Impaired autoregulation of cerebral blood flow in an experimental model of traumatic brain injury.

In order to study the pathophysiology and the intracranial hemodynamics of traumatic brain injury, we have developed a modified closed-head injury model of impact-acceleration that expresses several features of severe head injury in humans, including acute and long-lasting intracranial hypertension, diffuse axonal injury, neuronal necrosis, bleeding, and edema. In view of the clinical relevance of impaired autoregulation of cerebral blood flow after traumatic brain injury, and aiming at further characterization of the model, we investigated the autoregulation efficiency 24 h after experimental closed-head injury. Cortical blood flow was continuously monitored with a laser-Doppler flowmeter, and the mean arterial blood pressure was progressively decreased by controlled hemorrhage. Relative laser-Doppler flow was plotted against the corresponding mean arterial blood pressure, and a two-line segmented model was applied to determine the break point and slopes of the autoregulation curves. The slope of the curve at the right hand of the break point was significantly increased in the closed head injury group (0.751 +/- 0.966%/mm Hg versus -0.104 +/- 0.425%/mm Hg,p = 0.028). The break point tended towards higher values in the closed head injury group (62.2 +/- 20.8 mm Hg versus 46.9 +/- 12.7 mm Hg; mean +/- SD, p = 0.198). It is concluded that cerebral autoregulation in this modified closed head injury model is impaired 24 h after traumatic brain injury. This finding, in addition to other characteristic features of severe head injury established earlier in this model, significantly contributes to its clinical relevance.

Effect of repetitive umbilical cord occlusions on neuronal brain activity measured by the cerebral function analyzing monitor and histologic outcome in immature fetal sheep.

OBJECTIVE: To examine the effect of repetitive total umbilical cord occlusions on electrocortical brain activity as measured by cerebral function analyzing monitoring (CAFM) and the histologic outcome in immature sheep fetuses. STUDY DESIGN: We performed brief repeated total umbilical cord occlusions, two every 5 minutes, in 12 immature sheep fetuses (at 90 days of gestation, term 147 days) until fetal mean arterial pressure dropped below 50% of baseline value during two successive occlusions. A pair of electrodes was inserted on the parietal dura for recording of electrocortical brain activity (ECoG). Off-line ECoG signal processing consisted of amplitude integrated analysis (CFAM) and spectral analysis. Fetal blood gas analyses were performed at regular intervals just before subsequent umbilical cord occlusions. Three days after the occlusion neuronal damage was evaluated histologically in three regions of the fetal brain. RESULTS: CFAM amplitide parameters decreased significantly during the first occlusion and remained so during the entire repetitive occlusion period (analysis of variance [ANOVA]; P <.05). Spectral analysis of the ECoG signal demonstrated no changes in the distribution of frequency bands. Progressive acidemia and hypotension developed with ongoing occlusions. Five fetuses died at the end or shortly after the entire repetitive occlusion period. No neuronal damage or macroscopic intraventricular and/or germinal matrix hemorrhage was observed in the surviving fetuses. CONCLUSION: Repetitive umbilical cord occlusions in immature sheep fetuses resulted in functional, not structural changes of the fetal brain in surviving fetuses. At this gestational age, amplitude analysis is more sensitive than spectral analysis of the ECoG signal to functional changes of the compromised fetal brain.

Validation of a closed head injury model for use in long-term studies.

To study pharmacotherapy of traumatic brain injury in rats, a modified closed head injury model was used that expresses clinically relevant features including intracranial hypertension and morphological alterations. Long-term survival under ethically acceptable conditions would greatly improve its clinical relevance. To ensure this goal with great reproducibility, the experimental protocol was adapted, in particular the impact-acceleration kinetics. Variations in impact-acceleration conditions were obtained by modifying the stiffness of the impact site and changing the height of a 400 g weight dropped from 51.5 to 31.5 cm (51.5/400; 31.5/400). Impact and acceleration were measured with a force sensor incorporated in a rigid dummy-rat and an accelerometer mounted on the platform onto which the animals are positioned. Significant correlation was shown between impact and acceleration. Accelerations obtained in rats were significantly lower than those in the dummy. Unlike the 51.5/400 group, in the 31.5/400 group no mortality or cranial fractures were observed. In both groups intracranial pressure rose to pathological values immediately after trauma and remained elevated longer than 24 h. Diffuse axonal injury developed in all groups and remained present for at least 7 days. By reducing the impact-acceleration conditions, post-traumatic complications were diminished, while the clinically important features were maintained.

