Article Text
PDF
Abstract
Each year, approximately 2.5 million people experience some form of traumatic brain injury (TBI) in Europe. One million of these are admitted to hospital and 75 000 will die. TBI represents a major cause of death and disability, particularly among those of working age. Substantial investments have been made in an effort to improve diagnosis, management and survival in TBI, but with little success. The Collaborative European Neuro-Trauma Effectiveness Research in TBI (CENTER-TBI) study promises to use the natural variability seen in the management of TBI across Europe with the application of Comparative Effectiveness Research (CER). It will generate repositories of baseline and comprehensive TBI patient data, neuroimaging, neurogenetics and biomarkers, which aim to improve the diagnosis, stratification, management and prognostication of patients with TBI.
- TRAUMA MANAGEMENT
- Neurogenetics < NEUROLOGY
Statistics from Altmetric.com
Background
Each year, approximately 2.5 million people experience some form of traumatic brain injury (TBI) in Europe. One million of these are admitted to hospital and 75 000 will die.1 Many of these patients are young, and TBI is the most common cause of disability in young adults at a productive phase of their lives. Consequently, TBI represents a major cause of death and disability, with a high burden of suffering and cost, both on the individual and society. There has been a substantial investment of effort, both by public and commercial bodies, in an effort to improve the diagnosis and management of TBI. Outcomes from TBI have been documented for over a century, and improvements in systems of care and initial resuscitation have yielded substantial improvements in outcome over this period. However, when changes in outcome are examined over a shorter perspective of the last 25 years, the conclusions are different. The majority of research resource investment in TBI occurred over this 25-year period, with billions of dollars being spent on research and drug development. These investments resulted in substantial advances in our scientific understanding of the disease, and triggered the conduct of several high-quality trials (Table 1) based on extremely encouraging preclinical results. Unfortunately, none of these efforts has resulted in appreciable improvements in outcome.2
Recently investigated interventions in management of severe traumatic brain injury (TBI)
Several commentators have attempted to explain the mismatch between effort and improvement in clinical outcome, and several common themes have emerged.9 First, there is an appreciation that our current classification of TBI into mild, moderate and severe categories using the GCS provides a useful basis for triage, but the lack of pathoanatomical characterisation in this classification means that patients with perhaps widely differing disease mechanisms may be inappropriately lumped together in single trials. While calls have been made for using other schemes to assess coma level, such as the FOUR Score, these calls miss the critical consideration that categorisations based on coma depth (as a surrogate for TBI severity) cannot account for differences in disease mechanism. Emerging approaches to precision medicine seek to divide TBI based on mechanistically homogeneous endotypes, which may respond to targeted therapies regardless of TBI severity.
Second, there is a recognition that our conventional assessment of outcome, typically using the dichotomised five-category Glasgow Outcome Score (GOS), which is frequently used in randomised controlled trials (RCTs), may simply be too insensitive to detect small but meaningful improvements in outcome, especially the enormous burden of physical, psychiatric and psychological morbidity observed in many individuals who are deemed to have a ‘favourable’ outcome. While the modification of the GOS to the eight-category Extended GOS has provided some benefit, this scheme can still miss subtle but disabling cognitive and psychological problems, which preclude return to normal life and societal reintegration.
Third, there is recognition that our assessment of outcome needs to be matched to disease severity and likely prognosis, with the hurdles for defining treatment success scaled in proportion to expected outcome. These issues are addressed by more refined statistical approaches,10 such as proportional odds analysis and sliding dichotomy, which relate observed to expected outcome (measured using newer and more accurate prognostic schemes such as the CRASH and IMPACT prognostic schemes).11
Improvements in disease characterisation and prognostication, based on modern neuroimaging, genomics, physiological characterisation and protein biomarker measurements, provide enormous opportunities and allow us to address the key issues discussed above, and aid Precision Medicine12 approaches to TBI, consistent with a concept recently articulated by the US National Academy of Sciences. The application of targeted therapies to such well characterised patient subgroups can be tested using the emerging paradigm of Comparative Effectiveness Research (CER)—the generation and synthesis of evidence that compares the benefits and harms of alternative methods to prevent, diagnose, treat and monitor a clinical condition or to improve the delivery of care.13 There is a clear need for a paradigm shift in clinical research in TBI that takes account of the issues that are outlined above; the Collaborative European Neuro-Trauma Effectiveness Research in TBI (CENTER-TBI) study1 ,14 seeks to address this need.
CENTER-TBI trial
CENTER-TBI is a multicentre, longitudinal, prospective, observational trial that will generate a dataset of 5400 patients from across over 60 centres in more than 20 countries.1 The main aims of the study are:
to better characterise TBI as a disease, and describe it in a European context
to identify the most effective clinical interventions for managing TBI
to undertake CER.
Data collection
Patients with a clinical diagnosis of TBI who meet criteria for CT of their brain within 24 h of injury are eligible for enrolment into the CENTER-TBI core study. Patients with pre-existing neurological disorders that will confound outcome assessments are excluded. Following informed consent, participants’ data will also be entered into the CENTER-TBI Core dataset and repositories (blood, CT and MRI) throughout participation, to enable the use of data for legacy research. Broader registry level data collated for 15 000–25 000 patients will be entered into the CENTER-TBI Registry dataset to assess internal generalisability and representativeness of the core sample within participating centres. The CENTER-TBI Registry will be compared with participating national trauma registries (such as the UK Trauma Audit and Research Network; TARN) to examine the external generalisability of the CENTER-TBI study.
