Anne Pasco
Aram Ter Minassian Catherine Chapon Laurent Lemaire Florence Franconi Dana Darabi Christine Caron Jean-Pierre Benoit Jean-Jacques Le Jeune
Received: 2 June 2005 Revised: 17 October 2005 Accepted: 8 November 2005 Published online: 17 February 2006
#Springer-Verlag 2006
Dynamics of cerebral edema and the apparent diffusion coefficient of water changes
in patients with severe traumatic brain injury.
A prospective MRI study
Abstract The distinction between intracellular (ICE) and extracellular edema (ECE) has a crucial prognostic and therapeutic importance in patients with severe traumatic brain injury (STBI). Indeed, ICE usually leads to cellular death, and maintenance of a cerebral perfusion pressure (CPP) above 70 mmHg is still under debate since this practice may increase ECE.
The purpose of this study was to describe the ECE and ICE kinetics associated with STBI using quantita- tive diffusion MRI. Twelve patients were prospectively studied. The initial ADC in ICE measured on day 1.3±0.7 is significantly reduced compared to
normal-appearing parenchyma (0.51±
0.12 * 10−3mm2/s vs. 0.76±0.03 * 10−3mm2/s,n=12,P<0.0001) and reaches normality on MRI 3 per- formed on day 14.2±3.3. In patients presenting an extension of ICE on MRI 2 performed on day 6.7±1.4 (ADCMRI2=0.40±0.11 * 10−3mm2/s), ADC values in the extension area at the first MRI were slightly, but not significantly reduced compared to normal parenchyma (0.69±0.05 * 10−3mm2/s,P=0.29). Normalization occurred equally by day 14. ADC in ECE (1.34±0.22 * 10−3mm2/s) was elevated and stable with time under CPP therapy. Therefore, ECE is not worsened by CCP therapy, and ICE appears more relevant than ECE in STBI.
Keywords Traumatic brain injury . Edema . Magnetic resonance imaging . Diffusion . Apparent diffusion coefficient
Introduction
The acute pathological modifications of brain water movements (restricted vs. increased) are now well assessed by diffusion-weighed imaging (DWI) techniques, even if the exact mechanism of restricted diffusion is still not completely understood [1, 2]. The most successful application of this technique is in cases of acute cerebral ischemia, where the loss of ionic gradients because of
ischemia leads to a shift of water from the extracellular to the intracellular compartment and consequently to a restricted water diffusion [1,3].
In severe brain trauma (initial Glasgow Coma Scale≤8), the primary mechanical impact triggers a cascade of molecular and cellular events quite similar to those observed in arterial ischemia, leading to secondary injuries where edema is the main pathological manifestation [4].
Brain edema continues to represent the most challenging A. Pasco (*) .D. Darabi . C. Caron
Department of Radiology,
Larrey Hospital, Angers University, 49033 Cedex, France
e-mail: anpasco@chu-angers.fr Tel.: +33-2-41355262 Fax: +33-2-41355380 A. T. Minassian
Department of Anaesthesiology, Larrey Hospital, Angers University, 49033 Cedex, France
A. Pasco . C. Chapon . L. Lemaire . J.-P. Benoit . J.-J. Le Jeune
Inserm, U 646, 49100 Angers, France
A. Pasco . C. Chapon . L. Lemaire . J.-P. Benoit . J.-J. Le Jeune
Angers University, 49100 Angers, France F. Franconi
SCAS, Angers University, 49045 Angers, France
therapeutic target in patients with severe traumatic brain injury (STBI). Indeed, edema leads to intracranial hyper- tension and decreased cerebral perfusion pressure, the main causes of mortality and morbidity [5,6]. Therefore, as in stroke, diffusion MRI has a theoretically major potential application in STBI.