Vasogenic oedema and brain infarction in an experimental penumbra model.

The aim of this study was to modify the photochemical stroke model of Watson et al. [23] so as to make possible microscopical investigation of the so-called penumbra, a tissue zone at risk that surrounds an infarction. The idea was to minimize photochemical challenge to endothelial membranes in such a way that thrombotic vascular obstruction is avoided but destabilization of the blood-brain barrier is still obtained. Morphological examination of the challenged area revealed open blood vessels, overt blood-brain barrier leakage over the entire area, severely swollen glial cells and structurally intact neurons. The lesion expanded over time due to progressive extravasation, formation of perivascular edema and consequent development of secondary ischemia through mechanical compression and microvascular congestion. In contrast to a photothrombotic infarct, in which the ischemic insult is more severe and blood vessels are completely congested by aggregated platelets, with this approach blood flow is partially preserved. In this way, an ischemic penumbra is created that mimics pathologic conditions secondary to stroke and trauma. The model may be useful in studying effects of drugs on pathologic phenomena that are characteristic of a penumbra, e.g. vasogenic and cellular edema, inflammation and infarction.

Transient umbilical cord occlusion in late-gestation fetal sheep results in hippocampal damage but not in cerebral arteriovenous difference for nitrite, a stable end product of nitric oxide.

OBJECTIVE: To investigate the effect of total umbilical cord occlusion on cerebral arteriovenous difference for nitrite (a stable end product of nitric oxide) and neuronal outcome. METHOD: The cord was totally occluded for 10 minutes in 14 late-gestation (gestational age 113-120 days) chronically instrumented fetal sheep. Arterial (carotid artery) and venous (superior sagittal sinus) blood samples were taken at regular intervals for determination of acid-base status, glucose, lactate, and nitrite plasma levels. Three days after the occlusion period the fetal brain was perfusion fixed, and the parietal cortex, hippocampus, and cerebellum were scored for neuronal damage. RESULTS: Three fetuses died shortly after the occlusion period. Total umbilical cord occlusion resulted in a combined respiratory and metabolic acidosis as observed in carotid arterial blood gas samples (pH, 6.96 +/- 0.02; CaO2 [mmol/L], 0.43 +/- 0.9; PaCO2 [kPa], 13.46 +/- 0.38; base excess [mmol/L], -11.1 +/- 0.8; lactate [mmol/L], 10.57 +/- 0.95; bradycardia, 75 +/- 9 bpm; and hypotension, 29.85 +/- 3.00 mmHg) (n = 14, values are mean +/- standard error of the mean). Selective brain damage was observed in the hippocampus in 10 of the 11 surviving fetal sheep. No changes in arterial, venous, or cerebral arteriovenous difference for nitrite plasma levels were observed (n = 11). CONCLUSION: Total umbilical cord occlusion of 10 minutes in preterm fetal sheep results in hippocampal damage but not in changes of cerebral arteriovenous difference for nitrite plasma levels, a stable end product of nitric oxide.

Temporal changes in intracranial pressure in a modified experimental model of closed head injury.

OBJECT: The authors describe an experimental model of closed head injury in rodents that was modified from one developed by Marmarou and colleagues. This modification allows dual control of the dynamic process of impact compared with impulse loading that occurs at the moment of primary brain injury. The principal element in this weight-drop model is an adjustable table that supports the rat at the moment of impact from weights positioned at different heights (accelerations). The aim was to obtain reproducible pathological intracranial pressure (ICPs) while maximally reducing the incidence of mortality and skull fractures. METHODS: Intracranial pressure was investigated in different experimental settings, including two different rat strains and various impact-acceleration conditions and posttrauma survival times. Identical impact-acceleration injuries produced a considerably higher mortality rate in Wistar rats than in Sprague-Dawley rats (50% and 0%, respectively). Gradually increasing severity of impact-acceleration conditions resulted in findings of a significant correlation between the degree of traumatic challenge and increased ICP at 4 hours (p < 0.001, R2=0.73). When the impact-acceleration ratio was changed to result in a more severe head injury, the ICP at 4, 24, and 72 hours was significantly elevated in comparison with that seen in sham-injured rats (4 hours: 19.7+/-2.8 mm Hg, p=0.004; 24 hours: 21.8+/-1.1 mm Hg, p=0.002; 72 hours: 11.9+/-2.5 mm Hg, p=0.009). Comparison of the rise in ICP between moderate and severe impact-acceleration injury at 4 and 24 hours revealed a significantly higher value after severe injury (4 hours: p=0.008; 24 hours: p=0.004). Continuous recordings showed that ICP mounted very rapidly to peak values, which declined gradually toward a pathological level dependent on the severity of the primary insult. Histological examination after severe trauma revealed evidence of irreversible neuronal necrosis, diffuse axonal injury, petechial bleeding, glial swelling, and perivascular edema. CONCLUSIONS: This modified closed head injury model mimics several clinical features of traumatic injury and produces reliable, predictable, and reproducible ICP elevations with concomitant morphological alterations.