The enrolled patients are recruited into three strata (n=1800 in each stratum), based upon clinical management pathways, rather than conventional subdivisions of mild, moderate and severe TBI (based on admission Glasgow Coma Score (GCS)). These strata are:
ED Stratum—patients seen in the emergency department (ED) and subsequently sent home
Admission Stratum—patients admitted to the hospital, but not to intensive care
Intensive care unit (ICU) Stratum—patients admitted to the ICU during their hospital stay.
Follow-up for outcome assessment is performed as a cross-sectional study at the primary outcome time points (2–3 weeks, 3 and 6 months for ED stratum and 3, 6, 12 and 24 months for admission and ICU strata) and patients undergoing longitudinal MRI studies receive repeated neuropsychological testing. Longitudinal follow-up is administered by postal questionnaire or web-based completion. Recruitment is capped at 250 patients per centre (maximum 100 patients per stratum) to prevent over-representation. The data collected include baseline demographics and socioeconomics, medical history and drug history, mechanism of injury, prehospital clinical course, abbreviated injury score, injury severity score, brain CT within 24 h, ED, admission and/or ICU clinical course information, surgery and neuromonitoring data, reasons for clinical decision making, physician-based satisfaction with clinical care and prognostic estimates, admission and discharge data, discharge destination and acute care outcome evaluation.
Blood for genetics and baseline biomarker analysis is obtained within 24 h of TBI and further biomarker sample banking occurs depending on recruitment stratum and clinical course. ED stratum patients have blood biomarkers assessed at presentation, 2–3 weeks and 3 months. Admission stratum patients have blood biomarkers assessed at presentation then at 6, 12 and 24 months post injury. ICU stratum patients have blood biomarkers assessed at presentation, days 1, 3, 4, 5 and 2–3 weeks post injury and then at 6, 12 and 24 months post injury.
Extended studies are planned at specific volunteer clinical sites to assess MRI, extended biomarker and coagulation studies, high-resolution ICU monitoring, electrocorticographic monitoring and continuous electroencephalographic monitoring.
Computing platform and neuroinformatics resource
Acquisition, storage and analysis of common data element based clinical data has resulted in the development of an innovative informatics platform, with a vision that it will support ‘big data’ science, large-scale analytics and model building, and will form the prototype for future clinical studies on brain disorders and diseases.
Genetic associations
Previous research into genetic associations with TBI have been limited by small datasets and unsophisticated analysis. The use of conventional common disease-common variant approaches have been hindered as the admission phenotype was heavily influenced by the injury mechanism, rather than the host response.
Magnetic resonance imaging
It is now widely accepted that MRI is superior to CT in characterising the pathology of TBI. In current clinical practice, MRI is often used late in the decision-making process, and studies have focussed on comparing late MRI (at around 2 weeks post presentation) with admission CT imaging, rather that contemporaneous imaging. It is therefore difficult to establish the role of MRI in routine imaging, and especially in mild TBI. Changes in MRI findings in TBI are dynamic—advanced MRI techniques, including diffusion tensor imaging, resting state functional MRI and susceptibility weighted imaging, offer better detection, characterisation and quantification of TBI, but remain underused in routine clinical practice. CENTER-TBI will provide a repository of MRI images that aims to progress understanding and tracking of the disease process, guiding prognostication of cognitive outcome and improve mapping of covert cognition in disorders of consciousness. The ultimate aim is to deliver patient-tailored care pathways and to develop rehabilitation pathways that are better designed for individual patient needs (‘Precision Medicine’).
Biomarkers
Blood biomarkers have a possible role in diagnosing mild TBI, tracking disease progression and predicting outcome in TBI. However, studies have involved small patient numbers, and there has been no rigorous assessment of the adequacy of the established technique and models in larger patient numbers with good quality acute data and outcome documentation. There is also no evidence that these approaches can translate from technically demanding research tools to robust clinical tools that can be used in routine practice. There is a need for well-established laboratory markers that can be used for diagnosis, disease monitoring and prediction in TBI. Candidate biomarkers that CENTER-TBI is initially targeting include GFAP, S100, NSE and UCHL-1.
Comparative effectiveness research
Until recently, RCTs were considered the gold standard for investigating intervention benefits. RCTs generally involve strict enrolment criteria allowing assessment of the investigational intervention in an unbiased setting. However, this approach is restricted in validity as the results are only valid in selected subpopulations and translation to the ‘real-world’ context is limited. CER aims to identify the best treatment for the individual patient, with a specific type of injury, severity and comorbidities. The benefit is that it observes differences in care and outcome in observational studies, thus turning the natural variability into an asset, rather than a hindrance. CENTER-TBI actively exploits the existing heterogeneity in structure, process and outcome of the participating centres to compare interventions that are standard practice in some but not in others.
Outcome
Outcome after TBI is complex and multidimensional, and includes functional status, generic and disease-specific health related quality of life indicators, cognitive performance and emotional and psychosocial adjustment. CENTER-TBI examines the time course, correlates, predictors and moderators of different aspects of outcome and explores interdependencies to tailor outcome assessment after TBI. The overall aim is to develop a multidimensional European outcome classification for use in patients with TBI.
Future developments
CENTER-TBI commenced in December 2014 and is expected to report by 2017. It is contributing towards the overall goals of International Initiative on TBI Research (InTBIR), by identifying more efficient and effective treatments, improving outcome and reducing costs. The project will provide novel information on disease processes, treatment, outcome, and prognosis in TBI, and identify new therapeutic targets and therapies. CENTER-TBI data repositories will also ensure opportunities for legacy research. Following the initiation of CENTER-TBI and the ongoing research programme of InTBIR, satellite projects are being developed in Australia, India and China. Thus, the philosophy that TBI is a global problem that requires a global approach is now being translated into research practice.1
Footnotes
Contributors Both authors have contributed equally to the writing of the manuscript. JLCW is first author. DM is co-chairman of the management committee for CENTER-TBI.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.