Cerebral Microdialysis Effects of Propofolversus Midazolam in Severe Traumatic Brain Injury

Michele Tanguy,1 Philippe Seguin,1 Bruno Laviolle,2 Jean-Paul Bleichner,1

Xavier Morandi,3 and Yannick Malledant1

Abstract

Propofol, an anesthetic agent acting as an analogue of vitamin E, has been advocated to be an ideal neuro-protective agent both in animal models and in clinical practice, due to its positive effects on oxidative stress.Nevertheless, no studies have compared this agent to another sedative agent used for sedation after traumaticbrain injury (TBI). The objective was to compare the effects of propofol to midazolam on cerebral biomarkers atthe acute phase of severe TBI patients. Thirty patients aged 35 – 18 years were prospectively randomized toreceive propofol or midazolam and 29 were analyzed (n = 15 for propofol, and n = 14 for midazolam). A cerebralmicrodialysis catheter was used to measure the lactate:pyruvate (L:P) ratio, glutamate, glycerol, and glucose for72 h. No difference between groups was observed for the L:P ratio (time effect p = 0.201, treatment effect p = 0.401,time · treatment interaction p = 0.101). Similarly, no difference was observed for glutamate (time effect p = 0.930,treatment effect p = 0.651, time · treatment interaction p = 0.353), glycerol (time effect p = 0.223, treatment effectp = 0.922, time · treatment interaction p = 0.308), or glucose (time effect p = 0.116, treatment effect p = 0.088,time · treatment interaction p = 0.235). These results do not support a difference between propofol and mid-azolam for sedation for the cerebral metabolic profile in severe TBI.

Key words: lactate:pyruvate ratio; microdialysis; sedative agents; trauma brain injury

Introduction

There is growing evidence that the damaged perme-ability of mitochondria in response to post-traumatic

oxidative stress plays a pivotal role in the metabolic crisisobserved after a severe traumatic brain injury (TBI) (Jayaku-mar et al., 2008; Mustafa et al., 2010). This mitochondrialdysfunction leads to apoptotic cerebral cell death, especiallywhen oxidative stress is strongly expressed (i.e., in brain areaswith higher neuronal content and respiratory rate), and mayaffect neuronal recovery after TBI (Marcoux et al., 2008;Mazzeo et al., 2009). Microdialysis analysis assessment in vivoat the early phase post-TBI has advanced our understanding,and has increased interest in several biomarkers. Notably, el-evations of the lactate:pyruvate (L:P) ratio appear to be a keymarker of this cascade of events (Vespa et al., 2005). Severalpharmacological means of neuroprotection have been ex-plored ( Jayakumar et al., 2008; Mazzeo et al., 2008; Mustafaet al., 2010; Singh et al., 2006). Propofol, an anesthetic agentacting as an analogue of vitamin E, has been advocated as anideal neuroprotective agent both in animal models and inclinical practice, due to its positive effects on oxidative stress

(Nakahata et al., 2008; Tsuchiya et al., 2001,2002). Never-theless, no studies have compared this agent to another sed-ative agent used for post-TBI sedation. The main objective ofthis study was to compare the effects of propofol to midazolamon the cerebral L:P ratio in TBI patients during the first 72 hafter surgical intensive care unit (SICU) admission. The ef-fects of these agents on other cerebral biomarkers, outcome at12 months assessed by the Glasgow Outcome Scale (GOS), andmemory tests were also evaluated.

Methods

This prospective, randomized, single-blinded study wasperformed in a SICU of a university hospital from June 2006 toSeptember 2008. This unit acts as a referral center and receivessevere TBI patients from local and regional institutions. Theprotocol was approved by the Comite de Protection des Per-sonnes of Rennes, France on March 8, 2006, and written in-formed consent was obtained from each patient’s next of kin.