MRI has documented intracellular “cytotoxic” and extracellular “vasogenic” edema in experimental STBI models [7–11]. However, few studies are available in humans [12–14], and they do not offer the precise char- acterization and/or temporal evolution of these kinds of edema in the clinic. The differentiation between extra- and intracellular edema in severely head-injured patients is of crucial therapeutic importance. Indeed, the maintenance of a cerebral perfusion pressure around 70 mmHg is a current practice in intensive care units [15]. However, this practice is still debated because of the potential aggravating effect on extracellular edema [16].
The purpose of this study was thus to describe the nature of edema with diffusion MRI in the acute phase of STBI and to analyze its temporal and spatial evolution under CPP therapy.
Materials and methods
Patients
After the agreement of our local ethics committee and informed consent of the next of kin, 12 patients (10 men and 2 women; mean age: 27±2.8 years, range: 16 to 45 years) among 19 consecutive patients with STBI were included in this prospective study. The inclusion criteria were: (1) severe head trauma with Glasgow Coma Scale≤8 after initial resuscitation, (2) stable hemodynamic param- eters allowing the third criteria and (3) three MRI examinations in the first 3 weeks post-trauma with an initial MRI between day 0 and day 3 and a 4-day gap between each examination. After an initial computed tomographic (CT) scan at day 0, all patients were hospitalized in the intensive care unit and had continuous intracranial (ICP) and arterial pressure (MAP) monitoring by the intraparenchyma transducer (Codman, France) and radial catheter, respectively. Sedation with midazolam- morphine and mechanical ventilation were performed to achieve 100% arterial oxygen saturation and moderate hypocapnia (around 35 mmHg). According to the routine management protocol of our institution, patients with a cerebral perfusion pressure (CPP) of<70 mmHg were given a continuous infusion of norepinephrine in order to maintain the CPP above this threshold. MRI was only performed on ICP-stabilized patients under close anesthe- siologist and intensivist supervision for SpO2, FiO2, end- tidal CO2and MAP.
MRI protocol
The MR examinations (1.5 T Signa Echospeed Horizon, GE Medical System) included fluid-attenuated inversion recovery (FLAIR), T2 gradient-recalled echo (T2*) and diffusion echo-planar sequences. All images were acquired in the axial anterior commissure-posterior commissure plane (AC/PC) with 5-mm slice thickness, 1.5-mm gap, 24×24-cm2 or 28×28-cm2 field of view, and 20 slices enabling whole brain coverage. The acquisition parameters of the FLAIR and T2 gradient echo were as follows:
FLAIR sequence: repetition time (TR)/echo time (TE)/
inversion time (IT) 10,002/159.5/2,300 ms, matrix size 256×256, with 3’20 min imaging time and T2 gradient echo: TR/TE 700/20 ms, matrix 512×512 with 3’59 imaging time. Diffusion-weighted images were acquired with a single-shot, echo-planar, spin-echo sequence (TR/
TE 6,500/96 ms, matrix size 96×96, interpolated to 128×128) with a baseline T2 acquisition (b=0 s/mm2) and b=1,000 s/mm2. Diffusion gradients were successively and separately set in the three orthogonal directions for a total acquisition time of 0’26 min to cover the whole brain.
Trace images were then generated and apparent diffusion coefficient (ADC) maps were calculated on a pixel-by- pixel basis according to the Stejskal and Tanner equation [17] with a dedicated software program (Functool, General Electric, Buc, France). All MR images were reviewed and the ADC calculated prospectively by a trained neuroradiologist.
Lesion classification
On the initial CT scan, intra-axial lesions were classified as focal when at least a hemorrhagic lesion over 15 mm in diameter was observed, as diffuse axonal injury when CT appeared normal or showed petechial hemorrhages in typical locations (subcortical white matter, corpus callo- sum, basal ganglia, internal capsule and upper brain stem) and as mixed when both type of lesions were observed.
Lesions were classified on the first MRI also in three types depending on the size and extent. Focal (or multifocal) lesions corresponded to lesion(s) over 15 mm in diameter.
Diffuse axonal lesions corresponded to multiple lesions less than 10-15 mm in diameter along white matter tracts and at the interface of gray and white matter and/or in the corpus callosum and/or in the upper brain stem. The association of focal plus diffuse axonal lesions also defined mixed lesions.
