QUESTIONING THE PARADIGM, MICHAEL BELL, MD 1 Good morning to everyone, I wanted to be able to start to give you a talk today on pediatric traumatic brain injury in 2013, something that s called the ADAPT Trial which I ll talk about at the end of the talk, and questioning the paradigm. The goal of the current talk are to discuss the current state of care for children with severe traumatic brain injury or TBI, also to discuss how we might advance the field and how and how we can improve outcomes for children with this illness. I have several disclosures to make, I have no financial interest in any companies. I have been fortunate enough to have several NIH grants funded which are listed here, importantly the last grant on this list is the study we will be talking about during this talk and I ll talk about it in detail. I was a member of the Brain Trauma Foundation s Committee that wrote the Pediatric TBI Guidelines which will be discussed in this talk as well, and my biggest disclosure is I d like to know how to fix brains better. So why is pediatric TBI important? So based on the best CDC estimates from the U.S. there is about 7400 kids a year under 19 who die of traumatic brain injury. To put that in perspective, if you assume the mortality rate was 20% for severe TBI which is the basic rate that s been for most clinical trials, that means more than 35,000 kids per year suffer from severe TBI. In the most recent study that was performed about 50% of those kids had poor outcome at 6 months, and if you assume an average age of 9 years per kid for this disease, and a life expectancy of 78 years the overall productive life years that are affected by pediatric TBI each year is 1.3 million. To put it in a little more perspective this is about 20 children per day who die of TBI, which is about the same number
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 2 who died in the tragedy at Sandy Hook Elementary School and that s every single day. It s also three times the normal people who died in September 11 th terrorist activity in 2001. In contrast about 2900 children die each year from cancer, and so TBI is obviously a very important public health problem. Looking specifically by age group as you can see in this panel the blue panel is unintentional injuries and for the most part about 75% of those injuries are TBI, so it s the highest killer of kids 1 through 4, 5 through 9, 10 through 14 and all the way up to age 44 across in the U.S. If you look at that red bar that s homicides, again a high portion of homicides also occur by TBI and suicides are in green, so TBI in the U.S. is a major public health concern for young people and children in particular. So if your child has a severe traumatic brain injury and comes to Children s Hospital of Pittsburgh of UPMC you ll meet me and my team which includes neurosurgeons, trauma surgeons, intensivists, emergency dept. doctors, rehabilitation specialists, physical therapists, nurses, respiratory therapists, the whole conglomeration of people who come to help get your child through this injury as best as possible. We have a protocol which is listed in this slide which is extremely detailed about what to do in any given instance including which monitors to place to measure intracranial pressure, how to sedate your child, what machines are necessary to help keep your child safe, how to manage ventilation, how to start feeding, how to manage blood pressure and a whole host of medical things that are required to get your child through this injury with the best possible outcome. As you can see
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 3 from this slide it s extremely detailed and we include with every patient the ability to call me and other people on the team to try to get help when necessary. So what is evidence for all of this, this detail that I just talked about? Well really it s based on evidence based guidelines from the Brain Trauma Foundation that s been published both in 2003 and then updated in 2012. In these guidelines which are a panel of 15 experts including 3 statisticians, a forensic librarian and staff at the Oregon Health Science University went through a process of meetings and going over the literature to try to get the best evidence to how to improve outcome. They identified 15 topics that were believed to be highly associated with outcomes from severe TBI, they got peer review of the documents they generated and it was eventually endorsed by 10 societies including neurosurgical personnel, emergency medicine folks, critical care docs, neurological and anesthesia folks. To be included in the guidelines the patients need to be identified as having severe traumatic brain injury which currently is defined as GCS less than 9, identifiable population of children, kids less than 18 years of age, and a measurable relevant health outcome which usually means good or bad outcome based on relevant scores or mortality. Excluded manuscripts were ones that were not in English, if the sample contained less than 15% children or if they were case studies, editorials or expert opinions. The panel made several recommendations including Class I recommendations which are basically the highest recommendations which means that therapy must be done and generally that comes from successful randomized controlled trials or RCTs. Class II
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 4 recommendations are therapies that should be considered and these generally are for either poor quality randomized controlled trials or cohort studies with good controls. Other recommendations were Class III recommendations which was therapies that may be considered and that comes from all of the studies that qualify for the guidelines. The Guidelines Committee reviewed almost 5000 abstracts and 600 manuscripts over from 1966 to the current time and based and eventually led to 37 manuscripts being included in the final guideline. Disappointingly there were no Level I recommendations which means there was no therapies that clinicians could point to to say this thing must be done for every patient with severe TBI, which is a very big weakness in our field. There have been there are 4 Level II recommendations which again are the second strongest recommendations and three of them, outlined in red here, are negative recommendations including that steroids should be avoided, that hypothermia should be avoided and immune enhanced diet should be avoided. These are all from randomized controlled trials which show that these therapies were not helpful. The only Level II recommendation which was a positive was that hypertonic saline which is a salt solution should be considered when intracranial pressure is high. There was several Level III recommendations that are outlined in this slide and reading the wording of the recommendations is actually very instructive I think. ICP monitoring may be considered in severe TBI. Treatment of ICP may be considered at a threshold of 20. These are all very watered down recommendations because the literature is not sufficiently rigorous to compel clinicians to
