This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hasegawa, Y. Articles by Brayden, J. E. Search for Related Content PubMed PubMed Citation Articles by Hasegawa, Y. Articles by Brayden, J. E. Pubmed/NCBI databases Medline Plus Health Information Hypoglycemia

(Stroke. 1996;27:1648-1656.)
© 1996 American Heart Association, Inc.

Articles

Severe Transient Hypoglycemia Causes Reversible Change in the Apparent Diffusion Coefficient of Water

Yasuhiro Hasegawa, MD, PhD; James E. Formato, MS; Lawrence L. Latour, PhD; Jorge A. Gutierrez, MD; Kai-Feng Liu, MD; Julio H. Garcia, MD; Christopher H. Sotak, PhD Marc Fisher, MD

the Department of Neurology, Medical Center of Central Massachusetts–Memorial (Y.H., M.F.); Department of Biomedical Engineering, Worcester Polytechnic Institute (J.E.F., L.L.L., C.H.S.); and the Departments of Neurology (M.F.) and Radiology (C.H.S., M.F.), University of Massachusetts Medical School, Worcester, Mass; the Department of Neuropathology, Henry Ford Hospital, Detroit, Mich (J.A.G., K-F. L., J.H.G.); and the Cerebrovascular Division, Research Institute, National Cardiovascular Center, Osaka, Japan (Y.H.).

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References Introduction  References

 
Background and Purpose The aim of this study was to determine the effects of temporary severe hypoglycemia on the apparent diffusion coefficient (ADC) acquired by diffusion-weighted MRI of brain water with the use of serial multislice ADC mapping in rats. Severe hypoglycemia reduces the extracellular space volume, as does ischemia. Demonstrating a reduction of ADC with hypoglycemia should increase our understanding of the mechanisms underlying ADC changes in ischemia and other conditions.

Methods Fasted rats were given regular insulin (15 IU/kg IP). Rats were subjected to 15 minutes (n=5) and 50 minutes (n=5) of temporary severe hypoglycemia, causing a transiently isoelectric electroencephalogram (EEG). ADC mapping was performed every 30 seconds beginning at the onset of isoelectricity for 8.5 minutes. ADC maps were also obtained later during the isoelectric EEG period and 10, 20, 30, and 40 minutes after glucose infusion. Control images were obtained from a separate group of animals suffering cardiac arrest (n=5).

Results Abnormal ADC values were not observed before the onset of cerebral isoelectricity, except for isolated areas in the cortex and periventricular regions. Cortical ADC values globally declined at the onset of EEG isoelectricity. The ADC decline spread to subcortical regions within a few minutes. During the isoelectric period, significant declines of ADC values (27% to 45%) occurred in the entire brain. Glucose infusion normalized most of the ADC changes, even after a 50-minute period of isoelectricity.

Conclusions ADC mapping during hypoglycemia clearly demonstrates changes likely related to energy depletion. Most of these ADC declines were reversible. Hypoglycemia is a condition known to be associated with shrinkage of the extracellular space. These observations support the hypothesis that ADC reductions observed in ischemia are also related to shifts of water from the extracellular to the intracellular compartment.

Key Words: hypoglycemia • magnetic resonance imaging • rats

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References Introduction  References

 
The ADC of brain water significantly decreases within minutes after the onset of focal brain ischemia.^{1 2 3 4 5 6} The precise mechanism for this reduction of ADC values with focal brain ischemia remains uncertain; however, it has been attributed to intracellular edema and membrane permeability changes, caused by brain energy failure.^{5 7} Complete or partial recovery of such ADC reductions after blood flow restoration occurs in temporary focal ischemia models, suggesting that early ADC reductions after focal ischemia in part represent reversible ischemic tissue injury.^{4 8 9 10} Evaluation of ischemic damage with diffusion-weighted MRI during the potentially reversible phase of ischemia might help to develop new strategies for the treatment of ischemic stroke patients. ADC reductions similar to those seen after ischemia occur in status epilepticus,¹¹ spreading depression,^{12 13 14} and excitotoxic brain injury,^{15 16} pathological conditions characterized by a significant shrinkage of the ECS volume.^{16 17 18} However, status epilepticus and spreading depression in normal brain are conditions that are not associated with energy failure, and the ADC changes are reversible.

Glucose deprivation leads to severe brain energy failure and a reduction of cell membrane ionic pump activity, as does anoxia/ischemia, but the topographical and temporal evolution of hypoglycemic brain damage are different from anoxia/ischemia. A positive relationship between the density of neuronal necrosis and the duration of EEG isoelectricity is well documented in experiments of insulin-induced hypoglycemia.^{19 20 21 22 23 24 25} A drastic reduction of ECS volume to approximately 50% of normal (ie, the shift of water from the ECS to the ICS due to membrane ionic pump failure) occurs at the onset of the isoelectric EEG phase with insulin-induced hypoglycemia.²⁶ In this study we evaluated topographical and temporal changes in the ADC of brain water with the rat model of insulin-induced hypoglycemia to answer the following questions: (1) Do generalized cerebral ADC reductions coincide with cerebral isoelectricity? (2) Are the changes in ADC values reversible by restoring the glucose level after various durations of EEG isoelectricity? and (3) Would ADC measurements after recovery from hypoglycemia predict the extent of neuronal damage? To answer these questions, multislice ADC mapping was serially performed with the aid of echo-planar diffusion-weighted MRI in animals subjected to transient, severe hypoglycemia with concomitant continuous EEG monitoring.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References Introduction  References

 
Surgical Procedures and Experimental Groups
This study was approved by the Animal Research Committee of the University of Massachusetts Medical School (Animal Research Committee protocol A-818). Male Sprague-Dawley rats, weighing 300 to 360 g each, were fasted overnight before the experiment but were allowed access to water. Under chloral hydrate anesthesia (400 mg/kg IP), the left femoral artery and vein were cannulated with PE-50 polyethylene tubing. The tip of the venous catheter was located in the inferior vena cava. Experiments were performed only when the baseline physiological parameters remained within the normal range. The normal ranges of mean arterial blood pressure were set at 80 to 130 mm Hg; pH, 7.25 to 7.45; PaCO₂, 35 to 50 mm Hg; and PaO₂, 80 to 150 mm Hg. Animals were then divided into two experimental groups.

