The role of tissue oxygen monitoring in patients with acute brain injury

Transcription

1 British Journal of Anaesthesia 97 (1): (2006) doi: /bja/ael137 Advance Access publication June 3, 2006 The role of tissue oxygen monitoring in patients with acute brain injury J. Nortje 12 * and A. K. Gupta 12 1 Department of Anaesthesia and 2 Department of Neuro Critical Care, University of Cambridge, Addenbrooke s Hospital, Cambridge CB2 2QQ, UK *Corresponding author. jn254@cam.ac.uk Cerebral ischaemia is implicated in poor outcome after brain injury, and is a very common postmortem finding. The inability of the brain to store metabolic substrates, in the face of high oxygen and glucose requirements, makes it very susceptible to ischaemic damage. The clinical challenge, however, remains the reliable antemortem detection and treatment of ischaemic episodes in the intensive care unit. Outcomes have improved in the traumatic brain injury setting after the introduction of progressive protocol-driven therapy, based, primarily, on the monitoring and control of intracranial pressure, and the maintenance of an adequate cerebral perfusion pressure through manipulation of the mean arterial pressure. With the increasing use of multi-modal monitoring, the complex pathophysiology of the injured brain is slowly being unravelled, emphasizing the heterogeneity of the condition, and the requirement for individualization of therapy to prevent secondary adverse hypoxic cerebral events. Brain tissue oxygen partial pressure (Pb O2 ) monitoring is emerging as a clinically useful modality, and this review examines its role in the management of brain injury. Br J Anaesth 2006; 97: Keywords: brain, injury, ischaemia; monitoring, intensive care, oxygen; outcome; oxygenation, oxygen partial pressure Severe traumatic brain injury (TBI) has a mortality exceeding 40% in the UK; ischaemia playing a significant part Neuronal excitotoxicity, oxidative stress, mitochondrial dysfunction, lipid peroxidation with loss of membrane integrity and disturbances in ion homeostasis, free radical production, subsequent inflammation, necrosis and even apoptosis all play a role in the cascade of post-ischaemic events With targeted therapy having shown promising reductions in mortality, brain tissue oxygen monitoring may provide a further tool not only to elucidate the pathophysiology, but also individualise therapeutic targets. 65 Cerebral oxygenation and cerebral perfusion pressure targets Cerebral blood flow (CBF) is regulated by metabolic requirements under normal conditions, so-called flowmetabolism coupling, and ensures adequate cerebral oxygenation. Temporal patterns of CBF disturbances after traumatic brain injury (TBI) have been well documented, 258 with optimal therapy on the first post-injury day, not necessarily the most appropriate on subsequent days. High intracranial pressures (ICPs) and cerebral hypoxia show strong correlations with poor outcome, 47 making the control of these parameters with a sufficient cerebral perfusion pressure (CPP) critically important (see the review by Steiner and Andrews in this postgraduate issue). The exact level of CPP required following TBI has been subject to much debate, the latest definitive guidance being the downward revision of the Brain Trauma Foundation s guidelines on CPP targets (available from org/guidelines) in 2003, suggesting a CPP target of 60, rather than 70 mm Hg. The proviso is, however, that in selected patients, where there is evidence of regional or global ischaemia, the CPP target may need to be higher. Individualized CPP optimization, therefore, becomes dependent on, amongst other things, the monitored levels of brain oxygenation. Existing monitors of cerebral oxygenation Imaging Positron emission tomography (PET) Metabolic imaging of the brain after 15 O-radioisotope administration, allows the quantification of CBF, cerebral blood volume, the oxygen extraction fraction (OEF) and cerebral metabolic rate for oxygen (CMRO 2 ). Despite limitations, including the snapshot nature of the technique, limited availability, use of radiation, poor spatial resolution and Ó The Board of Management and Trustees of the British Journal of Anaesthesia All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 Nortje and Gupta the fact that the most unstable patients may not get scanned (selection bias), PET is still regarded as the gold standard for visualizing cerebral oxygen use. Early post-injury PET has identified the presence of regional ischaemia, with quantification of the ischaemic brain volume. Magnetic resonance spectroscopy (MRS) MRS is an application of magnetic resonance imaging (MRI) whereby spectra of metabolic changes in living tissue are obtained. The outcome predictive value of noninvasively measuring brain metabolites (biomarkers of injury processes) using MRS, is increasingly being demonstrated Measurement of adenosine triphosphate (ATP) using phosphorous spectroscopy ( 31 P-MRS) 69 and lactate using proton spectroscopy ( 1 H-MRS) 82 after brain injury can provide evidence of cerebral ischaemia, while N-acetyl aspartate represents neuronal integrity. The post-processing and availability of MRS preclude its routine clinical use at present. Bedside Jugular venous oximetry (S jo2 ) Catheterization of the internal jugular vein to measure the oxygen saturation of the effluent cerebral blood at the jugular bulb, allows assessment of the global oxygenation status of the brain, providing some insight into the adequacy of CBF, especially during manoeuvres such as hyperventilation. A significant association between jugular venous desaturation and poor neurological outcome exists, 24 with poor outcome in 55% of patients with no episodes of desaturation, 74% with one episode and 90% with multiple episodes. A shortcoming, however, is the inability of Sj O2 to detect regional ischaemia, with PET evidence 13 of approximately 13% of the brain being ischaemic before Sj O2 levels decrease below 50%. Near-infrared spectroscopy (NIRS) Penetration of human tissue by light in the near-infrared band and its resultant absorption and scatter allows assessment of cerebral changes in oxyhaemoglobin (HbO 2 ), deoxyhaemoglobin (Hb) and cytochrome oxidase. An attractive non-invasive regional monitor of cerebral oxygenation, NIRS has been beset by issues of extra-cranial blood contamination, light shielding, optimal optode placement, sample volume inaccuracies and the robustness of the derived algorithms. Spatial resolution and equipment improvements are addressing these issues, but to date NIRS has yet to find a clear role in routine clinical practice. 