The main objective of this study was to compare the effectsof propofol versus midazolam on the cerebral L:P ratio in TBIpatients during the first 72 h after SICU admission. Secondary

1Departement d’Anesthesie-Reanimation 1, 2Service de Pharmacologie et Centre d’Investigation Clinique, Universite Rennes 1, InsermU991, Service de Reanimation Chirurgicale, and 3Service de Neurochirurgie, Hopital de Pontchaillou, Cedex, France.

JOURNAL OF NEUROTRAUMA 29:1105–1110 (April 10, 2012)ª Mary Ann Liebert, Inc.DOI: 10.1089/neu.2011.1817

1105

objectives were to evaluate the effect of these drugs on cere-bral glutamate, glycerol, and glucose. Moreover, GOS andmemory tests were evaluated at 12 months.

Patients

All adult patients (age ‡ 18 years) with a severe and iso-lated closed head injury (Glasgow Coma Scale [GCS] score£ 8) requiring sedation, ventilation, and intracranial pressure(ICP) monitoring were eligible. Patients who had major focalinjuries ( > 25 cc), and/or who required neurosurgery forepidural and/or subdural hematomas, and/or who had co-agulation disorders and/or delayed admission to the SICU( > 12 h) were excluded.

All candidates were managed according to standardizedprotocols in order to maintain a mean arterial pressure (MAP)‡ 85 mm Hg, an ICP £ 20 mm Hg, and a cerebral perfusionpressure (CPP) ‡ 65 mm Hg (Vincent and Berre, 2005). Othertypes of therapy to avoid secondary insults were applied asrecommended (Vincent and Berre, 2005).

Data collection

The following data were recorded at inclusion: generalcharacteristics (age and sex), severity of illness as assessed bythe Simplified Acute Physiologic Score II (SAPS II) and GCS,core body temperature, vital signs (mean arterial pressure andoxygen saturation), blood glucose, carbon dioxide tension,and hemoglobin level. Cerebral hemodynamic parametersincluding ICP and CPP were noted, and computed tomog-raphy (CT) scan classification according to the TraumaticComa Data Bank (TCDB) was recorded (Marshall et al., 1992).During the study, core body temperature, MAP, cerebral he-modynamic parameters (ICP and CPP), arterial oxygen ten-sion, arterial carbon dioxide tension, and blood glucose wererecorded every 4 h.

At 12 months, a structured interview was used either withthe patient or with a caregiver or relative to allocate patients asassessed by the GOS into two outcome categories: A, goodrecovery or moderate disability; and B, death, vegetative state,or severe disability. A cognitive difficulties scale was also usedto assess memory complaints (Derouesne et al., 1993).

Investigated variables

All patients had an arterial catheter placed (Plastimed Di-vision, Prodimed, Le Plessis-Bouchard, France) to allow con-tinuous measurements of systemic arterial pressure. At thebedside of every patient a cerebral pressure sensor (CodmanMicrosensors ICP transducer; Codman & Shurtleff, Raynham,MA), and a cerebral microdialysis catheter (CMA 70; CMA/Microdialysis, Stockholm, Sweden) were inserted through aburr-hole into the frontal lobe and secured with a triple bolt.The pressure sensor allowed us to continuously monitor theICP. The microdialysis catheter was infused with Ringer’sHartmann solution at a rate of 0.3 lL/min via a pump (CMA106; CMA/Microdialysis). The L:P ratio, glutamate, glycerol,and glucose were measured using a CMA 600 bedside Mi-crodialysis Analyser (CMA/Microdialysis). The limit of de-tection was 0.1 mmol/L for lactate, 1 lmol/L for glutamate,10 lmol/L for pyruvate and glycerol, and 0.1 mmol/L forglucose. The normal values of the L:P ratio, glutamate,glycerol, and glucose were 23.0 – 4.0, 16.0 – 16.0 lmol/L,

82 – 44 lmol/L, and 1.7 – 0.9 mmol/L, respectively (Reinstrupet al., 2000).

