these processes. A series of state-of-the-art lectures presented at the Therapeutic Hypothermia and Temperature Management Meeting in Miami, Florida, brought together experts in the field of temperature management strategies.

From a military perspective, traumatic brain injury (TBI) is a serious problem in our wounded warriors. Concussion, moderate, and severe TBI occur in 10% to 20% of service members each year, with concussion being the most common injury. Dr. Donald Marion, Walter Reed Army Hospital, discussed the increased use of roadside bombs and the number of troops being injured and sustaining long-term functional problems. Cumulative brain injury resulting from multiple concussions has led to new policies regarding return to duty decision making. Routine cognitive testing and the use of a military acute concussion evaluation procedure is the current standard of care. While therapeutic hypothermia has been used previously to treat specific injuries caused by trauma in the military, it is not routinely used for moderate or severe TBI or refractory ICP except as second or third tier therapy.

Dr. Daniel Sessler, Department of Outcomes Research, Cleveland Clinic, discussed the challenges of producing hypothermia in anaesthetized patients. Prevention and treatment of shivering and drug-induced thermoregulatory impairments was highlighted. Treatments targeting shivering include skin and surface warming approaches as well as drugs and drug combinations to induce thermal tolerance. Use of meperidine, in combination with other drugs, seems to produce good outcomes, but some potential complications, including respiratory depression, have to be critically evaluated. Therapeutic hypothermia may also produce hypertension and blood loss and impair wound healing. Hypothermia may be induced using controlled methodologies with appropriate strategies to reduce risk. Finally, Dr. Dalton Dietrich, University of Miami Miller School of Medicine, discussed recent data emphasizing the detrimental effects of mild hyperthermia in mildly injured rats. Strategies to measure and record temperatures may be important in sporting or battlefield situations where athletes or soldiers may actually have elevated core temperatures because of ambient temperatures and strenuous activities associated with wearing protective body armor and carrying heavy weights. The following summary lists a series of questions and answers regarding these important topics.

Dr. Lyden: Two questions. First is on hyperthermia. The soldiers are walking around in 120° heat. But that doesn't mean their body temperature is elevated, right? They are compensating, so I wonder if they are at risk as you suggested they are?

Dr. Dietrich: That is an important question that we are currently addressing in our clinical studies. If the ambient temperature is 110° to 120° and they are exerting themselves, it may be that the autoregulatory mechanisms are not working as well as they could, and CNS temperature could be elevated. Recent clinical studies have reported that under adverse or strenuous conditions, rectal temperature can be significantly elevated. Does someone in the audience have anything to say about it? I guess it's something that waits to be proven.

Question: I was going to ask the same question about athletics, where soccer and football players have been sweating and have body temperature partially elevated. Has there been any comparison between concussions sustained at an athletic event versus a concussion from an accidental fall or a car crash? If you are not sweating is your body temperature actually less likely to be elevated?

Dr. Dietrich: During the NFL football season, when the players can be playing in some very warm temperatures, you see that when they sit on the bench their shoulder pads are actually being cooled. So they are trying to reduce body temperature in those athletes. So I think there is a possibility that core temperature is elevated during periods of extreme exertion. Does CNS temperature get raised as well? I just don't know.

Question: Do you know by any chance if the NFL is looking at this? Are they actually recording body temperature at the time these guys get a concussion and looking at if that impacts perhaps some of their short term psychological scores? We saw from your work that just a mild concussion with fever makes a difference.

Response: I do not know if this important question is being addressed. In car racing, they've actually recorded body (rectal) temperatures that are elevated in those athletes associated with increased cardiac function.

Dr. Sessler: Exercise increases core temperature by about 0.6°, which is actually a regulated increase. This happens even in a cold environment. In a warm environment, body temperature can go up, but people work hard to regulate their temperature. So, I think your point is valid. Just because people are walking around in a 100° environment, their core temperature does not normally increase. Normally, people regulate their temperatures pretty well. We need to be very careful in measuring temperature in unusual circumstances. Rectal temperature measurements are worthless in athletes, as the values are falsely high by many degrees. Probably the only reliable way to measure core temperature is with ingested capsules, which record true core temperature. This is an FDA-approved and readily available technology that could easily be used in football players and car racers.