4 Intracerebral microdialysis Microdialysis (the subject of a separate review by Smith and Tisdall in this postgraduate issue) has become a feasible clinical bedside technique in the intensive care unit (ICU), with metabolic patterns supplying valuable information on the adequacy of cerebral oxygenation. Brain tissue oxygen tension (P bo2 ) Improvements in technology, and the pitfalls of the techniques described above, have led to the introduction of brain tissue oxygenation monitoring. What is brain tissue oxygen partial pressure? Pb O2 is the partial pressure of oxygen in the extra-cellular fluid of the brain and reflects the availability of oxygen for oxidative energy (ATP) production. It represents the balance between oxygen delivery and consumption, and is influenced by changes in capillary perfusion. Distance from the supplying capillaries and possible barriers to oxygen diffusion 66 may be particularly important after injury. An experimental global ischaemia model 81 examining the relationships between Pb O2, local CBF and NIRS-derived CMRO 2 and AVDO 2 (arterio-venous difference in oxygen content) has also suggested that the relative predominance of arterial or venous vessels in the immediate proximity of the Pb O2 sensor may determine whether coupling exists between Pb O2 and CBF or between Pb O2 and OEF (AVDO 2 ), respectively. The equipment The two most commonly used systems to date are the Licox (GMS, Kiel-Mielkendorf, Germany) and the Neurotrend (Codman, Johnson & Johnson, Raynham, MA, USA). With good temporal resolution of acute biological changes, accomplished by the rapid response rates 35 of the Pb O2 systems, timely therapeutic interventions and assessment of the subsequent responses is possible. 0.8 mm 795 mv polarization e + O 2 1 Fig 1 Schematic representation of the tip of the Licox brain tissue oxygen sensor. 1, Polyethylene tube with diffusible membrane; 2, gold polarographic cathode; 3, silver polarographic anode; 4, electrolyte chamber; 5, brain parenchyma. (Adapted from the manufacturer s manual.) O

3 Tissue oxygen monitoring in acute brain injury 0.5 mm Fig 2 Schematic representation of the tip of the Neurotrend sensor demonstrating the microporous polyethylene tube (1), the three optical sensors, namely the Pb O2 sensor (2) (ruthenium dye in a silicone matrix), the Pb CO2 sensor (3) (phenol red in a bicarbonate solution), the ph sensor (4) (phenol red in a polyacrylamide gel) and the thermocouple (5) (copper and constantan wires) all suspended in phenol red and polyacrylamide gel (6) and implanted in the brain parenchyma (7). (Adapted from the manufacturer s manual.) Measurement principles The Licox system provides PO 2 measurement, with or without brain temperature (thermocouple), in an estimated mm 2 PO 2 -sensitive area. The PO 2 probe utilizes a closed polarographic (Clark-type) cell with reversible electrochemical electrodes (Fig. 1). Oxygen, which has diffused from the brain tissue across a semi-permeable membrane, is reduced by a gold polarographic cathode producing a flow of electrical current directly proportional to the oxygen concentration. This oxygen-consuming process is temperature-dependent. In addition to Pb O2, Neurotrend also offers Pb CO2, ph and temperature. Its predecessor, the Paratrend 7, was an intra-arterial monitor which was adapted for intracerebral use. 8 Although Neurotrend has been used both for research and clinical purposes, its manufacture has now been discontinued. The Neurotrend (Fig. 2) comprises three optical sensors (PO 2, PCO 2 and ph) and a thermocouple contained within the distal 25 mm of a 0.5 mm diameter microporous polyethylene tube. PO 2 measurement occurs by quenching (reduction) of the intensity of a fluorescent optical emission from an indicator (ruthenium) in the presence of oxygen (following light pulses from a blue light emitting diode). In contrast to the Licox, this process does not consume oxygen and does not affect the measured oxygen level. The ph sensor relies on optical absorption, the local ph affecting the intensity of light transmitted through an indicator (phenol red). The PCO 2 sensor, similarly, is a CO 2 -selective ph sensor. Calibration and insertion Using a sensor-specific pre-calibrated smart card, the Licox sensor can be inserted without delay, while the less userfriendly Neurotrend sensor requires a 31 min calibration in a chamber using three calibration gases before insertion. Fig 3 Computed tomography image showing the Neurotrend Pb O2 sensor (arrow) near the intra-parenchymal ICP monitor in the right frontal cerebral white matter. Sensors can be inserted, either via a cranial access device sited through a craniotomy, on the ICU, or under direct vision at surgery. Purpose-designed triple lumen cranial access devices 44 allow simultaneous ICP, intracerebral microdialysis and Pb O2 monitoring. Ideally, this should occur in an anatomically similar area of white matter where Pb O2 readings are likely to be more stable. Postinsertion CT confirmation of probe position in the brain parenchyma (Fig. 3) is important for interpretation of readings. Transiently increasing the FI O2 and observing the corresponding Pb O2 increase, is advised to exclude the presence of surrounding micro-haemorrhages or sensor damage at insertion. A run-in or equilibration time of up to a half hour is required before readings are stable. Adjustment of insertion depth (when used through an access device) is allowed by the Neurotrend, but not the Licox. Comparison Comparative studies have been carried out to evaluate their functioning and reliability under experimental 35 and clinical conditions. 45 In test conditions, the Licox was slightly more accurate, with the Neurotrend under-reading at low Pb O2. Both sensors displayed slight drift towards lower oxygen concentrations over time, but this was not thought to prohibit long-term use. The Neurotrend measured Pb CO2 and ph very accurately. Clinically, the Neurotrend sensor underread Pb O2 (in contrast to the reported overestimation 96 of the Paratrend 7), and was less robust than the Licox sensor. 97

