Measurement of changes in brain water in man by magnetic resonance imaging

John Hunter was undoubtedly aware of the water content of normal brain tissue, and described cerebral oedema. The advent of nuclear magnetic resonance (NMR) shed new light on brain water, and the derivation of spatial information and hence images from NMR signals, has permitted studies of regional brain water in man in vivo. The initial study described here tested whether NMR longitudinal relaxation time (T1) correlates with brain water content in the cerebral cortex and white matter in man, and significant relationships have been demonstrated in cortex (r = 0.65, P less than 0.002) and white matter (r = 0.94, P less than 0.0001), the latter having narrow 95% confidence limits. The residual variance allows the prediction of water content from the T1 of white matter, measured from the image of a single patient, with an accuracy of +/- 4% of total tissue water with 95% confidence. In the further study described, the effects of dexamethasone and an infusion of 20% mannitol on brain water content has been assessed in patients with intrinsic cerebral tumours. Dexamethasone had no significant effect on the T1 of normal brain, oedematous peritumoural white matter, or tumour tissue. It must be concluded that the water content of these tissues is not changed by dexamethasone and that the clinical improvement seen in patients with cerebral tumours immediately after dexamethasone has to be explained by some mechanism other than a reduction in cerebral oedema. Mannitol did reduce the T1 of oedematous peritumoural white matter, and the T1 of tumour tissue, but did not change the T1 of normal brain significantly.(ABSTRACT TRUNCATED AT 250 WORDS)     

A study on peritumoural brain oedema around meningiomas by CT and MRI scanning

Summary There is a great variability in the amount of peritumoural brain oedema accompanying meningiomas. In a previous study it was found that the degree of brain oedema in the white matter around meningiomas correlated with disruption of the layers (especially the cerebral cortex), which separate the tumour from the white matter, as well as with the size and histological subtype of the tumour.In the present study comprising 9 meningiomas, the volume of oedema was calculated by integration of the cross-sectional oedematous areas on serial MRI slices. The volume of oedema was zero in 3 cases and ranged from 11 to 176.4 ml in the other 6 cases. The MRI-scans also showed disruption of the cortex in all cases, ranging from slight to severe. T₁ and T₂ measurements were made at the level of maximum extension, using a mixed sequence at a field strength of 1.5 T. From the T₂ values tissue water content in % was calculated using the equations: WC=39.36/(R₂ + 37.2) for cortex, and WC=29.63/(R₂ + 27.8) for white matter. These had been obtained by correlating water content with relaxation rates, measured in vitro on human brain autopsy specimens which were subjected to hydration with distilled water or dehydration by hyperosmolar solutions. Mean water content amounted to 82.53% for normal cortex, 74.72% for normal white matter, and 84.59% for oedematous white matter around the tumour.On the assumption that the spread of contrast agent marks the advancement of the front of oedema produced by the tumour, CT-studies were made before, and at 1 1/2, 3 and 6h after contrast infusion. The increase in diameter of the contrast-stained area on the CT-scan allows calculation of the excess of oedema production per unit tumour volume. Of 6 tumours with oedema (mean peritumoural water content of 91% and mean volume of oedema of 69.2 ml) the production excess at the steady-state was 0.18–1.08 ml/h/cm³ tumour volume, whereas 3 tumours without associated oedema had a production excess of 0.03–0.12 ml/h/cm³. Moreover, penetration of the cortex seems to constitute a separate factor determinig the spread of oedema.     

Sodium nitroprusside (NPS) stimulates the lactate formation and causes intracellular electrolyte shifts in brain slices

Summary Nitroprusside sodium (NPS) was added to slices from grey and white rabbit brain matter. The respiration, consumption of glucose, formation of lactate and pyruvate, water content, size of the extracellular space, and total and intracellular sodium and potassium concentrations were determined and calculated.0.5 and 1 mM NPS stimulated the glycolytic glucose breakdown, and inhibited the respiration of grey and white slices. The respiration of white matter slices was stimulated by 0.05 and 0.01 mM NPS.The total and intracellular potassium concentrations in slices from grey and white matter were lowered markedly by 1 mM NPS. The total and the intracellular sodium concentrations in grey brain matter slices became higher in the presence of 1 mM NPS. A warning is given against uncritical clinical use of higher concentrations of NPS.Part of the Dissertation, Med. Fakultät Homburg/Saar, 1979.     

