Head injury: monitoring and optimising cerebral metabolism to improve outcome

The research program represents the next phase of our on-going studies into the pathophysiology of traumatic brain injury (TBI) and improving treatments for this common and disabling condition. We will apply medical and surgical interventions using novel strategies to address the previous problems of clinical trials. Specifically, we will use multimodality monitoring to refine treatment with the goal of improving outcome for these patients within five years of the end of the award. PATHOPHYSIOLOGY We will continue and expand our 13C microdialysis studies interrogating cerebral metabolism in the injured brain. Our previous studies have demonstrated that the injured human brain can utilise lactate as an energy substrate [1]. Ongoing 13C microdialysis studies are investigating the optimal concentration of glucose, and whether energy metabolism can be supplemented by administering pyruvate to bypass glycolysis, and administering succinate to enhance tricarboxylic acid (TCA) cycle activity. Our future plans include administering i.v. 13C labelled energy substrates to dose the whole brain in conjunction with in vivo 13C and 31P magnetic resonance spectroscopy (MRS) and microdialysis. In vivo MRS, through the labelling characteristics of downstream metabolites, enables both pathway tracking and the calculation of metabolic rates through biochemical pathways. This is performed by following the rate of change of fractional enrichment of a metabolic product, performed in real time. In this way, infusing 1-13C glucose and measuring the rate of fractional enrichment for glutamate C4 has been used to calculate TCA cycle activity in humans [2,3]. In vivo 13C MRS will be performed on the Siemens Verio scanner in the Wolfson Brain Imaging Centre using an existing 13C surface coil. To complement the above 13C measurements of TCA cycle, we will also address the effect of glucose infusion on ATP production in peri-contusional and non-contusioned areas of brain monitored using microdialysis, by the use of 31P in vivo MRS, using either an existing 31P surface coil, or a 31P head coil that has already been funded as part of our MRC research grant. We have longstanding experience of maintaining sedated, ventilated TBI patients throughout multi-scanning procedures on the WBIC’s scanners, which are located immediately adjacent to our Neurosciences Critical Care Unit. Hypothesis: Administration of glucose to patients with low brain glucose results in a reduction of the lactate / pyruvate ratio (microdialysis), enhanced TCA cycle activity (13C MRS) and ATP production (31P MRS). We propose to adminster 1-13C glucose intravenously, during in vivo 13C MRS (n=6 patients) and 31P MRS (n=6 patients) to determine TCA cycle flux and ATP production in regions of interest determined by microdialysis catheters placed in peri-contusional and non-contusioned areas of brain. The time course of the fractional enrichment of C4 glutamate during infusion of 1-13C glucose is related to the rate of glycolysis and the TCA cycle [3]. Patients will be administered an intravenous infusion of 1-13C glucose [4,5]. Prior to the commencement of 13C-glucose infusion, baseline 13C spectra / 31P spectra will be acquired. During the infusion spectra from peri-contusional and noncontusioned areas will be acquired. In order to calculate the fractional 1-13C glucose enrichment in plasma, blood samples will also be taken at similar intervals [3]. We also propose to administer i.v. 2-13C acetate to measure astroglial TCA activity specifically [2]. Since exogenous acetate acts an astrocyte-specific substrate it is useful for studying astroglial contributions to cerebral metabolism, especially as it is known to readily cross the blood-brain barrier . We propose to use 2-13C acetate to measure astroglial TCA cycle activity and determine the rate of glutamate/glutamine cycling, in 6 normal control subjects and 6 TBI patients. Hypothesis: IV infusion of acetate in TBI patients stimulates astroglial TCA cycle activity and increases glutamate/glutamine recycling. A 99% 13C enriched 2-13C acetate infusion will be administered. Prior to the infusion, baseline 13C spectra will be acquired. During the infusion spectra will be acquired. In order to calculate the fractional 2-13C acetate enrichment in plasma, blood samples will also be taken at similar intervals. The above two studies on 13C-labelled glucose and acetate intravenous infusions and the changes in brain chemistry they elicit, monitored by in vivo MRS, may ultimately enable better patient management strategies to be devised. This is especially important because despite modern protocol-driven therapy, some patients exhibit very low brain extracellular glucose concentration. Acetate infusion may offer the interesting prospect of ameliorating brain energy metabolism by feeding the astrocytes, which in turn may provide better support for neurons. Other beneficial effects of assisting the astroglia may be better blood-brain barrier integrity, reduced vasogenic oedema, and amelioration of the deterioration of white matter tracts, with overall potential to translate into better outcomes for TBI patients.TREATMENT 1. Overall Experimental design and methods (a) Trial design. Our approach to both medical and surgical interventions will be to address the previous problems of clinical trials Previous studies have used broad inclusion criteria including patients with a wide range of injury severity (mild, moderate, severe) and CT findings (mass lesion, diffuse injury). It is feasible that an intervention may have been effective for a particular pathological substrate of head injury e.g. glutamate antagonists and contusions but this effect was masked by including patients with types of injury (e.g. diffuse) that would not benefit. We will apply stringent inclusion criteria e.g. diffuse injury for IL-1ra studies and acute subdural haematoma