FLAIR, T2*-weighted and diffusion-weighted images with the calculation of apparent diffusion coefficient (ADC) sequences defined respectively the presence and location of edema, the presence and location of axonal injuries and hemorrhagic contusions and the distribution of edema (intracellular vs. extracellular).
Diffusion data processing
Diffusion sequences (the baseline T2 acquisition at b=0 s/mm2, the b=1,000 s/mm2 acquisition and the ADC maps) allowed the definition of edemas, both intracellular (ICE) and extracellular (ECE). ICE was defined on the ADC map by areas of low intensity (decrease diffusion) and ECE by an abnormally high intensity (increase diffusion). The largest lesion on the first MRI was defined as the main lesion.
The spatial evolution of edema was analyzed comparing the first two ADC maps and the last two ADC maps in the 12 patients. Four groups were then defined according to this evolution: “complete resolution” when edema dis- appeared, “partial resolution” when edema decreased,
“stable”when a minor extension or stabilization was ob- served and “enlargement” when the edema had at least doubled in size.
The absolute ADC values were obtained prospectively and after each examination by manual positioning of regions of interest (ROIs;≤14 mm2) over ICE and ECE of the main lesion and in visually healthy tissue (VHT) ipsi- and contralaterally to the main lesion. ADC in VHT was measured in four distinct neuroanatomic slices, i.e., a
“polygon” slice, a “Fleschig” slice, a “ventricular” slice and “semi-oval centrum” slice. The core of hemorrhagic lesions was excluded from ADC calculation. Despite the fact that imaging sessions were not performed under stereotactic conditions, the definition of the imaging plan along the AC/PC plan and the entire covering of the brain Fig. 1 Illustration of focal lesions.aFLAIR andbADC map of a
29-year-old woman presenting a right temporo-frontal lesion defined as a combination of extracellular (star) and intracellular (arrow) edemas on initial MRI performed 2 days post-trauma. The round hypointense signal within the lesion corresponds to a hematoma.c FLAIR and d ADC map of a 17-year-old man presenting a left temporo-occipital contusion defined as pure intracellular edema (arrow) on the ADC map
Fig. 2 Illustration of mixed lesions. The T2* GRE-weighted set of images (a) illustrate a left frontal hemorrhagic contusion plus multiple shearing injuries extending from the cortico-medullary junction to the corpus callosum and to the mesencephalon in a 22-
year-old man. In the presented case, intra- (hypointense area:arrow) and extracellular edema (hyperintense area: star) are observed around the focal lesion as defined on ADC maps (b), whereas DAIs are predominantly composed of intracellular edema (arrow)
Fig. 3 Illustration of extensive intracellular edema. The sets of ADC maps at day 1 (a) and day 6 (b) post-trauma show a large extension of an intracellular edema (hypointense signal) from the left temporo-occipital area to the left parieto-frontal region of a 17-year-old man Table 1 Spatial evolution of intra- and extracellular edema between MRI scans
Patients Lesion CT scan
Lesion MRI 1
Edema ICE ECE
MRI 1-2 (dates/trauma) MRI 2-3 (dates/trauma) MRI 1-2 MRI 2-3
Patient 1 DAI F ICE Enlargement
(days 1-6)
Partial resolution (days 6-13)
No signif. ECE No signif. ECE
Patient 2 F F ICE+ECE Enlargement
(days 2-8)
Partial resolution (days 8-15)
Partial resolution
Partial resolution Patient 3 F F ICE+ECE Stable (days 1-5) Partial resolution
(days 5-11)
Stable Stable
Patient 4 F F ICE+ECE Stable (days 1-4) Partial resolution (days 4-19)
Stable Partial resolution Patient 5 DAI DAI ICE Stable (days 2-6) Stable (days 6-10) No signif. ECE No signif. ECE Patient 6 DAI DAI ICE Stable (days 3-7) Partial resolution
(days 7-13)
No signif. ECE No signif. ECE Patient 7 DAI DAI ICE Partial resolution
(days 0-6)
Complete resolution (days 6-12)
No signif. ECE No signif. ECE
Patient 8 MX MX ICE+ECE Enlargement
(days 1-9)
Complete resolution (days 9-21)
Stable Partial resolution
Patient 9 F MX ICE+ECE Enlargement
(days 1-8)
Partial resolution (days 8-12)
Stable Stable
Patient 10 MX MX ICE+ECE Stable (days 2-7) Partial resolution (days 7-13)
Stable Stable
Patient 11 MX MX ICE+ECE Stable (days 1-8) Complete resolution (day 8-18)
Stable Partial resolution
Patient 12 F MX ICE+ECE Enlargement
(days 1-6)
Partial resolution (days 6-13)
Stable Partial resolution Notes:-ICE: intracellular edema;ECE: extracellular edema;signif.: significant;DAI: diffuse axonal injury;F: focal;MX: mixed (DAI+F)
with 20 slices allowed lesion relocalization and therefore ROI placement from one MR session to another.