Question: I enjoyed your talk, and the effect of mild hypothermia on blood loss during surgery is understandable. But then, to extend that to hemorrhagic stroke, I think, is maybe a little bit of a leap. The mechanisms are completely different. There is plenty of secondary injury around that hematoma, and I'm not sure it is an immediate translation to another disease, that we shouldn't be cooling patients with hemorrhagic stroke.

Dr. Sessler: Okay, perhaps I was too aggressive there. I'm not a neurologist or a stroke specialist. It would personally make me nervous cooling somebody who has a hemorrhagic stroke. Perhaps people who know better would feel comfortable with that. Fortunately, hemorrhagic strokes are rare since 90% are thrombotic.

Question: We are planning to do a stroke study in awake patients in which we will randomize the patients to surface cooling and to endovascular cooling. As far as I know, your volunteer studies were conducted with infusion of ice cold saline. Are you aware of data that compares surface cooling and endovascular cooling in terms of infusion of ice cold saline, and also in the amount of co-medication that you need to get to your target temperature?

Response: I don't know of any studies comparing intravenous cooling with endovascular cooling.

Comment: By intravenous, I mean just giving a few liters of ice cold saline.

Response: I think that's probably physiologically somewhat similar, except that with intravenous cooling you are giving a volume load and you have to stop after a few liters because people get over-hydrated.

Question: So what do you think about surface cooling then? Comparing surface cooling to endovascular cooling? The data might implicate that surface cooling is not as good as endovascular cooling because the side effects might appear more often.

Dr. Sessler: The data are too limited in contrasting endovascular and surface cooling. If you surface cool, you will trigger thermoregulatory defenses because humans sense the skin surface that is their early warning system for thermal perturbations. As you start cooling the surface, people will feel miserably cold, they will vasoconstrict, they will shiver. They will do so less with endovascular cooling, and they will do so even less if you combine endovascular cooling with surface warming, which you can do because surface warming transmits much less heat than the endovascular system. So, if you don't mind being invasive, from the purely thermoregulatory perspective, endovascular cooling is a more sophisticated way to approach the problem in terms of thermoregulatory defenses and patient comfort and the number and amount of drugs that you will need to defeat thermoregulatory defenses.

Question: So, in other words, you think if you use surface cooling, you might end up needing more medication against cold shivering, including side effects of opiates and other medications.

Dr. Sessler: Undoubtedly. For a given amount of core cooling, you will need to give more drugs with surface cooling than with endovascular cooling. I've done endovascular cooling with cold intravenous fluid myself many times, and it's not as bad as you would think. The remarkable thing about it is that you are lying on the table, shivering like crazy—the type of shivering that shakes the table—and it doesn't feel all that cold. Whereas, if you cool the surface even without changing core temperature, you feel miserably cold.

Question: Have you done that yourself?

Dr. Sessler: Yes, I participated in all the volunteer studies I showed you.

Question: Does it follow that an accidental hypothermia victim who is shivering should not be warmed at the surface?

Dr. Sessler: If you have an accidental hypothermia victim who is shivering, it depends on what your immediate goal is. If your goal is to rewarm them quickly, you can leave them shivering, because shivering is very effective and it will rewarm them as fast as surface cooling. If your goal is to control the metabolic stress to make them feel good, then a little bit of surface warming, or in fact aggressive surface warming, will stop the shivering. Because it is transferred and heated, they will rewarm at the same rate. So we actually did a study in which we took volunteers and in a crossover protocol, randomized them to rewarming by shivering or rewarming by surface warming. They rewarmed at exactly the same rate, but there was less stress with surface warming.

Question: I enjoyed the presentations very much. I have a couple of questions. Dr. Marion, can you address what the buy-in of the use of the Military Acute Concussion Evaluation (MACE) screening system has been at the level of the combatants?

Dr. Marion: It is used uniformly, and so I would say there is a great deal of buy in. They are told by their command to use the MACE. However, we need to continually improve on our record keeping and recording, and on completion of all parts of the MACE including the neurologic and symptoms scores. I'm surprised you asked that question.