4 Nortje and Gupta Table 1 Studies relating brain tissue oxygen partial pressure measurements to monitored variables of cerebral oxygenation and blood flow. Sj O2, jugular venous oximetry; NIRS, near-infrared spectroscopy; rcbf, regional cerebral blood flow; CT, computed tomography; PET, positron emission tomography; TBI, traumatic brain injury; SAH, subarachnoid haemorrhage; TOI, tissue oxygenation index; L/P, lactate/pyruvate Comparison modality Authors Study (pathology) Year Findings Direct local venous gas tensions Edelman and colleagues 20 Animal (normal) 2000 Local tissue and cerebral venous blood Po 2, PCO 2,pH were highly correlated (P<0.001) Sj O2 Kiening and colleagues 53 Human (TBI) 1996 Significant correlation between Sj O2 and Pb O2 (r 2 =0.71) Sj O2 Gupta and colleagues 29 Human (TBI) 1999 Significant correlation between Sj O2 and Pb O2 in areas without focal pathology (r 2 =0.69) Sj O2 Gopinath and colleagues 25 Human (TBI) 1999 Significant correlation between Sj O2 and Pb O2 (r 2 =0.58) NIRS and Sj O2 McLeod and colleagues 59 Human (TBI) 2003 Comparison of Pb O2, Sj O2 and TOI during step changes in FI O2 : similar patterns of response, but variable degree and speed rcbf using xenon CT Doppenberg and colleagues 19 Human (TBI) 1997 Pb O2 linearly related to regional CBF (r=0.74, P=0.0001) Local CBF using laser Critchley and Bell 15 Animal (SAH) 2003 Pb O2 correlated with local CBF (r=0.66, P=0.02) Doppler flow rcbf using thermal Jaeger and colleagues 46 Human 2005 Pb O2 correlated with rcbf (r=0.36, P<0.01) diffusion (TBI and SAH) PET Gupta and colleagues 30 Human (TBI and SAH) 2002 In normal areas: dpb O2 during hyperventilation correlates with dpv O2 (calculated from OEF) (r=0.78, P=0.0035) Microdialysis Zauner and colleagues 102 Human (TBI) 1997 Pb O2 correlated with brain glucose (P<0.01), but not lactate Microdialysis Meixensberger and colleagues 62 Human (TBI and SAH) 2001 Pb O2 only weak negative correlation with L/P ratio (SAH: r= 0.185, TBI: r= 0.358) Microdialysis Sarrafzadeh and colleagues 79 Human (TBI) 2003 Pb O mm Hg for >5 min: increased glutamate (P=0.03) Validation with existing techniques Establishing the relevance of a monitor requires validation of the measured variables and their relationships with the existing clinical and experimental techniques (Table 1). Clinical utility Normal values Normal brain gas tension measurements have been acquired experimentally, but human measurements have been restricted to normal values during neurosurgery and in normal-appearing brain after TBI. A feline study 100 revealed a normal Pb O2 of 42 (9) mm Hg [Pb CO2 of 59 (14) mm Hg, brain ph of 7.0 (0.2)], while a murine study 15 measured a normal Pb O2 of 29.4 (12.8) mm Hg. Human recordings have varied from a Pb O2 of 37 (12) mm Hg [Pb CO2 of 49 (5) mm Hg, brain ph of 7.16 (0.08)] 36 to a Pb O2 value of 48 mm Hg 60 in uncompromised patients undergoing cerebrovascular surgery. Hypoxic thresholds The identification of hypoxic tissue allows the institution of early potentially corrective interventions, and also provides meaningful therapeutic end-points. These values need to be considered in the context of probe type, probe site, underlying pathology and duration of hypoxia before irreversible damage occurs. Various Pb O2 hypoxic thresholds (Table 2) have been proposed. Pb O2 <10 mm Hg for >5 min: increased glutamate (P=0.007) and increased lactate (P=0.044) Microdialysis Hlatky and colleagues 34 Human (TBI) 2004 Pb O2 <10 mm Hg: increased lactate (P=0.015) Safety Initial concerns regarding the invasiveness of these intraparenchymal sensors and the risk of haemorrhage and infection, have proved unfounded. Eleven studies including 552 patients reported no infections and three iatrogenic haematomas with only one requiring surgical evacuation. Measurement accuracy with negligible zero drift was also a consistent finding Drawbacks Some of the reported problems include insertion trauma with subsequent gliosis and the ability to adequately position and secure sensors in position. The focal nature of these monitors must also be emphasized. Global assumptions of a focal monitor Jugular venous oxygen saturation (Sj O2 ) monitoring provides global cerebral oxygenation determination and can be used to calculate the arterio-venous oxygen content difference. Pb O2 sensors are extremely localized, only sampling approximately 15 mm 2 of tissue around the tip. The positioning of the sensor, however, becomes a vital question in the interpretation of the readings. In tissue at risk regions near focal pathology, global assumptions cannot be made and the monitor is purely focal, but when positioned in areas of seemingly normal tissue, or in areas of diffuse injury, the Pb O2 can be regarded as an 98

5 Tissue oxygen monitoring in acute brain injury Table 2 Human studies proposing hypoxic brain tissue oxygen thresholds. CT, computed tomography; Pv O2, cerebral venous/end-capillary P O2 ; PET, positron emission tomography; OEF, oxygen extraction fraction Sensor Authors Year Proposed threshold [mm Hg (kpa)] How the threshold was determined 1997, (3.3) Pb of 26 mm Hg CBF (xenon CT) < 18 ml per 100 g Paratrend 7 Zauner and colleagues 102 Doppenberg and colleagues 19 O2 per min; all patients with Pb O2 <25 mm Hg had a poor outcome Paratrend 7 Doppenberg and colleagues Between 19 and 23 (2.5 and 3) Combined above data with a feline MCA occlusion study and outcome Neurotrend Menon and colleagues (1.3) Significantly greater diffusion gradients for oxygen (Pv O2 Pb O2 )ifpb O2 <10 mm Hg Neurotrend Johnston and colleagues <14 (1.9) Significant linear relationship between Pb O2 and PET OEF (r 2 =0.21, P<0.05); mean normal OEF=40% associated with Pb O2 =14 mm Hg Licox Kiening and colleagues (1.1) Regression analysis: Sj O2 threshold of 50% correlated with Pb O2 of 8.5 mm Hg Licox van Santbrink and colleagues Between 10 and 15 (1.3 and 2) Significant difference in 6 month outcome at threshold <5 mmhg(p=0.04), suggested maintenance of Pb O2 between 10 and 15 mm Hg Licox Valadka and colleagues (2.7) [6 (0.8)] Tobit regression analysis relating the time below thresholds of with likelihood of death. Much greater likelihood of death, Licox van den Brink and colleagues <5 (0.6) for 30 min <10 (1.3) for 1 h 45 min <15 (2) for 4 h indicator of global oxygenation, allowing the use of Pb O2 as an endpoint in the optimization of CPP. Kiening and colleagues 53 compared the use of Sj O2 and Pb O2 in normal frontal brain white matter, in 15 patients with severe TBI showing a strong correlation (r 2 =0.71). Continuous Pb O2 could reliably be monitored twice as long as Sj O2, with good quality data acquisition 95% of the time with Pb O2, as opposed to the 43% for Sj O2. The requirement for repeated calibrations of the Sj O2 measurement system contrasted starkly with the lack of post-insertion calibration requirement of the Pb O2 technique. Dynamic indices In addition to static baseline Pb O2 readings, oxygen regulatory mechanisms challenged dynamically, may provide insight into the underlying pathophysiology and outcome