Therapeutic Hypothermia in Cardiac Arrest

During the 2018 Therapeutic Hypothermia and Temperature Management Meeting in Miami, a state-of-the-art round table discussion was presented on current advances of therapeutic hypothermia in cardiac arrest. Dr. Marko Noc, from the Center for Intensive Internal Medicine, Ljubljana, Slovenia moderated this interesting session. Dr. Hans Friberg from Lund University spoke on the current TTM2 study that will compare 33° hypothermia treatment with fever control in postcardiac patients. Dr. Chien-Hua Huang from National Taiwan University, Taipei, Taiwan provided a lecture on therapeutic hypothermia in postresuscitation in Asia providing very interesting registry data. Finally, Dr. Philip E. Empey from the Center for Clinical Pharmaceutical Sciences, University of Pittsburgh delivered an important lecture on the effect of targeted temperature management on drug deposition and response from cardiac arrest. Each of these three presentations was outstanding and generated considerable discussion during our question and answer session.

Question: First of all, congratulations on getting TTM2 launched. I think we are all excited about it. A lot of people are going to raise eyebrows about 37.8°C as potentially too high. Why not choose 37°C and stay with that and clamp the patients there because you are using a temperature controlled device. Second, you alluded to more conservative stopping rules with withdrawal of active therapy based on neurological features. I would love to hear more about that because I am not sure if I missed that in the talk. My last question is that I am using 36°C but many people are still using 32°C and I know there will be some debate about that later. What do you think about the possibility that people don't think it very sexy enough to go to 36 ° C because it doesn't feel like you are doing enough. Could you comment on that?

Dr. Hans Friberg: All three good questions. First question, why not stick to 37°C? This issue of what the control arm should look like was extensively discussed in our steering group. We have a very democratic system in our research group. There were those who proposed that we should go all the way and redo the Hypothermia After Cardiac Arrest (HACA) trial. There were strong proponents of that but there was also strong opposition. There were those who suggested what you have suggested. We talked at length at designing the TTM2 trial. At the end, it was a majority decision. There is no true answer. Also, the TTM2 trial will not be the last study investigating what target temperature is the optimal one. So for now, we agreed in the steering group on this design. We identified 37.8°C as the lowest definition of fever that we could find in the literature. That is the trigger. If you reach 37.8°C then you set the patient temperature at 37.5°C, which can certainly be discussed. Second question, regarding neuroprognostication in the TTM2 trial, the difference compared to the TTM1 trial is not huge. We assess neurological prognosis at 96 hours postcardiac arrest, performed by an independent physician not involved in direct patient care. The decision to prolong or withdraw care will be done by the clinical team no earlier than 96 hours after the arrest. In addition to deep coma without improvement, you will have to fulfill strong criteria for poor outcome to make a decision on withdrawal of care, just like in the TTM1 trial. It is not enough to have a loss of N20 potential in your somatosensory evoked potential (SSEP), lack of brainstem reflexes or lack of pupillary reflexes or diffuse ischemic injury in the MRI or CT, but you will have to fulfill at least two of these strong criteria for poor outcome to make a decision with regards to withdrawal of life support, in addition to deep coma without improvements. In addition, all criteria must point in the same direction, that is, indicating a very poor prognosis. This is a more conservative approach than we had in the TTM1 trial. Third question, I agree with you. There is so much emotion in this field. I am confident that you can choose either temperature, 33°C or 36°C, and it won't make a difference. Thirty-six degree celsius is probably easier to deliver but that doesn't matter either. As long as you treat strictly according to a protocol and follow that protocol you will find no difference between 33°C and 36°C. The TTM1 trial was a very good trial, but not perfect. We will try to do an even better TTM2 trial. For example, we will try to initiate more rapid cooling because that is potentially an important issue. We could discuss that too. Does postischemic cooling have an effect? This is something that we need to look into. What is the evidence for postischemic cooling? In the myocardium, postischemic cooling does not have an effect if you initiate cooling after reperfusion of the coronary arteries. Whether this is different in the human brain we don't know. We know postischemic cooling is neuroprotective in the rat but it has not been clearly shown in animals larger than the rat at clinically relevant delays.

Dr. Marko Noc: Hans, just one question regarding the speed of cooling. You mentioned quite properly that the cooling would be faster than in previous trials. How are you going to achieve that? Intravascular cooling?

Dr. Hans Friberg: We had hoped we could do that. There will be sites, for example the Essex Cardiothoracic Centre, where intravascular intervention will be used in all patients. We don't have the resources to do very rapid cooling in all patients like in the Acute Myocardial Infarction (AMI) trial. We are doing the best we can. We paralyze all patients in our unit. We give large amounts of cold fluids. We put patients on ice and cold air and activate a device quickly.

Question: Thank you for your presentation. I think that the most important thing should be the duration/maintenance. Why did you choose the 40 hours duration of cooling? Are you following the HACA trial or do you have evidence for the duration time?