Comment: Some colleagues of mine had similar experiences. Apparently, they were having a great deal of issues with it being aimed at the point of administration. So, on marches they would, as a march cadence, say the name of the objects they are supposed to remember so that everyone would memorize those objects. So we came up with three new versions. Apparently, those new versions are quite popular on the black market right now, especially with the marines.

Question: My second question: Dr. Sessler, again I enjoyed the presentation very much. It is interesting that meperedine now has a use that is not being condemned. As you know, in some hospitals they tried to actually remove it completely from the pharmacy. It is now starting to show some resurgence, much to some of our pharmacist's disdain. I wanted to ask you if you had any insight on the use of magnesium infusions as a control for shivering, and whether or not you think there may be some additional neuroprotective process from that.

Dr. Sessler: Magnesium is attractive because it is a slight muscle relaxant, so you might think that would be helpful for shivering, either by reducing the shivering threshold or by reducing the gain or maximum intensity of shivering. Magnesium is also believed to be neuroprotective, which would make it especially attractive. We thus formally tested it and, unfortunately, found that it doesn't work at all for this purpose.

Question: I come from an area which is quite cold and where we are used to seeing many winter sports injuries. We have observed within the past 10 or 15 years a change in the pattern of traumatic brain injury because of helmets and what equipment the skiers are carrying. But I don't think that it is only the protective effect of the helmets, because in former years we have seen lots of hemorrhagic contusions, lots of epidural and subdural hematomas. Now because of widespread helmet use, we see a lot of diffuse axonal injury in these patients. So could it be that wearing a helmet is also protecting the head from the cold outside temperature? It is very warm inside the helmet and could this alter the type of injury? So I want to ask Dr. Marion if they have seen a difference in Iraq compared to, I would say, any military injuries seen somewhere else.

Dr. Marion: I think the difference we have seen in Afghanistan compared to Iraq is a lot higher incidence of concussion and just incidents of TBI overall. But I would say not so much the type of injury: concussion has always been 70% to 80% of the injuries. In addition to the outer rigid shell of the helmet, it also is important to consider the padding on the inside that may serve as insulation.

Comment: May I comment on that? I don't think helmet insulation changes brain temperature. Brain temperature is controlled by core temperature, and it makes no difference how you insulate the helmet.

Dr. Marion: Let me just comment about that. Peter Safar did quite a bit of work years ago on dogs, and he showed that you could get some surface gradient temperature differences. I agree with you, you can't get core brain temperature change distinct from body temperature. But Safar first, and then some others I think, Linda Sternau in Miami, did the same thing. Linda had an ICP monitor with temperature thermistors at five different locations. She showed a gradient temperature in the human brain.

Comment: There might be a temperature gradient within the brain, and after an injury you definitely can have gradients because you have poorly perfused parts of the brain. But the normal brain, before injury, is at core temperature. When Zoll was Radiant, their original idea was to put an endovascular catheter into one carotid artery and to cool just one hemisphere. They finally did some calculations and realized that that's impossible, that all you are doing is cooling the whole body through the brain. But if you are going to cool the entire body anyway, it is a lot safer to do it from somewhere besides the carotid artery.

Question: Are there any attempts to treat those complications of hypothermia or prevent them, for example, giving transfusions of fresh frozen plasma (FFP) during the time of hypothermia, Factor 9 for problems of bleeding, or prophylactic treatment of infections by maintaining those patients on IV antibiotics? In terms of the myocardial depression, is there any use for pressors or any other type of medications that could prevent the complications?

Comment: I don't have any clinical experience regarding this question, but knowing what complications hypothermia poses, I suppose it would be prudent for clinicians to anticipate them and try to moderate or prevent them.

Comment: I don't think it's a bad idea to think about combination approaches, maybe directly targeting the temperature abnormality but also possibly a pharmacological agent that may target some detrimental process. So, I don't think that's a bad idea, and that's certainly where the preclinical research is going in terms of testing multiple pharmacological treatments, as well as combining hypothermia with drugs.

Comment: For example, it's quite likely that beta blockers reduce the myocardial risk associated with shivering and hypothermia. There is no evidence for this, but it is quite likely based on pharmacology that it does, and that would seem to be a reasonable thing to try. But currently, there is not a lot of evidence that any of these combinations work.