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 5 follow guidelines and so while our details on how we take care of kids is very detailed the overall guidelines are relatively weak, which is a very big weakness for the field. In particularly disappointing is that the randomized controlled trials of pediatric TBI have almost all failed. The ones listed here were published before 2013, there was a new publication in 2013 which is not included in the slide which I ll discuss in a second; but the only one that was positive was a study by Fisher from Seattle in 1992 and in this study hypertonic saline or 3% saline was shown to decrease ICP compared to normal saline in 18 children so that the data is very weak in order to compel clinicians to change their practices. The additional study that should be added to this list is a study by Dave Adelson from Phoenix Children s and he studied hypothermia which also did not show any benefit. So why do they all fail, why do all these studies fail? Is it just you know what are the arguments for that? So some have argued, without convincing to me anyway, that there are none of the therapies tested in children yet so we just find the right therapy we ll get it correct. There are others who argue that there are few animal models that are relevant for pediatric TBI so we need to get better animal models. There are some that argue that the number of children in RCTs are just simply too small to get an effect. And then there is another argument that there is a learning curve for investigators and pediatric TBI investigators are not experienced enough to do these studies yet. I dispute all of these arguments and basically will point to evidence by Andrew Maas in Europe to point to some of these problems.
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 6 Andrew Maas is an adult neurosurgeon who did a study called the IMPACT study which is outlined in brief here on these next few slides. In that study he looked at all the randomized controlled trials of adults to try to find out which therapies were best, most effective for adult TBI. As you can see in this busy slide there is a number of studies listed in the first column and their date of publication, the intervention is in the second column as to what medication they tried to use to help the patients with TBI, the number of centers is the third column and then it goes on from there across the slide. Importantly look at the fifth column which is the number of patients enrolled in these trials, and these are not small trials. The first study had 160 patients, the second 130, there is studies here that have 1000 patients or more in these studies. The important thing to note about Dr. Maas study is all the studies that had multiple center trials where the study was done over this large group of patients they all failed including this slide and this slide. And you can see in the most extreme case the Methylprednisolone study in the middle of the slide, the Edwards 2005 study had more than 10,000 patients, so we don t think that it s a lack of number of number of patients that are causing these studies to fail. And neither does Dr. Maas. In summary there are some successes in single randomized controlled trials, single center randomized controlled trials but none in multicenter trials. And this is a number of this paper is a few years old and these are a number of studies that are still going on and all of these have failed as well. So we started questioning well maybe the paradigm is incorrect. The general paradigm for medicine at this point is a clinician or a scientist sees a problem in the top circle in this slide and will make
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 7 observations and maybe try novel treatments, try a drug if they it s you know going to help, and if they see some success then they then try to do a standardized way and try to advance to Phase I clinical trial and eventually a Phase II trial which looks at safety, and then if they have some positive results from those trials they advance to a Phase III trial which is trying to establish the feasibility and efficacy of the trial. And then once a Phase III trial is successful then you try and implement that into clinical care. And this whole process tries to improve care throughout this whole process. There is one problem with this paradigm is that it assumes that all other aspects of care are identical except for the intervention. So the reason why Phase III trials fail in my opinion and in other people s opinion is you go from a Phase I trial at a single place where you have a relative homogeneous population where you treat people the same and these other aspects of care are generally controlled and then you add intervention and go to a multicenter trial where things are more chaotic. And so our hypothesis is that this is one of the problems. Dr. Maas IMPACT study looked at 13 large TBI studies in adult TBI victims including about 10,000 patients at 265 clinical centers of which about half the patients had poor outcome and half had good outcome and in his analysis the effect of clinical site as opposed to any other interventions was more important than any other single factor. And disappointingly for European sites the European sites performed worse than the U.S. sites with a 3.3-fold increased risk of poor outcome if you were in Europe compared to the U.S.
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 8 Dr. Maas took 3 years to combine the data from those studies due to imprecise definitions. So the NIH initiated effort called Common Data Elements which has now been published in both adults and pediatric TBI studies which includes core elements, supplemental elements and experimental elements so every study will have the same definition so you can combine your studies more readily and we are hoping that s a big advance in the field in the coming years. But our adult TBI colleagues have started to use this to try to redefine TBI stratification and they ve looked at this coordinated element set which is this is a website from the IMPACT website and you can then use these to make your definitions and improve outcome for adult TBI. So what about kids, maybe kids are not the same, maybe the outcomes are not different across different places in the U.S. Well we dispute that too. This is a study of 11 U.S. states showing that the outcomes from TBI are vastly different across the different states. You can see in the top panel in the red bars which is mortality rate for all children with pediatric TBI admitted to these states, if you are in New Jersey your mortality is about 1/5 of the mortality in North Carolina for whatever reason. If you look at the green bars mortality rate in pediatric patients with ventilation, which is another way to try to stratify kids who are a little more sick, again there are different mortality rates differ by 2.2 fold over the different states, and mortality rate for children with pediatric TBI and coma which is the bottom graph again shows marked variation between states. So the outcomes are different depending on where you get hurt.