Absence of neuronal damage after umbilical cord occlusion of 10, 15, and 20 minutes in midgestation fetal sheep.

OBJECTIVES: Our purpose was to determine whether neuronal damage results after total umbilical cord occlusion of increasing duration in midgestation fetal sheep. STUDY DESIGN: We performed total umbilical cord occlusion during 10 (n = 11), 15 (n = 8), or 20 (n = 4) minutes in chronically instrumented midgestation fetal sheep. Nine fetuses served as sham controls. During the experiment fetal blood pressure (mean arterial pressure) and heart rate were continuously recorded. Fetal blood gas analyses were performed at regular intervals before, during, and after the occlusion. Three days after the occlusion neuronal damage was evaluated histologically in three regions of the fetal brain. RESULTS: Total umbilical cord occlusion resulted in hypotension, bradycardia, severe mixed acidemia, hypoxia, and hypercapnia. All fetuses survived the occlusion. No neuronal damage nor macroscopic intraventricular or germinal matrix hemorrhages were observed in either group. CONCLUSION: Prolonging the duration of total umbilical cord occlusion in midgestation fetal sheep resulted in a progressive increase in the severity of asphyxia, not in neuronal damage.

Protection with lubeluzole against delayed ischemic brain damage in rats. A quantitative histopathologic study.

BACKGROUND AND PURPOSE: Cerebral ischemia may lead to glutamate-induced excitotoxic damage in vulnerable brain areas. Lubeluzole is not an N-methyl-D-aspartate antagonist but prevents postischemic increase in extracellular glutamate concentrations. The present study examined whether lubeluzole, administered after global incomplete ischemia in rats, is capable of preserving the structural integrity of CA1 hippocampus. METHODS: Ischemia was induced by bilateral carotid artery occlusion and severe hypotension for a duration of 9 minutes. Delayed neuronal cell death was histologically evaluated 7 days later. This was done by scoring acidophilic cell change and coagulative necrosis and by counting the number of surviving neurons in the CA1 subfield. Experiments were performed according to a paired design (13 animals per treatment group). RESULTS: Posttreatment with lubeluzole (0.31 mg/kg i.v. bolus at 5 minutes and 0.31 mg/kg i.v. infusion during 1 hour) resulted in significant neuroprotection. Whereas in the untreated rats there were 42 (median) viable neurons per millimeter CA1 layer in the left and 69 in the right hemisphere, in the drug-treated rats 99 viable neurons per millimeter were found in the left (P = .002) and 113 in the right hemisphere (P = .013). Histological scores, reflecting altered staining properties of the hippocampal cells, correlated strongly with the quantitative data, reflecting the structural integrity of CA1 pyramidal neurons. CONCLUSIONS: Lubeluzole, when administered after an ischemic insult in rats, protects vulnerable brain regions against delayed structural injury. The results support the potential clinical use of this new drug in stroke treatment.

Altered Na(+)-channel function as an in vitro model of the ischemic penumbra: action of lubeluzole and other neuroprotective drugs.