Question: I have a question for Dr. Sessler. Is there any role for neuromuscular blockers in attenuating the thermoregulatory response?

Dr. Sessler: If your patient is unconscious, then it is not usually difficult to induce hypothermia. In fact, unconscious patients often become hypothermic on their own. So what you are asking about is a patient who has a mild stroke or is otherwise conscious. You can always induce general anesthesia with or without a muscle relaxant and make people about as hypothermic as you want. The more difficult clinical situation is how to take somebody, say with a typical stroke, who would normally be sent to a ward, and make them hypothermic without keeping them anesthetized for maybe a couple of days, which would be a pretty serious effort. So yes, you can block shivering with a muscle relaxant, but that should not be your first approach.

Question: Dr. Sessler, I have a question for you. From the less sheltered classic teaching that spontaneous hypothermia is bad in a perioperative setting, what perioperative indications do you see for hypothermia? Sometimes at hypothermia meetings, we hear it is good for cardiac arrest patients. Where do you see the hypothesis that hypothermia is beneficial in the peri-operative setting or elective operative setting?

Dr. Sessler: The great hope for perioperative hypothermia was for cerebral aneurisms. That was supposed to be the best test since cerebral aneurisms are one of the rare circumstances where you know ahead of time that there will be brain ischemia. Consequently, you could induce hypothermia before the ischemic insult. But as we know, Mike Todd's multi-center trial didn't show benefit. So currently, there are no proven indications for therapeutic hypothermia in the perioperative situation. Brain trauma was another hope but, again, Guy Clifton's large randomized trials failed to demonstrate benefit. Hypothermia may yet prove helpful for other types of neurosurgery, but at the moment, that is completely speculative. In contrast, the potential complications associated with hypothermia are not speculative at all.

Comment: But the use of perioperative hypothermia for cardiac surgery is certainly seen.

Dr. Sessler: Undoubtedly, hypothermia is used in cardiac surgery. But I would point out that the current trend is away from hypothermia. When I started practicing, all patients going on bypass went to 28° and were kept there. Now, almost nobody uses that type of hypothermia. People use 32° to 34°, and there are many institutions that do bypass essentially normothermic. Texas Heart has been doing it for a long time and has as good an outcome as anyone. Also, the Cleveland Clinic now essentially does normothermic bypass for most cardiac operations. That said, hypothermia to about 18°C is still used for deep-hypothermic circulatory arrest cases, during reconstruction of the aortic arch.

Question: Question for Dr. Sessler. A few years ago there was a lot of talk in the media as well as in the scientific literature on assessing small molecules that affect hypothalamic function and induce states of hibernation in people—hydrogen sulfide, for example. What is the status of developing drugs that actually reset the thalamus so that you just think you are supposed to be at 32°?

Dr. Sessler: Are there drugs that can induce some sort of hibernation-like state in humans? In humans, no. In animals, hydrogen sulfide in tiny doses makes rodents profoundly hypothermic, which is attractive. But hydrogen sulfide is also terribly, terribly toxic. So it's not a drug I see very quickly coming into human use, even though the effect in animals is impressive. But there has always been a disconnect between animals and humans. Basically, every animal model in every species shows that hypothermia is beneficial for brain injury, results that have translated poorly into humans, proving once again that humans are a poor model for rats.

Question: I wanted to ask you a question along the lines of what you had been discussing with some of the complications of hypothermia. One of the problems that we have noticed is with patients for whom we are trying to maintain a normothermic situation by continuing the cooling in a post-injury status. After a few years of getting used to looking for fevers as a surrogate for infection, now once I have a patient who is effectively normothermic because of external efforts to cool them, the usual clues that we use clinically to indicate an aspiration pneumonia or some other infectious complication are masked. I'm wondering if you might address some of the issues that might go along there, if you had any experience with trying to identify the infection without the fever?

Dr. Sessler: The question is, if you use endovascular cooling to clamp somebody's core temperature, or for that matter you did it from a surface method, how would you detect infection if you no longer have fever? And so, maybe I could ask some of the folks in the audience, can the endovascular cooling device tell you energy input? Because energy input would be a surrogate for that.