Control/Cardiac Arrest Group
Five animals served as controls, and brain ADC values after cardiac arrest were measured for comparison with those in the hypoglycemic animals. After baseline physiological measurements, the head of the animal was fixed inside the ¹H "birdcage" coil with a tooth bar and ear bars. Neonatal electrocardiographic electrodes were attached to the left forelimb and both hindlimbs for electrocardiographic monitoring. Body temperature was controlled at 37±0.5°C with the use of a rectal probe and a feedback-regulated heated air flow system that completely surrounded the animal. Baseline multislice ADC mapping was performed before cardiac arrest. The animals were then killed in the imaging magnet by a rapid intravenous injection of 0.4 mL of 4 mol/L potassium chloride via the femoral catheter, and ADC mapping was performed 15 and 45 minutes after cardiac arrest. We documented brain temperature changes in a separate group of three rats after inducing cardiac arrest under the same experimental conditions without MRI. The frontoparietal cranium was exposed by a midsagittal incision, and two burr holes, 1.3 mm in diameter, were made in the right parietal skull at 0.5 mm caudal, 3.5 mm and 5.0 mm lateral to the bregma. Small copper-constantan thermocouples, 0.003 inch in diameter (type IT-23, Physitemp Instruments), were inserted into the right caudoputamen and the frontoparietal cortex and fixed to the frontoparietal bone with dental cement. The temperatures in these superficial and deep brain structures were continuously monitored for 45 minutes after cardiac arrest with the use of a scanning thermocouple thermometer with 0.1°C resolution (Cole-Parmer Instrument Co).

Hypoglycemia Groups
After arterial blood sampling for baseline glucose measurements, animals were given 15 IU/kg of regular insulin (NOVO Industries) intraperitoneally. Animals were then tracheostomized and intubated with PE-240 tubing. The parietal cranium was exposed by a midcoronal incision, and two burr holes, 1.2 mm in diameter, were made in the right and left parietal regions at 1.5 mm rostral and 2.5 mm lateral to the lambda. Ag-AgCl ball electrodes were placed in the burr holes and fixed with dental cement. The reference electrode was placed on the ear lobe, and the nasal base was grounded by an Ag-AgCl plate electrode. The head of the animal was then fixed inside a ¹H birdcage imaging coil with tooth bar and ear bars. Animals were immobilized with tubocurarine chloride (0.5 mg/kg IV) and mechanically ventilated at a rate adjusted to maintain PaO₂ of 90 mm Hg or greater and PaCO₂ of 35 to 45 mm Hg. To maintain the anesthetic state, animals inhaled 1% isoflurane throughout the experiment. The arterial pH was maintained between 7.35 and 7.45 with sodium bicarbonate.

The EEG was continuously monitored and graded as follows: stage I, normal; stage II, slowing to the theta to delta ranges; stage III, rhythmic high-amplitude delta slowing; stage IV, more slowing with a suppression phase; and stage V, isoelectricity.²⁵ We began ADC mapping 1.5 hours after the insulin injection. Then we waited for the development of EEG stages IV and V. At the onset of the stage IV EEG pattern, we administered atropine sulfate (0.1 mg IV) to prevent cardiac arrhythmia. When the suppression phase became prolonged to 2.0 seconds during stage IV, whole brain ADC mapping was performed every 30 seconds for 8.5 minutes to detect ADC changes at the onset of the isoelectricity. The blood glucose concentration was restored at 15 minutes (Iso₁₅ group; n=5) or 50 minutes (Iso₅₀ group; n=5) after the onset of EEG isoelectricity with an infusion of 0.4 mL of 50% glucose, given manually over 1 minute. Twenty-five minutes after the glucose injection, 0.3 mL of the 50% glucose solution was also given to avoid a delayed decline of the blood glucose level.

ADC maps were obtained 10 minutes after the onset of isoelectricity in the Iso₁₅ group and 15 and 45 minutes after the onset of isoelectricity in the Iso₅₀ group. During the recovery period, follow-up ADC maps were also obtained 10, 20, 30, and 40 minutes after the initial glucose infusion. When a blood pressure increase occurred from the resting mean value of 90 to 120 mm Hg, the mean blood pressure was held between 100 and 130 mm Hg by arterial bleeding. Blood was collected from the central venous catheter into a heparinized syringe and was immediately reinfused if the pressure fell. After the last ADC mapping during the recovery phase was accomplished, transcardiac perfusion fixation of the brain was performed with 4% paraformaldehyde. After immersion fixation of the decapitated head in 4% paraformaldehyde for 24 hours, the brain was removed. Brain specimens were sectioned coronally at 3-mm intervals before tissues were embedded in paraffin. Histology sections were cut in 5-µm slices and stained with hematoxylin-eosin and GFAP. Three additional rats without insulin administration were used as controls for histological analysis. After being anesthetized with chloral hydrate and isoflurane, their brains were harvested according to the same method used for the hypoglycemic animals. Neuronal injury, GFAP reactivity, and infiltration of polymorphonuclear leukocytes in each specimen were blindly assessed and scored by two of the authors (J.A.G. and K-F.L.) who did not know either the duration of hypoglycemia or the findings of ADC maps. The histopathological changes for each category were recorded according to a numerical grading scale (grade 0, normal/absent; 1, slight; 2, moderate; and 3, severe).

The body temperature was maintained at 37±0.5°C by a heated air flow system throughout the experiment. Arterial blood pressure was monitored continuously (Parametron 7150 Monitor, Roche Medical Electronics). Arterial blood gas analysis was performed intermittently with a blood gas analyzer (Corning 170 pH blood gas analyzer, Corning). Blood glucose was measured by the glucose oxidase method (Sigma Chemical Co) before injection of insulin (baseline), during EEG stage V, and 15 minutes and 35 minutes after the initial glucose infusion.