Extracellular space and electrolyte distribution in cortex and white matter of dog brain in cold induced oedema

Summary 24 hours after a circumscribed cold injury of the cortex dog brains were perfused from the lateral ventricle and the frontal subarachnoidal space to the cisterna magna with an artificial CSF containing trace amounts of³⁵S-labelled thiosulphate. Simultaneously the extracellular tracer was administered intravenously. Extracellular fluid volume was estimated and found to be increased from 10 to 15% in the oedematous cortex and from 10 to 27% in the oedematous white matter. The actual size of ECS in oedematous white matter, however, must be larger as indicated by the relative alterations of thiosulphate distribution, tissue water, sodium and chloride. Apparently a small part of the fluid accumulation affects the cellular compartment in oedematous white matter. It may be concluded from the close spatial correspondence of the spreading of I¹³¹ albumin and Evans blue, the increase in water and sodium content, and the enlargement of the TSS that the dilated extracellular channels are filled with a plasma like oedema fluid, derived from blood.This investigation was supported by a grant from the Deutsche Forschungsgemeinschaft.     

Spatial and quantitative distribution of human peritumoural brain oedema in computerized tomography

Computerized tomography (CT) was used to study the pathways of oedema spreading in man. Based on the assumption that local changes in CT numbers in oedematous white matter closely correspond to changes in tissue water content, CT numbers of consecutive tissue blocks of 3.0-3.6 mm were examined in the main directions of oedema spreading: towards the deep white matter, towards the cortex and towards the ventricle. Tumours with oedema grade II and III showed a reduction of CT number of 10 +/- 1.8. The corresponding increase in water content of about 10-12% seems to be an upper limit of fluid accumulation in the white matter. From this oedema centre, water content very slowly and gradually decreased along the oedema projection into the deep white matter. In contrast, if oedema reached the cortex of adjacent gyri, the decline in water content was very sharp. A similar observation was made in the external capsule where oedema sharply declined at the border to the adjacent grey matter, putamen and claustrum. Oedema projection towards the ventricle showed a nearly uniform magnitude from the centre to the ventricular lining, suggesting a certain resistance by a limited capacity of transependymal drainage of oedema fluid. It is assumed that the spatial distribution and extension of oedema around a brain tumour is determined by a system of differential resistance to fluid movement in the following order: grey matter--ventricular lining--white matter.     

Peritumoral oedema and lipid content

Summary The authors analyzed the water and lipid contents of oedematous tissues of white matter obtained during the operation, comparing with the absorption value in the CT scan. As the absorption value lowered its level, the water content increased, while the lipid decreased. However, after subtracting these two components, the content of protein also showed a significant decrement parallel to the absorption value despite the existence of extravasated protein. The decrement of lipid content was considered to be due to a dilution effect by an expanded volume of extravasated oedema fluid, as well as the decrease in protein, rather than the actual loss of tissue component; that is demyelination. It is the relative decrease in protein content which contributes to the low absorption value of oedematous tissue.     

Experimental transplantation gliomas in the adult cat brain

Experimental brain tumours were produced in adult cats by stereotactic xenotransplantation of the rat glioma clone F98. Regional ATP, glucose and lactate were measured after 2-4 weeks on coronal cryostat sections by substrate-induced bioluminescence, potassium content was imaged by the histochemical sodium cobaltinitrite method, and regional pH by incubating cryostat sections with the fluorescent pH-indicator umbelliferone. The regional biochemical alterations were correlated with magnetic resonance imaging and tissue water content. Biochemical changes were heterogeneous in tumours but exhibited a rather uniform pattern in peritumoural oedema. ATP was consistently reduced, glucose and lactate were increased and pH was more alkaline than in normal white matter. The decrease of ATP matched the increase of water, indicating that ATP decline represents fractional dilution in the oedematous tissue rather than break-down of energy metabolism. The increased lactate levels, therefore, may originate from the tumour and not from a metabolic disturbance in the peritumoural oedematous tissue. The implications of this interpretation for the pathogenesis of peritumoural oedema are discussed.     

