Accuracy of Brain Multimodal Monitoring to Detect Cerebral Hypoperfusion After Traumatic Brain Injury* Pierre Bouzat, MD, PhD1,2; Pedro Marques-Vidal, MD, MPH3; Jean-Baptiste Zerlauth, MD4; Nathalie Sala, MD1; Tamarah Suys, RN, MPH1; Patrick Schoettker, MD5; Jocelyne Bloch, MD6; Roy T. Daniel, MD6; Marc Levivier, MD6; Reto Meuli, MD4; Mauro Oddo, MD1
Objective: To examine the accuracy of brain multimodal monitoring—consisting of intracranial pressure, brain tissue Po2, and cerebral microdialysis—in detecting cerebral hypoperfusion in patients with severe traumatic brain injury. Design: Prospective single-center study. Patients: Patients with severe traumatic brain injury. Setting: Medico-surgical ICU, university hospital. Intervention: Intracranial pressure, brain tissue Po2, and cerebral microdialysis monitoring (right frontal lobe, apparently normal tissue) combined with cerebral blood flow measurements using perfusion CT. Measurements and Main Results: Cerebral blood flow was measured using perfusion CT in tissue area around intracranial monitoring (regional cerebral blood flow) and in bilateral supra-ventricular brain areas (global cerebral blood flow) and was matched to cerebral physiologic variables. The accuracy of intracranial monitoring to predict cerebral hypoperfusion (defined as an oligemic regional cerebral blood flow < 35 mL/100 g/min) was examined using area under the receiver-operating characteristic curves. Thirty perfusion CT scans (median, 27 hr [interquartile range, 20–45] after traumatic *See also p. 506. 1 Department of Intensive Care Medicine, Neuroscience Critical Care Research Group, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. 2 Joseph Fourier University, Grenoble, France. 3 Department of Internal Medicine, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. 4 Department of Radiology, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. 5 Department of Anesthesiology, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. 6 Clinical Neurosciences, Division of Neurosurgery, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. Dr. Bouzat received research grants from La Fondation des Gueules Cassées. Dr. Oddo received research grants from the Swiss National Science Foundation and the Novartis Foundation for Biomedical Research. The remaining authors have disclosed that they do not have any potential conflicts of interest. For information regarding this article, E-mail: [email protected] Copyright © 2015 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/CCM.0000000000000720
Critical Care Medicine
brain injury) were performed on 27 patients (age, 39 yr [24–54 yr]; Glasgow Coma Scale, 7 [6–8]; 24/27 [89%] with diffuse injury). Regional cerebral blood flow correlated significantly with global cerebral blood flow (Pearson r = 0.70, p < 0.01). Compared with normal regional cerebral blood flow (n = 16), low regional cerebral blood flow (n = 14) measurements had a higher proportion of samples with intracranial pressure more than 20 mm Hg (13% vs 30%), brain tissue Po2 less than 20 mm Hg (9% vs 20%), cerebral microdialysis glucose less than 1 mmol/L (22% vs 57%), and lactate/pyruvate ratio more than 40 (4% vs 14%; all p < 0.05). Compared with intracranial pressure monitoring alone (area under the receiver-operating characteristic curve, 0.74 [95% CI, 0.61–0.87]), monitoring intracranial pressure + brain tissue Po2 (area under the receiver-operating characteristic curve, 0.84 [0.74–0.93]) or intracranial pressure + brain tissue Po2+ cerebral microdialysis (area under the receiver-operating characteristic curve, 0.88 [0.79–0.96]) was significantly more accurate in predicting low regional cerebral blood flow (both p < 0.05). Conclusion: Brain multimodal monitoring—including intracranial pressure, brain tissue Po2, and cerebral microdialysis—is more accurate than intracranial pressure monitoring alone in detecting cerebral hypoperfusion at the bedside in patients with severe traumatic brain injury and predominantly diffuse injury. (Crit Care Med 2015; 43:445–452) Key Words: brain multimodal monitoring; brain oxygen; cerebral blood flow; intracranial pressure; microdialysis; neuromonitoring; traumatic brain injury
A
n important goal in patients with severe traumatic brain injury (TBI) is to manage secondary brain injury, that is, the number of pathological events (including cerebral ischemia, intracranial hypertension, and energy dysfunction) that occur after the primary cerebral insult and add further burden to patient outcome (1). Following these pathological mechanisms, cerebral blood flow (CBF) might be inadequate. Providing adequate CBF to the injured brain after TBI is an important goal of neurocritical care but can be challenging (2). Standard intracranial pressure (ICP)/cerebral perfusion pressure (CPP) monitoring remains a cornerstone in patients www.ccmjournal.org
445
Bouzat et al
with severe TBI (3, 4); however, it may not always detect cerebral ischemia (5, 6) and appears insufficient per se to improve outcome (7). Over the past decade, brain multimodality monitoring—consisting of monitoring of ICP with brain tissue Po2 (Pbto2) and cerebral microdialysis (CMD)—has increasingly developed (8, 9). Despite increasing utilization, data on the accuracy of brain multimodal monitoring to detect low CBF are controversial. Some studies reported a positive correlation between Pbto2 and CBF (10–12), but others did not (13), partly because Pbto2 may be affected but other variables than CBF, such as Pao2 (14). CMD has also been used to detect cerebral ischemia after severe TBI (15, 16), but data also are conflicting, some authors reporting a lack of correlation between elevated CMD lactate-to-pyruvate ratio (LPR) more than 40 and low CBF (17) or vasoconstriction (18), suggesting elevated LPR might not only be due to ischemia. An important criticism is that monitors are regional and representative of a small volume of brain tissue; thus, major concerns exist about their accuracy to detect changes of global CBF. Furthermore, the majority of these studies used variable techniques for CBF measurement— including local thermal diffusion probes, xenon CT, and PET scan—which are not widely available. Hemphill et al (13) in patients with severe TBI used perfusion CT (PCT), which is more readily available in this setting (19, 20) and equals xenon CT for the assessment of CBF (21, 22). The aim of this study was to examine in patients with severe TBI the accuracy of brain multimodal monitoring—consisting of ICP, Pbto2, and CMD—to detect cerebral hypoperfusion, using PCT to measure CBF regionally and globally, and to compare the accuracy of brain multimodal monitoring in detecting low CBF to that of ICP monitoring alone.
PATIENTS AND METHODS Patients This prospective observational study was conducted at the Department of Intensive Care Medicine, CHUV-University Hospital Centre, Lausanne, Switzerland, between May 2010 and November 2013. Patients were admitted after severe TBI (Glasgow Coma Scale [GCS], < 9) with abnormal brain CT (Marshall score, ≥ 2) (23) and monitored with ICP, Pbto2, and CMD, as part of standard care (24). The Ethical Research Committee of the University of Lausanne approved the study. Patients were treated according to a written protocol, according to international guidelines (4). Patients were sedated with propofol and sufentanil and mechanically ventilated aiming to Pao2 90–100 mm Hg and Paco2 35–40 mm Hg. Brain physiological targets were set to maintain ICP less than 20 mm Hg, CPP (CPP = mean arterial pressure, measured via an intraarterial catheter – ICP) more than 60 mm Hg, and Pbto2 more than 20 mm Hg. Intracranial Monitoring ICP was measured using a Codman probe (Raynham, MA), and Pbto2 was measured using a Licox catheter (Integra Neurosciences, Plainsboro, NJ). CMD technique consisted of a CMA 446
www.ccmjournal.org
70 catheter with 20 KDa cutoff (M Dialysis AB, Stockholm, Sweden), perfused with artificial cerebrospinal fluid via a CMA 106 pump (M Dialysis AB; 0.3 μL/min). Microdialysis samples were collected every hour and analyzed at the bedside for brain extracellular concentrations of glucose, lactate, pyruvate, and glutamate using a kinetic enzymatic analyzer (ISCUS Flex; M Dialysis AB). The three probes were inserted in the operating room by a neurosurgeon through a triple-lumen bolt (Integra Neurosciences) and placed into apparently normal brain parenchyma (subcortical white matter), in the right frontal lobe. A CT scan was performed at ≈ 24 hours to confirm the correct placement of intracranial monitors. CT Perfusion Brain PCT was performed using a multidetector row CT Lightspeed (GE Medical Systems, Milwaukee, WI). Scanning was initiated 5 seconds after injection of 50 mL of iohexol (300 mg/mL of iodine; GE Healthcare, Milwaukee, WI), at a rate of 5 mL/s, with the following variables: 80 kV, 240 mAs, 0.4 rotations/s, and total duration of 50 seconds. The series evaluated 16 adjacent 5-mm-thick sections of brain parenchyma. Postprocessing of PCT data were performed by two experienced neuroradiologists, using a dedicated software (Brilliance Workspace Portal; Philips Medical Systems, Cleveland, OH), which employs the central volume principle using deconvolution to measure the mean transit time (MTT); cerebral blood volume (CBV) is calculated from the time-enhancement curves, and CBF is derived from the equation CBF = CBV/MTT. For each PCT, one region of interest (ROI) was manually drawn around the probe (surface area, ~ 50 mm2) to calculate regional CBF, and two others ROI (one for each hemisphere) of approximately 250 mm2 were selected above the ventricular system and included anterior and middle cerebral artery territories (global CBF), as described previously (19) (Fig. 1A). Since probes were located in the white matter, supraventricular ROI was drawn in areas of predominant white matter to allow concordant measurements of global supratentorial CBF in the same type of tissue. Calculation of global and regional CBF was performed by an experienced neuroradiologist, blinded to intracranial monitoring data. Low regional cerebral blood flow (rCBF) was defined by an oligemic regional CBF less than 35 mL/100 g/min, that is, 2 sds below normal CBF, according to studies in patients with carotid stenosis (25) and TBI (26). The threshold of oligemic CBF used in this study was consistently higher than that of ischemic CBF, classically defined by a CBF less than 18–22 mL/100 g/min. Data Collection and Processing Demographic variables included age, gender, admission GCS, Marshall score (23), type of injury (diffuse vs focal), time from TBI to monitoring, and time from TBI to PCT. Main cerebral physiological variables (Pbto2, ICP, and CPP) were recorded every minute via a computerized medical chart system (Metavision; IMD soft, Tel-Aviv, Israël). As CMD samples were collected hourly, they were matched to Pbto2, ICP, and CPP collected during the hour previous to CMD sampling. Following data extraction, artifacts (e.g., periods of disconnection February 2015 • Volume 43 • Number 2
Neurologic Critical Care
intracranial hypertension (ICP > 20 mm Hg) and low CPP (CPP < 60 mm Hg). For CMD variables, thresholds of abnormalities were defined as a CMD glucose less than 1 mmol/L (29) and an LPR more than 40 (15, 17). To examine the relationship of brain multimodal monitoring with rCBF, CMD, Pbto2, and ICP variables were Figure 1. Correlation between regional and global cerebral blood flow (CBF). A, Unenhanced CT scan of matched to PCT data during one representative patient (left) showing the regions of interest (ROI) selected for the measure of global CBF the time window that we conand cerebral perfusion CT (right) in the same patient showing the ROI around intracranial monitoring (surface area, ~ 50 mm2; regional CBF). B, Positive linear correlation between regional and global CBF (Pearson linear sidered as the most representacorrelation coefficient r = 0.70, p < 0.01; n = 30 CT perfusion scans). tive of the patient brain state, from monitoring devices or flushing of arterial catheter) and that is, we included samples obtained at the time of PCT, plus values outside obvious range were manually eliminated. The samples from the 3 hours previous and the 3 hours following main threshold for brain hypoxic episodes was defined as a PCT time (total of seven epochs), according to Sala et al (24). Pbto2 less than 20 mm Hg for at least 5 minutes at any time Matching between brain multimodal monitoring and PCT during the preceding hour, in line with Doppenberg et al (27) data was double-blinded, that is, neuroradiologists who examand our previous studies (24, 28). This is also the threshold ined PCT were blinded to all cerebral physiological variables, to start treatment at our center. Additional hypoxic thresholds and ICU physicians who collected ICP, Pbto2, and CMD varitested were Pbto2 less than 15 and less than 10 mm Hg. The ables were blinded to PCT data. For the analysis, each PCT same process was used to determine and match episodes of (n = 30) was considered separately.