MRI Procedure
MR experiments were performed with the use of a General Electric CSI-II 2.0 T/45 cm imaging spectrometer operating 85.56 MHz for ¹H and equipped with ±20 G/cm self-shielding gradients. The multislice ADC mapping^{13 27} was performed on eight contiguous axial slices, encompassing the entire rat brain. An eight-slice spin-echo sequence²⁸ was used with echo-planar acquisition.²⁹ Each slice was 1.5 mm thick with a 25.6x25.6-mm field of view and 64x64 pixel resolution. Slice selection and diffusion gradients were applied along the z axis, ie, the anterior-posterior axis of the rat brain. The echo-planar acquisition time for each diffusion-weighted image was 65.54 milliseconds. To create ADC maps, diffusion sensitizing gradients were increased in amplitude in three different ways. In the control/cardiac arrest group, three ADC maps were acquired: while alive and at 15 minutes and 45 minutes after cardiac arrest. ADC maps were constructed with 20 different b values, ranging from 15 to 1700 s/mm². In the Iso₁₅ group, six ADC maps were acquired at 1.5 hours after insulin injection, 10 minutes after isoelectricity, and at 10, 20, 30, and 40 minutes after the initial glucose infusion. In the Iso₅₀ group, seven ADC maps were acquired at 1.5 hours after insulin injection, 15 minutes and 45 minutes after isoelectricity, and at 10, 20, 30, and 40 minutes after the initial glucose infusion. Because of the large amount of data, ADC maps in these hypoglycemia groups were constructed with 10 different b values, ranging from 62 to 1900 s/mm². When an increased time resolution was required to monitor sequential ADC changes during the transition to isoelectricity, only six b values were collected per map, ranging from 43 to 1500 s/mm².

The raw data were transferred to a Silicon Graphics Indigo R4000 workstation (Silicon Graphics) for off-line data processing. As determined by a linear regression algorithm, analysis of the natural logarithm of the signal intensity with respect to the series of b values yielded the ADC value on a pixel-by-pixel basis.^{27 30} There is a tradeoff between the time resolution gained and error in the fit when the number of b values is decreased. We empirically determined that six b values and a 30-second time resolution sufficiently met our requirements for time resolution and confidence in the quality of the fit. ADC values below 0.55x10^-3 mm²/s were considered abnormal diffusion coefficients based on prior studies in ischemic animals.^{27 31} Regions with abnormal diffusion coefficients were observed over time, and their volume was calculated at each time point.

ROI Analysis
ROI analysis was performed with a KBV image analysis program (Amerinex Artificial Intelligence) and the Silicon Graphics computer workstation. The ROI size was 3x3 pixels (1.2x1.2 mm). According to the stereotaxic coordinates,³² the following seven cerebral structures with a 5x5-pixel area on the imaging slices were selected: motor area in the frontoparietal cortex, somatosensory area in the frontoparietal cortex, caudoputamen, thalamus, lateral hypothalamic area, amygdaloid area, and hippocampus. We took another three ROIs in the dorsal lateral and medial brain stem, which roughly corresponded to superior colliculus, medial geniculate body, and substantia nigra (Fig 1).

View larger version (30K): [in this window] [in a new window]   Figure 1. Regions of interest. Ten regions of interest, 3x3 pixels in size, were taken from ADC map images: (1) motor area in the frontoparietal cortex (FrPaM), (2) somatosensory area in the frontoparietal cortex (FrPaSS), (3) caudoputamen (CPu), (4) thalamus (TH), (5) lateral hypothalamic area (LH), (6) amygdaloid area (AM), (7) hippocampus (Hi), (8) superior colliculus (SC), (9) medial geniculate body (MGB), and (10) substantia nigra (SN).

Statistical Analysis
Values presented in this study are mean±SD. For the relationship between two parametric variables, linear regression analysis was used. Student's t test was applied to test the statistical significance between ROI values in hypoglycemic animals and those in control animals. A statistically significant difference of mean ADC values over time or among different animal groups was evaluated by ANOVA and post hoc testing with the Bonferroni correction. A value of P<.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References Introduction  References

 
Control/Cardiac Arrest Group
The ADC maps in the control animals before cardiac arrest exhibited no abnormal ADC values, defined as less than the threshold value, 0.55x10^-3 mm²/s.²⁷ The baseline ADC values in each of the ROIs are shown in Table 1. There were statistically significant differences among the 10 ROIs because of the anisotropic nature of brain tissue. Fifteen minutes after cardiac arrest, ADC values were significantly decreased in every ROI with a 25.0% to 41.9% decline of mean values compared with baseline before cardiac arrest (P<.01, ANOVA with Bonferroni corrected t test). No further decline of ADC values was observed 45 minutes after the cardiac arrest. The direct brain temperature measurement demonstrated a gradual decline after cardiac arrest. However, the maximum declines in the parietal cortex and caudoputamen were only 0.4°C and 0.3°C at 45 minutes after cardiac arrest, respectively.

View this table: [in this window] [in a new window]   Table 1. Mean ADC Values in Each ROI

Hypoglycemia Groups
Physiological parameters are shown in Table 2. Mean arterial blood pressure, pH, PaCO₂, and PaO₂ were strictly controlled within a narrow range, and there were no statistically significant differences among the measured time points. Glucose levels significantly decreased during the stage V EEG phase. After the glucose infusion, blood glucose levels normalized in both the Iso₁₅ and Iso₅₀ groups. Although the blood glucose levels after glucose infusion in the latter group were slightly higher than those in the former group, these differences were not statistically significant.