for RESCUE-ASDH (see below). (b) Treatment effect. Previous studies have focussed on clinical outcome using relatively insensitive measures e.g the extended Glasgow Outcome Score (eGOS). In addition to clinical endpoints our approach is to determine surrogate endpoints i.e. monitor the physiological and chemical effect of the intervention. This is applied using multimodality monitoring and imaging. In the IL-1ra studies we will measure the extracellular concentration of downstream cytokines and other microdialysis markers of cerebral injury e.g. lactate / pyruvate ratio. In the decompressive craniectomy studies we will measure the effect of the intervention on intra-cranial pressure, microdialysis markers and tissue outcome using flumazenil PET [6]. (c) Drug site of action. We will apply microdialysis to measure drug penetration across the blood brain barrier into the extracellular space. We have undertaken successful pilot studies with vigabatrin and IL-1ra [7]. (d) Outcome measures. We will apply more sensitive measures of outcome including quality of life questionnaires (e.g. SF-36), neuropsychology testing and health economic analysis. 2. Medical Treatment strategies We have conducted preliminary studies using microdialysis to assist in the evaluation of potential neuroprotective agents. These are predicated on addressing two issues. Firstly, whether a given neurotropic drug crosses the BBB and enters the brain extracellular space and secondly, whether there are surrogate biochemical markers of the action of the drug such that dosages can be titrated to maximise potential drug efficacy. To this end, we have been investigating the effects of Interleukin-1 Receptor Antagonist (IL1ra) in severe TBI. The role of interleukin-1 (IL-1) in the pathogenesis of acute cerebral injury is well established. Animal models of neurodegeneration, including excitotoxic, ischaemic, hypoxic and traumatic insults have provided evidence for the detrimental effects of IL-1 in three ways. Firstly, expression of IL-1 (in particular IL-1ß) is rapidly increased in rodents in cerebral injury models that ultimately lead to neuronal death [8]. Secondly, injection of IL-1 into rodent brain exacerbates injury within these models [9]. Thirdly, inhibition of IL-1 in vivo by a range of mechanisms (including IL-1ra, a caspase inhibitor, IL-1ß antibody, or by deletion of the caspase 1 gene) reduces neuronal loss in rodents secondary to focal or global, and permanent or reversible cerebral ischaemia, infusion of excitotoxins and TBI [10,11]. An E-coli derived recombinant version of IL1ra (hrIL1ra, Anakinra (generic name) also known as Kineret® (trade name) is available for therapeutic use. hrIL1ra is a 17.3kDa protein that differs from native IL1ra by the addition of a single methionine residue at the Nterminus, in order to allow transcription within E. coli. IL1ra has been investigated extensively in animal models of brain injury and has been shown to reduce neuronal death [12]. We have administered IL1ra to a small number of patients with severe TBI with preliminary findings that the drug crosses the BBB and modulates the downstream cytokine response as gauged using cerebral microdialysis. These findings provide a paradigm for optimising the dose, timing and route of administration for this agent. We believe this translational approach will allow us to characterise the biology of this agent accurately in patients and will inform the rational design of a larger scale phase III study in due course. Hypothesis. Systemic administration of IL1ra to patients following severe TBI will modulate the downstream cytokine response in a predictable fashion, based on the timing and size of the dose administered. Methods. Our current studies have administered IL-1ra by a subcutaneous route of 100mg once daily for 5 days at a variable time from injury. We now propose a dose escalation study in 20 patients with early administration of the drug, ideally within 12 hours of injury, as this is the time frame we have identified in which IL1ß peaks. We have previously collaborated with the University of Manchester in which we delivered IL1ra at a dose of 500mg i.v. bolus in subarachnoid haemorrhage where it was found to be safe and well tolerated [13]. We will sequentially deliver 200mg, 300mg, 400mg and 500mg boluses of the drug to patients with severe diffuse TBI, with 5 patients in each group. The primary endpoint will be the IL1ra concentration of in the brain and the concentrations of cytokines downstream to the IL1-receptor which we have identified in our preliminary study. We will determine whether increasing doses of IL1ra modify the concentrations or pattern of production of these cytokines within the brain. Clinical outcome measures will include the eGOS and SF-36 at 6 months post-injury. In parallel to the microdialysis cytokine assay we will measure cytokines in blood and CSF (when an external ventricular drain is available).Impact and future developments. Focused clinical trials of this type with surrogate endpoints such as cerebral chemistry and determination of drug penetration across the BBB have the potential to screen effective agents before embarking on large phase III clinical trials. This methodology will allow us to address some of the pharmacokinetic issues that may have limited the translation of previous neuroprotective drug trials. 