Baseline Characteristics of Patients With Severe Traumatic Brain Injury
Table 1.
Variable
Value
Patient number
27
Age, yr Gender, female/male
39 (24–54) 6/21
Median admission Glasgow Coma Scale
7 (6–8)
Time from TBI to monitoring, hr
7 (6–15)
Time from TBI to perfusion CT, hr
27 (20–45)
Marshall CT classification 2
20
3
3
4
1
5
2
6
1
Glasgow Outcome Score at 6 mo 1 (death)
6
2 (vegetative state)
1
3 (severe disability)
6
4 (moderate disability)
9
5 (good recovery)
5
TBI = traumatic brain injury. Data are presented as median (interquartile range).
Critical Care Medicine
Statistical Analysis Statistical analysis was performed with STATA 11.0 software (Stata Corp, College Station, TX). A p value of less than 0.05 was considered statistically significant. Data were expressed as median and 10th–90th percentiles. Associations of cerebral variables with rCBF were analyzed with univariate analysis, using nonparametric Mann-Whitney test for continuous variables and chi-square test for categorical variables. To assess the diagnostic accuracy of brain multimodal monitoring in detecting low rCBF less than 35 mL/100 g/min, we used receiveroperating characteristic (ROC) curve analysis. The area under the ROC curve (AUC) for the different monitoring tested (ICP alone, ICP + Pbto2, ICP + CMD, and ICP + Pbto2 + CMD) was compared by use of a test for dependent ROC curves.
RESULTS Patient Characteristics A total of 27 patients were studied (Table 1). Monitoring was placed in the right frontal lobe (apparently normal brain parenchyma), was started 12 ± 12 hours from TBI, and lasted 6 ± 3 days. The majority of patients (24/27; 89%) had diffuse injury. A total of 30 PCTs were performed; therefore, three patients (all with diffuse injury) had two PCTs: 14 PCTs were classified as oligemic rCBF and 16 as normal rCBF. Oligemic rCBF was more frequently associated with poor outcome (Glasgow Outcome Score, 1–3; 8/14 PCT = 57%) than normal CBF (5/16 = 31%). All relevant systemic physiological variables around PCT were within normal ranges, and there was no difference in these variables between low and normal rCBF, except for Pao2 and Pao2/Fio2 ratio (higher in the low rCBF group) (Table 2). www.ccmjournal.org
447
Bouzat et al
Three patients had two PCTs: CBF did not change over time (normal CBF) in two patients, whereas in one patient, rCBF evolved from low to normal. Interestingly, in this latter patient, the first PCT with low rCBF was associated with elevated LPR more than 40, whereas LPR was always less than 40 during the second PCT showing normal rCBF. Regional CBF Around Brain Monitoring Correlated With Global CBF We first examined the correlation between regional CBF measured with PCT around the intracranial probe and global CBF measured in supraventricular white matter ROI (Fig. 1A). We found a statically significant correlation between regional and global CBF (Pearson linear correlation coefficient test r = 0.70; p < 0.01) (Fig. 1B). In all except two patients, rCBF was well matched to global CBF. In only two patients in whom data were discordant (i.e., low rCBF with concomitant normal global CBF), intracranial monitoring still revealed abnormal brain physiology: both patients had indeed ICP more than 20 mm Hg and Pbto2 less than 20 mm Hg, and one of them also had abnormal CMD values. Relationship Between Low rCBF and Brain Multimodal Monitoring Variables Differences of brain multimodal monitoring variables (total of 210 matched samples) between normal and low rCBF groups are illustrated in Table 3. Compared with the normal rCBF group, the low rCBF group had more episodes with elevated ICP more than 20 mm Hg, low Pbto2, low CMD glucose, and elevated LPR more than 40 (all p < 0.05). Median Pbto2 and CMD glucose were also lower in the low rCBF group. Average individual duration of low Pbto2 less than 20 mm Hg was longer in the low versus normal CBF group (47 ± 93 vs 13 ± 43 min). Average number of episodes with low CMD glucose/elevated LPR was also greater in the low versus normal CBF group (4 ± 4 vs 1 ± 2). When using a lower cutoff for abnormal LPR (> 25), the proportion of samples with LPR more than 25 did not differ between the two CBF groups. CPP and the proportion of samples with CPP less than 60 mm Hg did not differ significantly between the low and the normal rCBF groups.
Accuracy of Brain Multimodal Monitoring to Diagnose Cerebral Hypoperfusion Figure 2 shows the ROC curves to predict low rCBF for ICP monitoring data alone, ICP + Pbto2 monitoring data, and ICP + Pbto2 + CMD (including glucose and LPR) combined monitoring. Best AUC to predict cerebral hypoperfusion was that of monitoring combining ICP, Pbto2, and CMD (Table 4): compared with ICP monitoring alone (AUC, 0.74 [95% CI, 0.61–0.87]), monitoring ICP + Pbto2 (AUC, 0.84 [0.74–0.93]; p = 0.02 vs ICP alone) or ICP + Pbto2 + CMD (AUC, 0.88 [0.79–0.96]; p = 0.01 vs ICP alone) was significantly more accurate in predicting low rCBF. We also found a trend toward a better prediction of low rCBF for ICP + Pbto2 + CMD monitoring as compared with ICP + Pbto2 (AUC, 0.88 vs 0.84; p = 0.07). Figure 3 shows the ROC curves to predict low rCBF when using CMD alone (without Pbto2) in combination with ICP monitoring and analyzing separately ICP + CMD glucose and ICP + CMD glucose and LPR. The combination of ICP + CMD glucose and LPR provided higher AUC to predict low rCBF than ICP alone, although there was no statistically significant difference (0.79 [95% CI, 0.69–0.90] vs 0.74 [95% CI, 0.61–0.87] for ICP alone, p = 0.25).
DISCUSSION The main findings of our study are the following: 1) we found brain multimodal monitoring appears adequate to assess regional and global CBF at the bedside in patients with severe TBI, with predominantly diffuse injury type; 2) abnormalities in regional brain physiological and metabolic variables (ICP > 20 mm Hg, Pbto2 < 20 mm Hg, CMD glucose < 1 mmol/L, and CMD LPR > 40) were associated with low regional CBF; 3) brain multimodal monitoring—combining ICP, Pbto2, and CMD—was significantly more accurate than ICP monitoring alone in detecting cerebral hypoperfusion. Regional CBF Reflects Global Perfusion After Severe TBI Regional CBF correlated significantly with global CBF. Caution is warranted, however, because despite this correlation being statistically significant (p < 0.01), the strength of association
Table 2. Main Systemic Physiological Variables Around Perfusion CT Time According to Cerebral Perfusion Pattern Systemic Variables
Low rCBF (n = 14)
Normal rCBF (n = 16)
p
Pao2 (mm Hg)
129 (83–278)
106 (85–175)
< 0.01
Pao2/Fio2 (mm Hg)
381 (172–516)
283 (196–407)
< 0.01
36 (29–41)
37 (30–41)
0.34
Blood hemoglobin (g/L)
117 (92–146)
112 (96–130)
0.46
Blood glucose (mmol/L)
7.2 (5.7–9.3)
7.5 (5.7–9.0)
0.48
Paco2 (mm Hg)
rCBF = regional cerebral blood flow. Results are expressed as median (10th–90th percentiles). Perfusion CT was used to calculate rCBF in one region of interest that was manually drawn around intracranial monitoring (surface area, ~ 50 mm2, see also Fig. 1): cerebral hypoperfusion was defined as an oligemic rCBF < 35 mL/100 g/min as opposed to normal cerebral blood flow ≥ 35 mL/100 g/min.
