019_mulvey_az

Neurocritical CareCopyright 2004 Humana Press Inc. All rights of any nature whatsoever are reserved.
ISSN 1541-6933/04/3:XXX–XXX Translational Research
Multimodality Monitoring in Severe Traumatic Brain
Injury
The Role of Brain Tissue Oxygenation Monitoring
Jamin M. Mulvey,1*, Nicholas W.C. Dorsch,2 Yugan Mudaliar,1 and Erhard W Lang,2 1Department of Intensive Care, University of Sydney,Westmead Hospital,Westmead Australia, and2Department of Neurosurgery University of Sydney,Westmead Hospital,Westmead Australia Abstract
Traumatic brain injury (TBI) is a major cause of morbidity and mortalitywith widespread social, personal, and financial implications for those whosurvive. TBI is caused by four main events: motor vehicle accidents, sport-ing injuries, falls, and assaults. Similarly to international statistics, annu-al incidence reports for TBI in Australia are between 100 and 288 per 100,000.
Regardless of the cause of TBI, molecular and cellular derangements occurthat can lead to neuronal cell death. Axonal transport disruption, ionic dis-ruption, reduced energy formation, glutamate excitotoxicity, and free rad-ical formation all contribute to the complex pathophysiological process ofTBI-related neuronal death. Targeted pharmacological therapy has not *Correspondence and
proved beneficial in improving patient outcome, and monitoring and main- reprint requests to:
tenance of various physiological parameters is the mainstay of current therapy. Parameters monitored include arterial blood pressure, blood gases, intracranial pressure, cerebral perfusion pressure, cerebral blood flow, and brain tissue oxygenation. Currently, indirect brain oximetry is used for cerebral oxygenation determination, which provides some information regarding global oxygenation levels. Direct brain tissue oxygenation (ptiO2), a newly developed oximetry technique, has shown promising results for the early detection of cerebral ischaemia. ptiO2 monitoring provides a safe, easy, and sensitive method of regional brain oximetry, providing a greaterunderstanding of neurophysiological derangements and the potential forcorrecting abnormal oxygenation earlier, thus improving patient outcome.
This article reviews the current status of bedside monitoring for patientswith TBI and considers whether ptiO 2 ___________________________________________________________________________________Mulvey et al. Key Words: Brain tissue partial pressure of oxy-
Literature was identified through Medline and gen; intracranial pressure; cerebral blood flow PubMed searches using the key words autoreg- velocity; monitoring; severe head injury; cerebral ulation, brain tissue oxygen tension pressure, ischaemia; transcranial Doppler ultrasound; cerebral blood flow velocity, cerebrovascular perfusion, Licox, severe head injury, and tran-scranial Doppler ultrasound (TCD). A reference Introduction
library distributed by GMS (Kiel-Mielkendorf,Germany) and the senior author’s library was Injury to the brain causes significant mor- bidity and mortality through various mecha-nisms. Traumatic brain injury (TBI), regardless Mechanisms of Cellular Injury:
of the cause, has profound personal, social, and Primary and Secondary Injuries
financial implications to those directly and indi-rectly involved. TBI can be classified as mild, Research over the past 20–30 years has elicit- moderate, or severe. Severe TBI, which is the ed much information on the mechanisms lead- main focus of this review, is clinically defined ing to neuronal cell death. It has been shown as any head injury that results in a postresusci- that in both human and animal tissues, regard- tation Glasgow Coma Scale of 8 or less on admis- less of the precipitating factors (i.e., traumatic, sion or during the ensuing 48 hours (1). Studies ischaemic, hypoglycaemic), the basic mecha- of hospital admissions report that over 80% of nisms underlying neuronal degeneration and TBI admissions are for mild-to-moderate injury, eventual death share similar cellular and molec- whereas severe TBI accounts for 5–15% (2–4).
The overall mortality of patients with severe TBI The processes that contribute to neuronal Fig 1 who survive to reach hospital is between 25 and damage after injury can be classified into two main groups: the primary injury and the sec- ondary injury (17–19). Direct brain injury, or the and secondary injury allows intensive care primary injury, results from both the direct physicians and neurosurgeons to target thera- impact to the brain and the changing forces py (8). Monitoring devices are used to detect involved from the sudden deceleration at the disturbances of physiological parameters with- moment of impact. Large forces occur from in the brain. Based on data obtained by multi- acceleration, deceleration and rotation of the brain inside the cranium. Shearing forces occur measures may be used to correct abnormal val- between tissue planes of varying densities ues and potentially decrease patient morbidity (20–22). This leads to immediate primary injury and mortality. Because current neuroprotective at the moment of trauma. The traumatic forces, pharmacotherapy has not proven beneficial as well as causing immediate structural dam- (9–11), more emphasis is being placed on mon- age to the neurons, cause secondary disruptions itoring systemic and brain levels of physiolog- in membrane stability, intra-axonal cytoskele- ical parameters as well as substrate availability tal function, and axonal transport mechanisms (12–16). It is hypothesized that as monitoring (20). Data from experimental models of TBI have devices improve and by maintaining substrate shown that postevent impairment of antero- availability within the normal physiological grade axoplasmic transport occurs, leading to range, the extent of secondary injuries will be local axonal swelling (23–25). With disorgani- reduced and patient outcome will improve.
zation of microtubules and neurofilaments, con- The purpose of this article is to review the tinuation of this process leads to axonal current status of bedside monitoring in the man- disconnection, degradation, and distal degen- agement of patients with TBI and evaluate the role of direct brain tissue oxygenation moni- Many aspects of the primary injury are imme- toring (ptiO2) in the intensive care setting.
diate and irreversible, but it seems likely that a Neurocritical Care ♦ Volume 1, 2004 Monitoring Modalities in Traumatic Brain Injury _____________________________________________________3 Fig. 1. A schematic diagram representing the molecular events implicated in secondary neuronal injury caused
by ischemia. Regardless of the pathological etiology, the sequences of events are intimately related and lead to
neuronal death.
continuum exists between the primary injury flow. The aerobic metabolism of glucose includes and the development of the secondary injury the initial step of glycolysis, the tricarboxylic (8). Although currently elusive, treatment aimed acid cycle, and the electron transport chain.
Glucose is metabolized in the presence of oxy- injury, or even the earlier cessation of the pro- gen to produce a higher ATP yield than occurs gression of the primary injury, may influence under hypoxic conditions. For an in-depth the management and outcomes of TBI (10,11).
review of this topic, see ref. 26.
Secondary injury after insult is correlated to In an ischemic insult, loss of blood flow leads impaired cerebral metabolism, hypoxia, and to decreased availability of oxygen and glucose.
ischemia, and a complex series of events ensue.