In patients with a large spatial evolution of intracellular edema, additional ROIs were drawn within the extension area and back extrapolated on the previously acquired maps. The core was defined as the area of initial decreased ADC, and the extension area was defined as the area of decreased ADC on the following MRI. Statistical analysis of data was performed using a bi-factorial analysis of variance (ANOVA) with a multiple least square analysis.
Results
On the initial CTs, performed within the first hours post- TBI, five patients had focal lesions, four had diffuse axonal injuries and three had mixed lesions. Fig.1shows typical focal lesions as depicted in four patients on MRI 1 performed 1.3±0.7 days post-TBI. Those lesions were characterized on ADC maps as ICE in one case and as a combination of ICE/ECE in the three other cases. The combination of ICE/ECE was also observed on ADC maps Fig. 4 Illustration of extracel-
lular edema regression. The sets of ADC maps at day 9 (a) and day 21 (b) post-trauma show a partial regression of an extra- cellular edema (hyperintense signal) in the frontal region of a 22-year-old man. Note the re- gression of the intracellular edema (hypointense signal) during the same period of time
Fig. 5 ADC evolution of intracellular (■) and extracellular (Δ) edemas defined within the main lesion as a function of time after trauma.Dotted line(◊) corresponds to ADC evolution in contralat- eral visually healthy tissue. # Statistically different from VHH at the same time point (P<0.0001); § statistically different from first time point (P<0.0001);⌑ statistically different from second time point (P<0.002)
Fig. 6 ADC evolution in core (■) and extension area (●) of patients presenting extensive intracellular edema. Contralateral visually healthy tissue ADC (dotted line-◊) evolution is equally presented.
# Statistically different from VHH at the same time point (P<0.0001);
§ statistically different from first time point (P<0.0001);⌑statistically different from second time point (P<0.002); * statistically different from the extension area (*P<0.02, **P<0.0001)
for all patients showing mixed lesions on T2* at MRI 1 (Fig.2), whereas all DAI lesions detected on this sequence presented ICE on ADC maps. Pure ECE were never observed.
Table 1 presents the spatial evolution of ICE and ECE with time. Among the 12 patients included, 5 were characterized by an enlarged ICE (Fig. 3), 6 by a stable ICE and 1 by a partial resolution of a minor ICE within a DAI lesion between MRI 1 and MRI 2 performed 6.7±1.4 days post-TBI. Between MRI 2 and MRI 3 performed 14.2±
3.3 days post-TBI, a decrease or a partial resolution in ICE was observed in all patients, except in one case where the ICE remained stable. ECE detected on MRI 1 in 8 out of 12 patients remained stable on MRI 2 and tended to decrease between MRI 2 and MRI 3 in 5 cases (Fig.4).