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 9 What are the goals for various institutions? We talked about before how different goals can be affected by TBI, so we did an international survey of 32 clinical centers across the U.S. and Europe and found out that most sites target an ICP of about 20 and a lot of sites have age related CPP goals which cerebral perfusion pressure goals. About half the centers use CSF diversion to lower ICP, some use it continuously, some use it intermittently. About 30% use new PbO2 monitoring technology and no one thinks they hyperventilate patients. When it comes to nutrition and glucose the variability is marked in that there are some centers that use nutrition and glucose in the Emergency Dept. and some don t do it for more than a week. So this variability we believe has an effect on the randomized controlled trial s ability to find a difference in outcomes between the groups. Looking at a randomized controlled trial this is the data from the Cool Kids Trial looking at how the therapies differ between two groups, so CSF diversion, hyperosmolar therapies, prophylactic hyperventilation, PbO2 monitoring, nutritional start time, glucose start times were all variations that were all parts of the guidelines which show some variability. And as you can see from this table which is pretty busy the variation was quite marked across the different experimental groups. So what s our approach and what do we suggest that might be helpful? Well we always thought that insanity was doing the same thing over and over again and expecting different results and that came from Einstein; so we tried something different. So this is our basic protocol which we outlined in the first few slides of this talk and again it s very detailed. So we started thinking about how can we
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 10 advance the field? We thought intracranial pressure management seems pretty important and might be linked to outcomes so maybe that should be something we should study. We thought about Hypertonic Saline and Mannitol, they are both used for similar purposes, maybe we can find out which one is better. We looked at cerebrospinal fluid diversion as a tool to treat ICP, maybe that s worth the risk, maybe it s not, maybe that s a question we can answer. Hyperventilation seems bad but what is the proof for that? We thought about monitoring brain with brain oxygen monitors but what s the evidence for that? Nutritional support as I outlined a few slides ago seems like a good idea but how much is enough, when should you start? And glucose seems to be associated with outcome as well, how should we manage that? So these are all things that were going through our heads at the time and instead of looking at a randomized controlled trial which I outlined before earlier in this talk that have all failed, more than 40 have failed in adult TBI and 8 out of 9 have failed for pediatric TBI, instead of designing a randomized controlled trial where you hope clinical sites will perform the basic care in a standard way so that the experimental group can show differences, and you exclude children who might introduce variability we chose another technique called Comparative Effectiveness Research. In this sites are enrolling patients are enrolling patients in their own way, they take care of kids by their own paradigm and we use statistics to figure out which therapies are associated with best outcomes and they allow for a large sample size to get significant statistical power. So that s been our approach. So in this study which was once called the Multiple Medical Therapies for Pediatric TBI Proposal, and how it s called the ADAPT Study where approaches and decisions for acute PF TBI we are
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 11 looking to enroll 1000 patients in 32 sites across the U.S. and Europe. We have three aims and six sub-aims that will be tested including which CSF diversion technique is associated with best outcome; which hyperventilative therapy is associated with best outcome, is iatrogenic ischemia from hyperventilation without ICP a problem and how to treat hypoxia with PbO2 monitors; and also nutrition strategies and glucose strategies that we discussed on the previous slides. Our primary outcome is going to be a neurological outcome at 6 months along with a neuropsychological battery at 12 months. Our primary team includes me as the lead person for the clinical part and my partner in crime, Steven Wisniewski who is Associated Dean of the Graduate School of Public Health at the University of Pittsburgh. We ve assembled TBI experts across the world including Dave Adelson from Phoenix, Jamie Hutchinson from Toronto, Pat Kochanek from here in Pittsburgh, Robert Tasker from Cambridge UK as well as Boston and Monica Vavilala from Seattle to be the experts to try to adjudicate these strategies and try to figure out which is the best for kids going forward. We have statistical support from a variety of people including Tony Fabio, Cheryl Kelsey and Jerald Greenhouse. And we ve enlisted the help of sites from across the U.S. and across the world to try to enroll 1000 patients in this study in the next several years. They are all listed here, they are all valuable contributors and we are looking forward to working with all of them as this project goes forward. Importantly, we ve got initial indications from the NIH in the spring of 2013 that this will be funded, we hope to get this study started in the summer or fall or as soon as possible to try to help these children out as best we can.
QUESTIONING THE PARADIGM, MICHAEL BELL, MD 12 I d like to thank you for the attention. This is a picture of our beautiful city as well as our beautiful hospital on the bottom right corner, and thank you for all of your time.