Anaerobic metabolism is a largely inefficient Although a detailed outline of these processes form of energy production, and as a result, rapid are beyond the scope of this article, a brief syn- energy failure follows with decreased produc- opsis of the mechanisms involved are present- tion of ATP (27). With decreasing levels of ATP, ed, including mechanisms that may be clinically the physiological ionic homeostasis of the neu- monitored in the intensive care unit (ICU).
ron is lost. Changes in the intracellular concen-tration of sodium, potassium, and calcium occur, Cerebral Metabolism
leading to cellular injury and death. With pro- Oxygen delivery is paramount to the normal gressive switching to anaerobic metabolism, lac- metabolism of neurons. It is used in a variety of tate production rises sharply, as demonstrated reactions within different cellular components by the lactate/pyruvate ratio (28–31). Increased to ultimately generate energy in the form of lactate concentration and, therefore, tissue pH adenosine 5’-triphosphate (ATP) by aerobic glu- have been shown to correlate with a poor out- cose metabolism. Aerobic metabolism is the major source of energy formation in the brain, and neuronal survival relies on an adequate sup- ply of oxygen and glucose by cerebral blood deranged as a result of ischemic events, and Neurocritical Care ♦ Volume 1, 2004 4 ___________________________________________________________________________________Mulvey et al. regional hypo- and hypermetabolism are known mental studies (47,48). This effect was not repli- to occur (36). Depressed cerebral activity, mito- cated in TBI clinical trials using the calcium chan- autoregulatory capacity of metabolic activity significant improvement in outcome (16,49,50).
and substrate delivery have been strongly impli- Glutamate Excitotoxicity
Mitochondrial Dysfunction
during cerebral ischemia. Glutamate, an exci- Mitochondria, which house the machinery for tatory neurotransmitter, is released in larger aerobic energy production, play an important quantities during cerebral ischemia than dur- role in aerobic metabolism. Mitochondrial dys- ing normal physiological conditions and leads function has been implicated in the impaired to opening of glutamate receptors and further activation of ion channels. Of particular signif- episodes, including those resulting from TBI icance is the sodium/calcium antiporter ion (39,40). Although not completely understood, channel, which leads to an acute increase of both the contribution of mitochondria to cerebral cations intracellularly (51). The N-methyl-D- ischemic damage includes the impairment of aspartate (NMDA) and a-amino-3-hydroxy-5- ATP production, changes in mitochondrial per- meability, and the release of factors that con- glutamate receptors have been linked to the tribute to cell death (41). The most widely influx of calcium. The NMDA receptor directly accepted hypothesis regarding mitochondrial opens a calcium channel, allowing a rapid influx dysfunction relates to the mitochondrial per- of the calcium ion. The activated AMPA recep- meability transition (MPT) (42,43). MPT occurs tor opens a sodium channel allowing rapid as a result of the abnormal opening of protein influx of the sodium ion. Both ions, which are channels between the inner and outer mito- increased uncontrollably in ischemia, lead to the chondrial membrane secondary to ischemia.
physiological derangements previously out- This results in mitochondrial swelling, mem- lined. Increased intracellular calcium concen- brane depolarization, loss of oxidative phos- trations also stimulate glutamate release from phorylation, and the release of proapoptotic presynaptic vesicles, further potentiating the proteins (44). The ischemic induction of mito- pathological process (52). Although it would chondrial dysfunction is a potential target for seem plausible that interventions targeting the neuroprotective interventions and is currently glutamate excitotoxic cascade would improve outcomes in patients with TBI, clinical trialsusing the NMDA antagonist selfotel showed no Calcium-Induced Cellular Damage
significant improvement in the outcome of TBI Loss of calcium homeostasis, with calcium entry into injured neurons, has long been asso- Free Radical Formation
ciated with the process of delayed cell death(45,46). Calcium is physiologically important Reperfusion injury caused by the production because it acts as a messenger to regulate the of free radicals has been theorized to contribute activity of lipolytic enzymes, proteolytic to secondary injury and delayed cell death.
enzymes, protein kinases, protein phosphatases, Oxygen free radicals are formed by the reper- and gene activation/expression. During insults fusion-initiated metabolism of free fatty acids such as ischemia or TBI, intracellular calcium and arachidonic acid. The increased free radi- increases uncontrollably and induces abnormal cal formation leads to increased lipid peroxi- cellular machinery leading to neuronal death.
dation, protein oxidation, and DNA damage Calcium antagonism has shown its utility as a (57). The integrity of the cellular lipid membrane neuroprotective agent in preclinical experi- is compromised, which leads to failure of ionic Neurocritical Care ♦ Volume 1, 2004 Monitoring Modalities in Traumatic Brain Injury _____________________________________________________5 partitioning and general cellular functioning, known. With the high incidence of autoregula- contributing to cell death. Clinical trials target- tion dysfunction during TBI, global oxygena- ing the various pathological pathways described tion measurements may be in the normal range above have been investigated (58–63). Trials and not reflect abnormal regional differences.
using pegorgotein, tirilazad, or triamcinolone Probes can be used to measure regional val- have shown no significant improvement in over- ues of brain tissue oxygen tension, carbon diox- all morbidity or mortality in patients with TBI ide tension, and hydrogen ion concentrations (70,72–74). These multiparametric sensors areplaced adjacent to the ICP monitoring catheter The Utility of Combined Monitoring
in the brain tissue via a modified skull bolt. Two Overview
types of commercially available ptiO2 probes currently exist: Licox® and Neurotrend®. The venting or reducing secondary injury. Following Licox probe (GMS, Kiel-Mielkendorf, Germany) the poor results seen in pharmacotherapy clin- uses a polarographic cell in which oxygen dif- ical trials, current therapies focus on providing fuses from the tissue through a polyethylene an environment in which the body’s own cel- wall of the catheter into its inner electrolyte lular restorative processes are promoted.
chamber (Fig. 2A,B). Oxygen is transformed at Systemic physiological parameters, including the electrode, where it determines a current that blood pressure, blood sugar level, electrolytes, reflects the tissue partial pressure of oxygen.
and partial pressure of arterial dioxide (PaO The oxygen-sensitive sampling area of the polarographic gold cathode is approx 14 mm2.
addition, specific cerebral parameters are equal- ly important in neurologic intensive care.
