SGEM#336: You Can’t Always Get What You Want – TTM2 Trial

Date: July 1st, 2021 Guest Skeptic: Dr. Justin Morgenstern is an emergency physician and the creator of the #FOAMed project called First10EM.com. Reference: Dankiewicz et al: TTM2 Trial Investigators. Hypothermia versus Normothermia after Out-of-Hospital Cardiac Arrest. NEJM 2021 Case: A 58-year-old man collapsed in front of his family. When paramedics arrived, they found him to be in cardiac arrest, with ventricular fibrillation on the monitor. Paramedics managed to get return of spontaneous circulation after a single defibrillation, but the patient is still comatose on arrival. The charge nurse turns to you and asks: should I grab the ice packs? Background: Hypothermia has been a mainstay of post-arrest care after the publication of two trials in 2002 that both suggested a benefit. This trial by Bernard and colleagues randomized 77 patients with an initial cardiac rhythm of ventricular fibrillation who had achieved return of spontaneous circulation (ROSC) but were persistently comatose. The trial was not truly randomized, as the groups were based on the day of the month, and they also weren’t blinded. The results of this Australian trial seemed too good to be true. Hypothermia resulted in a large improvement in neurologic outcomes, defined as well enough to be sent home or to a rehab facility. It was 49% of the hypothermia group versus only 26% of the normothermia group. This gives a NNT of 4. The reported p value was borderline at 0.046, and when I re-calculate, it comes out as 0.06 (not statistically significant). The other trial was the Hypothermia After Cardiac Arrest (HACA), also published in NEJM 2002. They randomized 273 comatose adult patients out of 3,551 screened patients. These were witnessed OHCA who had a shockable rhythm, achieved ROSC, and had a short downtime. This trial used an air mattress to cool patients and was also not blinded. This second trial done in Europe also showed impressive results for favorable neurologic outcome. It was 55% in the hypothermia group vs 39% in the normothermia group (NNT 6). They also reported a 14% absolute decrease in mortality with therapeutic hypothermia post-OHCA. As a result of these two-small trials, hypothermia was widely adopted. However, there were many voices in the evidence-based medicine world that reminded us of the significant uncertainty that remained, and the weaknesses of these two trials. The SGEM covered a few  trials looking at therapeutic hypothermia for OCHA in the pre-hospital setting. The bottom line is there is not good evidence that therapeutic hypothermia is superior to usual care and cannot be recommended. SGEM#21: Ice, Ice, Baby (Hypothermia post Cardiac Arrest) SGEM#54: Baby It’s Cold Outside (Pre-Hospital Therapeutic Hypothermia in Out of Hospital Cardiac Arrest) SGEM#183: Don’t RINSE, Don’t Repeat Because of that uncertainty, a much larger, multi-center trial was run. This is the original Target Temperature Management (TTM) trial by Nielson et al NEJM 2013. As almost everyone knows, they compared two difference hypothermia targets, 33C and 36C. The result was no benefit for their primary outcome of mortality at the end of the trial and no benefit Cerebral Performance Category (CPC), modified Rankin Score (mRS) or mortality at 180 days. We did a structured critical appraisal of the TTM trial on SGEM#82. The bottom line was that the trial did not demonstrate a benefit of a targeted temperature of 33C vs. 36C for survival of OHCAs. But both groups in the TTM trial were hypothermic, so although it was the highest quality evidence available, it didn’t tell us whether hypothermia was any better than normothermia. Which is why the TTM2 trial was performed. Clinical Question: Does hypothermia result in improved survival after cardiac arrest as compared to controlled normothermia? Reference: Dankiewicz et al: TTM2 Trial Investigators. Hypothermia versus Normothermia after Out-of-Hospital Cardiac Arrest. NEJM 2021 Population: Comatose adult patients admitted to hospital after an OHCA, irrespective of initial rhythm. Intervention: Hypothermia with a target temperature of 33 degrees Celsius, maintained for 28 hours, and then slowly rewarmed by one degree every three hours. Comparison: Normothermia with an aim to maintain a temperature of 37.5C or less. Outcome: Primary Outcome: Death from any cause at six months Secondary Outcome: The main secondary outcome was a poor functional outcome at six months, defined as a score of four to six on the modified Rankin scale (mRS) Authors’ Conclusions: “In patients with coma after out-of-hospital cardiac arrest, targeted hypothermia did not lead to a lower incidence of death by 6 months than targeted normothermia.” Quality Checklist for Randomized Clinical Trials: The study population included or focused on those in the emergency