Examination of the metabolism of oedematous brain tissue

Summary Oedema were induced in rabbit brains by the application of cold metal to defined areas of intact dura for constant times. Trephine openings were made over the right cerebral hemispheres. 24 hours later the animals were killed, and identical pieces of occipital cortex from both hemispheres were removed. In all specimens we determined the water content, the water uptake of the sliced brain in vitro, and the glucose metabolism of the brain slices. Respiration, carbon dioxide production in the presence of bicarbonate, consumption of glucose, and production of lactate and pyruvate were measured. The incubations were made in physiological saline (with K⁺, Ca²⁺, Mg²⁺ and PO₄ ³- additions) in the presence of gas mixtures with different contents of oxygen.The mean difference of the water contents in vivo was 1.3%. The lowest water content after in vitro incubations was found in the presence of 95% O₂ and 5% CO₂.The oedematous tissue had a slightly stimulated respiration, and its aerobic glycolysis was markedly increased. These metabolic differences were observed only in the presence of 12% O₂ and 5% CO₂ and higher oxygen concentrations. In the presence of 4% O₂ and 5% CO₂ neither respiratory nor glycolytic differences existed between normal and oedematous tissue.A pronounced Pasteur effect was seen in all the experiments. Increase in glycolysis is a more sensitive indicator of oxygen lack than is decrease in respiration.During 150 min of incubation respiration and carbon dioxide production become slower. The decrease in total and glycolytic CO₂ production is greater in slices of control hemispheres than in slices of oedematous hemispheres.The best correlation between metabolic activities determined by manometry and metabolic activities determined by glucose, lactate and pyruvate determinations was found in experiments in the presence of 4% O₂ and 5% CO₂. It is concluded that unsliced brain in the presence of air, at room temperature and below, has nearly the same respiration and glycolysis within the whole glucose degradation process as have brain slices at 38 °C in the presence of 4% O₂ and 5% CO₂.The conclusion is made that in vitro the best metabolic situation for brain slices, normal and oedematous, is between 20% and 95% of oxygen.Supported by the Deutsche Forschungsgemeinschaft: He 573/3.With technical assistance from E. Seifert and H. Verfürth.     

Treatment of experimental brain oedema following sudden decompression,surgical wound,and cold lesion with vasoprotective drugs and the proteinase inhibitor “Trasylol”

The study was performed on 81 cats with three models of experimental brain oedema: sudden decompression, surgical wound, and cold injury. During the experiments blood pressure, central venous pressure, and intracranial pressure were recorded. The blood-brain-barrier was tested with Evans blue solution. The gray and white matter tissue was sampled at the end of the experiment, and the water content and sodium and potassium concentrations were determined. The animals with the same experimental model were divided into three groups: untreated, treated with the vasoprotective agents, and treated with the protease inhibitor Trasylol. In the sudden decompression model after balloon deflation, white matter haemorrhages and oedema development were found in gray matter and basal nuclei. In animals treated with the vasoprotective drugs, haemorrhages were not observed, and oedematous changes were less pronounced. The Trasylol effect on oedema development was not significant in this model. In the surgical wound model, oedematous changes were observed after 24 hours following the lesion. Oedema occurred in the white matter, as in the animals with cold lesions. In both models--surgical wound and cold lesion--the beneficial effect of Trasylol was shown, while the effect of Aescorin was less evident. The results obtained seemed to testify to the usefulness of both Trasylol and vasoprotective drugs in the prevention and treatment of brain oedema in neurosurgical patients.     