Study protocol

After baseline data were taken, thee patients were ran-domized to receive propofol or midazolam. In the absence ofrecommendations concerning sedation in TBI, our objectivewas to obtain a persistent Ramsay sedation level of 4–5(Bratton et al., 2007; Ramsay et al., 1974). Accordingly, pro-pofol was infused at an initial rate of 1 mg/kg.h - 1, and in-creased if necessary by increments of 1 mg/kg.h - 1, with anupper limit of 5 mg/kg.h - 1. Midazolam was infused at aninitial rate of 0.03 mg/kg.h - 1, with increments of 0.01 mg/kg.h - 1, to obtain the objective level. Both groups receivedfentanyl, infused at a rate of 1–3 lg/kg.h - 1. Concerning themicrodialysis analysis, the first dialysate sample was obtainedbefore the sedation was started. Thereafter, microvials of themicrodialysis analyzer were changed every 120 min, andsamples were analyzed every 120 min for L:P ratio, glutamate,glycerol, and glucose for 72 h. If refractory high cerebralpressure requiring thiobarbiturate or craniectomy occurred,the microdialysis analysis was stopped.

Statistical analysis

We planned to enroll 24 patients (12 per group). Thissample size allowed us to detect, in a two-sided test with aType I error of 5% and a power of 95%, a 30% decrease in thedialysate L:P ratio with propofol, assuming a mean referencevalue of 36 – 8 (Hutchinson et al., 2000). Due to the possiblevariability of the L:P ratio, a safety margin of 6 patients wasscheduled. Data are presented as mean – standard deviation(SD) unless otherwise noted for continuous variables, and asthe number (corresponding percentage) for categorical vari-ables. Statistical analysis was performed with SAS softwareversion 9.2 (SAS Institute, Cary, NC). Comparison of theevolution of the cerebral biomarkers between the two groupswas performed using a two-way (time · treatment) analysis ofvariance (ANOVA; mixed models). All other comparisonsbetween the two groups were performed using the Student’st-test or the Wilcoxon rank-sum test when needed for con-tinuous variables, and the chi-square test or the Fischer’s exacttest as appropriate for categorical variables. For all analyses,p values < 0.05 were considered significant.

Results

During the study period 94 adult patients with isolatedclosed severe head injury were admitted to the SICU. Fifty-three patients met exclusion criteria (emergent neurosur-gery [n = 20], major hematoma [n = 6], imminent braindeath [n = 19], contraindications to monitoring [n = 5], andother [n = 3]), and 11 patients could not be monitored dueto technical problems. Thirty patients aged 35 – 18 yearswere randomized, but in one patient we could not insertthe microdialysis catheter due to technical problems.Therefore 29 patients were analyzed (n = 15 for the pro-pofol group, and n = 14 for the midazolam group). Therewere no significant differences between the two groups atbaseline for general characteristics, severity, and cerebralhemodynamics (Table 1), as well as for microdialysis pa-rameters (Table 2).

1106 TANGUY ET AL.

Figure 1 shows the evolution over time of the L:P ratio,glutamate, glycerol, and glucose. No difference betweenpropofol and midazolam was observed for the evolution ofthe L:P ratio during the 72-h follow-up period (time effectp = 0.201, treatment effect p = 0.401, time · treatment interac-tion p = 0.101). Similarly, no significant difference was ob-served between the two groups for glutamate (time effectp = 0.930, treatment effect p = 0.651, time · treatment interac-tion p = 0.353), glycerol (time effect p = 0.223, treatment effectp = 0.922, time · treatment interaction p = 0.308), and glucose(time effect p = 0.116, treatment effect p = 0.088, time · treat-ment interaction p = 0.235). Although not significant, meancerebral glucose concentrations tended to be higher withpropofol, ranging between + 37% and + 114% between 6 and24 h after the start of treatment. During the study period, nodifference was observed between the two groups for corebody temperature, MAP, ICP, CPP, arterial oxygen and car-bon dioxide tension, blood glucose, and catecholamine and

mannitol use. Moreover, the number of patients who receivedbarbiturates and/or decompressive surgery was no differentbetween the two groups.