prediction. Pb O2 reactivity The increase in Pb O2 relative to an increase in arterial PO 2 is termed brain tissue oxygen reactivity. It is believed that this reactivity is controlled by an oxygen regulatory mechanism (cf. CBF autoregulation), and that this mechanism may be disturbed after brain injury. van Santbrink and colleagues 99 examined the brain tissue oxygen response (the degree of change in Pb O2 in response to changes in Pa O2 ) and showed that greater responsiveness in the first 24 h post-injury was associated with an unfavourable outcome (P=0.02); with multiple logistic regression analysis supporting its value as an independent predictor of unfavourable outcome (odds ratio 4.8). The effect of Pa CO2 on this mechanism was illustrated experimentally by Hoffman and colleagues 42 Pb O2 the longer the Pb O2 <20 mm Hg or any time of Pb O2 <6mmHg The relative risk of death was graded. Hypoxic thresholds are expressed as the depth and duration of hypoxia imparting a 50% risk of death in dogs where Pb O2 reactivity was attenuated during hypocapnoea, and emphasizing that when assessing Pb O2 reactivity, effects of Pa CO2 need to be considered. To understand the normal relationships of Pb O2 with mean arterial pressure (MAP), and with changing CO 2 concentrations in the uninjured brain, Hemphill and colleagues 32 studied 12 anaesthetized pigs. Pb O2 displayed a linear relationship with CO 2 (r 2 =0.70) and a sigmoid curve with MAP between 60 and 150 mm Hg (r 2 =0.72), and a linear correlation with CBF (measured using thermal diffusion probes) during CO 2 reactivity testing (r 2 =0.84). The conclusion was that Pb O2 is strongly influenced by factors regulating CBF, namely CO 2 and MAP. Pb O2 autoregulation Soehle and colleagues 84 introduced the concept of Pb O2 autoregulation, defined as the ability of the brain to maintain Pb O2 despite changes in CPP, thereby identifying appropriate individual CPP targets. Lang and colleagues 54 showed a significant correlation (r= 0.61) between static cerebral autoregulation (determined using CBF velocity in relation to changing CPP) and cerebral tissue oxygen reactivity (the rate of change of Pb O2 in relation to changing CPP) suggesting a close link between regulation of CBF and oxygenation. Following these findings, manipulation of Pb O2 by altering Pa O2 (by FI O2 increases) or altering the CPP (by MAP increases, ICP decreases, or both) have been investigated with the view to therapy optimization and potential prognostication. 99

6 Nortje and Gupta Applications of brain tissue oxygen monitoring In humans, Pb O2 monitoring has predominantly been applied to the investigation and management of subarachnoid haemorrhage (SAH) (both on the ICU and during operation) and severe TBI (see above) on the ICU. Other applications have included, arterio-venous malformation (AVM) resection, tumour resection and for studying the effects of anaesthetic agents. SAH and vasospasm on the ICU Despite its invasiveness, the application of Pb O2 monitoring seems attractive for the continuous surveillance and detection of delayed vasospasm-induced ischaemia in patients with SAH in the ICU (in contrast to the traditional snapshot use of TCD). Kett-White and colleagues 52 monitored 40 patients (35 after SAH and 5 after complex aneurysmal surgery) on the ICU with Pb O2 and intracerebral microdialysis, but only managed to show weak associations between episodes of low Pb O2, abnormal microdialysis and outcome. Possible reasons cited for the marked variability in measured Pb O2 readings were the variable distances from vasculature (oxygen gradients), grey/white matter influences (metabolic rates, varying depths attributable to gyri and sulci) and tissue heterogeneity. A study using Neurotrend in 10 SAH patients 7 (three of whom developed vasospasm), showed a significant decrease in ph and increase in Pb CO2 (P<0.001), but failed to show ischaemic Pb O2 levels. Meixensberger and colleagues 64 prospectively studied 42 patients presenting with severe grade SAH and failed to demonstrate the value of Pb O2 as an early predictor of non-survival, citing reasons such as small patient numbers, accuracy of probe placement in the affected cerebral territory, efficacy of the implemented triple H (hypervolaemia, hypertension, haemodilution) therapy, and the possibility that oxygen consumption may have been more depressed than CBF, resulting in unchanged Pb O2 values. The jury, therefore, remains out on the value of Pb O2 monitoring as an early warning of cerebral vasospasm in SAH patients on the ICU. Aneurysm surgery Intraoperative use of Pb O2 monitoring is both feasible and is a sensitive indicator of cerebral tissue at risk. Severe bleeds (Fisher grade 3) also significantly decrease Pb O2 (P<0.05). 43 Correctly positioned Pb O2 monitoring allows not only assessment of the effect and reversibility of temporary aneurysm clipping, but can also be indicative of the correct positioning of the subsequent permanent clip. 23 Hypoxic Pb O2 levels confirmed compromised perfusion detected on preoperative single photon emission computed tomography (SPECT) and cerebral angiography. 36 In a study of 46 patients undergoing craniotomy for aneurysm clipping, 51 the majority of 31 patients who required temporary clipping of the parent vessel, showed decreases in Pb O2, with a level of Pb O2 <8mm Hg for 30 min being predictive of cerebral infarction. Another study 92 found that Pb O2 monitoring during aneurysm clipping supplemented somatosensory evoked potential (SEP) monitoring in identifying ischaemia, especially in those patients where the baseline SEP was absent. AVM surgery Pb O2 measurement has been used to investigate the oxygenation of cerebral tissue supplied by vessels with AVMs. 38 In total, 13 patients undergoing resection of AVMs were compared with 8 non-ischaemic patients undergoing aneurysmal surgery (the controls). Low Pb O2,but normal Pb CO2 and ph (in contrast with raised Pb CO2 and acidosis seen in acute occlusive disease with ischaemia 39 ) before AVM resection suggested low perfusion and chronic hypoxia with possible metabolic adaptation and subsequent hypometabolism, while the marked Pb O2 increases post-resection indicate hyperperfusion with its attendant problems. Apart from enhancing the understanding of AVM pathophysiology, this study reaffirms the feasibility of intraoperative Pb O2 monitoring. Tumours The extent and nature of the effects of oedema on brain tissue oxygen surrounding tumours was investigated perioperatively in 19 patients. 