448
www.ccmjournal.org
February 2015 • Volume 43 • Number 2
Neurologic Critical Care
Table 3. Univariate Associations of Regional Cerebral Blood Flow Data With Brain Multimodal Monitoring Variables According to Cerebral Perfusion Pattern, Dichotomized as Low Versus Normal Brain Physiologic Variables
CMD glucose (mmol/L)
Low rCBF (< 35 mL/100 g/min)
Normal rCBF (≥ 35 mL/100 g/min)
p
0.95 (0.41–2.42)
1.38 (0.64–2.18)
< 0.01
57
22
< 0.01
CMD lactate (mmol/L)
3.15 (1.7–5.0)
3.26 (2.1–5.0)
0.22
CMD pyruvate (μmol/L)
106 (62–182)
118 (75–189)
0.15
CMD glucose < 1 mmol/ (% episodes)
CMD glutamate (μmol/L)
6.9 (1.1–37.2)
7.0 (2.4–26.4)
0.39
CMD LPR
30 (11–44)
28 (14–35)
0.62
CMD LPR > 40 (% episodes)
14
4
0.03
CMD LPR > 25 (% episodes)
70
61
0.24
21 (6–33)
27 (14–39)
< 0.01
Pbto2 < 20 mm Hg (% episodes)
20
9
0.04
Pbto2 < 15 mm Hg (% episodes)
16
5
0.01
Pbto2 < 10 mm Hg (% episodes)
7
0
0.01
Pbto2 (mm Hg)
a
ICP (mm Hg) ICP > 20 mm Hg (% episodes) CPP (mm Hg) CPP < 60 mm Hg (% episodes)
15 (3–24)
12 (0–18)
0.11
30
13
< 0.01
70 (63–83)
72 (63–83)
0.39
12
18
0.26
rCBF = regional cerebral blood flow, CMD = cerebral microdialysis, LPR = lactate-to-pyruvate ratio, Pbto2 = brain tissue oxygen pressure, ICP = intracranial pressure, CPP = cerebral perfusion pressure. a Pbto2 was corrected to Pao2. Perfusion CT was used to calculate rCBF in one region of interest that was manually drawn around intracranial monitoring (surface area, ~ 50 mm2): cerebral hypoperfusion was defined as an oligemic rCBF < 35 mL/100 g/min as opposed to normal CBF ≥ 35 mL/100 g/min. Results are expressed as median (10th–90th percentiles) or as percentages. Boldface data denote p < 0.05.
between the measures (R2 = 0.49) requires care to be taken in interpretation of data from focal monitors. This correlation was found within the acute phase (< 48 hr) of TBI, when providing adequate CBF is crucial (30): our data are in line with previous reports that found a good correlation of CBF with Pbto2 (10–12) or LPR (16) and further extend them, since they tested the value of monitoring using Pbto2 and CMD in combination. Although still debated (31, 32), at our center, the location of brain monitoring after severe TBI is in apparently normal brain parenchyma. The large proportion of patients with diffuse injury in our cohort justifies this approach. On the other extent, we cannot rule out regional disturbances of CBF attributable to focal injury in some patients (1, 33). Only two patients had regional CBF lower than 35 mL/100 g/min with concomitant normal global CBF. These two patients had abnormal cerebral physiology: this illustrates possible CBF heterogeneity and reinforces the fact that multimodal monitoring adequately detects CBF regional disturbances. Additional studies are needed to validate these findings in patients with focal injury. The Relationship Between Brain Multimodal Monitoring and CBF Abnormalities in ICP, Pbto2, and neurochemistry (cerebral glucose and LPR) were associated with low rCBF. These associations Critical Care Medicine
were observed despite moderate elevation of ICP in our cohort. In contrast, we did not find any association between low rCBF and CPP, indicating the poor value of CPP alone to assess cerebral perfusion (5) and the need for other modalities to interpret CPP, for example, the pressure reactivity index (34). Our findings support the use of ICP monitoring in after severe TBI (35) and that maintenance of ICP below 20 mm Hg is adequate to avoid cerebral hypoperfusion. Pbto2 less than 20 mm Hg was also a good marker of low rCBF, which is not unprecedented (10, 12, 14). Brain energy crisis (defined as reduced glucose and elevated LPR > 40 in CMD fluid) is considered as a marker of ischemia (15) and poor outcome after severe TBI (29): we found that both reduced cerebral glucose and elevated LPR were more likely to occur in conditions of low rCBF. Beyond ICP and Pbto2 monitoring, the contribution of CMD may be relevant at the bedside to detect inadequate CBF in a timely fashion after TBI. Intracranial Monitoring to Detect Cerebral Hypoperfusion After TBI at the Bedside Despite increasing utilization, data on the accuracy of brain multimodal monitoring to detect low CBF are limited. In addition, a relevant question from the clinical standpoint is to determine whether brain multimodal monitoring has any additive value over ICP monitoring alone in detecting low CBF www.ccmjournal.org
449
Bouzat et al
is therefore more than an ischemia monitoring in the setting of severe TBI.
Figure 2. Receiver-operating characteristic (ROC) curves of the different monitoring modalities to detect low regional cerebral blood flow (CBF). Low regional CBF, measured with perfusion CT, was defined as an oligemic CBF less than 35 mL/100 g/min. ROC curve for intracranial pressure (ICP) monitoring alone (light gray line), ICP + brain tissue oxygen pressure (Pbto2) monitoring (dark gray line), and ICP + Pbto2 + cerebral microdialysis (CMD) monitoring glucose and lactate/pyruvate ratio (black line) are represented on the same graph. Compared with ICP monitoring alone (area under the ROC curve [AUC], 0.74 [95% CI, 0.61–0.87]), ICP + Pbto2 monitoring (AUC, 0.84 [0.74–0.93]; p = 0.02) or ICP + Pbto2 + CMD monitoring (AUC, 0.88 [0.79–0.96]; p = 0.01) was significantly more accurate in predicting low regional CBF. We also found a trend toward a better prediction of low regional CBF for ICP + Pbto2 + CMD versus ICP + Pbto2 monitoring (p = 0.07).
and to further examine which one(s) of the monitored variables are the most accurate for this purpose. We found that the combination of ICP and Pbto2 was more accurate in predicting low rCBF than the monitoring of ICP alone. Combining CMD to ICP and Pbto2 monitoring further increased the performance of the AUC and was the most accurate to diagnose cerebral hypoperfusion. Using ICP plus CMD monitoring without Pbto2 was less accurate in predicting low rCBF than ICP + Pbto2 and CMD. This may be because CMD may not only be an ischemia monitor. Indeed, despite cerebral hypoperfusion was more frequently associated with low cerebral glucose and elevated LPR, other mechanisms than ischemia may cause cerebral glucose depletion or LPR elevation (e.g., increased glycolysis, energy dysfunction, barriers to oxygen diffusion, and mitochondrial failure) (36, 37). CMD monitoring
Study Limitations First, the study was single center and sample size was relatively small; therefore, data may not be generalized. However, patients were treated with a standardized algorithm, and the cohort was homogeneous, consisting predominantly of patients with diffuse injury, monitored in the right frontal lobe and apparently normal brain. Second, because patients had mainly diffuse injury and were monitored in normal brain regions, our findings cannot be extended to TBI patients with focal injuries. For the same reason, we cannot extrapolate that the observed relationship between CBF and brain multimodality monitoring variables is maintained when monitoring is located in pericontusional areas. Furthermore, diffuse injury induces moderate elevation of ICP (38), as in our study. Thus, the predictive value of ICP to predict low CBF may differ in patients with more severe ICP elevations. Time window for matching PCT with monitoring-derived variables (7 hr) may also be considered as arbitrary, but we considered this time window as the most representative of the patient brain state and also to take into account potential changes in brain physiology due to patient transport to the CT scan. Third, CBF was assessed using PCT, which only measures CBF in anterior and middle vascular territories; therefore, changes in posterior vascular territories may be missed. Finally, our definition of low rCBF deserves further discussion. We acknowledge the fact that although oligemic rCBF was about 2 sds below normal CBF, it could be considered as only marginally low and above the ischemic range. Although this may be a potential limitation, on the other extent, it may be an advantage: our threshold of rCBF less than 35 mL/100 g/min most likely reflects cerebral hypoperfusion, that is, a condition where therapy (e.g., CPP augmentation) might save brain tissue. In addition, although this study was not intended to examine outcome, the fact that low rCBF less than 35 mL/100 g/min was associated with worse outcome implies that this threshold might seem appropriate from the clinical standpoint.