Eleven patients (n = 7 for propofol, and n = 4 for mid-azolam) died during their SICU stay ( p = 0.316). All but twodied secondary to their brain trauma. One more patient diedin hospital after SICU discharge in the midazolam group. Onepatient was lost to follow-up at 12 months in the propofolgroup. The GOS score at 12 months did not differ between thetwo groups (propofol group A = 5 patients, B = 2 patients;midazolam group A = 6 patients, B = 3 patients; p = 1.000).Similarly, memory complaints at 12 months were no differentbetween the two groups (9 – 11 in the propofol group, and16 – 15 in the midazolam group; p = 0.440).

Discussion

The main result of this study is that in severe TBI patientswith a high L:P ratio, the variations in the extracellular L:Pratio were no different for the propofol and midazolamgroups over a 72-h period. This trial is, to the best of ourknowledge, the first to compare the effects of these sedativeagents on cerebral biomarkers in TBI.

This prospective and comparative study was undertaken todetect a putative neuroprotective effect of propofol in severeTBI. Indeed, the protective effects of propofol have been re-ported in numerous in vitro and in vivo models, due to itseffects on oxidative stress which are related to propofol’sphenolic chemical structure, similar to that of vitamin E,which is a major natural free radical scavenger (Kobayashiet al., 2008; Nakahata et al., 2008; Tsuchiya et al., 2001,2002).

Post-TBI increases in free radical generation have recentlyreceived considerable attention (Awasthi et al., 1997, Tyurinet al., 2000). Their proposed mechanisms include the arachi-donic acid cascade, the increased leakage of superoxide frommitochondrial electron transport, enhanced activity of xan-thine oxidase, auto-oxidation of catecholamines, activation ofneutrophils, and breakdown of hemoglobin with release ofiron (Awasthi et al., 1997). The brain contains a large amountof antioxidants. Lipophilic vitamin E acts at the cytoplasmicand membrane levels and is recycled by ascorbate (Li et al.,2003). Oxidative stress post-TBI may be synergistic or addi-tive with glutamate excitotoxicity (Peters et al., 2001; Sitaret al., 1999; Velly et al., 2003). Under normal circumstancesneurons are exposed to very brief pulses of glutamate, a majorexcitatory neurotransmitter. In TBI the interstitial compart-ment is at risk of glutamate flooding due to the inhibition ofglutamate reuptake carriers by oxidative stress (Peters et al.,2001; Sitar et al., 1999; Velly et al., 2003). This leads to thedestructive processes of the excitotoxic cascade, with exces-sive intracellular Ca2 + -induced mitochondrial dysfunctionand activation of pro-apoptotic mechanisms. Direct quantifi-cation of cerebral oxidative stress has been conducted in ex-perimental and animal studies (Awasthi et al., 1997; Tyurinet al., 2000). However, it is assumed that cerebral micro-dialysis can be indirectly used to assess oxidative stress(Vespa et al., 2005), and in a recent study performed in TBIpatients, a close relationship was demonstrated between adirect marker of oxidative stress and the routine microdialysismarkers of the post-TBI metabolic crisis (Clausen et al., 2011).

A clear link exists between severe TBI and a high L:P ratioin extracellular cerebral fluid (Hutchinson et al., 2000; Kerr

