Prevalence and determinants of subclinical brain infarction: the Northern Manhattan Study.

OBJECTIVE: Risk factors for subclinical brain infarcts (SBI) have not been well studied, especially in Hispanic and black populations who may be at higher risk for vascular disease. We examined the prevalence and determinants of SBI in a multiethnic community cohort. METHODS: The Northern Manhattan Study (NOMAS) includes 892 stroke-free participants who underwent brain MRI. Baseline demographic and vascular risk factor data were collected. The presence of SBI was determined from the size, location, and imaging characteristics of the lesion based on fluid attenuated inversion recovery (FLAIR) T1 and T2, and proton density MRI sequences. We calculated the prevalence of SBI and cross-sectional associations with sociodemographic and vascular risk factors, using logistic regression to adjust for relevant covariates. RESULTS: Among 892 subjects (mean age 71.3 years), 158 (17.7%) had SBI (13.5% had 1 lesion, 4.3% had >1 lesion). Of the total 216 infarcts, most were small (<1 cm, 82.4%) and subcortical (82.9%). SBI prevalence increased with age (<65: 9.7%; 65 to 75: 16.4%; >75: 26.1%), was increased among men (21.3% vs 15.2% in women), and was increased among blacks (24.0% vs 18.1% in whites and 15.8% in Hispanics). The presence of SBI was independently associated with older age (per year: OR 1.06, 95% CI 1.04 to 1.09), male sex (OR 1.79, 95% CI 1.22 to 2.61), and hypertension (OR 2.08, 95% CI 1.35 to 3.22) adjusting for age, sex, race-ethnicity, and vascular risk factors. A significant interaction (p = 0.002) between race and age was observed such that younger black subjects had greater odds of having SBI. CONCLUSIONS: SBI were detected in nearly 18% of subjects in a multiethnic community-based cohort. Age, male sex, and hypertension were independently associated with SBI. Subclinical cerebral infarcts are more prevalent than symptomatic infarcts and may increase the true public health burden of stroke.

Eosinophilic vasculitis in an isolated central nervous system distribution.

BACKGROUND: Eosinophilic vasculitis has been described as part of the Churg-Strauss syndrome, but affects the central nervous system (CNS) in <10% of cases; presentation in an isolated CNS distribution is rare. We present a case of eosinophilic vasculitis isolated to the CNS. CASE REPORT: A 39-year-old woman with a history of migraine without aura presented to an institution (located in the borough of Queens, New York, USA; no academic affiliation) in an acute confusional state with concurrent headache and left-sided weakness and numbness. Laboratory evaluation showed increased cerebrospinal fluid (CSF) protein level, but an otherwise unremarkable serological investigation. Magnetic resonance imaging showed bifrontal polar gyral-enhancing brain lesions. Her symptoms resolved over 2 weeks without residual deficit. After 18 months, later the patient presented with similar symptoms and neuroradiological findings involving territories different from those in her first episode. Again, the CSF protein level was high. She had a raised C reactive protein level and erythrocyte sedimentation rate. Brain biopsy showed transmural, predominantly eosinophilic, inflammatory infiltrates of medium-sized leptomeningeal arteries without granulomas. She improved, without recurrence, when treated with a prolonged course of corticosteroids. CONCLUSIONS: To our knowledge, this is the first case of non-granulomatous eosinophilic vasculitis isolated to the CNS. No aetiology for this patient's primary CNS eosinophilic vasculitis has yet been identified. Spontaneous resolution and recurrence of her syndrome is an unusual feature of the typical CNS vasculitis and may suggest an environmental epitope with immune reaction as the cause.

Circuit mechanisms underlying memory encoding and retrieval in the long axis of the hippocampal formation.