A quantitative approach of ADC evolution in ICE and ECE was performed and is presented in Fig.5. First of all, ADC values defined in visually healthy tissues in contra- or ipsilateral parenchyma were not statistically different from each other (P=0.99). Within the main lesion, the ADC value for the ICE on MRI 1 was equal to ADCICE MRI1= 0.51±
0.12 × 10−3mm2/s (n=12) and statistically reduced compared to the visually healthy contralateral hemisphere (VHH) ADCVHH MRI1= 0.76±0.03 × 10−3; mm2/s (P< 0.0001). ADC in ICE slightly increased between MRI 1 and 2, ADCICE MRI2= 0.55±0.16 × 10−3mm2/s (P= 0.4), to reach normality at MRI 3, ADCICE MRI3= 0.70±0.12 × 10−3mm2/s (P= 0.25 compared to VHH and P < 0.0001 compared to ADCICE MRI1).
For ECE, the ADC values of ADCECE MRI1¼1:34 0:22103mm2=sðn¼8Þ are calculated and are statis- tically increased compared to the visually healthy contra- lateral hemisphere (P<0.0001). The ECE remained stable and elevated all along the experimental time (Fig.5).
In the five patients presenting an extension of ICE between MRI 1 and MRI 2, the ADC values (Fig.6) within the ex- tension were not significantly different from the visually healthy contralateral hemisphere ADCextensionMRI1¼0:69 0:05103mm2=s P ¼0:29Þat the first MRI. On MRI 2, a significant reduction of ADC was measured com- pared to the visually healthy contralateral hemisphere ðADCextensionMRI2¼0:400:11103mm2=s;P<0:0001Þ;
and the value appeared to be significantly reduced com- pared to the initial ICE lesion (P<0.02). Normalization of all regions was observed on MRI 3.
Discussion
Brain trauma is a dynamic process characterized by two waves of lesions [6, 18]. The first wave encompasses the immediate mechanical damage to the central nervous system (CNS) that occurs at the moment of impact, and the second wave, initiated at the moment of the traumatic insult, will progress over a period of time ranging from hours to days after injury. This secondary injury to the CNS is a complex
and still incompletely understood process in which edema formation and intracranial hypertension can lead to critical ischemia [6,18,19]. These potentially reversible processes constitute a major therapeutic challenge.
Traumatic and ischemic brain edemas obviously have some similar pathophysiologic mechanisms [4]. Etiologies are nevertheless different. Cerebral ischemia results from a variety of causes that impairs cerebral blood flow and leads to the deprivation of both oxygen and glucose. An anoxic depolarization and a massive liberation of excitotoxic glutamate in the extracellular synaptic space are observed [20]. As a consequence, a massive influx of Na+and Ca++
and finally of osmotically obligated water occurs, resulting in glial and neuronal cell edema [4, 20]. In severe brain trauma, the mechanical impact itself triggers a massive release of glutamate and neuronal depolarization (traumatic depolarization) with secondary impairment of energy metabolism with similar consequences on ionic homeosta- sis and cellular edema as observed in ischemia [4, 6].
Furthermore, the decrease of cerebral perfusion secondary to intracranial hypertension promotes edema formation in a self-sustained way [5,21].
Quantitative diffusion MR imaging provides informa- tion about the distribution of water and is therefore a unique non-invasive tool to access intra- and extracellular conditions in ischemic stroke [3] and in brain trauma [1, 12–14, 22]. However, the methodological difficulties in performing MRI in the acute phase of severe human brain trauma explain the paucity of studies particularly within the first days post-trauma and the absence of kinetics studies of edema. Contrarily, animal studies of the evolution of post- traumatic brain edema are numerous and give kinetics data [7–11], but the uses of different traumatic models, associated to secondary insults or not (i.e., hypoxia and hypotension) [7, 23, 24], give rise to heterogeneous and sometimes conflicting results [7–11].
In our clinical study, intracellular edema (ICE) with decreased ADC values was associated with focal, diffuse axonal or mixed lesions and was observed as soon as 8 h post-trauma. STBI-related ICE appeared similar to arterial ischemia-related ICE [1, 3, 25, 26] in terms of ADC kinetics with pseudo-normalization around day 12, but also in terms of potential spatial and delayed extension of the main lesion. Indeed, huge extensions were observed in four patients between MRI 1 and MRI 2 with resolution by day 12, as usually noticed in ischemia [26]. Those “growing lesions”confirm the existence of a posttraumatic penumbra [6] as observed in ischemia [25–28].