Table 1. Baseline Demographic and Clinical

Characteristics in the Propofol

and Midazolam Groups

Propofol Midazolam(n = 15) (n = 14) p Value

Age, years 37 – 20 33 – 18 0.645Sex, male/female 10/5 13/1SAPS II score 39 – 12 42 – 10 0.525GCS score 5 – 2 5 – 2 0.687Core body temperature, �C 37.0 – 1.3 37.7 – 1.3 0.167MAP, mm Hg 91 – 11 100 – 16 0.072Spo2, % 97 – 2 98 – 1 0.082Paco2, mm Hg 38 – 7 35 – 10 0.388Blood glucose, mmol/L 8 – 2 8 – 4 0.633Hemoglobin, g/dL 12 – 2 13 – 2 0.280ICP, mm Hg, 19 – 12 20 – 12 0.759median [range] 17 [2–45] 18 [5–40]CPP, mm Hg, 71 – 14 73 – 11 0.645median [range] 68 [49–105] 77 [48–93]TCDB CT scan class

II 11 7III 3 5 0.408IV 1 2

Data are expressed as mean – standard deviation for continuousvariables, and as number for categorical variables, unless otherwisenoted.

SAPS II, Simplified Acute Physiologic Score II; GCS, GlasgowComa Scale; MAP, mean arterial pressure; Spo2, oxygen saturation;Paco2, arterial carbon dioxide tension; ICP, intracranial pressure;CPP, cerebral perfusion pressure; TCDB, Traumatic Coma DataBank; CT, computed tomography.

Table 2. Baseline Levels of Cerebral

Microdialysis Biomarkers

Propofol Midazolam(n = 15) (n = 14) p Value

Lactate:pyruvate ratio 53 – 38 (47) 57 – 53 (40) 0.844Glucose, mmol/L 0.8 – 0.7 (0.6) 0.6 – 0.5 (0.6) 0.544Glycerol, lmol/L 105 – 117 (77) 90 – 56 (82) 0.743Glutamate, lmol/L 38 – 35 (33) 43 – 45 (27) 0.948

Data are expressed as mean – standard deviation (median).

SEDATIVE AGENTS AND CEREBRAL MICRODIALYSIS IN HEAD INJURY 1107

et al., 2003; Vespa et al., 2005). In recent years the hypoxia/ischemic insult was thought to be the cause. This hypoxia/ischemic insult takes place after brain herniation, a pre-terminal event with major metabolic failure in which amarked increase in the L:P ratio and a near-zero level of in-tracerebral glucose are present. But in more common types ofhead trauma such as those explored here, the elevation of theL:P ratio may have another explanation (Hlatky et al., 2004;Vespa et al., 2005). Indeed, it could be related to the mito-chondrial alterations present after severe TBI, with causativesroles for oxidative stress and glutamate excitotoxicity. Con-sequently, the L:P ratio and extracellular glutamate kineticsappear to be relevant factors to use to compare the effects ofpropofol and midazolam on cerebral oxidative stress.

Our data do not support the view that cerebral oxidativestress would be better controlled with propofol than mid-azolam. The L:P ratio, a marker of mitochondrial dysfunction,as well as glutamate, demonstrated similar levels in the twosedation groups. Finally, with regard to the putative neuro-protective effect of propofol, the only biomarker that showeda positive effect was cerebral glucose. Extracellular brainglucose tended to be better preserved with propofol duringthe first 24 h. This could be indirect evidence of a lower ex-posure to glutamate with propofol.

The accuracy of glycerol as a microdialysis marker in TBI isa matter of debate. It is an end product of membrane phos-pholipid degradation, and favorable effects have been shownof antioxidant molecules on glycerol liberation after TBI(Hillered et al., 2005; Peerdeman et al., 2003). In our study,however, no difference between propofol and midazolam wasobserved for glycerol release. In addition to the absence of

acute effects on biochemical markers, we did not find anydifferences between the two studied groups for GOS scoreand memory complaints at 1 year. Nevertheless, the numberof patients analyzed was clearly too small to draw definitiveconclusions about these outcomes.