71 MRI-based stereotaxis was used to guide sensor placement into the peritumoral area before craniotomy and the effects of dural opening and resection on the Pb O2 was noted. In patients with swelling, Pb O2 increased significantly on dural opening (P<0.05), and postresection (P<0.05), implying the presence of ischaemic processes with oedema. This emphasizes the importance of maintaining adequate CPP in patients undergoing brain tumour surgery. Pb O2 monitoring during awake craniotomy for tumour resection has also been reported. 93 Anaesthetic drug pharmacodynamics Studies investigating the dose effects of anaesthetic agents such as isoflurane, 41 desflurane 37 and propofol 48 on cerebral autoregulation and oxygenation to levels of EEG burstsuppression have also utilized Pb O2 monitoring. The inhalational agents demonstrate a dose-related loss of autoregulation with corresponding increase in Pb O2 as long as the CPP is maintained, while propofol displays no such changes and flow-metabolism coupling remains intact. Therapeutic and research interventions The clinical utility and temporal responsiveness of Pb O2 to fraction of inspired oxygen (FI O2 ) increases, Pa CO2 changes (hyper- and hypoventilation) and haemorrhage to 70% blood loss or cardiac arrest with immediate subsequent resuscitation, were experimentally validated by Manley and colleagues. 57 In addition, investigation of pathophysiological 100

7 Tissue oxygen monitoring in acute brain injury 200 ABP mm Hg 0 50 ICP mm Hg CPP mm Hg 40 4 Pt O kpa mechanisms after brain injury in studies combining Pb O2 monitoring with PET, have proposed the presence of microcirculatory abnormalities with significant gradients for oxygen diffusion in injured tissue. 66 Structural evidence for these pathophysiological changes has also been well documented. 77 As more data are accumulated regarding hypoxic thresholds, and the clinical and outcome significance of Pb O2 in various clinical situations is established, the changes in baseline Pb O2 levels in response to potential therapeutic or research interventions gives weight to their clinical value. Interventions include the following. Elevation of CPP A comparative study using dopamine or norepinephrine to elevate CPP from 65 to 85 mm Hg, showed no significant 30/5 10:00 30/5 10:15 30/5 10:30 Fig 4 Bedside monitoring of a patient with severe TBI with the Neurotrend Pb O2 sensor placed near a cerebral contusion. Increasing the cerebral perfusion pressure (CPP) from around 70 mm Hg to above 100 mm Hg reverses the cerebral hypoxia (ABP, mean arterial blood pressure; ICP, intracranial pressure; Pt O2 = Pb O2 ; kpa, kilopascals). ABP 133 ICP 24.2 CPP 109 Pt O increase in the Pb O2 values monitored in predominantly normal-appearing brain. 49 However, targeted Pb O2 sensors in CT hypodense lesions in nine patients 89 (with hypoperfusion confirmed with SPECT) revealed significant improvements in Pb O2 readings with induced hypertension (r 2 =0.74). Significantly higher Pb O2 has been shown at a CPP >70 mm Hg than <70 mm Hg in TBI patients (P<0.001) 9 (cf. Fig. 4). Hyperventilation An experimental study in 12 healthy pigs 10 comparing Pb O2 (Licox), rcbf (thermal diffusion system) and metabolic microdialysis markers at baseline, during moderate (Pa O2 =30 mm Hg) and profound (Pa O2 =20 mm Hg) hyperventilation revealed that both moderate and profound 101

8 Nortje and Gupta hyperventilation may cause insufficient regional oxygen supply and anaerobic metabolism. The value of Pb O2 as a monitor of excessive hyperventilation (cf. Sj O2 ) was illustrated in a head injury study of 90 patients 6 where the Pa CO2 was decreased with hyperventilation and Pb O2 decreased significantly (P<0.001). Importantly, the risk of secondary cerebral ischaemia with hyperventilation increased over time. Hypothermia Despite the negative findings of the National Acute Brain Injury Study: Hypothermia 12 (no improvement in outcome using hypothermia to 33 C within 8 h of TBI) and the Intraoperative Hypothermia for Aneurysm Surgery Trial 94 (no improvement in neurological outcome using intraoperative hypothermia in good grade SAH), hypothermia is thought to reduce secondary brain injury following TBI by metabolic suppression and reduction of inflammation, free radicals, cytokines and excitatory amino acids. Pb O2 and direct brain temperature were measured in 58 patients after severe TBI, 85 and revealed decreases in the Pb O2 with mild hypothermia (34 36 C) accompanied by decreases in ICP and Pb CO2 and increases in brain ph. Decreased metabolic requirements and lowering of the Pb O2 critical threshold were concluded. Another Pb O2 study in 30 patients 28 indicated that 35 C might be the optimal temperature after severe TBI. Decompressive craniectomy In patients with intractable intracranial hypertension, despite maximal medical therapy, surgical (either bi-frontal or unilateral) removal of a part of the skull may be considered. The benefits of this technique and whether it offers any advantage in outcome and quality of outcome over the use of barbiturate coma, although not yet known, is currently the subject of the multi-centre RESCUE-ICP study (Randomized Evaluation of Surgery with Craniectomy for Uncontrollable Elevation of IntraCranial Pressure; available at Recent studies have added the use of Pb O2 monitoring to the traditional ICP/CPP parameters in assessing the impact of this last-resort procedure. A retrospective review 91 of 20 ICP and Pb O2 monitored SAH patients who ultimately underwent surgical hemicraniectomy for intracranial hypertension, revealed 12 patients in whom Pb O2 decreases to <10 mm Hg were the first sign of deterioration. Most notably, hypoxic Pb O2 readings were present 13.4 h (mean) before development of intracranial hypertension in nine patients hinting at an important potential therapeutic window. Another decompressive hemi-craniectomy study 87 revealed favourable decreases in ICP and increases in CPP, but also sustained improvement in cerebral oxygenation. Patients with severe brain hypoxia before surgery were likely to have poorer outcomes, raising the possibility of using Pb O2 monitoring to select the patients who might benefit most from surgical decompression. Subsequently, a study 75 used Pb O2 <10 mm Hg during refractory raised ICP as an indication for hemicraniectomy in five patients. It documented the significant reduction in ICP (P<0.039) and increase in Pb O2 (P<0.041) during the perioperative period, noting the importance of dural enlargement (by means of a dural graft) on the measured parameters. The authors proposed that hypoxic Pb O2 values may have heralded the onset of cytotoxic