View this table: [in this window] [in a new window]   Table 2. Physiological Parameters

EEG recording and ADC mapping were initiated 1.5 hours after insulin administration. Initial EEG findings in all animals were categorized as stage II or III, ie, moderate slowing or high-amplitude delta slowing. At this time point, ADC maps revealed small areas with abnormally decreased ADC values localized in the cortex and periventricular regions, mainly in the caudoputamen. The cortical abnormalities were multiple, and the location varied among animals. The periventricular lesions were localized in slices in which the largest sections of caudoputamen and thalamus appeared, ie, Vth and VIth slices in Figs 2 and 3. Consequently, mean ADC ROI values at this EEG stage were slightly lower than control values in the motor area in the frontoparietal cortex, caudoputamen, thalamus, and lateral hypothalamic area (Table 1). The mean volume of abnormal ADC values at this time point were 56±18.3 mm³ in the Iso₁₅ group and 60.6±18.7 mm³ in the Iso₅₀ group (Fig 4). Six of 10 hypoglycemic animals showed bilaterally symmetrical and synchronous EEG changes. In four animals (40%), however, isoelectricity did not occur simultaneously in the two hemispheres. ADC changes during the transition from stage IV to stage V were obtained by frequent mapping every 30 seconds. The ADC maps during EEG stage IV exhibited findings similar to those on the initial mapping described above. Then a marked ADC decline occurred coincident with the onset of EEG isoelectricity. Fig 5 shows the ADC maps and corresponding EEG traces in a rat with an asynchronous EEG change. Once the EEG activity ceased, definite ADC declines involving almost the whole brain were observed (Figs 2 and 3). The volume of brain with abnormally decreased ADC values 10 minutes after EEG cessation in the Iso₁₅ group was 498.9±225.9 mm³. The abnormal ADC region volumes in the Iso₅₀ group 15 and 45 minutes after the EEG cessation were 688.3±83.1 and 737.9±152.0 mm³, respectively (Fig 4). The tissue volume with abnormally reduced ADC values significantly increased in relation to the duration of EEG isoelectricity (P<.002, ANOVA with post hoc Bonferroni corrected t test). ROI values at 10 to 15 minutes after isoelectricity were significantly decreased in every ROI compared with those at the stage II or III EEG phase (Table 1). At 45 minutes after isoelectricity, further declines were found in lateral hypothalamic area, amygdaloid area, superior colliculus, medial geniculate body, and substantia nigra (P<.05, Student's t test), and the same magnitudes of ADC decline were observed in all brain structures as those in the animals with cardiac arrest (P=NS versus ADC values 45 minutes after the cardiac arrest in each ROI).

View larger version (51K): [in this window] [in a new window]   Figure 2. Serial multislice ADC maps in the Iso15 group. Eight contiguous slices of ADC maps, numbered from I to VIII, at six study time points are shown. A gray scale shows the absolute ADC values (x10-3 mm2/s). The regions with abnormal ADC pixel values, ie, below 0.55x10-3 mm2/s, are displayed as a white area. ADC maps during EEG slowing (stage II, III) exhibited small lesions in the cortices and in the periventricular regions. Ten minutes after the onset of EEG isoelectricity (stage V 10'), abnormal ADC declines were observed over most of the brain. After glucose infusion, these ADC declines normalized. During the recovery period (post i.v. 10', 20', 30', and 40'), residual lesions localized in the periventricular regions are seen.

View larger version (59K): [in this window] [in a new window]   Figure 3. Serial multislice ADC maps in the Iso50 group. The gray scale and display method are the same as in Fig 2.

View larger version (14K): [in this window] [in a new window]   Figure 4. Serial change in lesion volume with abnormally decreased ADC values in rats with transient severe hypoglycemia. *P<.01, P<.002 (ANOVA with post hoc Bonferroni corrected t test).

View larger version (32K): [in this window] [in a new window]   Figure 5. Serial ADC maps during the transition between stage IV and V EEG change. ADC maps with EEG tracings in both hemispheres are shown every 30 seconds. EEG scale shows 1 second and 0.3 mV. Arrow indicates the position of the EEG electrode. Regions of reduced ADC are dark.

Ten minutes after the glucose infusion, ADC values in all brain regions returned to normal except in the periventricular regions (Figs 2 and 3). The mean volumes of abnormal ADC values at 40 minutes after the glucose infusion were 26.7±15.3 mm³ in the Iso₁₅ group and 32.6±17.2 mm³ in the Iso₅₀ group. These values were significantly smaller than the initial lesion volumes obtained during EEG stage II or III (P<.001, Student's t test). There were no significant differences in the residual lesion volume between the Iso₁₅ and Iso₅₀ groups during the 40-minute recovery period (Fig 4). In two animals from the Iso₁₅ group, low-voltage slow EEG activity was observed after glucose infusion. However, none of the animals in the Iso₅₀ group showed EEG recovery.

Neuronal necrosis was obvious in the medial half of the CA1 sector of the hippocampus and the crest of the dentate gyrus in the Iso₁₅ group. In the Iso₅₀ group, neuronal necrosis was observed in the entire CA1 sector of the hippocampus, a widespread area of the dentate gyrus, striatum, and superficial layers of the neocortex. Histopathological damage scores are shown in Table 3. The scores of mean neuronal injury, GFAP reactivity, and polymorphonuclear leukocyte infiltration in the Iso₅₀ group were significantly higher than those in the Iso₁₅ and control groups (P<.05, ANOVA with Bonferroni corrected t test). The Iso₁₅ group was not significantly different from the controls.