Examination of the metabolism of oedematous brain tissue

Summary We induced oedema in rabbit brains by applying cold to a defined area of the intact dura for a constant time. 24 hours later the animals were killed, and a piece of brain cortex adjacent to the region of cold application and an exactly similar piece from the other hemisphere were removed. In both pieces we determined the water content, the water uptake of the sliced brain in vitro, and the glucose metabolism of the brain slices. Respiration, carbon dioxide production in the presence of bicarbonate, consumption of glucose, and production of lactate and pyruvate were measured. The incubations were made in physiological saline with different concentrations of glucose (2.5 ml, 5 mM, and 10 mM).The mean difference of the water contents in vivo was 1.9%. The water uptake in vitro did not depend on the glucose concentration.The oedematous tissue had a slightly stimulated respiration, and its aerobic glycolysis was markedly increased.The respiration was independent of the glucose concentration. The lactate: pyruvate quotient was somewhat higher on the oedematous side, and it became greater on both sides with increasing glucose concentration.The glycolysis increased considerably between 2.5 mM and 5 mM glucose, especially in the oedematous brain. Between 5 mM and 10 mM glucose the aerobic glycolysis was accelerated only slightly by increasing glucose concentration.We think that the increased glycolysis is an expression of a preceding energy deficit. By a simple enhancement of the glucose concentration to more than 100 mg/100 ml the metabolism is not stimulated further, and the water content is not diminished.Supported by the Deutsche Forschungsgemeinschaft: H. 573/3.Part of the Dissertation, Med. Fakultät Homburg (Saar), 1972.     

Changes in the components and content of biological water in the brain of experimental hydrocephalic rabbits

Changes in biological water components and their respective content in the cortical gray matter and periventricular white matter were studied in rabbits rendered hydrocephalic by intracisternal kaolin injection. There was no change in either total water content or free or bound water content in the cortical gray matter at the various stages of hydrocephalus development. While there was no significant change in total water content in the periventricular white matter at any stage of hydrocephalus, free water content was significantly elevated and bound water content was decreased at the acute and subacute stages, with a return to relatively normal levels at the chronic stage. It is concluded that in the periventricular white matter, free water enters the brain across the ependymal lining during the acute and subacute stages of experimental hydrocephalus with a simultaneous reduction in the bound water and that there is some recovery at the chronic stage. It is suggested that alternative drainage pathways may develop in chronic hydrocephalus allowing drainage of free water in the periventricular white matter, which in turn permits bound water to return to relatively normal levels.     

Correlations between the apparent diffusion coefficient, water content, and ultrastructure after induction of vasogenic brain edema in cats

OBJECT: The authors examined the correlation between changes in the apparent diffusion coefficient, regional water content, and tissue ultrastructure after vasogenic brain edema. METHODS: Vasogenic edema was induced in the white matter of six cats by cortical cold lesioning. The trace of diffusion tensor (Trace[D]) obtained from magnetic resonance imaging to measure the orientationally averaged water diffusibility was compared with the corresponding tissue water content determined by gravimetric studies and with ultrastructural water localization. Edema fluid had spread to the subcortical and deep white matter by 4.5 hours postlesioning. The increase in Trace(D) showed a significant linear correlation with the increase in tissue water content, both in the subcortical and deep white matter as follows: y = 45.5x - 2367 (r = 0.94) and y = 37.0x - 1769 (r = 0.93), respectively, where x is the water content (gram water/gram tissue) and y the Trace(D) (x 10(-6) mm2/second). On histological examination, nerve fibers were found to be dissociated in the white matter and the extracellular space was markedly enlarged with protein-rich fluid. No noticeable hydropic swelling of the cellular components was observed. CONCLUSIONS: A linear correlation was observed between increases in Trace(D) and increases in extracellular water volume in in vivo vasogenic brain edema. A similar correlation between the subcortical and deep white matter showing different arrangements of nerve fibers (parallel compared with intermingled, respectively) indicated that measurement of Trace(D) is a suitable parameter for the evaluation of vasogenic brain edema.     