brain oedema, highlighting the role of Pb O2 in the timing of surgical intervention. Hyperoxia The therapeutic use of normobaric hyperoxia, to counter the effects of cerebral ischaemia by facilitating oxidative metabolism, has received increasing attention in the management of severe TBI Although attractive in its ease of use (by increasing the FI O2 in ventilated patients), the target population, the dose and the duration of hyperoxia have yet to be defined. While studies utilizing Pb O2 and microdialysis monitoring agree universally on the elevation of the Pb O2 with this intervention, and have documented the resultant biochemical effects, the clinical benefit of this therapy needs further elucidation. Other cerebral tissue parameters This review has focused on oxygen partial pressure in the tissue, but the use of the Neurotrend sensor also allowed Pb CO2 and ph monitoring. The relationships of these variables to other brain parameters have been well documented in a study by Clausen and colleagues 9 where they demonstrated that Pb CO2 was significantly higher at 6 h post-injury in patients with poor outcomes compared with patients with good outcomes (P<0.05). They also showed that Pb CO2 was significantly higher at a CPP below 70 mm Hg as opposed to higher (P<0.0001), and also significantly higher at a Pb O2 below 10 mm Hg as opposed to higher (P<0.0005). These findings were in the face of a stable Pa CO2. Brain ph differences at these CPP and Pb O2 thresholds were also significant (both P<0.0001). Low brain ph has also been correlated with adverse outcome (P=0.003). 31 These parameters may, therefore, also provide useful end-points in optimization of therapy. Brain temperature measurement (available with both Neurotrend and Licox) has allowed investigation of hyperthermia and hypothermia. Cerebral oxygenation and outcome prediction The theme of brain oxygen manipulation, and its effect on patient outcome, was first addressed in 1998 when Cruz 16 demonstrated a significant improvement in 6 months outcome in a group of 178 patients who underwent jugular oximetry monitoring and manipulation of oxygen extraction, compared with a group of 175 patients receiving ICP/CPP-guided treatment alone (P< ). However, 102

9 Tissue oxygen monitoring in acute brain injury Table 3 TBI studies demonstrating the value of brain tissue oxygen tension in outcome prediction (dichotomized outcome=alive vs dead/persistent vegetative state) Authors Year No. of patients Sensor Outcome prediction van Santbrink and colleagues Licox Pb O2 <5 mm Hg in the first 24 h post-injury, negatively correlated with 6 month dichotomized outcome (P=0.04, Fisher s exact test). Oxygen reactivity was also significantly lower in the good outcome group (P=0.02, Mann Whitney test) Doppenberg and colleagues Paratrend All patients with a Pb O2 <25 mm Hg, measured at a mean of 27 h post-injury, died or remained in a vegetative state Zauner and colleagues Paratrend If mean Pb O2 <15 mm Hg, patients died within 36 h. Pb O2 was the strongest predictor of outcome (multiple logistic regression analysis) Valadka and colleagues Licox Mortality increased with increasing time spent with a Pb O2 <15 mm Hg, or any Pb O2 values <6 mm Hg (Tobit regression analysis using maximum likelihood methods) van den Brink and colleagues Licox Pb O2 <10 mm Hg for >30 min was an independent predictor of death or unfavourable outcome (multiple logistic regression analysis). Depth and duration of tissue hypoxia are independently related to 6 month outcome to enable reliable continuous Sj O2 readings unaffected by head movement and motion artifacts, all of the patients in this group wore cervical collars. Pb O2 monitors are much less prone to accidental displacement and are not affected by head movement, resulting in considerably less maintenance post-insertion. The value of adding Pb O2 measurement to the standard ICP/CPP-guided therapy has been assessed in two prospective observational studies Meixensberger and colleagues 61 compared the effect of keeping Pb O2 above 1.33 kpa (10 mm Hg) as a secondary therapeutic target in ICP/CPP-guided management algorithms in 53 patients, to a historical control group of 40 patients receiving ICP/CPP-guided management alone. Pb O2 was successfully maintained at higher levels in the test group, and although there was no statistically significant difference in 6 month outcome, there was a positive trend in the Pb O2 -guided group. Stiefel and colleagues 88 similarly compared a traditional ICP/ CPP-guided group (25 patients) to a Pb O2 -guided group (28 patients) targeting a Pb O2 of greater than 3.33 kpa (25 mm Hg). They evaluated hospital outcome and demonstrated a significant reduction in mortality in the Pb O2 - guided group (P<0.05), and better functional outcomes in this group. Meixensberger and colleagues 63 also showed a correlation between low Pb O2 in the acute post-injury phase with poor neuropsychological performance 2 3 yr later in 40 patients. The possible value of Pb O2 -guided therapy may also have been demonstrated by a study 80 comparing 36 patients with isolated severe head injuries to 44 patients with severe head injuries and severe extra-cranial injuries; both groups managed with ICP/CPP/Pb O2 protocols and late surgical intervention for extra-cranial injuries. Revealing no significant difference in 6 month or 1 yr outcome, it contrasted with previous reports of increased mortality in head injured patients with significant co-existing extra-cranial injuries. Numerous severe head injury studies have correlated low with adverse outcome (Table 3). There are no Pb O2 randomized controlled trials of Pb O2 -guided therapy to date. Given the clinical feasibility and the increasing evidence of worsened outcome with low brain Pb O2, prospective randomized Pb O2 -directed studies are now essential. The future The invasiveness of the Pb O2 technique will always be an issue and may limit its usefulness in patients with coagulopathy, for instance. Combining parameters such as ICP and Pb O2 into a single probe may reduce the number of probes inserted, but the ideal remains a non-invasive monitor. Better integration of the collected data with continuous online derived indices at the bedside may also facilitate patient optimization. Routine use of dynamic challenges (e.g. increases in CPP or FI O2 when cerebral hypoxia presents), may identify individualized therapeutic targets. Conclusion Brain tissue oxygen partial pressure measurement contributes to the prevention of delayed cerebral damage after TBI and SAH. The integrated, continuous monitoring of Pb O2 is now accepted as being safe and feasible. Reliable bedside Pb O2 monitoring, both at baseline, and during the course of interventions such as CPP manipulation and hyperventilation, complements the use of existing cerebral monitors, and with the increasingly multi-modal approach of cerebral monitoring and data recording, may allow appropriate individualization of therapy. The rapid response of Pb O2 sensors to altering tissue oxygen levels, together with further Pb O2 threshold data, will allow more accurate identification of adverse cerebral conditions. Coupled with modalities such as ICP, transcranial Doppler and brain chemistry (microdialysis), more timely intervention and prognostication may be allowed. Pb O2 measurement, therefore, is emerging as a useful clinical tool and, taken within the context of 103