View this table: [in this window] [in a new window]   Table 3. Histopathological Score

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References Introduction  References

 
The topographical and temporal changes in the ADC of rat brains with temporary severe hypoglycemia can be summarized as follows: (1) Widespread ADC reductions occur when the EEG becomes isoelectric; (2) global ADC reductions rapidly normalize with glucose infusion, except in the periventricular regions; and (3) the residual regions of ADC decline are not good predictors of the pathologically determined sites of injury. Tissue energy stores are known to be progressively depleted during hypoglycemia, but marked depletion occurs only when EEG activity ceases.^{33 34} Thus, the sudden ADC reductions observed over the whole brain after the onset of EEG isoelectricity are likely related to marked brain energy depletion. These ADC declines seen in our hypoglycemic animals are unrelated to incidental brain ischemia, because cerebral blood flow increases during hypoglycemia.³⁵ The increase in blood flow might influence ADC values through a relative increase in local blood oxygenation. However, a higher than "normal" ADC would then be expected because of the decline in static field inhomogenieties caused by the relatively smaller amount of paramagnetic deoxygenated hemoglobin. Therefore, the ADC decline observed in the hypoglycemic brain is probably not an artifact caused by internal gradient effects. The density of neuronal necrosis correlates with the number of minutes of cerebral isoelectricity.²⁵ Ten minutes of cerebral isoelectricity produces only very minimal histological damage. Sixty minutes of isoelectricity causes widespread neuronal necrosis in the cerebral cortex, hippocampus, caudate nucleus, and spinal cord. We chose 15-minute and 50-minute periods of temporary hypoglycemia with EEG isoelectricity, and our qualitative histopathological analysis agrees with these earlier observations. However, there were neither qualitative nor quantitative differences in the reversal of decreased ADC values between the Iso₁₅ and Iso₅₀ groups. Although the mean residual lesion volume on ADC mapping in the Iso₅₀ group was larger than that in the Iso₁₅ group, the difference was not statistically significant. The observed discrepancy between the final ADC maps and the histological analysis may be partially due to the resolution of the ADC map (400 µm in the imaging plane), which was too insensitive to evaluate small regions, such as the CA1 sector or the cortical layers, because of partial volume effects. This discrepancy indicates that ADC measurements made shortly after recovery from hypoglycemia might not predict the precise resulting neuronal damage. This contrasts with the situation in temporary focal ischemia models, in which there is good correlation between the location of the ischemic lesion on ADC maps and postmortem.^{4 8 9}

In contrast to the generalized ADC decline during isoelectricity, the minimal changes that appeared on the ADC maps before cessation of EEG activity may not be caused by widespread brain energy depletion but by localized imbalances between energy supply and demand, spreading depression,³⁶ or an excitotoxic mechanism.²⁵ One or two periods of transient impedance increase with ionic change before isoelectricity occurs in the insulin-induced hypoglycemia model, and such a phenomenon is assumed to represent spreading depression.^{26 36} It is likely that the early ADC changes observed in this experiment represent effects of spreading depression. The asymmetry of the ADC changes and isoelectricity in 40% of the animals may be related to an asymmetrical onset of spreading depression in severe hypoglycemia; and this asymmetry also occurred in a histopathological study.²⁰ A specific location of brain lesions after hypoglycemic injury, ie, periventricular regions, was documented by histopathological analyses; involvement by excitotoxins, such as aspartate and glutamate, in the generation of hypoglycemic neuronal damage at selected sites was suggested.^{25 37} Supporting evidence for this hypothesis stems from pharmacological experiments^{37 38 39} and a study of glutamatergic corticostriatal projections.⁴⁰ A significant ADC drop occurs in the caudoputamen after the local injection of N-methyl-D-aspartate.^{15 16} Further studies with ADC mapping are necessary to identify the mechanism of these early ADC changes in the preisoelectric phase.

The local ADC values during isoelectricity are similar to those observed after cardiac arrest. Brain temperature decreased by 0.3°C to 0.4°C after the cardiac arrest, and normal brain ADC changed 0.013x10^-3 mm²/s per 1°C change of brain temperature.³¹ The observed temperature change after cardiac arrest could not explain our results. We reported a similar magnitude of ADC decline in spreading depression, as observed with hypoglycemia.^{13 14} In all these different pathological states, ie, global ischemia, hypoglycemia, and spreading depression, an abrupt reduction of the ECS volume up to 50% of normal has been documented,^{19 20 27} although the mechanisms inducing this ECS volume change differ. Similar ADC declines occur in the cortex during biccuculine-induced status epilepticus and in the striatum of neonatal rats injected with N-methyl-D-aspartate, other conditions in which the ECS volume decreases.^{11 16} Temporal changes in ADC values and ECS volume were observed in a global ischemia reperfusion model.⁴¹ An analysis of the trace of the diffusion tensor of the ischemic cat brain showed that the changes in water diffusion during acute stroke are directly related to changes in the relative volume of the ICS and ECS.⁴² Although the exact mechanisms for the ADC declines in various conditions are not completely understood, our findings support the concept that ADC measurements likely detect water shifts between ECS and ICS in a variety of conditions, including hypoglycemia, that are caused by a variety of mechanisms. It is therefore likely that ADC declines are a marker of ECS shrinkage that is nonspecific in its origin.

	Selected Abbreviations and Acronyms

 

ADC = apparent diffusion coefficient ECS = extracellular space EEG = electroencephalogram GFAP = glial fibrillary acidic protein ICS = intracellular space Iso15 group = group in which blood glucose concentration was restored at 15 minutes Iso50 group = group in which blood glucose concentration was restored at 50 minutes ROI = region of interest

	Acknowledgments

 
This study was supported in part by the Harrington Neurological Research Fund. The authors thank Dr Bernard J. Dardzinski and Dr Kentaro Takano for valuable discussion and support and Vanessa Brown for assistance. For the glucose measurements, we thank Dr Sunil Mukhopadhyay and Dr M. Sawkut Anwer of Tufts School of Veterinary Medicine.

	Footnotes

 
Reprint requests to Marc Fisher, MD, Department of Neurology, Medical Center of Central Massachusetts–Memorial, 119 Belmont St, Worcester, MA 01605-2982.

Received February 27, 1996; revision received April 22, 1996; accepted June 4, 1996.