Cerebral distribution of immunoconjugate after treatment for neoplastic meningitis using an intrathecal radiolabeled monoclonal antibody

A detailed autopsy and autoradiographic study was performed after the death of a patient undergoing intrathecal, antibody-guided irradiation for carcinomatous meningitis. The results demonstrated tumor cells infiltrating the surface meninges and a severe astrocytic reaction associated with oedema in the periventricular and brain stem subpial white matter. This was not seen in cortical or other gray matter structures. Autoradiographic examination correlated well, demonstrating isotope within the oedematous areas of the white matter in addition to the expected concentration in the leptomeningeal layers. These findings are discussed in the context of antibody binding to tumor tissue and the possible benefits conferred in the treatment of infiltrating tumor cells.     

Experimental carbon dioxide laser brain lesions and intracranial dynamics: Part 2. Effect on brain water content and its response to acute therapy

Experimental brain lesions were created over the left parietooccipital cortex of the albino rabbit through the intact dura mater with high radiating carbon dioxide laser energy (40-W impact, 0.5-second duration, for a total time of 4 seconds on a 12.5-mm surface). The brain water content was studied 2, 6, and 24 hours after the insult. Another two groups of animals received acute therapy with either dexamethasone (1 mg/kg) or furosemide (1 mg/kg). In all groups, Evans blue extravasation uniformly extended from the impact crater into the surrounding white matter. The brain water content in the gray matter was elevated from the control value by 2 hours after impact (P less than 0.005) and remained elevated at 6 and 24 hours. The white matter brain water content did not increase until 6 hours after impact and remained elevated in the 24-hour group (P less than 0.005). After dexamethasone treatment, there was a significant decrease of water in the gray matter (P less than 0.01), but not in the white matter. With furosemide therapy, there was no reduction of gray or white matter brain water.     

Examination of the metabolism of oedematous brain tissue

Summary A circumscribed brain oedema was caused by the localized application of cold in the rabbit's brain. The metabolism of oedematous brain slices and analogous areas from the non-lesioned hemisphere were examined with the new 2-vessel method ofWarburg. In comparison with the uninjured brain, aerobic and anaerobic glycolyses were strongly increased in oedematous tissue, the respiration was slightly increased. These findings may be interpreted as the expression of an elimination of a previously existing energy deficit.The work was supported by D.F.G. (German Res. Community) H. 573/1.     

The effect of mannitol on cerebral white matter water content

The authors have studied the effect of a low-dose (0.28 gm/kg) bolus infusion of mannitol on brain water in man. In eight patients with severe head injury, small pieces of subcortical white matter were taken at craniotomy both before and after infusion of mannitol. The tissue specific gravity was measured using a graduated specific-gravity column, and from it the brain water content was calculated. White matter specific gravity rose from a mean (+/- standard error of the mean) of 1.0325 +/- 0.0012 before mannitol infusion to 1.0352 +/- 0.0011 after mannitol administration, and the brain water content fell from a mean of 80.94% +/- 2.5% to 75.28% +/- 2.3%. The differences were significant (p less than 0.01). This study shows that, after head injury in man, mannitol increases the white matter specific gravity and probably does so by reducing brain water.     

In vivo estimation of water content in cerebral white matter of brain tumour patients and normal individuals: Towards a quantitative brain oedema definition

Summary The building stones in MR imaging, namely the relaxation times (RT) correlate in certain circumstances well with water content of cerebral white matter (WM). It is, therefore, possible to estimate regional in vivo water content using RT imaging with an accuracy within ±2–4%. The purpose of the present prospective investigation was to study the RT's of cerebral WM and thereby estimate normal in vivo water content. Furthermore, the method was used to study whether water content was different in the apparently normal cerebral hemisphere of brain tumour patients compared to normal individuals. RT data of the non-tumour bearing hemisphere in 42 brain tumour patients were compared to 35 normal individuals. Results: 1) Water content was significantly higher in the posterior WM as compared to frontal WM in both normal individuals and apparently normal WM in brain tumour patients and this appears to be a true physiological difference. 2) Water content of normal WM correlated positively with age. 3) There was no significant difference between RT data from the normal hemisphere in histologically different types of brain tumours. 4) Water content was significantly higher in the apparently normal WM of brain tumour patients compared to normal individuals. 5) Based on the present results, a definite lower threshold of white matter oedema could be established as 75% water content, corresponding to a T1 of 0.850 seconds.     