MA) uses optical sensors where dye, embedded The neurological monitoring modalities cur- in a plastic matrix, is connected to a fibreoptic rently available can be classified into three types: cable. Depending on the gas concentration and pressure, flow, and oxygenation. Monitoring pH of the surrounding tissues, the dye alters its modalities include intracranial pressure (ICP) properties, changing light transmission and monitoring, TCD, and jugular venous oximetry reflecting tissue partial pressure of oxygen. The Neurotrend probe is comprised of four sensors 2). A new modality, which is still largely used as an experimental modality, is ptiO and is able to measure ptiO2, ptiCO2, pH, and define ptiO2 as (32,66,71). The physiological data gathered by temperature. The sampling area of the Au: Pls using these monitoring modalities may allow Neurotrend probe is approximately 2 mm2.
greater understanding of the complex sequence ptiO2 probes generally are placed in the right of events that influence the final outcome in TBI.
frontal lobe white matter in diffuse brain injury, ICP and cerebral perfusion pressure (CPP) are or on the affected side in a hemispheric injury, the most important monitoring parameters on and remain in situ for as long as ICP measure- which therapeutic interventions are instituted.
ments are required (69,75). ptiO2 probes are read- However, both reveal little in terms of cerebral ily identified on computed tomography (CT) scanning (Fig. 3). This allows for correct place-ment and the accurate detection of oxygenation Invasive Cerebral Tissue Oxygen
in either normal or pericontusional brain.
Monitoring
Currently available monitoring methods of cerebral oxygenation and cerebral blood flow models (76). Studies have shown that in TBI, detect a “global” measurement. The data ptiO2 values in patients with normal ICP and obtained imply that the brain acts as a homog- CPP are between 25 and 30 mmHg (77,78). The enous structure; however, the heterogeneity of critical threshold for ischemic damage and a brain activity and substrate utilization is well- poorer outcome has been proposed at ptiO2 val- Neurocritical Care ♦ Volume 1, 2004 6 ___________________________________________________________________________________Mulvey et al. Fig. 2. (A) A schematic diagram of the Licox polarographic oxygenation probe.The numbered components of
the diagram are: (1) polyethylene tube diffusion membrane; (2) polarographic gold cathode; (3) polarographic sil-
ver anode; (4) cell filled with electrolyte; and (5) cerebral tissue. (B) A schematic diagram of the Licox probe
illustrating placement via a cranial bolt into the cerebral tissues. Placement is similar to ICP monitoring and is
often used through the same bolt.
ues of 10–15 mmHg (69,77,79,81). Critical thresh- microenvironment, with low velocities show- old is not the only factor that is important in ing the highest variability in terms of oxygena- terms of outcome; the duration spent below that tion differences (83). At times, the disparity between the different probe types can be appre- The metabolic heterogeneity of different tis- ciated, because sampling areas are quite differ- sue types is well-known. It is important to fac- tor the heterogeneous nature of the brain when compensated for by a sufficiently large sensor interpreting oximetry data. Experiments on rats have demonstrated the differing ptiO2 within Comparative Studies
the cortex depending on the depth of probeplacement (82). It was proposed that the differ- ing base levels related to the metabolism, micro- Cerebral blood flow is physiologically regu- circulation, and overall microstructure of each lated by several factors, including pressure of environment. Furthermore, depending on the blood flow, the pressures within the cranial probe’s relationship to the arterial microvessels, vault, and vascular autoregulatory processes.
a gradient within the tissues can exist with oxy- Following TBI, alterations in ICP and CPP are gen levels decreasing from artery to venous cir- commonplace. A few studies have investigated culation. The microenvironment is influenced the association between CPP, ICP, and ptiO2. A by the cerebral blood flow velocity of each prospective study of 23 patients with TBI inves- Neurocritical Care ♦ Volume 1, 2004 Monitoring Modalities in Traumatic Brain Injury _____________________________________________________7 associated with higher ptiO2. This suggests thatthe critical threshold of CPP is 60 mmHg andthat ptiO2 is more sensitive than SjvO2 to changesin CPP. In contrast, Hartl et al. (85) report thattreatment of ICP with mannitol was not asso-ciated with improvements in ptiO2. However,it should be noted that in this study, ICP wastreated before it was severely raised (23 ± 1mmHg), and initial CPP before treatment was68 ± 2 mmHg.
Focal ischemic tissue may at times have nor- mal CPP but decreased ptiO2. In a prospectivestudy of nine patients who demonstrated acutefocal lesions on CT scan and/or single photonemission computed tomography (SPECT) fromeither subarachnoid hemorrhage (SAH), TBI, ormeningioma, changes in ptiO2 were investigat-ed in relation to increased MAP and CPP (86).
ptiO2 increased from 24 ± 13 mmHg to 31 ± 13 Au: Pls mmHg in a positive linear fashion when CPP define MAPincreased from initial values of 77 ± 9 mmHg to96 ± 11 mmHg (r2 = 0.74). However, in somepatients with an initial ptiO2 below 20 mmHg,CPP was considered to be already within thenormal range. These data suggest that althoughCPP values above 60 mmHg are usually asso- Fig. 3. A computer tomography image demonstrating
the position of a Licox oxygenation probe in the frontal cortex of a patient with TBI. Oxygenation probes are always accurate enough to assess brain tissue readily identifiable on scanning modalities, illustrating the position relative to contusional tissue and regions form a major focus of current treatment inpatients with TBI. Although severe alterations tigated the effects of aggressive treatment of CPP of ICP and CPP are correlated with poor out- when below 60 mmHg. Dopamine infusion was come, studies suggest that other methods of always associated with an increase in ptiO2 (66).
monitoring would provide additional, and at Intervention led to significant elevations of CPP times more sensitive, information regarding from 32 ± 2 to 67 ± 4 mmHg and of ptiO2 from cerebral blood flow and substrate availability.
13 ± 2 to 19 ± 3 mmHg. When initial CPP exceed- Changes in ptiO2 are often detected concurrently ed 60 mmHg, further CPP elevation did not sig- with changes in CPP, but ptiO2 can be low (or nificantly improve ptiO2, suggesting a plateau even within the hypoxic range) even with nor- phase of oxygenation. Another prospective mal values of CPP (86). Arecent study has shown study , comparing different methods of oxy- that in 18 of 26 patients after aneurysmal SAH genation monitoring in 17 patients with TBI or severe TBI who had a unilateral decompres- showed that decreases in CPP below 60 mmHg sion hemicraniectomy for extensive cerebral were significantly correlated with decreases in oedema, pathological monitoring trends always ptiO2 (84). Furthermore, changes in SjvO2 were proceeded clinical deterioration (87). In 9 of 20 not significant when correlated with decreased patients with SAH, decreases in ptiO2 occurred CPP, and CPP values above 60 mmHg were not several hours before neurological deterioration Neurocritical Care ♦ Volume 1, 2004 8 ___________________________________________________________________________________Mulvey et al. or ICP increase. This was not always the case to CBF monitoring would provide increased for patients with TBI. It is plausible that multi- accuracy in interpreting CBF values.