CONCLUSIONS Brain multimodal monitoring—consisting of ICP, Pbto2, and CMD—gives appropriate estimation of global CBF in patients
Table 4. Receiver-Operating Characteristic Curve Analysis of the Different Brain Monitoring Modalities to Predict Cerebral Hypoperfusion (Regional Cerebral Blood Flow < 35 mL/100 g/min) Variable
Area Under the ROC Curve
95% CI
ROC curve for ICP monitoring alone
0.74
0.61–0.87
ROC curve for ICP and CMD monitoring
0.79
0.69–0.90
ROC curve for ICP and Pbto2 monitoring
0.84
0.74–0.93
ROC curve for ICP, Pbto2, and CMD monitoring
0.88
0.79–0.96
ROC = receiver-operating characteristic, ICP = intracranial pressure, CMD = cerebral microdialysis (including CMD glucose and lactate/pyruvate ratio), Pbto2 = brain tissue oxygen pressure.
450
www.ccmjournal.org
February 2015 • Volume 43 • Number 2
Neurologic Critical Care
Figure 3. Receiver-operating characteristic (ROC) curves comparing intracranial pressure (ICP) monitoring alone with ICP + cerebral microdialysis (CMD) to detect low regional cerebral blood flow (CBF). Low regional CBF, measured with perfusion CT, was defined as an oligemic CBF less than 35 mL/100 g/min. ROC curves for ICP monitoring alone (light gray line), ICP + CMD glucose monitoring (dark gray line), and ICP + CMD glucose and lactate/pyruvate ratio (LPR) monitoring (black line) are represented on the same graph. Area under the curve (AUC) is illustrated. The combination of ICP + CMD glucose and LPR provided higher AUC to predict low regional CBF than ICP alone, although there was no statistically significant difference (0.79 [95% CI, 0.69–0.90] vs 0.74 [95% CI, 0.61–0.87]; p = 0.25).
with severe TBI and predominantly diffuse injury. Multimodal monitoring with ICP, Pbto2, and CMD was significantly more accurate in predicting cerebral hypoperfusion than ICP monitoring alone. Our findings suggest that beyond ICP, monitoring of Pbto2 and CMD provides a better assessment of cerebral perfusion state at the bedside after severe TBI.
REFERENCES
1. Maas AI, Stocchetti N, Bullock R: Moderate and severe traumatic brain injury in adults. Lancet Neurol 2008; 7:728–741 2. Dagal A, Lam AM: Cerebral blood flow and the injured brain: How should we monitor and manipulate it? Curr Opin Anaesthesiol 2011; 24:131–137 3. Bratton SL, Chestnut RM, Ghajar J, et al: Guidelines for the management of severe traumatic brain injury. VIII. Intracranial pressure thresholds. J Neurotrauma 2007; 24(Suppl 1):S55–58 4. Bratton SL, Chestnut RM, Ghajar J, et al: Guidelines for the management of severe traumatic brain injury. IX. Cerebral perfusion thresholds. J Neurotrauma 2007; 24(Suppl 1):S59–64 5. Eriksson EA, Barletta JF, Figueroa BE, et al: Cerebral perfusion pressure and intracranial pressure are not surrogates for brain tissue oxygenation in traumatic brain injury. Clin Neurophysiol 2012; 123:1255–1260 6. Oddo M, Levine JM, Mackenzie L, et al: Brain hypoxia is associated with short-term outcome after severe traumatic brain injury independently of intracranial hypertension and low cerebral perfusion pressure. Neurosurgery 2011; 69:1037–1045 7. Chesnut RM, Temkin N, Carney N, et al; Global Neurotrauma Research Group: A trial of intracranial-pressure monitoring in traumatic brain injury. N Engl J Med 2012; 367:2471–2481 8. Oddo M, Villa F, Citerio G: Brain multimodality monitoring: An update. Curr Opin Crit Care 2012; 18:111–118 9. Bouzat P, Sala N, Payen JF, et al: Beyond intracranial pressure: Optimization of cerebral blood flow, oxygen, and substrate delivery after traumatic brain injury. Ann Intensive Care 2013; 3:23 10. Doppenberg EM, Zauner A, Bullock R, et al: Correlations between brain tissue oxygen tension, carbon dioxide tension, pH, and cerebral
Critical Care Medicine
blood flow—A better way of monitoring the severely injured brain? Surg Neurol 1998; 49:650–654 11. Valadka AB, Hlatky R, Furuya Y, et al: Brain tissue PO2: Correlation with cerebral blood flow. Acta Neurochir Suppl 2002; 81:299–301 12. Jaeger M, Soehle M, Schuhmann MU, et al: Correlation of continuously monitored regional cerebral blood flow and brain tissue oxygen. Acta Neurochir (Wien) 2005; 147:51–56 13. Hemphill JC 3rd, Smith WS, Sonne DC, et al: Relationship between brain tissue oxygen tension and CT perfusion: Feasibility and initial results. AJNR Am J Neuroradiol 2005; 26:1095–1100 14. Rosenthal G, Hemphill JC 3rd, Sorani M, et al: Brain tissue oxygen tension is more indicative of oxygen diffusion than oxygen delivery and metabolism in patients with traumatic brain injury. Crit Care Med 2008; 36:1917–1924 15. Hillered L, Vespa PM, Hovda DA: Translational neurochemical research in acute human brain injury: The current status and potential future for cerebral microdialysis. J Neurotrauma 2005; 22:3–41 16. Sarrafzadeh AS, Sakowitz OW, Callsen TA, et al: Bedside microdialysis for early detection of cerebral hypoxia in traumatic brain injury. Neurosurg Focus 2000; 9:e2 17. Vespa P, Bergsneider M, Hattori N, et al: Metabolic crisis without brain ischemia is common after traumatic brain injury: A combined microdialysis and positron emission tomography study. J Cereb Blood Flow Metab 2005; 25:763–774 18. Asgari S, Vespa P, Hu X: Is there any association between cerebral vasoconstriction/vasodilatation and microdialysis lactate to pyruvate ratio increase? Neurocrit Care 2013; 19:56–64 19. Wintermark M, Chiolero R, Van Melle G, et al: Cerebral vascular autoregulation assessed by perfusion-CT in severe head trauma patients. J Neuroradiol 2006; 33:27–37 20. Wintermark M, van Melle G, Schnyder P, et al: Admission perfusion CT: Prognostic value in patients with severe head trauma. Radiology 2004; 232:211–220 21. Wintermark M, Thiran JP, Maeder P, et al: Simultaneous measurement of regional cerebral blood flow by perfusion CT and stable xenon CT: A validation study. AJNR Am J Neuroradiol 2001; 22:905–914 22. Huang AP, Tsai JC, Kuo LT, et al: Clinical application of perfusion computed tomography in neurosurgery. J Neurosurg 2014; 120:473–488 23. Marshall LF, Marshall SB, Klauber MR, et al: The diagnosis of head injury requires a classification based on computed axial tomography. J Neurotrauma 1992; 9(Suppl 1):S287–S292 24. Sala N, Suys T, Zerlauth JB, et al: Cerebral extracellular lactate increase is predominantly nonischemic in patients with severe traumatic brain injury. J Cereb Blood Flow Metab 2013; 33:1815–1822 25. Marshall RS, Lazar RM, Pile-Spellman J, et al: Recovery of brain function during induced cerebral hypoperfusion. Brain 2001; 124:1208–1217 26. Kelly DF, Martin NA, Kordestani R, et al: Cerebral blood flow as a predictor of outcome following traumatic brain injury. J Neurosurg 1997; 86:633–641 27. Doppenberg EM, Zauner A, Watson JC, et al: Determination of the ischemic threshold for brain oxygen tension. Acta Neurochir Suppl 1998; 71:166–169 28. Oddo M, Levine JM, Frangos S, et al: Brain lactate metabolism in humans with subarachnoid hemorrhage. Stroke 2012; 43:1418–1421 29. Timofeev I, Carpenter KL, Nortje J, et al: Cerebral extracellular chemistry and outcome following traumatic brain injury: A microdialysis study of 223 patients. Brain 2011; 134:484–494 30. Robertson CS, Valadka AB, Hannay HJ, et al: Prevention of secondary ischemic insults after severe head injury. Crit Care Med 1999; 27:2086–2095 31. Bhatia A, Gupta AK: Neuromonitoring in the intensive care unit. II. Cerebral oxygenation monitoring and microdialysis. Intensive Care Med 2007; 33:1322–1328 32. Andrews PJ, Citerio G, Longhi L, et al; Neuro-Intensive Care and Emergency Medicine (NICEM) Section of the European Society of Intensive Care Medicine: NICEM consensus on neurological monitoring in acute neurological disease. Intensive Care Med 2008; 34:1362–1370 www.ccmjournal.org
451
Bouzat et al 33. Longhi L, Pagan F, Valeriani V, et al: Monitoring brain tissue oxygen tension in brain-injured patients reveals hypoxic episodes in normal-appearing and in peri-focal tissue. Intensive Care Med 2007; 33:2136–2142 34. Aries MJ, Czosnyka M, Budohoski KP, et al: Continuous determination of optimal cerebral perfusion pressure in traumatic brain injury. Crit Care Med 2012; 40:2456–2463 35. Stein SC, Georgoff P, Meghan S, et al: Relationship of aggressive monitoring and treatment to improved outcomes in severe traumatic brain injury. J Neurosurg 2010; 112:1105–1112
452
www.ccmjournal.org
36. Menon DK, Coles JP, Gupta AK, et al: Diffusion limited oxygen delivery following head injury. Crit Care Med 2004; 32:1384–1390 37. Carre E, Ogier M, Boret H, et al: Metabolic crisis in severely headinjured patients: Is ischemia just the tip of the iceberg? Front Neurol 2013; 4:146 38. Cooper DJ, Rosenfeld JV, Murray L, et al; DECRA Trial Investigators; Australian and New Zealand Intensive Care Society Clinical Trials Group: Decompressive craniectomy in diffuse traumatic brain injury. N Engl J Med 2011; 364:1493–1502
February 2015 • Volume 43 • Number 2
Copyright of Critical Care Medicine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Objective: To examine the accuracy of brain multimodal monitoring—consisting of intracranial pressure, brain tissue Po2, and cerebral microdialysis—in detecting cerebral hypoperfusion in patients with severe traumatic brain injury. Design: Prospective single-center study. Patients: Patients with severe traumatic brain injury. Setting: Medico-surgical ICU, university hospital. Intervention: Intracranial pressure, brain tissue Po2, and cerebral microdialysis monitoring (right frontal lobe, apparently normal tissue) combined with cerebral blood flow measurements using perfusion CT. Measurements and Main Results: Cerebral blood flow was measured using perfusion CT in tissue area around intracranial monitoring (regional cerebral blood flow) and in bilateral supra-ventricular brain areas (global cerebral blood flow) and was matched to cerebral physiologic variables. The accuracy of intracranial monitoring to predict cerebral hypoperfusion (defined as an oligemic regional cerebral blood flow < 35 mL/100 g/min) was examined using area under the receiver-operating characteristic curves. Thirty perfusion CT scans (median, 27 hr [interquartile range, 20–45] after traumatic *See also p. 506. 1 Department of Intensive Care Medicine, Neuroscience Critical Care Research Group, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. 2 Joseph Fourier University, Grenoble, France. 3 Department of Internal Medicine, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. 4 Department of Radiology, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. 5 Department of Anesthesiology, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. 6 Clinical Neurosciences, Division of Neurosurgery, CHUV-Lausanne University Hospital, Faculty of Biology and Medicine, Lausanne, Switzerland. Dr. Bouzat received research grants from La Fondation des Gueules Cassées. Dr. Oddo received research grants from the Swiss National Science Foundation and the Novartis Foundation for Biomedical Research. The remaining authors have disclosed that they do not have any potential conflicts of interest. For information regarding this article, E-mail: [email protected] Copyright © 2015 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/CCM.0000000000000720