Dr. Marion: Yes, in the study that we did in Pittsburgh, you could tell when the machine was working harder, and that was an indicator of elevated body temperature.

Answer: That's what I would suggest. Use an energy output into, say, an endovascular cooling system as a surrogate for fever, and you can plot it just like a fever curve. And then, you'll have to tell us whether it actually works.

Dr. Dietrich: So Don, has Guy Clifton finally found a pearl in this complicated problem of traumatic brain injury where patients who are cooled early, prior to decompression surgeries, seem to benefit? That might be consistent with Dr. Bush's hypothesis regarding myocardial infarction in terms of targeting reperfusion injury.

Dr. Marion: I do not believe so … Guy's latest notion is that people with diffuse swelling and diffuse injury don't benefit. It is people who have surgical mass lesions that do. He needs to now conduct the sub-trial, including patients who are cooled and have a surgical mass lesion removal.

Comment: I think so, too. Actually, in all honesty, the TBI trial that I think still hasn't been done, and really should be done, is a trial of endovascular rapid cooling.

Dr. Marion: Well, it's not just TBI, it's also for aneurisms and for acute myocardial infarction. The trials that so far have been completed do not support the use of hypothermia for those indications. That is not to say that they proved that it doesn't work. There were serious limitations to all those trials.

Question: For out-of-hospital cardiac arrest or for TBI?

Answer: For TBI, cerebral aneurisms, and for acute myocardial infarction.

Comment: All of those were negative trials, but they all had serious limitations, including the fact that many subjects did not reach target temperature, for example.

Dr. Marion: I interpret that result in two ways. I think that one could also argue that if your treatment wasn't good enough to show an effect with the multi-center trial, with all the trials and tribulations of doing a multi-center trial, and all the differences you see, it's not a very good treatment, or it's not a very robust treatment. And so, although I certainly have read and reread that Journal of Neurosurgery article that Guy wrote a year after the New England Journal study in 2001, I am well aware of the hypotensive differences and the morphine differences and all the other intercenter variations in that trial of which we were a part. To me, a larger problem there was still the 8 to 10 hour delay in getting to target temperature. If you had an endovascular cooling device that could get you there in two hours, and keep you at that target temperature across the board that everyone can reach, that would be an important study to conduct.

Comment: Absolutely, and the same with the aneurism trial. Many of the patients were not at target temperature at the time of the treatment of the aneurisms. With the acute myocardial infarction trial, many of the patients were not cooled to target temperature by the time they did the stenting. So, none of these trials are perfect, but currently available data do not support using therapeutic hypothermia for those indications in humans, but the currently available data doesn't rule it out either. We just have a big paucity of data, and we need good large outcome trials for each of those.

Comment: In Sweden, they are having second thoughts about out-of-hospital cardiac arrest and thinking that they need to take a relook at even that indication.

Comment: Frankly, I think that would be a good idea. As I do more and more studies, I'm increasingly leery about small trials, because small trials so often turn out to be wrong. And while I appreciate the enormous effort that goes into a 200- or 300-patient trial, it is still relatively small when you are looking at a relatively rare dichotomous outcome. The chances of a spurious outcome are non-trivial for these studies. They should be redone with larger populations.

Question: So I have a question for all of you. Are infections a major complication of hypothermia? I ask this question because I have taken some time going through some clinical trials, and the number of pneumonias for example is not stated very well. Pneumonia may be a side effect of hypothermia, or is it an effect of co-medication or medical treatment?

Answer: No, it's probably the hypothermia, per se. For the type of patient who is treated with therapeutic hypothermia, I think it's a reasonable thing to say that infection is the most important of the potential complications.

Dr. Lyden: In planning a large multi-site trial of hypothermia for stroke, we've put in our protocol meperidine, for shivering control. We're having a difficult time with pharmacies in most of our sites because of a fear of seizures induced by these 24 hours of meperidine infusions. I have read the review articles that mention this, and I found the warnings that mention this. However, I'm having a hard time putting my hand on the data that support, in fact, more than isolated case reports of seizures with prolonged infusions of this drug. Have you seen it, and can you help us understand how common it is?