department. Yes The patients were adequately randomized. Yes The randomization process was concealed. Yes The patients were analyzed in the groups to which they were randomized (ITT analysis). Yes The study patients were recruited consecutively (i.e. no selection bias). Yes The patients in both groups were similar with respect to prognostic factors. Yes All participants (patients, clinicians, outcome assessors) were unaware of group allocation. No All groups were treated equally except for the intervention. Unsure Follow-up was complete (i.e. at least 80% for both groups). Yes All patient-important outcomes were considered. Yes The treatment effect was large enough and precise enough to be clinically significant. No Results: They assessed 4,355 patients for eligibility in the trial and randomized 1,900. The mean age was 64 years and 80% were male. Most (91%) were bystander witnessed arrest, bystander CPR was performed in 80% of the arrests and three-quarters had shockable rhythms. Key Result: No statistical difference between hypothermia vs normothermia Primary Outcome: At six months, 50% of the hypothermia group and 48% of the normothermia group had died. This gives a relative risk (RR) of 1.04 (95% CI; 0.94 to 1.14); P=0.37 Secondary Outcomes: At six months, 55% of both groups had a poor functional outcome (mRS 4-6). RR 1.00 (95% CI; 0.92 to 1.09) Harms: Arrythmias resulting in hemodynamic compromise were more common in the hypothermia group (24% versus 16%, p<0.001) 1. Blinding: Keeping everyone perfectly blinded would have been very difficult to do in a hypothermia trial, so it makes sense that this is an open label study. However, the lack of blinding certainly increases the risk of bias, especially for more subjective outcomes like neurologic function. We are unsure which direction the bias would have been. Did the clinicians and researchers think that hypothermia was going to have a benefit or not? It would have been interesting to survey them before the trial to see what their thought of this treatment a priori.  2. Death: This outcome is not as objective as we sometimes think. When a study is unblinded, you generally want to see an objective outcome being used. Death is about as objective as it gets – you are either alive or dead. There generally isn’t much debate. However, in a modern ICU, we have a lot of control of when someone dies. If the doctors truly believed in hypothermia, they might have decided to keep people alive longer, to give them a better chance, biasing the results. These authors tried to account for that by ensuring that an independent, blinded physician performed a standard assessment for neuroprognostication on all patients. Although it is worth discussing, I don’t think bias is playing a hug role in the results we are seeing here. 3. Harms: Although I anticipate that many people will focus on the harm from arrythmias, I don’t think it was ultimately all that important. Arrythmias resulting in hemodynamic compromise certainly sound important, but ultimately that is a monitor-oriented outcome (MOO). The reason we care about arrythmias is that patients might die or have poor neurologic outcomes, and those patient-oriented outcomes (POOs) that are more important. In this trial the POOs were not different. So ultimately, I don’t think this is truly a harmful practice, but rather a neutral practice. 4. Control Group: Did they choose the best control group? Both groups in this trial had their temperature controlled to some extent, but was that necessary? Although it is a common hypothesis that avoiding fever is important in this patient population, there are no RCTs that demonstrate a benefit from fever control. This study still leaves us with the question of whether any temperature control is necessary. Adding a third group without any temperature control could have settled this issue once and for all. 5. Importance of Skepticism: This story arc around therapeutic hypothermia for OHCA is a great reminder of why we should be skeptical of any claim. The time to accept a claim is when there is sufficient evidence. There are many examples in medicine that show we can fall prey to intervention bias (Foy 2013). We can be overly optimistic about small studies that are nowhere close to scientifically definitive, and quickly crown borderline practices as “silver bullets” or “standard of care”. Think thrombolytics for acute ischemic stroke, TXA for just about anything, or cardiac stress testing. Science is a process that requires replication. Single studies rarely, if ever, define the “truth”. The overzealous adoption of medical practices can hurt our patients. Time and time again, we adopt medical interventions that are unproven. In many cases, the harms end up outweighing the benefits, but we only acknowledge this decades later, when replication studies are finally done.