modal monitoring of ICP, CPP, and ptiO2 could Although investigators have reported on the improve the sensitivity of detection of decreased validity of cerebrovascular autoregulation cerebral blood flow and substrate availability.
assessment and its prognostic relationship to Therefore, early treatment interventions should outcome, particularly related to CPP and CBFV, increase the viability of injured and noninjured few have compared the correlation between neuronal tissue, thereby improving patient out- autoregulation (88–90). It appears thatCBF/CBFV and ptiO but during autoregulatory dysfunction and Various investigators have studied the cor- relation between CBF and ptiO2, particularly in would at times provide misleading information the initial periods of TBI when derangements regarding potential ischemic episodes. A recent in both CBF and ptiO2 are often at their great- publication by of one the present authors study- est. In considering these two clinical variables, ing autoregulatory function of ptiO2 in 14 it is important to remember that ptiO2 reflects patients with TBI, demonstrated a plateau phase regional values, whereas CBF, depending on the for the CPP–ptiO2 relationship similar to the modality used, may reflect either macro- or autoregulatory plateau seen in the relationship between CPP and CBFV (71). When autoregu- Doppenburg et al. investigated the correla- lation was impaired, ptiO2 increased in a linear tions between CBF (Xenon computed tomogra- fashion with increases in CPP. If autoregulation phy technique) and ptiO2 in 25 patients with TBI remained intact, then increases in CPP had min- and described a significant linear relationship imal effect on ptiO2. It was concluded that between the two modalities (r = 0.74, p = 0.0001) manipulation of CPP was only of potential ben- (32). Patients with increased CBF showed high- efit in increasing brain oxygenation if autoreg- er ptiO2, whereas those with decreased CBF had ulatory mechanisms were dysfunctional.
a lower ptiO2, below 26 mmHg. All patients in Furthermore, they suggested that continuous this study with ptiO2 below 25 mmHg either ptiO2 monitoring would provide more sensitive information on the integrity of autoregulation Dings et al. investigated the relationship after TBI, directing accurate therapy.
between ptiO2, CBF velocity (CBFV), and CO2 Cerebral oxygen reactivity/autoregulation reactivityin 17 patients with TBI (78). Low mean has been assessed in patients with TBI by chang- values for both ptiO2 and CBFV were seen on ing the fractional inspired oxygen concentration the day of injury (7.7 ± 2.6 mmHg and 60.5 ± (FiO2) (33,91). The ability to increase ptiO is par- 32.0 cm/second, respectively). Both variables ticularly useful in conditions where normal increased, and by day 4 ptiO2 was 31.5 ± 10.0 autoregulatory function is impaired. By increas- mmHg and CBFV was 87.9 ± 21.0 cm/second.
ing FiO2 from 35 to 100%, ptiO2 is able to be The authors concluded that although ptiO2 and increase to supranormal levels, allowing for aer- obic metabolism. It has been proposed that FiO2 increased further, suggesting vasospasm and manipulation can improve oxygenation better uncoupling of flow and metabolism. To further than CPP manipulation; however, patients with support these findings, they discovered that at a high oxygen reactivity (indicating a signifi- times during increased CBFV, both CPP and cant disturbance in autoregulation) have a poor- ptiO2 were seen to decrease, indicating uncou- pling or dysfunction of autoregulation. Thissuggests that ptiO2 monitoring as an adjuvant Neurocritical Care ♦ Volume 1, 2004 Monitoring Modalities in Traumatic Brain Injury _____________________________________________________9 stability was improved. Gopinath et al. also global brain tissue oxygenation monitoring of ately after insertion; however, values usually patients with TBI since the early 1980s (92–94).
stabilized within 60 minutes (67). van den Brink et al. reported low sensitivity drift (0 ± 6%) and being used to detect ischemic episodes in negligible zero drift in ptiO2 (68). All authors patients with TBI (67,69,70,62,80,85). ptiO ures direct regional oxygen tension levels, and was a reliable method of detecting brain tissue investigators have compared the utility of ptiO ischemia over a prolonged period of time and try, it is important to consider (a) calibration, (b) The efficiency and quality of information the time of good-quality data (TGQD), and (c) gathered by the different methods of oximetry can be quantified and compared. One methodis through the function of TGQD, expressed by the equation: TGQD (%) = 100 – [time of arte- The initial calibration of any monitoring facts (minute) × 100/total monitoring time device is crucial to obtaining accurate and reli- able data. Based on the manufacturer’s recom- In an investigation comparing ptiO2 and SjvO2 mendations, ptiO2 catheters are calibrated monitoring in 15 patients with TBI and altered before insertion and after withdrawal from the CPP, TGQD and the total duration of monitor- brain tissue; no intramonitoring calibration is ing differed greatly between the two oximetry possible. Two calibration parameters have been methods (77). The median duration of moni- described for ptiO2 catheters: sensitivity cali- toring reported was 9 days (range: 5–12) for bration and zero drift (68,70,77). Sensitivity cal- ptiO2 and 4 days (range: 3–7) for SjvO2. TGQD ibration is defined as the difference in measured was reported at 95% (2491 hours total) and 43% oxygen tension when room oxygen is measured, (607 hours) for ptiO2 and SjvO2, respectively.
and zero drift is the difference in an oxygen-free This difference in the SjvO2 arm was attributed solution. Calibration of SjvO2 is based on co- to poor light intensity in the system, repetitive calibrations, and dislocations. Meixensberger et every 10–12 hours for the duration of its usage.
al. reported similar disparities of TGQD between ptiO2 and SjvO2 monitoring (96). This prospec- have minimal drift during continuous moni- tive study of 45 patients with TBI reported toring. In a prospective study of 15 patients with TGQD for ptiO2 and SjvO2 at 95 and 40–50%, TBI comparing ptiO2 and SjvO2 monitoring, respectively. Only five patients were monitored ptiO2 monitoring showed low variability (3.7% with SjvO2 for comparison because of increas- sensitivity drift) and greater reliability over time ing technical difficulties and poor reliability.