Critical Care Medicine
brain injury) were performed on 27 patients (age, 39 yr [24–54 yr]; Glasgow Coma Scale, 7 [6–8]; 24/27 [89%] with diffuse injury). Regional cerebral blood flow correlated significantly with global cerebral blood flow (Pearson r = 0.70, p < 0.01). Compared with normal regional cerebral blood flow (n = 16), low regional cerebral blood flow (n = 14) measurements had a higher proportion of samples with intracranial pressure more than 20 mm Hg (13% vs 30%), brain tissue Po2 less than 20 mm Hg (9% vs 20%), cerebral microdialysis glucose less than 1 mmol/L (22% vs 57%), and lactate/pyruvate ratio more than 40 (4% vs 14%; all p < 0.05). Compared with intracranial pressure monitoring alone (area under the receiver-operating characteristic curve, 0.74 [95% CI, 0.61–0.87]), monitoring intracranial pressure + brain tissue Po2 (area under the receiver-operating characteristic curve, 0.84 [0.74–0.93]) or intracranial pressure + brain tissue Po2+ cerebral microdialysis (area under the receiver-operating characteristic curve, 0.88 [0.79–0.96]) was significantly more accurate in predicting low regional cerebral blood flow (both p < 0.05). Conclusion: Brain multimodal monitoring—including intracranial pressure, brain tissue Po2, and cerebral microdialysis—is more accurate than intracranial pressure monitoring alone in detecting cerebral hypoperfusion at the bedside in patients with severe traumatic brain injury and predominantly diffuse injury. (Crit Care Med 2015; 43:445–452) Key Words: brain multimodal monitoring; brain oxygen; cerebral blood flow; intracranial pressure; microdialysis; neuromonitoring; traumatic brain injury
A
n important goal in patients with severe traumatic brain injury (TBI) is to manage secondary brain injury, that is, the number of pathological events (including cerebral ischemia, intracranial hypertension, and energy dysfunction) that occur after the primary cerebral insult and add further burden to patient outcome (1). Following these pathological mechanisms, cerebral blood flow (CBF) might be inadequate. Providing adequate CBF to the injured brain after TBI is an important goal of neurocritical care but can be challenging (2). Standard intracranial pressure (ICP)/cerebral perfusion pressure (CPP) monitoring remains a cornerstone in patients www.ccmjournal.org
445
Bouzat et al
with severe TBI (3, 4); however, it may not always detect cerebral ischemia (5, 6) and appears insufficient per se to improve outcome (7). Over the past decade, brain multimodality monitoring—consisting of monitoring of ICP with brain tissue Po2 (Pbto2) and cerebral microdialysis (CMD)—has increasingly developed (8, 9). Despite increasing utilization, data on the accuracy of brain multimodal monitoring to detect low CBF are controversial. Some studies reported a positive correlation between Pbto2 and CBF (10–12), but others did not (13), partly because Pbto2 may be affected but other variables than CBF, such as Pao2 (14). CMD has also been used to detect cerebral ischemia after severe TBI (15, 16), but data also are conflicting, some authors reporting a lack of correlation between elevated CMD lactate-to-pyruvate ratio (LPR) more than 40 and low CBF (17) or vasoconstriction (18), suggesting elevated LPR might not only be due to ischemia. An important criticism is that monitors are regional and representative of a small volume of brain tissue; thus, major concerns exist about their accuracy to detect changes of global CBF. Furthermore, the majority of these studies used variable techniques for CBF measurement— including local thermal diffusion probes, xenon CT, and PET scan—which are not widely available. Hemphill et al (13) in patients with severe TBI used perfusion CT (PCT), which is more readily available in this setting (19, 20) and equals xenon CT for the assessment of CBF (21, 22). The aim of this study was to examine in patients with severe TBI the accuracy of brain multimodal monitoring—consisting of ICP, Pbto2, and CMD—to detect cerebral hypoperfusion, using PCT to measure CBF regionally and globally, and to compare the accuracy of brain multimodal monitoring in detecting low CBF to that of ICP monitoring alone.
PATIENTS AND METHODS Patients This prospective observational study was conducted at the Department of Intensive Care Medicine, CHUV-University Hospital Centre, Lausanne, Switzerland, between May 2010 and November 2013. Patients were admitted after severe TBI (Glasgow Coma Scale [GCS], < 9) with abnormal brain CT (Marshall score, ≥ 2) (23) and monitored with ICP, Pbto2, and CMD, as part of standard care (24). The Ethical Research Committee of the University of Lausanne approved the study. Patients were treated according to a written protocol, according to international guidelines (4). Patients were sedated with propofol and sufentanil and mechanically ventilated aiming to Pao2 90–100 mm Hg and Paco2 35–40 mm Hg. Brain physiological targets were set to maintain ICP less than 20 mm Hg, CPP (CPP = mean arterial pressure, measured via an intraarterial catheter – ICP) more than 60 mm Hg, and Pbto2 more than 20 mm Hg. Intracranial Monitoring ICP was measured using a Codman probe (Raynham, MA), and Pbto2 was measured using a Licox catheter (Integra Neurosciences, Plainsboro, NJ). CMD technique consisted of a CMA 446
www.ccmjournal.org
70 catheter with 20 KDa cutoff (M Dialysis AB, Stockholm, Sweden), perfused with artificial cerebrospinal fluid via a CMA 106 pump (M Dialysis AB; 0.3 μL/min). Microdialysis samples were collected every hour and analyzed at the bedside for brain extracellular concentrations of glucose, lactate, pyruvate, and glutamate using a kinetic enzymatic analyzer (ISCUS Flex; M Dialysis AB). The three probes were inserted in the operating room by a neurosurgeon through a triple-lumen bolt (Integra Neurosciences) and placed into apparently normal brain parenchyma (subcortical white matter), in the right frontal lobe. A CT scan was performed at ≈ 24 hours to confirm the correct placement of intracranial monitors. CT Perfusion Brain PCT was performed using a multidetector row CT Lightspeed (GE Medical Systems, Milwaukee, WI). Scanning was initiated 5 seconds after injection of 50 mL of iohexol (300 mg/mL of iodine; GE Healthcare, Milwaukee, WI), at a rate of 5 mL/s, with the following variables: 80 kV, 240 mAs, 0.4 rotations/s, and total duration of 50 seconds. The series evaluated 16 adjacent 5-mm-thick sections of brain parenchyma. Postprocessing of PCT data were performed by two experienced neuroradiologists, using a dedicated software (Brilliance Workspace Portal; Philips Medical Systems, Cleveland, OH), which employs the central volume principle using deconvolution to measure the mean transit time (MTT); cerebral blood volume (CBV) is calculated from the time-enhancement curves, and CBF is derived from the equation CBF = CBV/MTT. For each PCT, one region of interest (ROI) was manually drawn around the probe (surface area, ~ 50 mm2) to calculate regional CBF, and two others ROI (one for each hemisphere) of approximately 250 mm2 were selected above the ventricular system and included anterior and middle cerebral artery territories (global CBF), as described previously (19) (Fig. 1A). Since probes were located in the white matter, supraventricular ROI was drawn in areas of predominant white matter to allow concordant measurements of global supratentorial CBF in the same type of tissue. Calculation of global and regional CBF was performed by an experienced neuroradiologist, blinded to intracranial monitoring data. Low regional cerebral blood flow (rCBF) was defined by an oligemic regional CBF less than 35 mL/100 g/min, that is, 2 sds below normal CBF, according to studies in patients with carotid stenosis (25) and TBI (26). The threshold of oligemic CBF used in this study was consistently higher than that of ischemic CBF, classically defined by a CBF less than 18–22 mL/100 g/min. Data Collection and Processing Demographic variables included age, gender, admission GCS, Marshall score (23), type of injury (diffuse vs focal), time from TBI to monitoring, and time from TBI to PCT. Main cerebral physiological variables (Pbto2, ICP, and CPP) were recorded every minute via a computerized medical chart system (Metavision; IMD soft, Tel-Aviv, Israël). As CMD samples were collected hourly, they were matched to Pbto2, ICP, and CPP collected during the hour previous to CMD sampling. Following data extraction, artifacts (e.g., periods of disconnection February 2015 • Volume 43 • Number 2
Neurologic Critical Care
intracranial hypertension (ICP > 20 mm Hg) and low CPP (CPP < 60 mm Hg). For CMD variables, thresholds of abnormalities were defined as a CMD glucose less than 1 mmol/L (29) and an LPR more than 40 (15, 17). To examine the relationship of brain multimodal monitoring with rCBF, CMD, Pbto2, and ICP variables were Figure 1. Correlation between regional and global cerebral blood flow (CBF). A, Unenhanced CT scan of matched to PCT data during one representative patient (left) showing the regions of interest (ROI) selected for the measure of global CBF the time window that we conand cerebral perfusion CT (right) in the same patient showing the ROI around intracranial monitoring (surface area, ~ 50 mm2; regional CBF). B, Positive linear correlation between regional and global CBF (Pearson linear sidered as the most representacorrelation coefficient r = 0.70, p < 0.01; n = 30 CT perfusion scans). tive of the patient brain state, from monitoring devices or flushing of arterial catheter) and that is, we included samples obtained at the time of PCT, plus values outside obvious range were manually eliminated. The samples from the 3 hours previous and the 3 hours following main threshold for brain hypoxic episodes was defined as a PCT time (total of seven epochs), according to Sala et al (24). Pbto2 less than 20 mm Hg for at least 5 minutes at any time Matching between brain multimodal monitoring and PCT during the preceding hour, in line with Doppenberg et al (27) data was double-blinded, that is, neuroradiologists who examand our previous studies (24, 28). This is also the threshold ined PCT were blinded to all cerebral physiological variables, to start treatment at our center. Additional hypoxic thresholds and ICU physicians who collected ICP, Pbto2, and CMD varitested were Pbto2 less than 15 and less than 10 mm Hg. The ables were blinded to PCT data. For the analysis, each PCT same process was used to determine and match episodes of (n = 30) was considered separately.