3. Surgical treatment strategies We have established the infrastructure to initiate and proceed with an international multicentre study of secondary decompressive craniectomy for refractory intracranial hypertension, which we will complete over the next three years. In addition we aim to commence a trial of primary decompressive craniectomy for acute subdural haematoma. We will use multimodality monitoring and imaging to obtain secondary endpoints. Secondary decompressive craniectomy Secondary decompressive craniectomy (DC; removal of part of the skull) can be used as a last-tier therapy when standard medical therapies to reduce ICP have failed. Although DC is currently performed there is no consensus if and when to proceed with surgery. There have been several observational studies demonstrating a range of outcomes and two randomised clinical trials: the DECRA trial (Melbourne) [14] the RESCUEicp trial (Cambridge) [15]. DECRA recruited 155 patients and failed to show clinical benefit with early/neuroprotective DC for patients with diffuse TBI. RESCUEicp is investigating the use of DC as a last-tier therapy for refractory raised ICP and has currently recruited 360 patients (target 400) from 50 units in 18 countries. Hypothesis. The use of DC as a last-tier therapy for refractory intracranial hypertension results in improved clinical outcome compared to advanced medical management. Methods. The RESCUEicp study is a randomised trial comparing DC with advanced medical management for the refractory ICP following TBI. The target study group includes ventilated, ICP-monitored patients with refractory ICP. The two arms of the trial are continuation of optimal medical management (with barbiturates) versus surgery (DC). The required sample size is 400 (200 in each arm of the study) for a 15% difference in dichotomised outcome (power=80%, 2-sided significance 0.05). The primary endpoint is outcome is assessed at 6 months using the eGOS. Secondary endpoints include eGOS at 1 and 2 years and the SF-36 quality of life questionnaire, ICP control and health-economic analysis (collaboration with the Health Economics Research Group - Brunel University). Primary decompressive craniectomy RESCUEicp focuses on the role of secondary DC for patients with predominantly diffuse injury and uncontrolled ICP. One of the other major controversies in the surgical management of TBI is whether to replace or leave out the bone flap after removing an acute subdural haematoma, the commonest mass lesion in TBI patients. This condition is treated with elevation of the bone flap and evacuation of the haematoma. Controversy persists as to whether to replace the bone flap at the end of the procedure. Leaving out the bone flap (primary DC) in theory helps to control brain swelling and ICP but exposes patients to the risk of the complications of DC and requires a second skull reconstruction operation. Replacing the bone flap may exacerbate brain swelling, necessitating a second operation to remove the flap in the ensuing hours / days. Hypothesis. The use of DC as opposed to craniotomy as treatment for acute subdural haematoma results in improved clinical outcome. Methods. We are therefore proposing a multicentre study (RESCUE-ASDH) whereby patients with acute subdural haematoma are randomised to either craniotomy or craniectomy. The required sample size is 600 (300 in each arm of the study) for a 8% difference in dichotomised outcome (power=80%, 2-sided significance 0.05). The primary endpoint is outcome is assessed at 6 months using the eGOS. Secondary endpoints will include eGOS at 1 and 2 years, quality of life questionnaires, ICP control and health economic analysis. Physiological effects of decompressive craniectomy – the role of monitoring and imaging in clinical trials. In parallel with the clinical trials we are studying the physiological effects of DC on the injured human brain. Both intracranial hypertension and DC have a number of effects on cerebral metabolism, venous outflow, cerebral perfusion and autoregulation. The objectives of these physiological studies will be firstly to determine the role of impaired cranial venous outflow (venous sinus narrowing) in the pathophysiology of raised ICP following TBI. Secondly, to characterise the effects of DC on cranial venous outflow and cerebral perfusion. Thirdly, to investigate the effects of DC on cerebral autoregulation. Fourthly, to determine the effects of DC on the energy metabolism of the injured brain. Finally, to determine the relationship between the diameter/location of the craniectomy and the deformation of the brain tissue after craniectomy. Studying these effects will enhance our understanding of the exact role of DC after TBI and allow us to develop surrogate end-points for future studies. Hypotheses. 1 Impaired cranial venous outflow due to venous sinus narrowing contributes to raised ICP following TBI. 2 Persistently raised ICP following DC is due to venous sinus narrowing and thrombosis.Methods. Cambridge patients will undergo monitoring to characterise the physiological and biochemical effects of DC. Computed Tomography Venograms (CTV) and venous transcranial Doppler ultrasonography studies allow delineation of the role of cerebral venous system in the pathophysiology of raised ICP. Pre-craniectomy and post-craniectomy CTVs reveal structural abnormalities of the cranial venous system. We aim to validate a number of metrics (diameter /stenosis) and assess how they change post-craniectomy. TCD studies measure venous blood flow velocities and determine their correlation with ICP and outcome. Computed Tomography Perfusion scans (CTP) and middle cerebral artery TCD studies performed at the same time-points enable us to characterize the changes in cerebral perfusion following DC. ICP waveform analysis allows us to study the effect of DC on cerebral autoregulation. Microdialysis to evaluates energy metabolism and we propose to use Flumazenil-PET to determine neuronal loss in the parenchyma deep to the site of DC. Finally, in collaboration with Dr M Sutcliffe (Engineering), we are developing a parametric finite element model to elucidate brain tissue deformation in relation to the size and location of DC. Impact and future developments. DC in the management of TBI remains controversial. It has been in and out of vogue for many years with no clear consensus as to whether and when it should be undertaken. RESCUEicp and RESCUEASDH are randomised studies which will provide a definitive answers to the questions of primary and secondary DC. LaySummary>