The effect of craniectomy on the biomechanics of normal brain

Does an open skull alter the fundamental biomechanical properties of normal brain tissue? This question was studied in 32 anesthetized cats, 16 of which underwent a standard craniectomy (2.5 X 2.0 cm) in the left frontoparietal region. Brain tissue pressure, regional cerebral blood flow (rCBF), and brain water content were measured from the same area of cortical gray and white matter, and intracranial pressure (ICP) was recorded from the cisterna magna. Brain tissue resistance, tissue compliance, and the pressure-volume index were analyzed in response to a bolus injection of saline into brain tissue or the cisterna magna. Cerebrovascular resistance was also calculated. In craniectomized animals 2 hours after surgery, ICP had fallen to 3.75 +/- 0.39 mm Hg, and cortical gray and white matter tissue pressure had fallen to 3.19 +/- 0.47 and 4.69 +/- 0.54 mm Hg, respectively (mean +/- standard error of the mean); these variables did not fall further over 4 hours. The pressure-volume index in the same animals increased significantly from 0.67 +/- 0.01 to 0.86 +/- 0.04 ml. Tissue compliance rose in the cortical gray matter but tissue resistance fell, approximating that found in subjacent white matter. There was no significant difference between animals with and without craniectomy in rCBF, cerebrovascular resistance, or brain water content in either gray or white matter. These findings indicate that in the cat craniectomy causes an increase in the compensatory capacity of the intracranial cavity to increased volume. The data also indicate that cortical tissue has high hydraulic conductivity and compliance when the skull is opened.     

Study of brain edema by an infusion edema Model--the method and characteristics of the model

In this report, we have described the way of making the infusion edema model, physiological changes of various parameters during this procedure, distribution of water content in white and gray matter and the light and electron microscopic findings of this edema model, for the further understanding of vasogenic edema of the brain. To make the infusion edema model, 25-G needle was stereotaxically inserted into the left frontal white matter of the cat brain. Through the polyethylene catheter with three way stop cock, this catheter was connected to the pressure transducer and slow infusion pump. By this way, we can monitor the pressure of infusing fluid into the white matter. Normal saline was infused with initial rate of 0.75 microliter/min for the first 2 hours. The inflow rate was increased to 1.5 microliter/min for the next one hour, and then changed to 3.0 microliters/min for maintenance inflow rate. The total amount of infused volume was 0.5 ml in this study. During making the infusion edema model, blood pressure and PaCO2 changed little. Intracranial pressure slightly increased from 5.8 to 15.1 mmHg. Pressure volume index (PVI) changed from 0.74 to 0.64, suggesting the changes of intracranial compliance. The water content measured by specific gravimetric technique showed nearly the same water contents and distribution of edema fluid in the white matter of the cat as in the cryogenic injury model. Pathological findings of this infusion edema model demonstrated that the infused liquid was accumulated in the extracellular space of white matter without damaging the tight junction, and endothelial cells.(ABSTRACT TRUNCATED AT 250 WORDS)     

A study of electrical impedance of oedematous cerebral tissue during operations

Summary In 11 patients, most of whom had cerebral tumours, electrical impedance was recorded during the introduction of an impedance electrode during operation. After removal of the lesion impedance was measured in the tissue adjacent to the walls of the wound cavity. The readings were correlated with water and electrolyte content and with histological findings in the specimens taken from the same places. Electrical impedance appeared to be lowered in regions invaded by tumour and also in the surrounding oedematous areas of white matter. Unless primarily affected, cerebral cortex usually did not show oedema.The increase of water in the oedematous area was accompanied by a proportional increase of tissue Na⁺ and a decrease of tissue K⁺. The method of impedance measurement does not allow the delineation of a tumour from the surrounding oedema.