(77). SjvO2 monitoring required a total of 170 Similar problems for SjvO2 have been reported calibrations over 7 days, with 55% of calibra- tions showing an increased drift (>5%) when Dings et al. have also studied the reliability compared with co-oximetry. In a study by Dings of ptiO2 (70). Investigating the technical and et al. reporting on the stability and complica- diagnostic reliability of ptiO2 monitoring, 118 tions of ptiO2 monitoring in 70 patients with catheter probes were used in 101 patients with either TBI or SAH, 54 Licox catheters showed a TBI. The TGQD was 99.2%, with artifacts relat- drift of –6.2 ± 11.9% (95). Sensitivity drift was ed to transport, positioning of the patient, and greatest in situ during the first 4 days, after which displacement of the catheter or the bolt. Dings Neurocritical Care ♦ Volume 1, 2004 10 __________________________________________________________________________________Mulvey et al. et al. concluded that ptiO2 was a safe and reli- The technique of ptiO2 probe placement is able technique for monitoring cerebral oxy- almost identical to ICP monitor placement. Thus, it seems plausible that ptiO2 probe could be However, not all studies have found ptiO2 to inserted by practitioners other than neurosur- be superior to SjvO2 in the detection of critical geons. Aretrospective study looking at the com- ischemic episodes. A prospective study com- plication rates of ICP probe insertion by paring the utility of the methods in 65 patients neurosurgeons, general surgical registrars, and with TBI concluded that both modalities should intensivists found no significant difference in be used in conjunction and that neither identi- complication rates between the different groups fies all episodes of cerebral ischemia (67). Of 65 (98). They concluded that the use of non-neu- patients, 7 were unable to have ptiO2 data col- rosurgeons for the placement of probes could lected because of technical difficulties. Of those provide the prompt and early monitoring of monitored, no significant difference was found high-risk patients. We propose that it would be in the TGQD, with values of 90 and 88% for SjvO2 safe practice to utilize non-neurosurgeons for and ptiO2 monitoring, respectively (p = 0.524).
ptiO2 probe insertion; however, a neurosurgeon Decreases in oxygenation were detected simul- should be on standby if complications occur.
taneously in 90% of episodes; however, only 66% Complication rates for SjvO2 monitoring are of these episodes saw both modalities below similarly low. Gopinath et al. reported zero com- plications related to SjvO2 monitoring in 58patients (67). Kiening et al. reported dislodge- ment as a main complication but did not quan- Oximetry is an invasive procedure and car- tify the rate (77). In a prospective study of 44 ries a potential for complications related to inser- patients admitted to ICU for TBI, SAH, or stroke tion and continuous monitoring. For routine and requiring SjvO2 monitoring, complication ICU purposes, probes are inserted via single or rates were below 5% and were clinically insignif- multiple lumen bolts if other monitoring modal- ities are combined. Depending on the hospital’spolicies, bolts can be inserted in the ICU. In oper- Brain Oxygenation and Outcome
ative cases, probes can be inserted directly dur- No randomized control trials have been con- ducted to demonstrate improved outcome with In studies to date, complication rates for both one monitoring modality over another. ICP mon- ptiO2 and SjvO2 are low. Numerous studies itoring has become routine practice in the neuro- using ptiO2 have reported complication rates ICU worldwide, although it has never been below 3% (67,70,95). These complications subjected to randomized controlled trials.
involved secondary hematoma formation, none Uncontrolled intracranial hypertension is neg- of which required treatment. The insertion trau- atively correlated with outcome (100–102). It ma can cause microhemorrhages and an odema also seems plausible that reduced brain oxy- zone around the probe tract (97). This has min- genation would be correlated with a poorer out- imal effect on measurements and does not com- promise accuracy. Complications were related Regardless of the lack of controlled trial data, to technical issues such as the accidental removal current clinical trials investigating the utility of of the catheter during transport, broken catheter ptiO2 suggest that prolonged periods of hypox- cables, or unidentified technical problems. These ia correlate with a poor outcome. van Santbrink technical problems were reduced with experi- et al. studied the utility of ptiO2 in 22 patients ence, and larger studies have reported zero com- with TBI and showed that hypoxic periods in plication rates for the insertion of ptiO2 catheters the acute posttraumatic phase was common (69).
More than 80% of patients showed prolonged Neurocritical Care ♦ Volume 1, 2004 Monitoring Modalities in Traumatic Brain Injury ____________________________________________________11 hypoxic periods less than 20 mmHg in the first improved significantly over the past 30 years, 24 hours postinjury. In five patients, ptiO2 fell mortality is still alarmingly high in those who below 5 mmHg within the first 24 hours, and survive to hospital. Because the pharmacologi- four of those were either dead or partial vege- cal management of TBI is currently poor and tative state at 6 months. In the patients who had still under extensive research, the integration monitoring within the acute phase without a and management of physiological variables ptiO2 drop below 5 mmHg, 15 had good out- remain the mainstay of current therapy.
come measures, and only 1 died or was vegeta- tive at 6 months. ptiO2 was found to be strongly attention and has generated regular interna- tional meetings. Although it has become a rou- Kiening et al. have also demonstrated poor tine monitoring tool in several neurosurgical outcome with reduced brain oxygenation in TBI and neurological ICUs, it is still considered (66). In 16 patients followed for 6 months postin- experimental in other centers. Based on the data jury, the number of ischemic episodes was asso- available, it has been shown to provide a safe, ciated with outcome. An ischemic episode was easy-to-use, and accurate method of cerebral defined as a ptiO2 less than 10mmHg for longer oximetry determination. It can provide addi- than 15 minutes. In the first week postinjury, the tional, sensitive information regarding brain numbers of ischemic episodes were always asso- oxygen availability, autoregulation, and brain ciated with a poorer outcome on the Glasgow perfusion in patients with TBI. Compared to Outcome Scale (GOS). Interestingly, absence of other oximetry methods, ptiO2 has minimal episodic hypoxia did not ensure a favorable out- complications, increased accuracy, and greater come. Bardt et al. also demonstrated poor out- in situ monitoring time. ptiO2 often provides come with prolonged ischemic episodes (81). In more sensitive information than current moni- 35 patients with TBI, analysis of data showed toring methods regarding regional CBF, CPP, significant differences in outcome measures ICP, and oxygen availability. Indeed, some cur- rent therapeutic interventions used to manipu- than 30 minutes. In patients with less than 30 minutes of hypoxia during the monitoring peri- oxygenation may, in fact, cause hypoxia. Brain od, GOS analysis at discharge demonstrated that tissue oxygenation monitoring has the poten- 80% were either vegetative or severely disabled, tial to detect early ischemic injury before alter- 20% had a favorable outcome, and no patients ations in other variables occur and may improve died acutely. In this same group, GOS at 6 months showed that 72.8% had a favorable out-come, 18.2% were vegetative or severely dis- References
abled, and 9% died. In contrast, in patients with 1. Teasdale G, Jennett B. Assessment of coma and all refer- more than 30 minutes of hypoxia, 48% died impaired consciousness. Apractical scale. Lancet ences, acutely and 52% had an unfavorable GOS at dis- charge. In those discharged, GOS at 6 months 2. Hillier SL, Hiller JE, Metzer J. Epidemiology of first 6 showed that 55.6% had died, 22.2% were severe- traumatic brain injury in South Australia. Brain but et al.
ly disabled, and 22.2% had a favorable outcome.