Dr. Hans Friberg: We have to choose a time. Now we know there is the trial comparing 24–48 hours. There is a nonsignificant trend of benefit of prolonged cooling in that small trial but to show a difference in a trial like that you would have to include 4000–5000 patients. We have 28 hours (24 hours plus 4 hours) and a total of 40 hours for the intervention period, which includes 12 hours of rewarming at 0.33°C per hour. So we have decided to stick with the 24 hours at target temperature in accordance with the present guidelines with slower rewarming rate than in the TTM1 trial.

Dr. Chien-Hua Huang: When we tried to apply the TTM1 trial in our clinical practice, from your evidence, I mean the study also has some inclusion criteria in the TTM trials, you don't include unwitnessed patients, we do find significant numbers in this category. Another thing I wanted to say that in the TTM1 trial, you have the exclusion criteria of patients with systolic BP less than 80 mmHg.

Dr. Hans Friberg: In the TTM1 trial we excluded patients with hemodynamic shock that could not be reversed within the inclusion time. We don't exclude these patients in the TTM2 trial.

Dr. Chien-Hua Huang: So how low is the blood pressure in the TTM2 trial?

Dr. Hans Friberg: Low blood pressure is not an exclusion criteria in the TTM2 trial.

Dr. Chien-Hua Huang: In the TTM1 trial, most patients started cooling within 4 hours from arrest.

Dr. Hans Friberg: No, this is something that some people have misunderstood but the fact is that the median time to inclusion was 130 minutes.

Dr. Chien-Hua Huang: That is inclusion but how about induction?

Dr. Hans Friberg: The median time to inclusion and induction of cooling was 130 minutes, a temperature below 34°C was reached after another 150 minutes, but that can probably be improved too.

Dr. Chien-Hua Huang: How about witnessed asystolic patients?

Dr. Hans Friberg: Witnessed asystolic patients were included but we have very few asystolic patients who survive. Do they survive in your unit?

Dr. Chien-Hua Huang: Yes, it is relatively few.

Question: I have a question for Dr. Friberg. Until patients are randomized to the study, will they be initiating cooling from the emergency department (ED) until they get randomized, then will they stop cooling depending on group placement? How do you expect that time of admission to randomization to work?

Dr. Hans Friberg: This is a very relevant question. This is a pragmatic trial so it is up to each center how they organize this. As long as they document what they do, I would say that inclusion and randomization in the ED is the way to go and to initiate the treatment immediately. We have stressed this very much in the TTM2 trial that you should initiate cooling rapidly in the 33°C-treatment arm and that this should be done in the ED.

Dr. Marko Noc: One question related to what Dr. Empey showed. You mentioned that you have neuroprognostication at 96 hours and you have two groups, one is 33°C and the other is 37.8°C. In view of his presentation describing decreased drug clearance including sedatives/analgesics in hypothermia, is there not a bias against the hypothermia group if you do neuroprognostication so early?

Dr. Hans Friberg: I think the previous lecture was very relevant. We plan to perform a predesigned sub-study in the TTM2 trial where we actually do collect blood samples. We will have blood samples at the end of the intervention at 40 hours, at 72 hours, and at 96 hours, which is the time of prognostication. Importantly, prognostication at 96 hours does not mean that you withdraw care! This is only the time that you make a structured neurological assessment, performed by an external and blinded physician. Then the treating team based on that assessment can make any decision they find feasible. However, to make a decision of withdrawal of care, there are very strict criteria that must be fulfilled.

Dr. Marko Noc: For example, in our center, we admit about 100 out-of-hospital cardiac arrests per year and cool to 33°C. We published two years ago as an abstract showing that by day 5, 20% of our comatose survivors who were a CPC 3/4, ultimately walked out of the hospital. This was in a way scary for us because if you stop active treatment too early, you will lose these patients. I really appreciate your pharma talk because we are now starting to think on delayed sedative clearance as possible reason for delayed awakening.

Dr. Hans Friberg: I agree with you.

Comment: I can just add that we talked about this a little bit yesterday how gender, sex, and age can affect the distribution of drug into the patient. I wonder about the pharmacokinetics and temperature sensitivity in different sexes and ages. In some of the pediatric hypothermia trials for example, has that been looked at as critically as it really needs to? I think based on your recommendation in terms of more measurements and utilizing some of your modeling systems may help in this regard.

Dr. Philip E. Empey: Thank you for that. I think it is important in regards to prognostication. I think it is just making all of the clinicians who are taking care of these patients aware of these considerations. On the research side, I think collecting the data is essential. We are certainly aware of differences in ethnicity and sex in pharmacokinetics. I think ethnicity is boiling down to looking more deeply at genetics. If you just look at the clopidogrel examples, our institution is systematically genotyping in the cardiac catheterization lab now. I do think we need a greater appreciation for the variability in drug response. For example, with the majority of drugs that I talked about this morning, we do not perform routine therapeutic drug monitoring. We do not follow midazolam levels, fentanyl levels, propofol levels, or dexmedetomidine levels. I think in the future we will have models that will be able to predict levels with some accuracy because they have been tested and validated or we will end up doing more therapeutic drug monitoring. I think it is prudent. Pharmacokinetic modeling may help you know where your patient is regarding drug exposure.