10 Nortje and Gupta underlying pathology, probe positioning, and responses and trends, there is mounting evidence that Pb O2 -guided therapy may bring us a step closer to the goal of outcome improvement after brain injury. Acknowledgement J.N. is funded by a British Journal of Anaesthesia/Royal College of Anaesthetists Fellowship. References 1 Alves OL, Daugherty WP, Rios M. Arterial hyperoxia in severe head injury: a useful or harmful option? Curr Pharm Des 2004; 10: Bouma GJ, Muizelaar JP, Choi SC, Newlon PG, Young HF. Cerebral circulation and metabolism after severe traumatic brain injury: the elusive role of ischemia. J Neurosurg 1991; 75: Bramlett HM, Dietrich WD. Pathophysiology of cerebral ischemia and brain trauma: similarities and differences. J Cereb Blood Flow Metab 2004; 24: Buchner K, Meixensberger J, Dings J, Roosen K. Near-infrared spectroscopy not useful to monitor cerebral oxygenation after severe brain injury. Zentralbl Neurochir 2000; 61: Bullock MR. Hyperoxia: good or bad? J Neurosurg 2003; 98: , discussion Carmona Suazo JA, Maas AI, van den Brink WA, van Santbrink H, Steyerberg EW, Avezaat CJ. CO 2 reactivity and brain oxygen pressure monitoring in severe head injury. Crit Care Med 2000; 28: Charbel FT, Du X, Hoffman WE, Ausman JI. Brain tissue PO 2, PCO 2, and ph during cerebral vasospasm. Surg Neurol 2000; 54: 432 7, discussion Charbel FT, Hoffman WE, Misra M, Hannigan K, Ausman JI. Cerebral interstitial tissue oxygen tension, ph, HCO 3,CO 2. Surg Neurol 1997; 48: Clausen T, Khaldi A, Zauner A, et al. Cerebral acid base homeostasis after severe traumatic brain injury. J Neurosurg 2005; 103: Clausen T, Scharf A, Menzel M, et al. Influence of moderate and profound hyperventilation on cerebral blood flow, oxygenation and metabolism. Brain Res 2004; 1019: Clayton TJ, Nelson RJ, Manara AR. Reduction in mortality from severe head injury following introduction of a protocol for intensive care management. Br J Anaesth 2004; 93: Clifton GL, Miller ER, Choi SC, et al. Lack of effect of induction of hypothermia after acute brain injury. N Engl J Med 2001; 344: Coles JP, Fryer TD, Smielewski P, et al. Incidence and mechanisms of cerebral ischemia in early clinical head injury. J Cereb Blood Flow Metab 2004; 24: Coles JP, Fryer TD, Smielewski P, et al. Defining ischemic burden after traumatic brain injury using 15O PET imaging of cerebral physiology. J Cereb Blood Flow Metab 2004; 24: Critchley GR, Bell BA. Acute cerebral tissue oxygenation changes following experimental subarachnoid hemorrhage. Neurol Res 2003; 25: Cruz J. The first decade of continuous monitoring of jugular bulb oxyhemoglobin saturation: management strategies and clinical outcome. Crit Care Med 1998; 26: Dings J, Meixensberger J, Jager A, Roosen K. Clinical experience with 118 brain tissue oxygen partial pressure catheter probes. Neurosurgery 1998; 43: Doppenberg EM, Zauner A, Watson JC, Bullock R. Determination of the ischemic threshold for brain oxygen tension. Acta Neurochir Suppl (Wien) 1998; 71: Doppenberg EMR, Zauner A, Bullock PD, Ross JD, Fatouros PP, Young HF. Correlations between brain tissue oxygen tension, carbon dioxide tension, ph, and cerebral blood flow a better way of monitoring the severely injured brain? Surg Neurol 1998; 49: Edelman GJ, Hoffman WE, Rico C, Ripper R. Comparison of brain tissue and local cerebral venous gas tensions and ph. Neurol Res 2000; 22: Elf K, Nilsson P, Enblad P. Outcome after traumatic brain injury improved by an organized secondary insult program and standardized neurointensive care. Crit Care Med 2002; 30: Garnett MR, Blamire AM, Corkill RG, Cadoux-Hudson TA, Rajagopalan B, Styles P. Early proton magnetic resonance spectroscopy in normal-appearing brain correlates with outcome in patients following traumatic brain injury. Brain 2000; 123: Gelabert-Gonzalez M, Fernandez-Villa JM, Ginesta-Galan V. Intra-operative monitoring of brain tissue O 2 (PtiO2) during aneurysm surgery. Acta Neurochir (Wien) 2002; 144: 863 6, discussion Gopinath SP, Robertson CS, Contant CF, et al. Jugular venous desaturation and outcome after head injury. J Neurol Neurosurg Psychiatry 1994; 57: Gopinath SP, Valadka AB, Uzura M, Robertson CS. Comparison of jugular venous oxygen saturation and brain tissue PO 2 as monitors of cerebral ischemia after head injury. Crit Care Med 1999; 27: Graham DI, Adams JH, Doyle D. Ischaemic brain damage in fatal non-missile head injuries. J Neurol Sci 1978; 39: Graham DI, Lawrence AE, Adams JH, Doyle D, McLellan DR. Brain damage in fatal non-missile head injury without high intracranial pressure. J Clin Pathol 1988; 41: Gupta AK, Al-Rawi PG, Hutchinson PJ, Kirkpatrick PJ. Effect of hypothermia on brain tissue oxygenation in patients with severe head injury. Br J Anaesth 2002; 88: Gupta AK, Hutchinson PJ, Al-Rawi P, et al. Measuring brain tissue oxygenation compared with jugular venous oxygen saturation for monitoring cerebral oxygenation after traumatic brain injury. Anesth Analg 1999; 88: Gupta AK, Hutchinson PJ, Fryer T, et al. Measurement of brain tissue oxygenation performed using positron emission tomography scanning to validate a novel monitoring method. J Neurosurg 2002; 96: Gupta AK, Zygun DA, Johnston AJ, et al. Extracellular brain ph and outcome following severe traumatic brain injury. J Neurotrauma 2004; 21: Hemphill JC, 3rd, Knudson MM, Derugin N, Morabito D, Manley GT. Carbon dioxide reactivity and pressure autoregulation of brain tissue oxygen. Neurosurgery 2001; 48: , discussion Hillered L, Persson L, Ponten U, Ungerstedt U. Neurometabolic monitoring of the ischaemic human brain using microdialysis. Acta Neurochir (Wien) 1990; 102: Hlatky R, Valadka AB, Goodman JC, Contant CF, Robertson CS. Patterns of energy substrates during ischemia measured in the brain by microdialysis. J Neurotrauma 2004; 21:

11 Tissue oxygen monitoring in acute brain injury 35 Hoelper BM, Alessandri B, Heimann A, Behr R, Kempski O. Brain oxygen monitoring: in-vitro accuracy, long-term drift and response-time of Licox- and Neurotrend sensors. Acta Neurochir (Wien) 2005; 147: Hoffman WE, Charbel FT, Edelman G. Brain tissue oxygen, carbon dioxide, and ph in neurosurgical patients at risk for ischemia. Anesth Analg 1996; 82: Hoffman WE, Charbel FT, Edelman G. Desflurane increases brain tissue oxygenation and ph. Acta Anaesthesiol Scand 1997; 41: Hoffman WE, Charbel FT, Edelman G, Ausman JI. Brain tissue gases and ph during arteriovenous malformation resection. Neurosurgery 1997; 40: , discussion Hoffman WE, Charbel FT, Edelman G, Ausman JI. Brain tissue oxygenation in patients with cerebral occlusive disease and arteriovenous malformations. Br J Anaesth 1997; 78: Hoffman WE, Charbel FT, Gonzalez-Portillo G, Ausman JI. Measurement of ischemia by changes in tissue oxygen, carbon dioxide, and ph. Surg Neurol 1999; 51: Hoffman WE, Edelman G. Enhancement of brain tissue oxygenation during high dose isoflurane anesthesia in the dog. J Neurosurg Anesthesiol 2000; 12: Hoffman WE, Edelman G, Wheeler P. Cerebral oxygen reactivity in the dog. Neurol Res 2000; 22: Hoffman WE, Wheeler P, Edelman G, Charbel FT, Torres NJ, Ausman JI. Hypoxic brain tissue following subarachnoid hemorrhage. Anesthesiology 2000; 92: Hutchinson PJ, Hutchinson DB, Barr RH, Burgess F, Kirkpatrick PJ, Pickard JD. A new cranial access device for cerebral monitoring. Br J Neurosurg 2000; 14: Jaeger M, Soehle M, Meixensberger J. Brain tissue oxygen (PtiO 2 ): a clinical comparison of two monitoring devices. Acta Neurochir Suppl 2005; 95: Jaeger M, Soehle M, Schuhmann MU, Winkler D, Meixensberger J. Correlation of continuously monitored regional cerebral blood flow and brain tissue oxygen. Acta Neurochir (Wien) 2005; 147: 51 6, discussion Jiang JY, Gao GY, Li WP, Yu MK, Zhu C. Early indicators of prognosis in 846 cases of severe traumatic brain injury. J Neurotrauma 2002; 19: Johnston AJ, Steiner LA, Chatfield DA, et al. Effects of propofol on cerebral oxygenation and metabolism after head injury. Br J Anaesth 2003; 91: Johnston AJ, Steiner LA, Chatfield DA, et al. Effect of cerebral perfusion pressure augmentation with dopamine and norepinephrine on global and focal brain oxygenation after traumatic brain injury. Intensive Care Med 2004; 30: Johnston AJ, Steiner LA, Coles JP, et al. Effect of cerebral perfusion pressure augmentation on regional oxygenation and metabolism after head injury. Crit Care Med 2005; 33: , discussion Kett-White R, Hutchinson PJ, Al-Rawi PG, et al. Cerebral oxygen and microdialysis monitoring during aneurysm surgery: effects of blood pressure, cerebrospinal fluid drainage, and temporary clipping on infarction. J Neurosurg 2002; 96: Kett-White R, Hutchinson PJ, Al-Rawi PG, Gupta AK, Pickard JD, Kirkpatrick PJ. Adverse cerebral events detected after subarachnoid hemorrhage using brain oxygen and microdialysis probes. Neurosurgery 2002; 50: , discussion Kiening KL, Unterberg AW, Bardt TF, Schneider GH, Lanksch WR. Monitoring of cerebral oxygenation in patients with severe head injuries: brain tissue PO 2 versus jugular vein oxygen saturation. J Neurosurg 1996; 85: Lang EW, Czosnyka M, Mehdorn HM. Tissue oxygen reactivity and cerebral autoregulation after severe traumatic brain injury. Crit Care Med 2003; 31: Longhi L, Stocchetti N. Hyperoxia in head injury: therapeutic tool? Curr Opin Crit Care 2004; 10: Magnoni S, Ghisoni L, Locatelli M, et al. Lack of improvement in cerebral metabolism after hyperoxia in severe head injury: a microdialysis study. J Neurosurg 2003; 98: Manley GT, Pitts LH, Morabito D, et al. Brain tissue oxygenation during hemorrhagic shock, resuscitation, and alterations in ventilation. J Trauma 1999; 46: Marion DW, Darby J, Yonas H. Acute regional cerebral blood flow changes caused by severe head injuries. J Neurosurg 1991; 74: McLeod AD, Igielman F, Elwell C, Cope M, Smith M. Measuring cerebral oxygenation during normobaric hyperoxia: a comparison of tissue microprobes, near-infrared spectroscopy, and jugular venous oximetry in head injury. Anesth Analg 2003; 97: Meixensberger J, Dings J, Kuhnigk H, Roosen K. Studies of tissue PO 2 in normal and pathological human brain cortex. Acta Neurochir Suppl (Wien) 1993; 59: Meixensberger J, Jaeger M, Vath A, Dings J, Kunze E, Roosen K. Brain tissue oxygen guided treatment supplementing ICP/CPP therapy after traumatic brain injury. J Neurol Neurosurg Psychiatry 2003; 74: Meixensberger J, Kunze E, Barcsay E, Vaeth A, Roosen K. Clinical cerebral microdialysis: brain metabolism and brain tissue oxygenation after acute brain injury. Neurol Res 2001; 23: Meixensberger J, Renner C, Simanowski R, Schmidtke A, Dings J, Roosen K. Influence of cerebral oxygenation following severe head injury on neuropsychological testing. Neurol Res 2004; 26: Meixensberger J, Vath A, Jaeger M, Kunze E, Dings J, Roosen K. Monitoring of brain tissue oxygenation following severe subarachnoid hemorrhage. Neurol Res 2003; 25: Menon DK. Procrustes, the traumatic penumbra, and perfusion pressure targets in closed head injury. Anesthesiology 2003; 98: Menon DK, Coles JP, Gupta AK, et al. Diffusion limited oxygen delivery following head injury. Crit Care Med 2004; 32: Menzel M, Doppenberg EM, Zauner A, Soukup J, Reinert MM, Bullock R. Increased inspired oxygen concentration as a factor in improved brain tissue oxygenation and tissue lactate levels after severe human head injury. J Neurosurg 1999; 91: Nordstrom CH, Reinstrup P, Xu W, Gardenfors A, Ungerstedt U. Assessment of the lower limit for cerebral perfusion pressure in severe head injuries by bedside monitoring of regional energy metabolism. Anesthesiology 2003; 98: Noseworthy MD, Bray TM. Effect of oxidative stress on brain damage detected by MRI and in vivo 31P-NMR. Free Radic Biol Med 1998; 24: Patel HC, Menon DK, Tebbs S, Hawker R, Hutchinson PJ, Kirkpatrick PJ. Specialist neurocritical care and outcome from head injury. Intensive Care Med 2002; 28: Pennings FA, Bouma GJ, Kedaria M, Jansen GF, Bosch DA. Intraoperative monitoring of brain tissue oxygen and carbon dioxide pressures reveals low oxygenation in peritumoral brain edema. J Neurosurg Anesthesiol 2003; 15: Persson L, Hillered L. Chemical monitoring of neurosurgical intensive care patients using intracerebral microdialysis. J Neurosurg 1992; 76: Reinert M, Barth A, Rothen HU, Schaller B, Takala J, Seiler RW. Effects of cerebral perfusion pressure and increased fraction of 105