Dr. Sessler: Pharmacies have been eliminating meperidine because there is otherwise not much of an indication for it. It's not the first-line opioid for anything except therapeutic hypothermia. The concern about meperidine is that it is metabolized to normeperidine, which is an excitatory neurotoxin and can cause seizures. When metabolism is slow, it probably takes even more than 24 hours. It probably takes many days of high dose meperidine to cause a problem. I'm not aware that it's been a problem in any of the therapeutic hypothermia trials. Remember, the alternative is not meperidine versus some drug without complications; it is meperidine versus drugs that have different and perhaps worse complications. Certainly, using another opioid is not attractive, because they don't stop shivering nearly as well as meperidine does.

Question: We have a case in which the patient developed seizures after surgery from meperidine and died. We think that it was because he had renal failure. How can we move forward in finding drugs that target sweating, vasoconstriction, and shivering? Do you think we can design drugs to induce hypothermia that will no longer require physical cooling?

Answer: Somebody this morning showed me an animal study of a receptor antagonist that, at least in rodents, seems to work, so there might be better drugs coming. But it will be a difficult sell to develop a new drug for this purpose, because it is a pretty small indication now. So you'd have to find some company that is willing to kick in 100 million dollars and try to get it through the FDA. That's why we looked at drugs that are already approved right now.

Question: Question for Dr. Marion. So the Clifton studies continued to show no benefit in all types of head injury within a certain Glasgow Coma Scale (GCS), yet at the same time, there are the two very large studies that showed the opposite effect? In fact, when you titrated temperature into intracranial pressure (ICP) and then weaned by responsiveness, their outcomes were better. So why should we conclude that the U.S. study, using a finite time in a certain population, was actually the true outcome measure of the effect of hypothermia versus the Chinese study, which showed that when you titrated to ICP and weaned that way, the effect actually showed a beneficial outcome? Could it be just that, again, this wasn't so much a failure of hypothermia as inadequate study design?

Dr. Marion: I think that is a good question, and there are several factors to consider. There are different ideas about how to use hypothermia. One is to try to treat the secondary injury, which is the reason I used hypothermia in my trial and why I think Guy designed the trials he designed the way he did. The second idea is to use it to control brain swelling, and especially for brain swelling refractory to other kinds of therapies. Our Japanese colleagues have used it a lot in that kind of a study design. I think you are talking about the Jiang study in China, which is the large single-center trial of 395 patients. It was slightly larger than the National Acute Brain Injury Study: Hypothermia (NABIS:H) study. They found a very positive treatment effect. There was a statistically significant improvement in the group that got hypothermia in that trial. It wasn't just 48-hour treatment protocols; hypothermia was continued as long as necessary. The Japanese used the same approach and continual cooling as long as necessary, sometimes a couple of weeks. They didn't find a benefit, so it is confusing.

Comment: One of the constants, and I've got a slide tomorrow to show, is most of the single-center studies showed an effect, and most of the multi-center studies in Japan and this country did not. I don't know what the explanation for that is. I don't think it's as simple as a lot of inter-center differences in control of volume or blood pressure. I think it is probably more complicated than that. But again, my approach today would be to look at intravascular cooling and to try to get these people as cold as quickly as possible, and tightly maintain their temperatures with a strict critical care protocol. Strictly monitoring intravascular volume, use of narcotics, cerebral perfusion pressure, and so forth. As for how long you should cool them, I don't know the answer to that. I guess I wouldn't have a problem with titrating that a little bit according to their ICP. I certainly wouldn't recommend discontinuing hypothermia at 48 hours, as you watch the ICP rapidly increase with rewarming.

Comment: I like the question of duration.

Question: Dr. Armando gave a great talk at the National Neurotrauma Meeting this year, and he discussed using hypothermia as an adjunct therapy in select cases targeting, for example, vasospasm. Is that correct?

Dr. Marion: He does occasionally use hypothermia, sometimes for blast injuries that may have associated vasospasm. He thought he saw more vasospasm in blast injury than in non-blast. It turns out, it's not really more common.

Disclaimer

Dr. Marion's comments are his own views, and do not reflect official policy or views of the U.S. Army, the U.S. Dept. of Defense, The Department of Veterans Affairs, or the U.S. Government.