Veratridine blocks Na(+)-channel inactivation and causes a persistant Na(+)-influx. Exposure of hippocampal slices to 10 microM veratridine led to a failure of synaptic transmission, repetitive spreading depression (SD)-like depolarizations of increasing duration, loss of Ca(+)-homeostasis, a large reduction of membrane potential, spongious edema and metabolic failure. Normalization of the amplitude of the negative DC shift evoked by high K+ ACSF 80 min after veratridine exposure was taken as the primary endpoint for neuroprotection. Compounds whose mechanisms of action includes Na(+)-channel modulation were neuroprotective (IC50-values in microM): tetrodotoxin 0.017, verapamil 1.18, riluzole 1.95, lamotrigine > or = 10, and diphenylhydantoin 16.1. Both NMDA (MK-801 and PH) and non-NMDA (NBQX) excitatory amino acid antagonists were inactive, as were NOS-synthesis inhibitor (nitro-L-arginine and L-NAME) Ca(2+)-channel blockers (cadmium, nimodipine) and a K(+)-channel blocker (TEA). Lubeluzole significantly delayed in time before the slices became epileptic, postponed the first SD-like depolarization, allowed the slices to better recover their membrane potential after a larger number of SD-like DC depolarizations, preserved Ca2+ and energy homeostasis, and prevented the neurotoxic effects of veratridine (IC50-value 0.54 microM). A concentration of lubeluzole, which was 40 x higher than its IC50-value for neuroprotection against veratridine, had no effect on repetitive Na(+)-dependent action potentials induced by depolarizing current in normal ACSF. The ability of lubeluzole to prevent the pathological consequences of excessive Na(+)-influx, without altering normal Na(+)- channel function may be of benefit in stroke.

Intracranial pressure in a modified experimental model of closed head injury.

Intracranial pressure (ICP) was studied in a modified experimental model of closed head injury, in which the dynamic process of impact versus impulse loading was separately controlled. In this model, mortality of Wistar rats was considerably higher as compared to Sprague-Dawley rats subjected to similar traumatic conditions. Therefore Sprague-Dawley rats were used for all further experiments. Twenty-four rats, divided into 4 groups, underwent either sham or gradually increasing impact-acceleration trauma. Four hours after closed head injury, ICP measurements showed a significant correlation between the severity of the traumatic challenge and the resultant pressure rise (r2 = 0.731; p < 0.001). At the moment of impact there was a momentary blood pressure peak immediately followed by a transient period of hypotension. ICP measurements following directly to an impact-acceleration trauma, revealed an abrupt rise in ICP reaching pathological levels within 5 minutes. In conclusion, this modified model of closed head injury produces a predictable and reproducible pathologic ICP in Sprague-Dawley rats.

Lubeluzole blocks increases in extracellular glutamate and taurine in the peri-infarct zone in rats.

A microdialysis probe was positioned inside the peri-infarct zone of a photochemically induced neocortical infarct in rats. Extracellular glutamate rose within 20 min after the start of infarct induction and continued to increase during the 5 h observation period to 5.5-fold the pre-infarct baseline value of 0.8 +/- 0.4 micromol/l. Glutamine increased only 1.4-fold. Changes in peri-infarct glutamate were preceded by steep rises in taurine (a 3.9-fold increase from the baseline value of 2.8 +/- 0.7 micromol/l), which coincided with spreading depressions during infarct induction. Post-treatment with lubeluzole ((S)-4-(2-benzothiazolylmethylamino)-alpha-[(3,4-difluoro-phenoxy) methyl]-1-piperidineethanol, 1.25 mg/kg i.v.), a new cerebroprotective drug, blocked the peri-infarct increases of glutamate and taurine, whereas the R-enantiomer was ineffective. Since lubeluzole has previously been shown to stereospecifically decrease glutamate-activated nitric oxide (NO) toxicity in vitro, the present in vivo stereospecific effect of lubeluzole may be related to modulation of the cascade of NO toxicity, thus preventing NO toxicity-mediated increases in extracellular glutamate. Blockade of the peri-infarct taurine response suggests that lubeluzole also may have reduced cellular osmotic stress in the peri-infarct zone.

In vivo noninvasive determination of abnormal water diffusion in the rat brain studied in an animal model for multiple sclerosis by diffusion-weighted NMR imaging.