	References

Top Abstract Introduction Materials and Methods Results Discussion References Introduction  References

 

1. Moseley ME, Kucharczyk J, Kurhanewicz J, Norman D. Diffusion-weighted MR imaging of acute stroke: correlation with T2-weighted and magnetic susceptibility-enhanced MR imaging in cats. AJNR Am J Neuroradiol.. 1990;11:423-429.[Abstract]
2. Mintorovitch J, Moseley ME, Chileuitt L, Shimizu H, Cohen Y, Weinstein PR. Comparison of diffusion- and T2-weighted MRI for the early detection of cerebral ischemia and reperfusion in rats. Magn Reson Med.. 1991;18:39-50.[Medline] [Order article via Infotrieve]
3. Minematsu K, Li L, Fisher M, Sotak CH, Davis MA, Fiandaca MS. Diffusion weighted magnetic resonance imaging: rapid and quantitative detection of focal brain ischemia. Neurology.. 1992;42:235-240.[Abstract/Free Full Text]
4. Minematsu K, Li L, Sotak CH, Davis MA, Fisher M. Reversible focal ischemic injury demonstrated by diffusion-weighted magnetic resonance imaging. Stroke.. 1992;23:1304-1311.[Abstract/Free Full Text]
5. Benveniste H, Hedlund LW, Johnson GA. Mechanism of detection of acute cerebral ischemia in rats by diffusion-weighted magnetic resonance microscopy. Stroke.. 1992;23:746-754.[Abstract/Free Full Text]
6. Davis D, Ulatowski J, Eleff S, Izuta M, Mori S, Shungu D, van Zijl DCM. Rapid monitoring of changes in water diffusion coefficients during reversible ischemia in cat and rat brain. Magn Reson Med.. 1994;31:454-460.[Medline] [Order article via Infotrieve]
7. Moseley ME, Cohen Y, Mintorovitch J, Chileuitt L, Shimizu H, Kucharczyk J, Kurhanewicz J, Wendland MF, Weinstein PR. Early detection of regional cerebral ischemia in cats: comparison of diffusion- and T2-weighted MRI and spectroscopy. Magn Reson Med.. 1990;14:330-346.[Medline] [Order article via Infotrieve]
8. Minematsu K, Fisher M, Li L, Sotak CH. Diffusion and perfusion MRI studies to evaluate a non-competitive NMDA antagonist and reperfusion in experimental stroke. Stroke.. 1993;24:2074-2081.[Abstract/Free Full Text]
9. Hasegawa Y, Fisher M, Latour LL, Dardzinski BJ, Sotak CH. MRI diffusion mapping of reversible and irreversible ischemic injury in focal brain ischemia. Neurology.. 1994;44:1484-1490.[Abstract/Free Full Text]
10. Hoehn-Berlage M, Norris DG, Kohno K, Mies G, Leibfritz D, Hossmann K-A. Evolution of regional changes in apparent diffusion coefficient during focal ischemia of rat brain: the relationship of quantitative diffusion NMR imaging to reduction in cerebral blood flow and metabolic disturbances. J Cereb Blood Flow Metab.. 1995;15:1002-1011.[Medline] [Order article via Infotrieve]
11. Zhong J, Petroff OAC, Prichard JW, Gore JC. Changes in water diffusion and relaxation properties of rat cerebrum during status epilepticus. Magn Reson Med.. 1993;30:241-246.[Medline] [Order article via Infotrieve]
12. Gyngell ML, Back T, Hoehn-Berlage M, Kohno K, Hossmann KA. Transient cell depolarization after permanent middle cerebral artery occlusion: an observation by diffusion-weighted MRI and localized 1H-MRS. Magn Reson Med.. 1994;31:337-341.[Medline] [Order article via Infotrieve]
13. Latour LL, Hasegawa Y, Formato JE, Fisher M, Sotak CH. Spreading waves of decreased diffusion coefficient after cortical stimulation in the rat brain. Magn Reson Med.. 1994;32:189-198.[Medline] [Order article via Infotrieve]
14. Hasegawa Y, Latour LL, Formato JE, Sotak CH, Fisher M. Spreading waves of a reduced diffusion coefficient of water in normal and ischemic rat brain. J Cereb Blood Flow Metab.. 1995;15:179-187.[Medline] [Order article via Infotrieve]
15. Verheul HB, Balazs R, Berkelbach van der Sprenkel JW, Tulleken CAF, Nicolay K, Tamminga KS, van Lookeren Campagne M. Comparison of diffusion-weighted MRI with changes in cell volume in a rat model of brain injury. NMR Biomed.. 1994;7:96-100.[Medline] [Order article via Infotrieve]
16. van Lookeren Campagne M, Verheul JB, Nicoley K, Balazs R. Early evolution and recovery from excitotoxic injury in the neonatal rat brain: a study combining magnetic resonance imaging, electrical impedance, and histology. J Cereb Blood Flow Metab.. 1994;14:1011-1023.[Medline] [Order article via Infotrieve]
17. Hansen AJ, Olsen CE. Brain extracellular space during spreading depression and ischemia. Acta Physiol Scand.. 1980;108:355-365.[Medline] [Order article via Infotrieve]
18. van Harreveld A, Khattab FI. Changes in cortical extracellular space during spreading depression investigated with electron microscope. J Neurophysiol.. 1967;30:911-929.[Free Full Text]
19. Agardh C-D, Kalimo H, Olsson Y, Siesjo BK. Hypoglycemic brain injury, I: metabolic and light microscopic findings in rat cerebral cortex during profound insulin-induced hypoglycemia and in the recovery period following glucose administration. Acta Neuropathol (Berl).. 1980;50:31-41.[Medline] [Order article via Infotrieve]