Despite the fact that qualitative assessment of MRI 1 failed to find pathological modification in the extension areas, ADC values were not strictly normal. Indeed, the mean value of 0.69±0.05×10−3mm2/s in the extension area on MRI 1 corresponds to the lower threshold of 2 standard deviations of normal ADC values. This suggests that the traumatic penumbra could be detected by early ADC value changes. However, this remains to be demonstrated on a
larger series. Furthermore, we have to be cautious in the interpretation of the ADC changes. Many factors are involved in the restricted diffusion of water. Besides the widely accepted theory of disrupted energy metabolism and ion-homeostasis and the reduction of extracellular space by cellular swelling, many other factors are thought to contribute to the reduced mobility of free water, including increased tortuosity of the extracellular and intracellular spaces and increased intracellular viscosity [1]. Furthermore, ICE and ECE often coexist in brain trauma [7,8,12,13], and as they have their own kinetics, they may make variable contributions to the ADC values.
Complementary perfusion-weighted MRI could then effi- ciently sustain diffusion-weighted MRI despite the fact that ICE surrounding intracranial hematoma is not necessarily associated with decreased cerebral blood flow [29].
Another issue of this work deals with the controversial role of ECE in brain trauma. On one hand, ECE is, by definition, associated with a BBB breakdown [30] and therefore interstitial plasmatic leakage that could be increased in CPP-associated therapy [16]. On the other hand, as ECE totally resorbs without brain sequela, its role may be overestimated compared to ICE [8, 30]. In our study, ECE was found only in association with intrapa- renchymal hemorrhage, not extensively and definitely never predominantly, as in Hergan’s study [14], despite the CPP therapy performed at our institution. Experimental studies support that ECE is related to post-traumatic microvascular section [8, 10]. Dye studies showed a biphasic profile for BBB permeability with a first phase within minutes after STBI and a second phase about day 3 post-trauma [10,31].
Contrast-enhanced MRI using DOTA-gadolinium, a diffusive contrast agent, showed on an impact-acceleration model that BBB permeability occurs very early and transiently (no more than 30 mn) [32], and the use of quantitative diffusion MRI showed an increase of ADC values during the first 60 mn, indicating ECE formation [8]. However, the subsequent decrease in ADC associated with the continuous increase in cerebral water content indicates ICE formation and suggests the absence of worsening of the ECE.
Another potential cause of ECE development is the presence of intraparenchymal blood itself. Indeed, animal models have shown a critical role of extravascular blood in the initiation of the ECE via a direct [33] or an indirect effect [34] of thrombin on the BBB breakdown and on the inflammatory reaction. The osmotic effect of intraparenchy-
mal clots has also been advocated in ECE development. In a recent study on human brain hemorrhage, Carpuapoma et al.
[35] found a correlation between the size of hematomas and the severity of ECE development that seems to indicate a
“dose-dependent”relationship between the concentration of osmotically active substances and the severity of BBB disruption and subsequently ECE.
The present study has however some limitations. First, the relatively small number of patients only permits the description of trends. Secondly, DAI lesions are character- ized by small-sized lesions with a hemorrhagic component, and therefore ADC calculations were usually performed on small ROI. There are also possible pitfalls of ADC calculations due to signal alterations secondary to local field inhomogeneities (hemorrhage) or susceptibility artifacts (near the skull base or due to metallic implants).
Finally, a quantitative measurement of the volume of both types of edema would permit a better refinement of kinetic study, and coupling with hemodynamic and metabolic studies would allow the definition of the “traumatic penumbra,”which is a crucial therapeutic target.