In vivo NMR images of the rat brain were obtained using a NMR microscope (7 T) from SMIS (England). Four animals were imaged every 3-4 days during a pathological cycle (starting after induction and up to 37 days) of experimental allergic encephalomyelitis (EAE), an animal model for multiple sclerosis. The EAE rats were weighted and clinically scored daily. We aimed at measuring the apparent diffusion coefficient (ADC) or the mean diffusivity (D) with a high accuracy, and within a reasonable experimental time frame, because of the clinical situation of the animals. Therefore, we fitted the ADC value from five diffusion-weighted images--with an experimental time of 17 min/image--and chose to apply diffusion-sensitizing gradients in a direction intersecting all fiber directions of the external capsule. With this, we also obtained high b-values. For the control rats, we obtained a statistical mean value of ADC = (388 +/- 16) 10(-12) m2/s for gray matter and a statistical mean value of (D) of (750 +/- 30) 10(-12) m2/s for white matter, measured in the external capsule. For the EAE rats, no alterations in ADC values of gray matter with increasing clinical scores were observed. Concerning white matter, as determined in the external capsule, there were no significant differences in (D) values between controls and EAE rats before clinical signs occurred. However, when clinical signs were observed, we could demonstrate a significant positive correlation between the clinical score and the (D) values in the external capsule. As the clinical signs became more severe, we measured a rise in water diffusion (increase in (D)) in the external capsule, which was accompanied by the occurrence of interstitial edema as revealed by a complementary histological study.

Histopathological characterization of photochemical damage in nervous tissue.

This paper discusses histological and ultrastructural changes produced by dye-sensitized photoreactions in the central and peripheral nervous system. Particular attention has been given to morphological outcome in experimental models which reproduce widespread clinical pathologies, e.g. stroke, spinal cord injury and peripheral neuropathy. Evaluation of structural alterations may not only help to characterize the evolution of these disease processes but also allow us to study possibilities of therapeutic intervention.

Effects of asphyxia on the fetal lamb brain.

OBJECTIVE: Our purpose was to study the effect of fetal asphyxia on the release of hypoxanthine and xanthine in cerebrospinal fluid and on brain histologic characteristics. STUDY DESIGN: In seven fetal lambs (3 to 5 days after surgery, gestational age 124.3 +/- 2.6 days) asphyxia was induced by restriction of uterine blood flow. RESULTS: Fetal pH and base excess were reduced to 6.99 +/- 0.02 and -17.6 +/- 0.9 mmol/L, respectively. Cerebral blood flow increased during asphyxia and returned to normal in the recovery phase. Maximum concentrations of cerebrospinal fluid hypoxanthine and xanthine were reached in the normoxemic recovery phase. This high level of substrates during normoxemia facilitates oxygen free radical formation and may thus aggravate postasphyctic brain damage. Histologic evaluation of the brain 3 days after the insult showed a variable degree of edema. Coagulative neuronal changes, characteristic of irreversible cell death, were only occasionally detected. These changes were most obvious in the Purkinje cells of the cerebellum. CONCLUSIONS: Fetal asphyxia induced by uterine blood flow restriction is associated with high levels of cerebrospinal fluid hypoxanthine and xanthine in the recovery phase. Microscopically detectable brain damage, although not extensive, is mainly located in the cerebellum.

Possible neuroprotective properties of flunarizine infused after asphyxia in fetal lambs are not explained by effects on cerebral blood flow or systemic blood pressure.

Neuroprotective properties of the calcium channel blocker flunarizine have been reported after hypoxic-ischemic insults in immature, infant, and adult rats. However, its effect on fetal regional cerebral blood flow (rCBF) and systemic blood pressure after severe asphyxia is not known. In 15 fetal lambs (3 to 5 d after surgery; gestational age at the experiment, 123.2 +/- 2.5 d), arterial oxygen content was progressively reduced to 30% by restriction of uterine blood flow with an inflatable balloon occluder around the maternal common internal iliac artery. The rCBF was measured with radioactive microspheres at baseline condition, after 1 h of severe asphyxia, and at 30 and 120 min in the recovery phase. Immediately after the end of the occlusion period, fetuses randomly received either flunarizine or its solvent (0.5 mg/kg estimated fetal weight). No differences in rCBF changes between groups were observed during and after asphyxia. Changes in arterial blood pressure or fetal heart rate due to flunarizine could not be demonstrated either. Only five fetuses (33%) survived this degree of asphyxia longer than 24 h: four of the flunarizine-treated group and one of the control group. It is unlikely that this possible protective property of the drug is caused by its influence on rCBF, arterial blood pressure, or fetal heart rate in the phase immediately after asphyxia.

A new model of photochemically induced acute and reversible demyelination in the peripheral nervous system.