Bardt et al. concluded that even short periods 3. Jennett B. Epidemiology of head injury. J Neurol if there of hypoxia adversely affected outcome and that Neurosurg Psychiatry 1996;60:362–369.
functional recovery is possible in the prolonged 4. Sorensen SB, Kraus JF. Occurrence, severity and outcomes of brain injury. J Head Trauma Rehab1991;6:1–10.
Conclusions
5. Murray GD, Teasdale GM, Braakman R, et al.
The European Brain Injury Consortium survey TBI is a major cause of morbidity and mor- of head injuries. Acta Neurochir 1999;141: tality worldwide. Although management has Neurocritical Care ♦ Volume 1, 2004 12 __________________________________________________________________________________Mulvey et al. 6. Lang EW, Pitts LH, Damron SL, Rutledge R.
20. Povlishock JT, Christman CW. The pathobiolo- Outcome after severe head injury: an analysis of gy of traumatically induced axonal injury in ani- prediction based upon comparison of neural net- mals and humans: a review of current thoughts.
work versus logistic regression analysis. Neurol 21. Jennett B, Teasdale G. Structural Pathology. In: 7. Marshall L, Gautille T, Klauber M, et al. The out- Jennett B, Teasdale G, eds. Management of Head come of severe closed head injury. J Neurosurg Injuries. Vol. 20. Philadelphia: F.A. Davis 8. Reilly PL. Brain injury: the pathophysiology of 22. Maxwell WL, Watt C, Graham DI, Gennarelli TA.
the first hours. ‘Talk and die revisited’. J Clin Ultrastructural evidence of axonal shearing as a result of lateral acceleration of the head in non- 9. Faden AI. Neuroprotection and traumatic brain injury: the search continues. Arch Neurol 2001; 23. Erb DE, Povlishock JT. Axonal damage in severe 10. Clausen T, Bullock R. Medical treatment and neu- traumatic brain injury: an experimental study in roprotection in traumatic brain injury. Curr cat. Acta Neuropathol 1988;76:347–358.
24. Maxwell WL, Irvine A, Graham DI, et al. Focal 11. Maas AI. Neuroprotective agents in traumatic axonal injury: the early axonal response to brain injury. Expert Opin Investig Drugs 2001; stretch. J Neurocytol 1991;20:157–164.
25. Povlishock JT, Becker DP, Cheng CLY, Vaughan 12. Narayan RK, Michel ME, Ansell B, et al. Clinical GW. Axonal change in minor head injury. J trials in head injury. J Neurotrauma 2002; 19:503- Neuropath Exp Neurol 1983;42:225–242.
26. Zauner A, Daugherty WP, Bullock MR, Warner 13. Morris GF, Bullock R, Marshall SB, Marmarou DS. Brain oxygenation and energy metabolism: A, Maas AIR, Marshall LF. Failure of the com- Part I-biological function and pathophysiology.
petitive N-methyl-d-aspartate antagonist Selfotel (CGS 19755) in the treatment of severe head 27. Siesjo BK. Cerebral circulation and metabolism.
injury: results of two phase III clinical trials. The Selfotel Investigators. J Neurosurg 1999;91:737–743.
28. Zauner A, Clausen T, Alves OL, et al. Cerebral metabolism after fluid-percussion injury and 14. Marshall LF, Maas A, Marshall S, et al. A multi- centre trial on the efficacy of using tirilazad mesy- late in cases of head injury. J Neurosurg 1998;89:519–525.
29. Valadka AB, Goodman JC, Gopinath SP, Uzura 15. Miller JD, Marshall LF. Are steroids useful in the M, Robertson CS. Comparison of brain tissue treatment of head-injured patients? Surg Neurol oxygen tension to microdialysis-based measures of cerebral ischemia in fatally head-injuredhumans. J Neurotrauma 1998;15:509–519.
16. Maas AI, Steyerberg EW, Murray GD, et al. Why have recent trials of neuroprotective agents in 30. Rehncrona S, Rosen I, Siesjo BK. Brain lactic aci- head injury failed to show convincing efficacy? dosis and ischemic cell damage. I. Biochemistry A pragmatic analysis and theoretical considera- and neurophysiology. J Cereb Blood Flow Metab tions. Neurosurgery 1999;44:1286–1298.
17. Graham DI, Adams JH, Doyle D. Ischaemic brain 31. Pulsinelli WA, Levy D, Duffy TE. Regional cere- damage in fatal non-missile head injuries. J bral blood flow and glucose metabolism fol- lowing transient forebrain ischemia. Ann Neurol 18. Chesnut RM, Marshall LF, Klauber MR, et al. The role of secondary brain injury in determining 32. Doppenberg EMR, Zauner A, Bullock MR, Ward outcome from severe head injury. J Trauma JD, Fatouros PP, Young HF. Correlations between brain tissue oxygen tension, carbon dioxide, pH, 19. Chesnut RM. Secondary brain insults after head and cerebral blood flow—a better way of moni- injury: clinical perspectives. New Horiz 1995 toring the severely injured brain? Surg Neurol Neurocritical Care ♦ Volume 1, 2004 Monitoring Modalities in Traumatic Brain Injury ____________________________________________________13 33. Menzel M, Doppenberg E, Zauner A, Soukup J, 46. Siesjo BK, Bengtsson F. Calcium fluxes, calcium Reinert M, Bullock R. Increased inspired oxygen antagonists, and calcium-related pathology in concentration as a factor in improved brain tis- brain ischemia, and spreading depression: a uni- sue oxygenation and tissue lactate levels after fying hypothesis. J Cereb Blood Flow Metab 1989; severe human head injury. J Neurosurg 1999; 47. Kazda S, Garthoff B, Krause HP, Schlossmann K.
34. Menzel M, Doppenberg EM, Zauner A, et al.
Cerebrovascular effects of the calcium antago- Cerebral oxygenation in patients after severe nistic dihydropyridine derivative nimodipine in head injury: monitoring and effects of arterial animal experiments. Arzneimittelforschung hyperoxia on cerebral blood flow, metabolism and intracranial pressure. J Neurosurg Anesthe- 48. Archer DP, Pappius HM. Effects of two dihy- dropyridine calcium channel blockers on cere- 35. Voldby B, Enevoldsen EM, Jensen FT.
bral metabolism and blood flow in traumatized Cerebrovascular reactivity in patients with rup- rat brain. Neurochem Pathol 1986;5:117–130.
tured intracranial aneurysms. J Neurosurg 49. Teasdale G, Bailey I, Bell A, et al. A randomized trial of nimodipine in severe head injury: HIT I.