Conclusion
Cerebral edema development is a critical event after severe head injury that can be characterized over time using non- invasive quantitative diffusion MRI. Indeed, on this follow-up of patients with severe head injury over 2 weeks, intracellular and extracellular associated edemas were kinetically described. Intracellular edema appeared to evolve with respect to MR characteristics and to the extent, whereas extracellular edema did not. Therefore, intracel- lular edemas may be more relevant for the clinician, and the lack of worsening in extracellular edema despite CPP therapy may argue positively for this practice. Despite their different etiologies, severe cerebral trauma and ischemia have obvious similarities in edema kinetics. These preliminary results with diffusion imaging constitute a step toward improving the target definition and the monitoring of anti-edematous treatment.
Acknowledgements The authors thank Pierre Gautronneau for his technical assistance. They also gratefully acknowledge the teams of the Departments of Radiology and Anesthesiology for their help during the study.
References
1. Huisman TA (2003) Diffusion- weighted imaging: basic concepts and application in cerebral stroke and head trauma. Eur Radiol 13:2283-2297
2. Sorensen AG (2002) Apparently, dif- fusion coefficient value and stroke treatment remains mysterious. AJNR Am J Neuroradiol 23:177-178
3. Warach S, Moseley M, Sorensen A, Koroshetz WJ (1996) Time course of diffusion imaging abnormalities in human stroke. Stroke 27:1254-1256
4. Leker RR, Shohami E (2002) Cerebral ischemia and trauma-different etiolo- gies yet similar mechanisms: neuro- protective opportunities. Brain Res Brain Res Rev 39:55-73
5. Rosner MJ, Rosner SD, Johnson AH (1995) Cerebral perfusion pressure:
management protocol and clinical re- sults. J Neurosurg 83:949-962 6. Sahuquillo J, Poca MA, Amoros S (2001)
Current aspects of pathophysiology and cell dysfunction after severe head injury.
Curr Pharm Des 7:1475-1503
7. Ito J, Marmarou A, Barzo P, Fatouros P, Corwin F (1996) Characterization of edema by diffusion-weighted imaging in experimental traumatic brain injury.
J Neurosurg 84:97-103
8. Barzo P, Marmarou A, Fatouros P, Hayasaki K, Corwin F (1997) Con- tribution of vasogenic and cellular edema to traumatic brain swelling measured by diffusion-weighted imag- ing. J Neurosurg 87:900-907
9. Albensi BC, Knoblach SM, Chew BG, O’Reilly MP, Faden AI, Pekar JJ (2000) Diffusion and high resolution MRI of traumatic brain injury in rats:
time course and correlation with his- tology. Exp Neurol 162:61-72 10. Beaumont A, Marmarou A, Hayasaki K
et al (2000) The permissive nature of blood brain barrier (BBB) opening in edema formation following traumatic brain injury. Acta Neurochir Suppl 76:125-129
11. Schneider G, Fries P, Wagner-Jochem D et al (2002) Pathophysiological changes after traumatic brain injury:
comparison of two experimental animal models by means of MRI. Magma 14:233-241
12. Marmarou A, Fatouros PP, Barzo P et al (2000) Contribution of edema and cerebral blood volume to traumatic brain swelling in head-injured patients.
J Neurosurg 93:183-193
13. Marmarou A, Portella G, Barzo P et al (2000) Distinguishing between cellular and vasogenic edema in head injured patients with focal lesions using mag- netic resonance imaging. Acta Neurochir Suppl 76:349-351
14. Hergan K, Schaefer PW, Sorensen AG, Gonzalez RG, Huisman TA (2002) Diffusion-weighted MRI in diffuse ax- onal injury of the brain. Eur Radiol 12:2536-2541
15. Bullock R, Chesnut RM, Clifton G et al (1996) Guidelines for the management of severe head injury. Brain Trauma Foun- dation. Eur J Emerg Med 3:109-127 16. Miller JD (1994) Vasoconstriction as
head injury treatment–right or wrong?