Acute photochemical demyelination accompanied by minor axonal degeneration was produced in rat sciatic nerve after topical application of the photosensitive dye Rose Bengal and focal illumination with cold light. Animals were sacrificed at different time periods after challenge and the exposed nerves prepared for light microscopic and ultrastructural evaluation. Important structural changes were already observed at 4-6 h. These included endoneurial swelling, diapedesis of neutrophils and monocytes, vacuolization and vesicularization of Schwann cell cytoplasm, lamellar separation of myelin sheaths, disintegration of axonal microtubules, and accumulation of vesicular material and mitochondria in the axoplasm. Disrupted myelin fragments were phagocytosed by macrophages which penetrated Schwann tubes at Day 3. Schwann cells proliferated and started to enwrap denuded segments of the axon. They were surrounded by redundant basal lamina, thrown into deep folds. Axons remained partly hypertrophic and contained many neurofilaments. A minority showed signs of degeneration. At Days 5-7 denudation was almost complete in the light-exposed nerve area but also in small distal nerve fascicles. After 1 month, axons in the illuminated area and distal to it were completely remyelinated although they had thinner sheaths. Exposure to increased light intensity resulted in deeper lesions and more extended anterograde damage, which also recovered within 1 month. All animals showed rapid functional deterioration which correlated with the severity and extent of structural damage. Recovery was slow and also depended on the degree of histologic damage. Neither control nerves nor sham-exposed nerves revealed signs of structural or functional changes.(ABSTRACT TRUNCATED AT 250 WORDS)

The effect of adenosine transport inhibition on cardiovascular function and survival after severe asphyxia in fetal lambs.

When the energy demand exceeds the energy supply, anaerobic metabolism takes over and the ATP catabolite adenosine is generated. Adenosine acts as a coronary vasodilator, thereby increasing the oxygen supply to the heart. Its potential, however, is poorly exploited due to extensive catabolism. R-75231 inhibits transport of adenosine into endothelial cells, where it is catabolized, resulting in an elevation of interstitial adenosine concentrations. In 14 fetal lambs (3 to 5 d after surgery, gestational age 124.1 +/- 1.1 d), seven fetuses were pretreated with R-75231 (0.1 mg/kg estimated fetal weight as a bolus injection in the inferior vena cava), whereas the other seven served as controls. After 1 h of severe asphyxia, induced by restriction of uterine blood flow, those fetuses treated with R-75231 showed a faster normalization of aortal pH and, in contrast to the control group, did not develop tachycardia. The percentage increase in myocardial blood flow during asphyxia, measured with radioactive microspheres, was significantly higher in the R-75231-treated group compared with the control group (437 and 284%, respectively). In the control group, only three fetuses recovered and survived, whereas in the R-75231 group, all seven animals recovered after severe asphyxia. It is concluded that fetal lambs pretreated with R-75231 before the onset of severe asphyxia have an enhanced increase in myocardial blood flow during asphyxia, recover faster, and survive longer.

Learning performance at different time periods after hypoxia in infant rats.

Infant rats were exposed to hypoxia. At different time periods after the hypoxic aggression the animals were subjected to a one-way active avoidance task consisting of 3 daily training sessions. Learning performance of hypoxic rats, expressed as the number of avoidances per 10 trials, was significantly decreased in all the sessions at posthypoxic day 35 (P < 0.001). The rate to which learning behaviour was affected by hypoxia was comparable among all 3 sessions in both groups. Histological evaluation of hippocampus and cortex at day 7 and day 35 after hypoxic exposure, did not reveal detectable structural changes. It can be concluded from this study that hypoxia, induced in 21 day old rats leads to learning performance deficits in the adult animal.

Photochemically-induced cerebral infarction in the rat: comparison of NMR imaging and histologic changes.

The evolution of a photochemically induced cerebral thrombotic infarction was followed in rats during the first week after the insult by means of NMR imaging and histology. Heavily T2-weighted images provided an excellent lesion detection and a high specificity for the discrimination of different histological abnormalities. The T2-weighted images showed a brain lesion evolving during the first 24 h from a homogeneous hyperintense area, histologically corresponding to diffuse vasogenic and cytotoxic oedema with concomitant neuronal necrosis, to an iso-intense area with a hyperintense seam, which microscopically correlated with increased vascular permeability at the periphery of the lesion. The hyperintense seam was observed up to day 7, but at that time coincided with gliomesodermal repair reaction which could be verified histochemically and ultrastructurally. It may be concluded that NMR-micro-imaging at a moderately high field, enables early detection and adequate follow-up of small cerebral infarctions in rats.