Baseline Characteristics of Patients With Severe Traumatic Brain Injury
Table 1.
Variable
Value
Patient number
27
Age, yr Gender, female/male
39 (24–54) 6/21
Median admission Glasgow Coma Scale
7 (6–8)
Time from TBI to monitoring, hr
7 (6–15)
Time from TBI to perfusion CT, hr
27 (20–45)
Marshall CT classification 2
20
3
3
4
1
5
2
6
1
Glasgow Outcome Score at 6 mo 1 (death)
6
2 (vegetative state)
1
3 (severe disability)
6
4 (moderate disability)
9
5 (good recovery)
5
TBI = traumatic brain injury. Data are presented as median (interquartile range).
Critical Care Medicine
Statistical Analysis Statistical analysis was performed with STATA 11.0 software (Stata Corp, College Station, TX). A p value of less than 0.05 was considered statistically significant. Data were expressed as median and 10th–90th percentiles. Associations of cerebral variables with rCBF were analyzed with univariate analysis, using nonparametric Mann-Whitney test for continuous variables and chi-square test for categorical variables. To assess the diagnostic accuracy of brain multimodal monitoring in detecting low rCBF less than 35 mL/100 g/min, we used receiveroperating characteristic (ROC) curve analysis. The area under the ROC curve (AUC) for the different monitoring tested (ICP alone, ICP + Pbto2, ICP + CMD, and ICP + Pbto2 + CMD) was compared by use of a test for dependent ROC curves.
RESULTS Patient Characteristics A total of 27 patients were studied (Table 1). Monitoring was placed in the right frontal lobe (apparently normal brain parenchyma), was started 12 ± 12 hours from TBI, and lasted 6 ± 3 days. The majority of patients (24/27; 89%) had diffuse injury. A total of 30 PCTs were performed; therefore, three patients (all with diffuse injury) had two PCTs: 14 PCTs were classified as oligemic rCBF and 16 as normal rCBF. Oligemic rCBF was more frequently associated with poor outcome (Glasgow Outcome Score, 1–3; 8/14 PCT = 57%) than normal CBF (5/16 = 31%). All relevant systemic physiological variables around PCT were within normal ranges, and there was no difference in these variables between low and normal rCBF, except for Pao2 and Pao2/Fio2 ratio (higher in the low rCBF group) (Table 2). www.ccmjournal.org
447
Bouzat et al
Three patients had two PCTs: CBF did not change over time (normal CBF) in two patients, whereas in one patient, rCBF evolved from low to normal. Interestingly, in this latter patient, the first PCT with low rCBF was associated with elevated LPR more than 40, whereas LPR was always less than 40 during the second PCT showing normal rCBF. Regional CBF Around Brain Monitoring Correlated With Global CBF We first examined the correlation between regional CBF measured with PCT around the intracranial probe and global CBF measured in supraventricular white matter ROI (Fig. 1A). We found a statically significant correlation between regional and global CBF (Pearson linear correlation coefficient test r = 0.70; p < 0.01) (Fig. 1B). In all except two patients, rCBF was well matched to global CBF. In only two patients in whom data were discordant (i.e., low rCBF with concomitant normal global CBF), intracranial monitoring still revealed abnormal brain physiology: both patients had indeed ICP more than 20 mm Hg and Pbto2 less than 20 mm Hg, and one of them also had abnormal CMD values. Relationship Between Low rCBF and Brain Multimodal Monitoring Variables Differences of brain multimodal monitoring variables (total of 210 matched samples) between normal and low rCBF groups are illustrated in Table 3. Compared with the normal rCBF group, the low rCBF group had more episodes with elevated ICP more than 20 mm Hg, low Pbto2, low CMD glucose, and elevated LPR more than 40 (all p < 0.05). Median Pbto2 and CMD glucose were also lower in the low rCBF group. Average individual duration of low Pbto2 less than 20 mm Hg was longer in the low versus normal CBF group (47 ± 93 vs 13 ± 43 min). Average number of episodes with low CMD glucose/elevated LPR was also greater in the low versus normal CBF group (4 ± 4 vs 1 ± 2). When using a lower cutoff for abnormal LPR (> 25), the proportion of samples with LPR more than 25 did not differ between the two CBF groups. CPP and the proportion of samples with CPP less than 60 mm Hg did not differ significantly between the low and the normal rCBF groups.
Accuracy of Brain Multimodal Monitoring to Diagnose Cerebral Hypoperfusion Figure 2 shows the ROC curves to predict low rCBF for ICP monitoring data alone, ICP + Pbto2 monitoring data, and ICP + Pbto2 + CMD (including glucose and LPR) combined monitoring. Best AUC to predict cerebral hypoperfusion was that of monitoring combining ICP, Pbto2, and CMD (Table 4): compared with ICP monitoring alone (AUC, 0.74 [95% CI, 0.61–0.87]), monitoring ICP + Pbto2 (AUC, 0.84 [0.74–0.93]; p = 0.02 vs ICP alone) or ICP + Pbto2 + CMD (AUC, 0.88 [0.79–0.96]; p = 0.01 vs ICP alone) was significantly more accurate in predicting low rCBF. We also found a trend toward a better prediction of low rCBF for ICP + Pbto2 + CMD monitoring as compared with ICP + Pbto2 (AUC, 0.88 vs 0.84; p = 0.07). Figure 3 shows the ROC curves to predict low rCBF when using CMD alone (without Pbto2) in combination with ICP monitoring and analyzing separately ICP + CMD glucose and ICP + CMD glucose and LPR. The combination of ICP + CMD glucose and LPR provided higher AUC to predict low rCBF than ICP alone, although there was no statistically significant difference (0.79 [95% CI, 0.69–0.90] vs 0.74 [95% CI, 0.61–0.87] for ICP alone, p = 0.25).
DISCUSSION The main findings of our study are the following: 1) we found brain multimodal monitoring appears adequate to assess regional and global CBF at the bedside in patients with severe TBI, with predominantly diffuse injury type; 2) abnormalities in regional brain physiological and metabolic variables (ICP > 20 mm Hg, Pbto2 < 20 mm Hg, CMD glucose < 1 mmol/L, and CMD LPR > 40) were associated with low regional CBF; 3) brain multimodal monitoring—combining ICP, Pbto2, and CMD—was significantly more accurate than ICP monitoring alone in detecting cerebral hypoperfusion. Regional CBF Reflects Global Perfusion After Severe TBI Regional CBF correlated significantly with global CBF. Caution is warranted, however, because despite this correlation being statistically significant (p < 0.01), the strength of association
Table 2. Main Systemic Physiological Variables Around Perfusion CT Time According to Cerebral Perfusion Pattern Systemic Variables
Low rCBF (n = 14)
Normal rCBF (n = 16)
p
Pao2 (mm Hg)
129 (83–278)
106 (85–175)
< 0.01
Pao2/Fio2 (mm Hg)
381 (172–516)
283 (196–407)
< 0.01
36 (29–41)
37 (30–41)
0.34
Blood hemoglobin (g/L)
117 (92–146)
112 (96–130)
0.46
Blood glucose (mmol/L)
7.2 (5.7–9.3)
7.5 (5.7–9.0)
0.48
Paco2 (mm Hg)
rCBF = regional cerebral blood flow. Results are expressed as median (10th–90th percentiles). Perfusion CT was used to calculate rCBF in one region of interest that was manually drawn around intracranial monitoring (surface area, ~ 50 mm2, see also Fig. 1): cerebral hypoperfusion was defined as an oligemic rCBF < 35 mL/100 g/min as opposed to normal cerebral blood flow ≥ 35 mL/100 g/min.