20. Auer RN, Wieloch T, Olsson Y, Siesjo BK. The distribution of hypoglycemic brain damage. Acta Neuropathol (Berl).. 1984;64:177-191.[Medline] [Order article via Infotrieve]
21. Auer RN, Kalimo H, Olsson Y, Siesjo BK. The temporal evolution of hypoglycemic brain damage, I: light- and electron-microscopic findings in the rat cerebral cortex. Acta Neuropathol (Berl).. 1985;67:13-24.[Medline] [Order article via Infotrieve]
22. Auer RN, Kalimo H, Olsson Y, Siesjo BK. The temporal evolution of hypoglycemic brain damage, II: light and electron microscopic findings in hippocampal gyrus and subiculum of the rat. Acta Neuropathol (Berl).. 1985;67:25-36.[Medline] [Order article via Infotrieve]
23. Kalimo H, Agardh C-D, Olsson Y, Siesjo BK. Hypoglycemic brain injury, II: electron-microscopic findings in rat cerebral cortical neurons during profound insulin-induced hypoglycemia and in the recovery period following glucose administration. Acta Neuropathol (Berl).. 1980;50:43-52.[Medline] [Order article via Infotrieve]
24. Kalimo H, Auer RN, Siesjo BK. The temporal evolution of hypoglycemic brain damage, III: light and electron microscopic findings in the rat caudoputamen. Acta Neuropathol (Berl).. 1985;67:37-50.[Medline] [Order article via Infotrieve]
25. Auer RN, Olsson Y, Siesjo BK. Hypoglycemic brain injury in the rat: correlation of density of brain damage with the EEG isoelectric time: a quantitative study. Diabetes.. 1984;33:1090-1098.[Abstract]
26. Pelligrino D, Almquist LO, Siesjo BK. Effects of insulin-induced hypoglycemia on intracellular pH and impedance in the cerebral cortex of the rat. Brain Res.. 1981;221:129-147.[Medline] [Order article via Infotrieve]
27. Dardzinski BJ, Sotak CH, Fisher M, Hasegawa Y, Li L, Minematsu K. Apparent diffusion coefficient mapping of experimental focal cerebral ischemia using diffusion-weighted echo-planar imaging. Magn Reson Med.. 1993;30:318-325.[Medline] [Order article via Infotrieve]
28. Stejskal EO, Tanner JE. Spin diffusion measurements: spin echoes in the presence of a time-dependent field gradient. J Chem Phys.. 1965;42:288-292.
29. Mansfield P. Multi-planar image formation using NMR spin-echoes. J Phys C.. 1977;10:L55-L58.
30. Le Bihan D. Molecular diffusion nuclear magnetic resonance imaging. Magn Reson Q.. 1991;7:1-30.[Medline] [Order article via Infotrieve]
31. Hasegawa Y, Latour LL, Sotak CH, Dardzinski BJ, Fisher M. Temperature dependent change of the apparent diffusion coefficient of water in normal and ischemic brain of rat. J Cereb Blood Flow Metab.. 1994;14:383-390.[Medline] [Order article via Infotrieve]
32. Paxinos G, Watson C, eds. The Rat Brain in Stereotaxic Coordinates. New York, NY: Academic Press; 1982.
33. Lewis LD, Ljunggren B, Ratcheson RA, Siesjo BK. Cerebral energy state in insulin-induced hypoglycemia, related to blood glucose and to EEG. J Neurochem.. 1974;23:673-679.[Medline] [Order article via Infotrieve]
34. Ratcheson RA, Blank AC, Ferrendelli JA. Regionally selective metabolic effects of hypoglycemia in brain. J Neurochem.. 1981;36:1952-1958.[Medline] [Order article via Infotrieve]
35. Abdul-Rahman A, Agardth C-D, Siesjo BK. Local cerebral blood flow in the rat during severe hypoglycemia and in the recovery period following glucose injection. Acta Physiol Scand.. 1980;109:307-314.[Medline] [Order article via Infotrieve]
36. Harris RJ, Wieloch T, Symon L, Siesjo BK. Cerebral extracellular calcium activity in severe hypoglycemia: relation to extracellular potassium and energy state. J Cereb Blood Flow Metab.. 1984;4:187-193.[Medline] [Order article via Infotrieve]
37. Wieloch T. Hypoglycemia-induced neuronal damage prevented by an N-methyl-D-aspartate antagonist. Science.. 1985;230:681-683.[Abstract/Free Full Text]
38. Westerberg E, Kehr J, Ungerstedt U, Wieloch T. The NMDA-antagonist MK-801 reduces extracellular amino acid levels during hypoglycemia and prevents striatal damage. Neurosci Res Communicator.. 1988;3:151-158.
39. Nellgard B, Wieloch T. Cerebral protection by AMPA- and NMDA-receptor antagonists administered after severe insulin-induced hypoglycemia. Exp Brain Res.. 1992;92:259-266.[Medline] [Order article via Infotrieve]
40. Linden T, Kalimo H, Wieloch T. Protective effect of lesion to the glutamatergic corticostriatal projections on the hypoglycemic nerve cell injury in rat striatum. Acta Neuropathol (Berl).. 1987;74:335-344.[Medline] [Order article via Infotrieve]
41. Hossmann KA, Fischer M, Bockhorst K, Hoehn-Berlage M. NMR imaging of the apparent diffusion coefficient (ADC) for the evaluation of metabolic suppression and recovery after prolonged cerebral ischemia. J Cereb Blood Flow Metab.. 1994;14:723-731.[Medline] [Order article via Infotrieve]
42. van Gelderen P, de Vleeschouwer MHM, Des Pres D, Pekar J, van Zijl PCM, Moonen CTW. Water diffusion and acute stroke. Magn Reson Med.. 1994;31:154-163.[Medline] [Order article via Infotrieve]