Our study has several limitations. First, it was a single-blinded trial, but therapeutic goals and sedation levelswere independent from microdialysis biomarkers, be-cause the physicians were unaware of the biochemicalvalues. Second, the ability of microdialysis to analyzethese effects is obviously limited, and oxidative stress ispredominantly present in areas with high neuronal con-tent and O2 demands. Thus we cannot exclude that abeneficial effect of propofol might be observed in theseareas. The concentration of propofol used may have beeninsufficient to produce antioxidant effects; however,higher dosages are not recommended. Indeed, it is wellknown that propofol can induce propofol infusion syn-drome in head trauma patients receiving the drug forprolonged periods (Smith et al., 2009). In our study, weused a level of sedation that did not induce EEG sup-pression, so the infusion rate was limited to 5 mg/kg.h - 1,as recommended. Third, a 30% reduction in the L:P ratioplanned in our statistical methods may have been toolarge and negatively impact our results. Finally, it couldbe argued that midazolam possesses itself neuroprotec-tive effects. Nevertheless, the antioxidant effect of mid-azolam is weak, and is largely inferior to that reported forpropofol (Tsuchiya et al., 2001), although it has recently beenshown that benzodiazepines may exert a mitochochondrialprotective effect through other mechanisms. In consequence,

FIG. 1. Changes seen over time for the cerebral lactate:pyruvate ratio, glutamate, glycerol, and glucose. Data aremean – standard deviation.

1108 TANGUY ET AL.

they may interfere with the biomarkers we have chosen tostudy (Sarnowska et al., 2009; Tanabe et al., 2011).

In conclusion, our results indicate that there is no differ-ence between the effects of propofol and midazolam sedationon the cerebral metabolic profile during the acute phase ofsevere TBI. Accordingly, the use of propofol as a sedativeagent in TBI and its neuroprotective effects warrant furtherinvestigation.

Acknowledgments

The study was supported by a grant from ProgrammeHospitalier de Recherche Clinique (PHRC) of Rennes, October2005 (R0903).

Author Disclosure Statement

No competing financial interests exist.

References

Awasthi, D., Church, D.F., Torbati, D., Carey, M.E., and Pryor,W.A. (1997). Oxidative stress following traumatic brain injuryin rats. Surg. Neurol. 47, 575–581.

Bratton, S.L., Chestnut, R.M., Ghajar, J., McConnell Hammond,F.F., Harris, O.A., Hartl, R., Manley. G.T., Nemecek, A.,Newell, D.W., Rosenthal, G.,. Schouten, J., Shutter, L., Tim-mons, S.D., Ullman, J.S., Videtta, W., Wilberger, J.E., andWright, D.W. (2007). Guidelines for the management of severetraumatic brain injury. XI. Anesthetics, analgesics, and seda-tives. J. Neurotrauma 24, S71–S76.

Clausen, F., Marklund, N., Lewen, A., Enblad, P., Basu, S., andHillered, L. (2011). Interstitial F(2)-Isoprostane 8-Iso-PGF(2a)as a biomarker of oxidative stress after severe human trau-matic brain injury. J. Neurotrauma [Epub ahead of print].

Derouesne, C.D.M., Boyer, P., Lubin, S., Piette, F., Kohler, F., andAlperovitch, A. (1993). Empirical evaluation of the ‘cognitivedifficulties scale’ for assessment of memory complaints ingeneral practice: a study of 1628 cognitively normal subjectsaged 45–75 years. Int. J. Geriatr. Psych. 8, 599–607.

Hillered, L., Vespa, P.M., and Hovda, D.A. (2005). Translationalneurochemical research in acute human brain injury: thecurrent status and potential future for cerebral microdialysis.J. Neurotrauma 22, 3–41.

Hlatky, R., Valadka, A.B., Goodman, J.C., Contant, C.F., andRobertson, C.S. (2004). Patterns of energy substrates duringischemia measured in the brain by microdialysis. J. Neuro-trauma 21, 894–906.

Hutchinson, P.J., al-Rawi, P.G., O’Connell, M.T., Gupta, A.K.,Maskell, L.B., Hutchinson, D.B., Pickard, J.D., and Kirkpatrick,P.J. (2000). On-line monitoring of substrate delivery and brainmetabolism in head injury. Acta Neurochir. Suppl. 76, 431–435.