36. Ritter AM, Robertson CS. Cerebral metabolism.
British/Finnish Co-operative Head Injury Trial Neurosurg Clin N Am 1994;5:633–645.
Group. J Neurotrauma 1992;9:S545–S550.
37. Sutton RL, Hovda DA, Adelson PD, Benzel EC, 50. Harders A, Kakarieka A, Braakman R. Traumatic Becker DP. Metabolic changes following cortical subarachnoid hemorrhage and its treatment with contusion: relationship to edema and morpho- nimodipine. German tSAH study group. J Neuro- logical changes. Acta Neurochir (suppl) 1994; 51. Choi DW. Ionic dependance of glutamate neu- 38. Mulvey JM, Renshaw GMC. Brainstem neuronal rotoxicity. J Neurosci 1987;7:369–379.
oxidative hypometabolism in response to hypox- 52. White BC, Sullivan JM, DeGracia DJ, et al. Brain ic pre-conditioning in the epaulette shark ischemia and reperfusion: molecular mecha- (Hemiscyllium ocellatum). Neurosci Lett 2000 nisms of neuronal injury. J Neurol Sci 2000; 39. Bouma GJ, Muizelaar JP, Choi SC, Newlon PG, 53. McCulloch J, Ozyurt E, Park CK, Nehls DG, Young HF. Cerebral circulation and metabolism Teasdale GM, Graham DI. Glutamate receptor in severe traumatic brain injury: the elusive role antagonists in experimental focal cerebral of ischemia. J Neurosurg 1991;75:685–693.
ischaemia. Acta Neurochir Suppl (Wien) 1993; 40. Xiong Y, Peterson PL, Lee CP. Alterations in cere- bral energy metabolism induced by traumatic 54. Park CK, Nehls DG, Graham DI, Teasdale GM, brain injury. Neurol Res 2001;23:129–138.
McCulloch J. Focal cerebral ischaemia in the cat: 41. Sims NR, Anderson MF. Mitochondrial contri- treatment with the glutamate antagonist MK-801 butions to tissue damage in stroke. Neurochem after induction of ischaemia. J Cereb Blood Flow 42. Cromptom M. The mitochondrial permeability 55. Panter SS, Faden AI. Pretreatment with NMDA transition pore and its role in cell death. Biochem antagonists limits release of excitatory amino acids following traumatic brain injury. Neurosci 43. Blomgren K, Zhu C, Hallin U, Hagberg H.
Mitochondria and ischemic reperfusion damage 56. Tsuchida E, Rice M, Bullock R. The neuropro- in the adult and in the developing brain. Biochem tective effect of the forebrain-selective NMDA antagonist CP101,606 upon focal ischemic brain 44. Neumar RW. Molecular mechanisms of ischemic damage caused by acute subdural hematoma in the rat. J Neurotrauma 1997;14:409–417.
57. Siesjo BK. Basic mechanisms of traumatic brain 45. Goldberg MP, Choi DW. Combined oxygen and damage. Ann Emerg Med 1993;22:959–969.
glucose deprivation in cortical cell culture: cal- 58. Smith SL, Andrus PK, Gleason DD, Hall ED.
Infant rat model of the shaken baby syndrome: mechanisms of neuronal injury. J Neurosci 1993; preliminary characterization and evidence for the role of free radicals in cortical hemorrhaging Neurocritical Care ♦ Volume 1, 2004 14 __________________________________________________________________________________Mulvey et al. 71. Lang EW, Czosnyka M, Mehdorn HM. Tissue oxygen reactivity and cerebral autoregulation 59. Clark WM, Hazel JS, Coull BM. Lazaroids. CNS after severe traumatic brain injury. Crit Care Med 72. Gupta AK, Hutchinson P, Al-Rawi P, et al.
60. Wienkers LC, Steenwyk RC, Mizsak SA, Pearson Measuring brain tissue oxygenation compared PG. In vitro metabolism of tirilazad mesylate in with jugular venous oxygen saturation for mon- itoring cerebral oxygenation after traumatic cytochrome P4502C11 and delta 4-5 alpha-reduc- brain injury. Anesthesia and Analgesia 1999; tase. Drug Metab Dispos 1995;23:383–392.
61. Smith SL, Andrus PK, Zhang JR, Hall ED. Direct 73. Zauner A, Bullock R, Di X, Young HF. Brain oxy- measurement of hydroxyl radicals, lipid perox- gen, CO2, pH and temperature monitoring: eval- idation, and blood-brain barrier disruption fol- uation in the feline brain. Neurosurgery 1995 lowing unilateral cortical impact head injury in the rat. J Neurotrauma 1994;11:393–404.
74. Zauner A, Doppenberg EMR, Woodward JJ.
62. Hall ED, Andrus PK, Yonkers PA. Brain hydrox- Multiparametric continuous monitoring of brain yl radical generation in acute experimental head metabolism and substrate delivery in neurosur- injury. J Neurochem 1993;60:588–594.
gical patients. Neurol Res 1997;19:265–273.
63. Sanada T, Nakamura T, Nishimura MC, Isayama 75. Meixensberger J, Baunach S, Amschler J, Dings K, Pitts LH. Effect of U74006F on neurologic func- J, Roosen K. Influence of body position on tis- tion and brain edema after fluid percussion injury in rats. J Neurotrauma 1993;10:65–71.
intracranial pressure in patients with acute brain 64. Grumme T, Baethmann A, Kolodziejczyk D, et injury. Neurol Res 1997;19:249–253.
al. Treatment of patients with severe head injury 76. Maas AIR, Fleckenstein W, de Jong DA.
by triamcinolone: a prospective, controlled mul- Monitoring cerebral oxygenation: experimental ticenter clinical trial of 396 cases. Res Exp Med studies and preliminary clinical results of con- tinuous monitoring of cerebrospinal fluid and 65. Young B, Runge JW, Harrington T, et al. Effects brain tissue oxygen tension. Acta Neurochir of pegorgotein on neurologic outcome of patients with severe head injury. A multicenter, random- 77. Kiening KL, Unterberg AW, Bardt TF, Schneider ized controlled trial. JAMA 1996;276:538–543.