448
www.ccmjournal.org
February 2015 • Volume 43 • Number 2
Neurologic Critical Care
Table 3. Univariate Associations of Regional Cerebral Blood Flow Data With Brain Multimodal Monitoring Variables According to Cerebral Perfusion Pattern, Dichotomized as Low Versus Normal Brain Physiologic Variables
CMD glucose (mmol/L)
Low rCBF (< 35 mL/100 g/min)
Normal rCBF (≥ 35 mL/100 g/min)
p
0.95 (0.41–2.42)
1.38 (0.64–2.18)
< 0.01
57
22
< 0.01
CMD lactate (mmol/L)
3.15 (1.7–5.0)
3.26 (2.1–5.0)
0.22
CMD pyruvate (μmol/L)
106 (62–182)
118 (75–189)
0.15
CMD glucose < 1 mmol/ (% episodes)
CMD glutamate (μmol/L)
6.9 (1.1–37.2)
7.0 (2.4–26.4)
0.39
CMD LPR
30 (11–44)
28 (14–35)
0.62
CMD LPR > 40 (% episodes)
14
4
0.03
CMD LPR > 25 (% episodes)
70
61
0.24
21 (6–33)
27 (14–39)
< 0.01
Pbto2 < 20 mm Hg (% episodes)
20
9
0.04
Pbto2 < 15 mm Hg (% episodes)
16
5
0.01
Pbto2 < 10 mm Hg (% episodes)
7
0
0.01
Pbto2 (mm Hg)
a
ICP (mm Hg) ICP > 20 mm Hg (% episodes) CPP (mm Hg) CPP < 60 mm Hg (% episodes)
15 (3–24)
12 (0–18)
0.11
30
13
< 0.01
70 (63–83)
72 (63–83)
0.39
12
18
0.26
rCBF = regional cerebral blood flow, CMD = cerebral microdialysis, LPR = lactate-to-pyruvate ratio, Pbto2 = brain tissue oxygen pressure, ICP = intracranial pressure, CPP = cerebral perfusion pressure. a Pbto2 was corrected to Pao2. Perfusion CT was used to calculate rCBF in one region of interest that was manually drawn around intracranial monitoring (surface area, ~ 50 mm2): cerebral hypoperfusion was defined as an oligemic rCBF < 35 mL/100 g/min as opposed to normal CBF ≥ 35 mL/100 g/min. Results are expressed as median (10th–90th percentiles) or as percentages. Boldface data denote p < 0.05.
between the measures (R2 = 0.49) requires care to be taken in interpretation of data from focal monitors. This correlation was found within the acute phase (< 48 hr) of TBI, when providing adequate CBF is crucial (30): our data are in line with previous reports that found a good correlation of CBF with Pbto2 (10–12) or LPR (16) and further extend them, since they tested the value of monitoring using Pbto2 and CMD in combination. Although still debated (31, 32), at our center, the location of brain monitoring after severe TBI is in apparently normal brain parenchyma. The large proportion of patients with diffuse injury in our cohort justifies this approach. On the other extent, we cannot rule out regional disturbances of CBF attributable to focal injury in some patients (1, 33). Only two patients had regional CBF lower than 35 mL/100 g/min with concomitant normal global CBF. These two patients had abnormal cerebral physiology: this illustrates possible CBF heterogeneity and reinforces the fact that multimodal monitoring adequately detects CBF regional disturbances. Additional studies are needed to validate these findings in patients with focal injury. The Relationship Between Brain Multimodal Monitoring and CBF Abnormalities in ICP, Pbto2, and neurochemistry (cerebral glucose and LPR) were associated with low rCBF. These associations Critical Care Medicine
were observed despite moderate elevation of ICP in our cohort. In contrast, we did not find any association between low rCBF and CPP, indicating the poor value of CPP alone to assess cerebral perfusion (5) and the need for other modalities to interpret CPP, for example, the pressure reactivity index (34). Our findings support the use of ICP monitoring in after severe TBI (35) and that maintenance of ICP below 20 mm Hg is adequate to avoid cerebral hypoperfusion. Pbto2 less than 20 mm Hg was also a good marker of low rCBF, which is not unprecedented (10, 12, 14). Brain energy crisis (defined as reduced glucose and elevated LPR > 40 in CMD fluid) is considered as a marker of ischemia (15) and poor outcome after severe TBI (29): we found that both reduced cerebral glucose and elevated LPR were more likely to occur in conditions of low rCBF. Beyond ICP and Pbto2 monitoring, the contribution of CMD may be relevant at the bedside to detect inadequate CBF in a timely fashion after TBI. Intracranial Monitoring to Detect Cerebral Hypoperfusion After TBI at the Bedside Despite increasing utilization, data on the accuracy of brain multimodal monitoring to detect low CBF are limited. In addition, a relevant question from the clinical standpoint is to determine whether brain multimodal monitoring has any additive value over ICP monitoring alone in detecting low CBF www.ccmjournal.org
449
Bouzat et al
is therefore more than an ischemia monitoring in the setting of severe TBI.
Figure 2. Receiver-operating characteristic (ROC) curves of the different monitoring modalities to detect low regional cerebral blood flow (CBF). Low regional CBF, measured with perfusion CT, was defined as an oligemic CBF less than 35 mL/100 g/min. ROC curve for intracranial pressure (ICP) monitoring alone (light gray line), ICP + brain tissue oxygen pressure (Pbto2) monitoring (dark gray line), and ICP + Pbto2 + cerebral microdialysis (CMD) monitoring glucose and lactate/pyruvate ratio (black line) are represented on the same graph. Compared with ICP monitoring alone (area under the ROC curve [AUC], 0.74 [95% CI, 0.61–0.87]), ICP + Pbto2 monitoring (AUC, 0.84 [0.74–0.93]; p = 0.02) or ICP + Pbto2 + CMD monitoring (AUC, 0.88 [0.79–0.96]; p = 0.01) was significantly more accurate in predicting low regional CBF. We also found a trend toward a better prediction of low regional CBF for ICP + Pbto2 + CMD versus ICP + Pbto2 monitoring (p = 0.07).
and to further examine which one(s) of the monitored variables are the most accurate for this purpose. We found that the combination of ICP and Pbto2 was more accurate in predicting low rCBF than the monitoring of ICP alone. Combining CMD to ICP and Pbto2 monitoring further increased the performance of the AUC and was the most accurate to diagnose cerebral hypoperfusion. Using ICP plus CMD monitoring without Pbto2 was less accurate in predicting low rCBF than ICP + Pbto2 and CMD. This may be because CMD may not only be an ischemia monitor. Indeed, despite cerebral hypoperfusion was more frequently associated with low cerebral glucose and elevated LPR, other mechanisms than ischemia may cause cerebral glucose depletion or LPR elevation (e.g., increased glycolysis, energy dysfunction, barriers to oxygen diffusion, and mitochondrial failure) (36, 37). CMD monitoring
Study Limitations First, the study was single center and sample size was relatively small; therefore, data may not be generalized. However, patients were treated with a standardized algorithm, and the cohort was homogeneous, consisting predominantly of patients with diffuse injury, monitored in the right frontal lobe and apparently normal brain. Second, because patients had mainly diffuse injury and were monitored in normal brain regions, our findings cannot be extended to TBI patients with focal injuries. For the same reason, we cannot extrapolate that the observed relationship between CBF and brain multimodality monitoring variables is maintained when monitoring is located in pericontusional areas. Furthermore, diffuse injury induces moderate elevation of ICP (38), as in our study. Thus, the predictive value of ICP to predict low CBF may differ in patients with more severe ICP elevations. Time window for matching PCT with monitoring-derived variables (7 hr) may also be considered as arbitrary, but we considered this time window as the most representative of the patient brain state and also to take into account potential changes in brain physiology due to patient transport to the CT scan. Third, CBF was assessed using PCT, which only measures CBF in anterior and middle vascular territories; therefore, changes in posterior vascular territories may be missed. Finally, our definition of low rCBF deserves further discussion. We acknowledge the fact that although oligemic rCBF was about 2 sds below normal CBF, it could be considered as only marginally low and above the ischemic range. Although this may be a potential limitation, on the other extent, it may be an advantage: our threshold of rCBF less than 35 mL/100 g/min most likely reflects cerebral hypoperfusion, that is, a condition where therapy (e.g., CPP augmentation) might save brain tissue. In addition, although this study was not intended to examine outcome, the fact that low rCBF less than 35 mL/100 g/min was associated with worse outcome implies that this threshold might seem appropriate from the clinical standpoint.
CONCLUSIONS Brain multimodal monitoring—consisting of ICP, Pbto2, and CMD—gives appropriate estimation of global CBF in patients
Table 4. Receiver-Operating Characteristic Curve Analysis of the Different Brain Monitoring Modalities to Predict Cerebral Hypoperfusion (Regional Cerebral Blood Flow < 35 mL/100 g/min) Variable
Area Under the ROC Curve
95% CI
ROC curve for ICP monitoring alone
0.74
0.61–0.87
ROC curve for ICP and CMD monitoring
0.79
0.69–0.90
ROC curve for ICP and Pbto2 monitoring
0.84
0.74–0.93
ROC curve for ICP, Pbto2, and CMD monitoring
0.88
0.79–0.96
ROC = receiver-operating characteristic, ICP = intracranial pressure, CMD = cerebral microdialysis (including CMD glucose and lactate/pyruvate ratio), Pbto2 = brain tissue oxygen pressure.