Int Care Med 20:249-250
17. Stejskal EO, Tanner JE (1965) Spin diffusion measurements: spin-echoes in the presence of a time-dependent field gradient. J Chem Phys 42:288-292 18. Laurer HL, Lenzlinger PM, McIntosh
TK (2000) Models of traumatic brain injury. Eur J Trauma 26:95-1000 19. Graham DI, Adams JH, Doyle D
(1978) Ischaemic brain damage in fatal non-missile head injuries. J Neurol Sci 39:213-234
20. Dirnagl U, Iadecola C, Moskowitz MA (1999) Pathobiology of ischaemic stroke: an integrated view. Trends Neurosci 22:391-397
21. Bouma GJ, Muizelaar JP, Choi SC, Newlon PG, Young HF (1991) Cerebral circulation and metabolism after severe traumatic brain injury: the elusive role of ischemia. J Neurosurg 75:685-693 22. Parizel PM, Van Goethem JW, Özsarlak
Ö, Maes M, Phillips CD (2005) New developments in the neuroradiological diagnosis of craniocerebral trauma. Eur Radiol 15:569–581
23. Beaumont A, Hayasaki K, Marmarou A, Barzo P, Fatouros P, Corwin F (2000) The effects of dopamine on edema formation in two models of traumatic brain injury. Acta Neurochir Suppl 76:147-151
24. Portella G, Beaumont A, Corwin F, Fatouros P, Marmarou A (2000) Char- acterizing edema associated with corti- cal contusion and secondary insult using magnetic resonance spectroscopy.
Acta Neurochir Suppl 76:273-275 25. Baird AE, Benfield A, Schlaug G et al
(1997) Enlargement of human cerebral ischemic lesion volumes measured by diffusion-weighted magnetic resonance imaging. Ann Neurol 41:581-589 26. Schwamm LH, Koroshetz WJ, Sorensen
AG et al (1998) Time course of lesion development in patients with acute stroke: serial diffusion- and hemody- namic-weighted magnetic resonance imaging. Stroke 29:2268-2276
27. Sharp FR, Lu A, Tang Y, Millhorn DE (2000) Multiple molecular penumbras after focal cerebral ischemia. J Cereb Blood Flow Metab 20:1011-1032 28. Witte OW, Bidmon HJ, Schiene K,
Redecker C, Hagemann G (2000) Functional differentiation of multiple perilesional zones after focal cerebral ischemia. J Cereb Blood Flow Metab 20:1149-1165
29. Zazulia AR, Diringer MN, Videen TO et al (2001) Hypoperfusion without ischemia surrounding acute intracere- bral hemorrhage. J Cereb Blood Flow Metab 21:804-810
30. Klatzo I (1967) Neuropathological aspects of brain edema. J Neuropath Exp Neurol 26:1-14
31. Baskaya MK, Rao AM, Dogan A, Donaldson D, Dempsey RJ (1997) The biphasic opening of the blood-brain barrier in the cortex and hippocampus after traumatic brain injury in rats.
Neurosci Lett 226:33-36
32. Barzo P, Marmarou A, Fatouros P, Corwin F, Dunbar J (1996) Magnetic resonance imaging-monitored acute blood-brain-barrier changes in experi- mental traumatic brain injury.
J Neurosurg 85:1113-1121
33. Lee KR, Kawai N, Kim S, Sagher O, Hoff JT (1997) Mechanisms of edema formation after intracerebral hemor- rhage: effects of thrombine on cerebral blood flow, blood-brain barrier perme- ability, and cell survival in a rat model.
J Neurosurg 86:272-278
34. Sansing LH, Kaznatcheeva EA, Perkins CJ, Komaroff E, Gutman FB, Newman GC (2003) Edema after intracerebral hemorrhage: correlations with coagula- tion parameters and treatment.
J Neurosurg 98:985-992
35. Carhuapoma JR, Barker PB, Hanley DF, Wang P, Beauchamp NJ (2002) Human brain hemorrhage: quantifica- tion of perihematoma edema by use of diffusion-weighted MR imaging. AJNR Am J Neuroradiol 23:1322-1326