Question: With the concerns about the TTM1 trial and the major limitation at least in my mind, the external validity, and the fact that TTM1 patients were pretty healthy compared to what I typically see, is there anything being done to try and expand or try to prevent that weakness in the TTM2 trial to try and be more inclusive of a sicker patient population who has been down for 20 or 30 minutes before anyone gets to them? In terms of pharmacokinetics, given that we have a lot of information to come in future trials, what do you personally recommend now for medication adjustments for example, if you look at the European Association of Endoscopic Surgery (EAES) table and recommended medication doses, are you recommending any changes based on target temperature or do anything pragmatically now with the limited information that we do have.

Dr. Hans Friberg: With regards to external validity, circumstances differ a lot between different countries. We had different outcomes between different countries in the TTM1 trial. We have done those analyses and looked at the outcome between countries and they differ. With regards to what patients will be included in the TTM2 trial, there are similar inclusion/exclusion criteria as in the TTM1 trial. One difference is that we will not exclude patients in shock in the TTM2 trial. So, there will be sicker patients who will be included in the TTM2 trial with regards to hemodynamic failure.

Dr. Philip E. Empey: With regard to medications, we haven't made any specific changes. I think that the biggest realization is that standardized dosing and standardized caps on dosing are probably inappropriate. For example, when we first started using dexmedetomidine, we had hard-coded limits and now we know we can exceed those in many patients without having bradycardia concerns. I think, in general, for drug therapy in the intensive care unit we need to think more individualized. There are some medications that we have to be careful about. There are medications that when a pathway of elimination is altered could have a different side effect profile, so watching for those is important.

Question: Dr. Huang, in your studies where you looked at mean arterial pressure and cardiac index, were those measurements done with or without pressors and ionotropes onboard?

Dr. Chien-Hua Huang: All of these patients received vasopressors and ionotropic agents as needed. In our study, we also put a cardiovascular sequential organ failure assessment (SOFA) score similar to that used in the TTM trial. The TTM trial also had results about hemodynamic response and they also found cardiovascular SOFA score and the dosage of vasopressor is related to outcome. In our study we found that cardiac output, heart rate, and mean arterial pressure are even more important than cardiovascular SOFA score. The difference may come from a different study population but these are prospective observational studies. I think that is the initial step but maybe the next step will be to try and do some design to do the interventional study to try to keep the cardiac output or the blood pressure to see whether you improve the outcome or not.

Dr. Marko Noc: Additional question. I really appreciate your work on the hemodynamics in postresuscitation shock that is pathophysiologically different from cardiogenic shock in acute coronary syndrome (ACS) without cardiac arrest. The balloon pump in the Intra-Aortic Balloon Pump (IABP) Shock trial that included about 40% of patients with previous cardiac arrests was negative. I noticed on one of your slides that a proportion of patients on the balloon pump was 40%. Not that I would not agree with such strategy because we do the same, but not 40%, probably around 10–15%, so how do you explain this 40%?

Dr. Chien-Hua Huang: I think the 40% is from the study in Japan (J-PULSE Hypo study) and they had a combined treatment of extracorporeal circulation support with balloon pump, so they have the high percentage. In our hospital, when a patient goes to extracorporeal membrane oxygenation (ECMO), I think only 10% need an additional balloon pump. For routine postcardiac arrest, with adequate hemodynamics, systole above 90, we don't use the balloon pump.

Dr. Marko Noc: So pretty much you are using IABP in ECMO patients for left ventricular (LV) unloading.

Dr. Chien-Hua Huang: Yes

Dr. Hans Friberg: Again, with regards to external validity, one issue of importance is the frequency of bystander cardiopulmonary resuscitation (CPR) that is known to be strongly associated with improved survival. I was impressed to see that in your comparison between Asian countries, Taiwan differed from the rest. Taiwan had a bystander CPR rate similar to the Scandinavian rate, close to 70%. How did you manage to get the general public to perform CPR at such high levels?

Dr. Chien-Hua Huang: Around 70% is for the population receiving TTM but for the whole cardiac arrest patients, the number is around 36% for bystander CPR. It is not easy to increase bystander CPR. About 10–15 years ago, we had only 20%. We have increased our public outreach for individuals to be trained in CPR.

Dr. Marko Noc: I think this is very important because whatever study I read, basic life support comes up as an independent predictor of good outcome, despite everything we do in postresuscitation medicine. The message to intensively work on lay knowledge and performance is very clear.

Dr. Chien-Hua Huang: I have a comment about witnessed asystole. We still have a percentage of patients who are unwitnessed or presenting with asystole. There are more asystole patients in Asian countries, so I think that is the difference.