Editorial Comment

Joseph E. Brayden, PhD, Guest Editor

Department of PharmacologySmooth Muscle Ion Channel GroupUniversity of Vermont Medical Research FacilityColchester, Vt

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References Introduction  References

 
Hypoglycemia is quite different from ischemia with respect to brain damage. The best measure of the degree of a hypoglycemic insult is measurement of the duration of EEG isoelectricity. The brain can tolerate much greater durations of hypoglycemic "coma" than it can global cerebral ischemia. In the accompanying article the authors use 15 minutes and 50 minutes of hypoglycemic insult (mild and severe injuries, respectively) to show that the ADC of water, a measure of the freedom of water molecules in the ECS, transiently declines during hypoglycemia. This does not imply global cerebral swelling but a reduction in the ECS probably due to water moving into cells across the cell membrane. Importantly, the authors document reversibility even after the longer, more severe insult of 50 minutes.

A novel finding is that ADC changes began in cortical regions, spread to the subcortex, and were often asymmetrical (40% of animals) at onset. This implies that hypoglycemia may cause perturbations that do not take place simultaneously throughout the brain. The authors note that early ionic shifts in hypoglycemia resemble spreading depression, and this process may not only spread throughout the hemisphere but may be initiated asymmetrically between the hemispheres. This may explain a longstanding conundrum in hypoglycemia, ie, how a symmetrical reduction in blood sugar supply to the brain could lead to asymmetrical necrosis (see Fig 9 in Reference 1)^1R .

The authors could extend their present observations into areas not known to be damaged by even the more severe of the two insults that they studied.

The cerebellum and brain stem, for example, are notoriously resistant to hypoglycemic brain damage compared with the cerebrum. This may be related to a better influx of glucose across the capillary into those brain regions when supply is short.^2R If so, one would expect a lesser degree of alteration in ADC using the methods in the present study. Alternatively, if such "resistant" brain regions in hypoglycemia show identical changes in ADC, then other factors would have to be invoked to explain the resistance of the cerebellum and brain stem to a hypoglycemic insult.

	Selected Abbreviations and Acronyms

 

ADC = apparent diffusion coefficient ECS = extracellular space EEG = electroencephalogram GFAP = glial fibrillary acidic protein ICS = intracellular space Iso15 group = group in which blood glucose concentration was restored at 15 minutes Iso50 group = group in which blood glucose concentration was restored at 50 minutes ROI = region of interest

PMNs indicates polymorphonuclear leukocytes.

*Score of histopathological damage ranged from 0 (normal/absent) to 3 (severe).

P<.05 vs Iso¹⁵ and control groups (one-way ANOVA with Bonferroni corrected t test in each column).

	References

Top Abstract Introduction Materials and Methods Results Discussion References Introduction  References

 

1. Auer RN, Wieloch T, Olsson Y, Siesjo BK. The distribution of hypoglycemic brain damage. Acta Neuropathol (Berl).. 1984;64:177-191.
2. LaManna JC, Harik SI. Regional comparisons of brain glucose influx. Brain Res.. 1985;326:299-305.[Medline] [Order article via Infotrieve]

This article has been cited by other articles:

				C. Enzinger, F. Thimary, P. Kapeller, S. Ropele, R. Schmidt, F. Ebner, and F. Fazekas Transient Global Amnesia: Diffusion-Weighted Imaging Lesions and Cerebrovascular Disease Stroke, August 1, 2008; 39(8): 2219 - 2225. [Abstract] [Full Text] [PDF]

				C.H. Vite, S. Magnitsky, D. Aleman, P. O'Donnell, K. Cullen, W. Ding, S. Pickup, J.H. Wolfe, and H. Poptani Apparent Diffusion Coefficient Reveals Gray and White Matter Disease, and T2 Mapping Detects White Matter Disease in the Brain in Feline Alpha-Mannosidosis AJNR Am. J. Neuroradiol., February 1, 2008; 29(2): 308 - 313. [Abstract] [Full Text] [PDF]

				J Maruya, H Endoh, H Watanabe, H Motoyama, and H Abe Rapid improvement of diffusion-weighted imaging abnormalities after glucose infusion in hypoglycaemic coma J. Neurol. Neurosurg. Psychiatry, January 1, 2007; 78(1): 102 - 103. [Full Text] [PDF]

				S. Albayram, H. Ozer, S. Gokdemir, F. Gulsen, G. Kiziltan, N. Kocer, and C. Islak Reversible reduction of apparent diffusion coefficient values in bilateral internal capsules in transient hypoglycemia-induced hemiparesis. AJNR Am. J. Neuroradiol., September 1, 2006; 27(8): 1760 - 1762. [Abstract] [Full Text] [PDF]

				G. Hagemann, H.-J. Mentzel, H. Weisser, A. Kunze, and C. Terborg Multiple Reversible MR Signal Changes Caused by Epstein-Barr Virus Encephalitis AJNR Am. J. Neuroradiol., August 1, 2006; 27(7): 1447 - 1449. [Abstract] [Full Text] [PDF]

				L. Lo, C.H.A. Tan, T. Umapathi, and C.C. Lim Diffusion-Weighted MR Imaging in Early Diagnosis and Prognosis of Hypoglycemia AJNR Am. J. Neuroradiol., June 1, 2006; 27(6): 1222 - 1224. [Abstract] [Full Text] [PDF]

				D. Le Bihan, S.-i. Urayama, T. Aso, T. Hanakawa, and H. Fukuyama Direct and fast detection of neuronal activation in the human brain with diffusion MRI PNAS, May 23, 2006; 103(21): 8263 - 8268. [Abstract] [Full Text] [PDF]

				C. Cordonnier, C. Oppenheim, C. Lamy, J. -F. Meder, and J. -L. Mas Serial diffusion and perfusion-weighted MR in transient hypoglycemia Neurology, July 12, 2005; 65(1): 175 - 175. [Full Text] [PDF]

				H Axer, A Ragoschke-Schumm, J Bottcher, C Fitzek, O W Witte, and S Isenmann Initial DWI and ADC imaging may predict outcome in acute disseminated encephalomyelitis: report of two cases of brain stem encephalitis J. Neurol. Neurosurg. Psychiatry, July 1, 2005; 76(7): 996 - 998. [Abstract] [Full Text] [PDF]

				J. Bottcher, A. Kunze, C. Kurrat, P. Schmidt, G. Hagemann, O.W. Witte, and W.A. Kaiser Localized Reversible Reduction of Apparent Diffusion Coefficient in Transient Hypoglycemia-Induced Hemiparesis Stroke, March 1, 2005; 36(3): e20 - e22. [Abstract] [Full Text] [PDF]