450
www.ccmjournal.org
February 2015 • Volume 43 • Number 2
Neurologic Critical Care
Figure 3. Receiver-operating characteristic (ROC) curves comparing intracranial pressure (ICP) monitoring alone with ICP + cerebral microdialysis (CMD) to detect low regional cerebral blood flow (CBF). Low regional CBF, measured with perfusion CT, was defined as an oligemic CBF less than 35 mL/100 g/min. ROC curves for ICP monitoring alone (light gray line), ICP + CMD glucose monitoring (dark gray line), and ICP + CMD glucose and lactate/pyruvate ratio (LPR) monitoring (black line) are represented on the same graph. Area under the curve (AUC) is illustrated. The combination of ICP + CMD glucose and LPR provided higher AUC to predict low regional CBF than ICP alone, although there was no statistically significant difference (0.79 [95% CI, 0.69–0.90] vs 0.74 [95% CI, 0.61–0.87]; p = 0.25).
with severe TBI and predominantly diffuse injury. Multimodal monitoring with ICP, Pbto2, and CMD was significantly more accurate in predicting cerebral hypoperfusion than ICP monitoring alone. Our findings suggest that beyond ICP, monitoring of Pbto2 and CMD provides a better assessment of cerebral perfusion state at the bedside after severe TBI.
REFERENCES
1. Maas AI, Stocchetti N, Bullock R: Moderate and severe traumatic brain injury in adults. Lancet Neurol 2008; 7:728–741 2. Dagal A, Lam AM: Cerebral blood flow and the injured brain: How should we monitor and manipulate it? Curr Opin Anaesthesiol 2011; 24:131–137 3. Bratton SL, Chestnut RM, Ghajar J, et al: Guidelines for the management of severe traumatic brain injury. VIII. Intracranial pressure thresholds. J Neurotrauma 2007; 24(Suppl 1):S55–58 4. Bratton SL, Chestnut RM, Ghajar J, et al: Guidelines for the management of severe traumatic brain injury. IX. Cerebral perfusion thresholds. J Neurotrauma 2007; 24(Suppl 1):S59–64 5. Eriksson EA, Barletta JF, Figueroa BE, et al: Cerebral perfusion pressure and intracranial pressure are not surrogates for brain tissue oxygenation in traumatic brain injury. Clin Neurophysiol 2012; 123:1255–1260 6. Oddo M, Levine JM, Mackenzie L, et al: Brain hypoxia is associated with short-term outcome after severe traumatic brain injury independently of intracranial hypertension and low cerebral perfusion pressure. Neurosurgery 2011; 69:1037–1045 7. Chesnut RM, Temkin N, Carney N, et al; Global Neurotrauma Research Group: A trial of intracranial-pressure monitoring in traumatic brain injury. N Engl J Med 2012; 367:2471–2481 8. Oddo M, Villa F, Citerio G: Brain multimodality monitoring: An update. Curr Opin Crit Care 2012; 18:111–118 9. Bouzat P, Sala N, Payen JF, et al: Beyond intracranial pressure: Optimization of cerebral blood flow, oxygen, and substrate delivery after traumatic brain injury. Ann Intensive Care 2013; 3:23 10. Doppenberg EM, Zauner A, Bullock R, et al: Correlations between brain tissue oxygen tension, carbon dioxide tension, pH, and cerebral
Critical Care Medicine
blood flow—A better way of monitoring the severely injured brain? Surg Neurol 1998; 49:650–654 11. Valadka AB, Hlatky R, Furuya Y, et al: Brain tissue PO2: Correlation with cerebral blood flow. Acta Neurochir Suppl 2002; 81:299–301 12. Jaeger M, Soehle M, Schuhmann MU, et al: Correlation of continuously monitored regional cerebral blood flow and brain tissue oxygen. Acta Neurochir (Wien) 2005; 147:51–56 13. Hemphill JC 3rd, Smith WS, Sonne DC, et al: Relationship between brain tissue oxygen tension and CT perfusion: Feasibility and initial results. AJNR Am J Neuroradiol 2005; 26:1095–1100 14. Rosenthal G, Hemphill JC 3rd, Sorani M, et al: Brain tissue oxygen tension is more indicative of oxygen diffusion than oxygen delivery and metabolism in patients with traumatic brain injury. Crit Care Med 2008; 36:1917–1924 15. Hillered L, Vespa PM, Hovda DA: Translational neurochemical research in acute human brain injury: The current status and potential future for cerebral microdialysis. J Neurotrauma 2005; 22:3–41 16. Sarrafzadeh AS, Sakowitz OW, Callsen TA, et al: Bedside microdialysis for early detection of cerebral hypoxia in traumatic brain injury. Neurosurg Focus 2000; 9:e2 17. Vespa P, Bergsneider M, Hattori N, et al: Metabolic crisis without brain ischemia is common after traumatic brain injury: A combined microdialysis and positron emission tomography study. J Cereb Blood Flow Metab 2005; 25:763–774 18. Asgari S, Vespa P, Hu X: Is there any association between cerebral vasoconstriction/vasodilatation and microdialysis lactate to pyruvate ratio increase? Neurocrit Care 2013; 19:56–64 19. Wintermark M, Chiolero R, Van Melle G, et al: Cerebral vascular autoregulation assessed by perfusion-CT in severe head trauma patients. J Neuroradiol 2006; 33:27–37 20. Wintermark M, van Melle G, Schnyder P, et al: Admission perfusion CT: Prognostic value in patients with severe head trauma. Radiology 2004; 232:211–220 21. Wintermark M, Thiran JP, Maeder P, et al: Simultaneous measurement of regional cerebral blood flow by perfusion CT and stable xenon CT: A validation study. AJNR Am J Neuroradiol 2001; 22:905–914 22. Huang AP, Tsai JC, Kuo LT, et al: Clinical application of perfusion computed tomography in neurosurgery. J Neurosurg 2014; 120:473–488 23. Marshall LF, Marshall SB, Klauber MR, et al: The diagnosis of head injury requires a classification based on computed axial tomography. J Neurotrauma 1992; 9(Suppl 1):S287–S292 24. Sala N, Suys T, Zerlauth JB, et al: Cerebral extracellular lactate increase is predominantly nonischemic in patients with severe traumatic brain injury. J Cereb Blood Flow Metab 2013; 33:1815–1822 25. Marshall RS, Lazar RM, Pile-Spellman J, et al: Recovery of brain function during induced cerebral hypoperfusion. Brain 2001; 124:1208–1217 26. Kelly DF, Martin NA, Kordestani R, et al: Cerebral blood flow as a predictor of outcome following traumatic brain injury. J Neurosurg 1997; 86:633–641 27. Doppenberg EM, Zauner A, Watson JC, et al: Determination of the ischemic threshold for brain oxygen tension. Acta Neurochir Suppl 1998; 71:166–169 28. Oddo M, Levine JM, Frangos S, et al: Brain lactate metabolism in humans with subarachnoid hemorrhage. Stroke 2012; 43:1418–1421 29. Timofeev I, Carpenter KL, Nortje J, et al: Cerebral extracellular chemistry and outcome following traumatic brain injury: A microdialysis study of 223 patients. Brain 2011; 134:484–494 30. Robertson CS, Valadka AB, Hannay HJ, et al: Prevention of secondary ischemic insults after severe head injury. Crit Care Med 1999; 27:2086–2095 31. Bhatia A, Gupta AK: Neuromonitoring in the intensive care unit. II. Cerebral oxygenation monitoring and microdialysis. Intensive Care Med 2007; 33:1322–1328 32. Andrews PJ, Citerio G, Longhi L, et al; Neuro-Intensive Care and Emergency Medicine (NICEM) Section of the European Society of Intensive Care Medicine: NICEM consensus on neurological monitoring in acute neurological disease. Intensive Care Med 2008; 34:1362–1370 www.ccmjournal.org
451
Bouzat et al 33. Longhi L, Pagan F, Valeriani V, et al: Monitoring brain tissue oxygen tension in brain-injured patients reveals hypoxic episodes in normal-appearing and in peri-focal tissue. Intensive Care Med 2007; 33:2136–2142 34. Aries MJ, Czosnyka M, Budohoski KP, et al: Continuous determination of optimal cerebral perfusion pressure in traumatic brain injury. Crit Care Med 2012; 40:2456–2463 35. Stein SC, Georgoff P, Meghan S, et al: Relationship of aggressive monitoring and treatment to improved outcomes in severe traumatic brain injury. J Neurosurg 2010; 112:1105–1112
452
www.ccmjournal.org
36. Menon DK, Coles JP, Gupta AK, et al: Diffusion limited oxygen delivery following head injury. Crit Care Med 2004; 32:1384–1390 37. Carre E, Ogier M, Boret H, et al: Metabolic crisis in severely headinjured patients: Is ischemia just the tip of the iceberg? Front Neurol 2013; 4:146 38. Cooper DJ, Rosenfeld JV, Murray L, et al; DECRA Trial Investigators; Australian and New Zealand Intensive Care Society Clinical Trials Group: Decompressive craniectomy in diffuse traumatic brain injury. N Engl J Med 2011; 364:1493–1502
February 2015 • Volume 43 • Number 2
Copyright of Critical Care Medicine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