Jayakumar, A.R., Rao, K.V., Panickar, K.S., Moriyama, M.,Reddy, P.V., and Norenberg, M.D. (2008). Trauma-inducedcell swelling in cultured astrocytes. J. Neuropathol. Exp.Neurol. 67, 417–427.

Kerr, M.E., Ilyas Kamboh, M., Yookyung, K., Kraus, M.F., Puc-cio, A.M., DeKosky, S.T., and Marion, D.W. (2003). Relation-ship between apoE4 allele and excitatory amino acid levelsafter traumatic brain injury. Crit. Care Med. 31, 2371–2379.

Kobayashi, K., Yoshino, F., Takahashi, S.S., Todoki, K., Maehata,Y., Komatsu, T., Yoshida, K., and Lee, M.C. (2008). Direct

assessments of the antioxidant effects of propofol mediumchain triglyceride/long chain triglyceride on the brain ofstroke-prone spontaneously hypertensive rats using electronspin resonance spectroscopy. Anesthesiology 109, 426–435.

Li, X., Huang, J., and May, J.M. (2003). Ascorbic acid sparesalpha-tocopherol and decreases lipid peroxidation in neuronalcells. Biochem. Biophys. Res. Commun. 305, 656–661.

Marcoux, J., McArthur, D.A., Miller, C., Glenn, T.C., Villablanca,P., Martin, N.A., Hovda, D.A., Alger, J.R., and Vespa, P.M.(2008). Persistent metabolic crisis as measured by elevatedcerebral microdialysis lactate-pyruvate ratio predicts chronicfrontal lobe brain atrophy after traumatic brain injury. Crit.Care Med. 36, 2871–2877.

Marshall, L.F., Marshall, S.B., Klauber, M.R., Van Berkum Clark,M., Eisenberg, H., Jane, J.A., Luerssen, T.G., Marmarou, A.,and Foulkes, M.A. (1992). The diagnosis of head injury re-quires a classification based on computed axial tomography. J.Neurotrauma 9 Suppl. 1, S287–S292.

Mazzeo, A.T., Alves, O.L., Gilman, C.B., Hayes, R.L., Tolias, C.,Niki Kunene, K., and Ross Bullock, M. (2008). Brain metabolicand hemodynamic effects of cyclosporin A after human severetraumatic brain injury: a microdialysis study. Acta Neurochir.(Wien.) 150, 1019–1031.

Mazzeo, A.T., Beat, A., Singh, A., and Bullock, M.R. (2009). Therole of mitochondrial transition pore, and its modulation, intraumatic brain injury and delayed neurodegeneration afterTBI. Exp. Neurol. 218, 363–370.

Mustafa, A.G., Singh, I.N., Wang, J., Carrico, K.M., and Hall,E.D. (2010). Mitochondrial protection after traumatic braininjury by scavenging lipid peroxyl radicals. J. Neurochem.114, 271–280.

Nakahata, K., Kinoshita, H., Azma, T., Matsuda, N., Hama-Tomioka, K., Haba, M., and Hatano, Y. (2008). Propofolrestores brain microvascular function impaired by high glu-cose via the decrease in oxidative stress. Anesthesiology 108,269–275.

Peerdeman, S.M., Girbes, A.R., Polderman, K.H., and Vander-top, W.P. (2003). Changes in cerebral interstitial glycerolconcentration in head-injured patients; correlation with sec-ondary events. Intensive Care Med. 29, 1825–1828.

Peters, C.E., Korcok, J., Gelb, A.W., and Wilson, J.X. (2001).Anesthetic concentrations of propofol protect against oxida-tive stress in primary astrocyte cultures: comparison withhypothermia. Anesthesiology 94, 313–321.

Ramsay, M.A., Savege, T.M., Simpson, B.R., and Goodwin, R.(1974). Controlled sedation with alphaxalone-alphadolone. Br.Med. J. 2, 656–659.