GH, Lanksch WR. Monitoring of cerebral oxy- 66. Kiening KL, Hartl R, Unterberg AW, Schneider genation in patients with severe head injuries: itoring in comatose patients: implications for uration. J Neurosurg 1996;85:751–757.
therapy. J Neurol Res 1997;19:233–240.
78. Dings J, Meixensberger J, Amschler J, Hamelbeck 67. Gopinath SP, Valadka AB, Uzura M, Robertson B, Roosen K. Brain tissue pO2 in relation to cere- CS. Comparison of jugular venous oxygen sat- bral perfusion pressure, TCD-findings and TCD- 2 reactivity after severe head injury. Acta bral ischemia after head injury. Crit Care Med 79. Dings J, Jager A, Meixensberger J, Roosen K. Brain 68. van den Brink WA, van Santbrink H, Steyerberg tissue pO2 after severe head injury. Neurol Res WW, et al. Brain oxygen tension in severe head injury. Neurosurgery 2000;46:868–878.
80. Valadka AB, Gopinath SP, Contant CF, Uzura M, 69. van Santbrink H, Maas A, Avezaat C. Continuous Robertson CS. Relationship of brain tissue pO2 monitoring of partial pressure of brain tissue to outcome after severe head injury. Crit Care oxygen in patients with severe head injury.
81. Bardt TF, Unterberg AW, Hartl R, Kiening KL, 70. Dings J, Meixensberger J, Jager A, Roosen K.
Scheider GH, Lanksch WR. Monitoring of brain Clinical experience with 118 brain tissue oxygen tissue pO2 in traumatic brain injury: effect of partial pressure catheter probes. Neurosurgery cerebral hypoxia on outcome. Acta Neurochir Neurocritical Care ♦ Volume 1, 2004 Monitoring Modalities in Traumatic Brain Injury ____________________________________________________15 82. Baumgartl H, Lubbers DW. Microcoaxial needle gen during CO2 and O2 reactivity tests in severe- sensors for polarographic measurements of local ly head-injured patients. Acta Neurochir (Wien) O2 in the cellular range of living tissue. In: Gnaiger E, Forstner H, eds. Polarographic 92. Robertson CS, Narayan RK, Contant CF, et al.
Oxygen Sensors. New York: Springer Verlag, Clinical experience with a continuous monitor of intracranial compliance. J Neurosurg 1989;71: 83. Lubbers DW, Baumgartl H. Heterogeneities and profiles of oxygen pressure in brain and kidney 93. Cruz J, Miner ME, Allen SJ, Alves WM, T.A. G.
as examples of pO2 distribution in the living tis- Continuous monitoring of cerebral oxygenation sue. Kidney International 1997;51:372–380.
in acute brain injury: assessment of cerebral 84. Sarrafzadeh AS, Unterberg AW, Keining KL, hemodynamic reserve. Neurosurgery 1991;29: Bardt T, Schneider GH, Lanksch WR. Monitoring of cerebral oxygenation in traumatic brain 94. Cruz J, Raps EC, Hoffstad OJ, Jaggi JL, Gennarelli injured patients. In: Bauer BL, Kuhn TJ, eds.
TA. Cerebral oxygenation monitoring. Crit Care Severe Head Injuries. Berlin: Springer Verlag, 95. Dings J, Meixensberger J, Roosen K. Brain tissue 85. Hartl R, Bardt T, Keining KL, Sarrafzadeh AS, pO2 monitoring: catheter stability and compli- Schneider GH, Unterberg AW. Mannitol decreas- cations. Neurol Res 1997;19:241–245.
es ICP but does not improve brain tissue pO2 in 96. Meixensberger J, Jager A, Dings J, Baunach S, severely head-injured patients with intracranial Roosen K. Quality and therapeutic advances in hypertension. Acta Neurochir 1997;70(suppl): multimodality neuromonitoring following head injury. In: Bauer BL, Kuhn TJ, eds. Severe Head 86. Stocchetti N, Chieregato A, De Marchi M, Croci Injuries. Berlin: Springer Verlag, 1997:109–120.
M, Benti R, Grimoldi N. High cerebral perfusion 97. van den Brink WA, Haitsma IK, Avezaat CJ, pressure improves low values of local brain tis- sue O2 tension (ptiO2) in focal lesions. Acta Neurochir 1998; 71(suppl):162–165.
histopathological study in the rat. J Neurotrauma 87. Strege RJ, Lang EW, Stark AM, et al. Cerebral edema leading to decompressive craniectomy: 98. Kaups KL, Parks SN, Morris CL. Intracranial An assessment of the preceding clinical and neu- pressure monitor placement by midlevel practi- romonitoring trends. Neurol Res 2003:accepted.
tioners. J Trauma 1998;45:884–886.
88. Czosnyka M, Smielewski P, Kirkpatrick P, Menon 99. Coplin WM, O’Keefe GE, Grady MS, et al.
DK, Pickard JD. Monitoring of cerebral autoreg- Thrombotic, infectious, and procedural compli- cations of the jugular bulb catheter in the inten- sive care unit. Neurosurg 1997;41:101–109.
89. Bouma G, Muizelaar J. Cerebral blood flow, cere- 100.Lang EW, Chesnut RM. Intracranial pressure.
bral blood volume, and cerebrovascular reac- Monitoring and management. Neurosurg Clin tivity after severe head injury. J Neurotrauma 90. Lang E, Mehdorn H, Dorsch N, Czosnyka M.
101.Narayan RK, Kishore PR, Becker DP, et al.
Intracranial pressure: to monitor or not to mon- autoregulation: a validation study. J Neurol itor? Areview of our experience with severe head Neuros Psychiatry 2002;72:583–586.
injury. J Neurosurg 1992;56:650–659.
91. Fandino J, Stocker R, Prokop S, Imhof HG.
102.Zink BJ. Traumatic brain injury outcome: con- Correlation between jugular bulb oxygen satu- ration and partial pressure of brain tissue oxy- Neurocritical Care ♦ Volume 1, 2004 e suitable for long term oxygenation monitoring was placed in an ischemic but salvageable part of the below 10 mmHg for longer than 10 min was always associ could be performed for twice the duration of SjvO Neither modality detected all ischemic episodes Both modalities would compliment each other Similar duration of monitoring between ptiO equent episodes of hypoxia associated with a poor esholds seen in 50% cases on day 1 posttrauma >20 mmHg, mannitol infusion significantly changed No significant change was detected with ptiO 18 __________________________________________________________________________________Mulvey et al. Traumatic brain injurySubarachnoid hemorrhageStrokeTumor Neurocritical Care ♦ Volume 1, 2004

Source: http://www.dabasons.com.br/downloads/O2%20in%20TBI%20lang%20et%20al.pdf