Reinstrup, P., Stahl, N., Mellergard, P., Uski, T., Ungerstedt, U.,and Nordstrom, C.H. (2000). Intracerebral microdialysis inclinical practice: baseline values for chemical markers duringwakefulness, anesthesia, and neurosurgery. Neurosurgery 47,701–709.

Sarnowska, A., Beresewicz, M., Zablocka, B., and Domanska-Janik, K. (2009). Diazepam neuroprotection in excitotoxic andoxidative stress involves a mitochondrial mechanism addi-tional to the GABAAR and hypothermic effects. Neurochem.Int. 55, 164–173.

Singh, I.N., Sullivan, P.G., Deng, Y., Mbye, L.H., and Hall, E.D.(2006). Time course of posttraumatic mitochondrial oxidativedamage and dysfunction in a mouse model of focal traumaticbrain injury: implications for neuroprotective therapy. J. Cer-eb. Blood Flow Metab. 26, 1407–1418.

Sitar, S.M., Hanifi-Moghaddam, P., Gelb, A., Cechetto, D.F.,Siushansian, R., and Wilson, J.X. (1999). Propofol prevents

SEDATIVE AGENTS AND CEREBRAL MICRODIALYSIS IN HEAD INJURY 1109

peroxide-induced inhibition of glutamate transport in cul-tured astrocytes. Anesthesiology 90, 1446–1453.

Smith, H., Sinson, G., and Varelas, P. (2009). Vasopressors andpropofol infusion syndrome in severe head trauma. Neurocrit.Care 10, 166–172.

Tanabe, K., Kozawa, O., and Iida, H. (2011). Midazolam sup-presses interleukin-1b-induced release from rat glial cells. J.Neuroinflammation 8, 68.

Tsuchiya, M., Asada, A., Kasahara, E., Sato, E.F., Shindo, M.,and Inoue, M. (2002). Antioxidant protection of propofol andits recycling in erythrocyte membranes. Am. J. Respir. Crit.Care Med. 165, 54–60.

Tsuchiya, M., Asada, A., Maeda, K., Ueda, Y., Sato, E.F., Shindo,M., and Inoue, M. (2001). Propofol versus midazolam re-garding their antioxidant activities. Am. J. Respir. Crit. CareMed. 163, 26–31.

Tyurin, V.A., Tyurina, Y.Y., Borisenko, G.G., Sokolova, T.V.,Ritov, V.B., Quinn, P.J., Rose, M., Kochanek, P., Graham,S.H., and Kagan, V.E. (2000). Oxidative stress followingtraumatic brain injury in rats: quantitation of biomarkers anddetection of free radical intermediates. J. Neurochem. 75,2178–2189.

Velly, L.J., Guillet, B.A., Masmejean, F.M., Nieoullon, A.L.,Bruder, N.J., Gouin, F.M., and Pisano, P.M. (2003). Neuro-protective effects of propofol in a model of ischemic corticalcell cultures: role of glutamate and its transporters. Anesthe-siology 99, 368–375.

Vespa, P., Bergsneider, M., Hattori, N., Wu, H.M., Huang, S.C.,Martin, N.A., Glenn, T.C., McArthur, D.L., and Hovda, D.A.(2005). Metabolic crisis without brain ischemia is commonafter traumatic brain injury: a combined microdialysis andpositron emission tomography study. J. Cereb. Blood FlowMetab. 25, 763–774.

Vincent, J.L., and Berre, J. (2005). Primer on medical manage-ment of severe brain injury. Crit. Care Med. 33, 1392–1399.

Address correspondence to:Philippe Seguin, M.D.

Hopital PontchaillouService d’Anesthesie-Reanimation 1

2 Rue Henri Le GuillouxCedex 9, France

E-mail: philippe.seguin@chu-rennes.fr

1110 TANGUY ET AL.