Circuits within the hippocampal formation are active during memory processing. Here we used functional magnetic resonance imaging (fMRI) to examine multiple sites across the long axis of the hippocampal formation while subjects performed different phases of an associative memory task, learning to associate faces with names. Viewing faces and hearing names in isolation resulted in separate hippocampal activation patterns. Pairing faces with names resulted a spatially redistributed activation pattern, rather than a simple summation of the activation patterns resulting from viewing faces and hearing names in isolation. Recalling names when cued with faces reactivated a pattern similar to that found during paired training. Finally, the activation patterns representing faces and names were found to be experience dependent, emerging with repeated exposure. Interpreted in the context of hippocampal anatomy and physiology, these findings reveal hippocampal circuit mechanisms that underlie memory encoding and retrieval.

A modified transorbital baboon model of reperfused stroke.

BACKGROUND AND PURPOSE: Although pathophysiological studies of focal cerebral ischemia in nonhuman primates can provide important information not obtainable in rodent models, primate experimentation is limited by considerations of cost, availability, effort, and ethics. A reproducible and quantitative model that minimizes the number of animals necessary to detect differences between treatment groups is therefore crucial. METHODS: Eight male baboons (weight, 22+/-2 kg) underwent left transorbital craniectomy followed by 1 hour of temporary ipsilateral internal carotid artery occlusion at the level of the anterior choroidal artery together with bilateral temporary occlusion of both anterior cerebral arteries (A1) proximal to the anterior communicating artery. A tightly controlled nitrous oxide-narcotic anesthetic allowed for intraoperative motor evoked potential confirmation of middle cerebral artery (MCA) territory ischemia. Animals survived to 72 hours or 10 days if successfully self-caring. Outcomes were assessed with a 100-point neurological grading system, and infarct volume was quantified by planimetric analysis of both MRI and triphenyltetrazolium chloride-stained sections. RESULTS: Infarction volumes (on T2-weighted images) were 32+/-7% (mean+/-SEM) of the ipsilateral hemisphere, and neurological scores averaged 29+/-9. All animals demonstrated evidence of hemispheric infarction, with damage evident in both cortical and subcortical regions in the MCA vascular territory. Histologically determined infarction volumes differed by <3% and correlated with absolute neurological scores (r=0.9, P:=0.003). CONCLUSIONS: Transorbital temporary occlusion of the entire anterior cerebral circulation with strict control of physiological parameters can reliably produce reperfused MCA territory infarction. The magnitude of the resultant infarct with little interanimal variability diminishes the potential number of animals required to distinguish between 2 treatment regimens. The anatomic distribution of the infarct and associated functional deficits offer comparability to human hemispheric strokes.

Evaluating the function of hippocampal subregions with high-resolution MRI in Alzheimer's disease and aging.

Memory ability declines in older age groups. There is a growing list of physiological processes that target the hippocampal formation in an age-related fashion, and some might underlie the hippocampal component of memory decline. The hippocampal formation is comprised of separate subregions, and physiological processes differentially target these subregions. The ability to evaluate the functional integrity of individual subregions-performing subregional analysis-is a major clinical goal since it can aid in the diagnosis of memory decline, as well as in elucidating mechanisms of disease and testing potential interventions. Because of its superior spatial resolution, magnetic resonance imaging (MRI) is best suited to accomplish this goal. Despite limited success, most functional MRI (fMRI) protocols have difficulty in performing complete subregional analysis of the hippocampal formation. Here we address sources of difficulty by (1) generating T2* -weighted maps of the hippocampal formation with sub-millimeter resolution; and (2) by adapting an approach used by animal investigators to identify the hippocampal subregions using anatomical landmarks. The protocol is tested in patients with Alzheimer's disease and in healthy controls, in an effort to determine whether it can detect neuronal dysfunction. Results showed diminished signal in the hippocampal formation of patients with Alzheimer's disease (AD) compared to controls, and multivariate analysis showed that this difference was most prominent in the entorhinal cortex. The protocol can be used to perform subregional analysis of the hippocampal formation. Testing the protocol in other clinical populations is needed to demonstrate its efficacy in evaluating the neuronal integrity of all hippocampal subregions.

Interhemispheric transfer of language in patients with left frontal cerebral arteriovenous malformation.

Cerebral arteriovenous malformations (AVMs) are frequently evaluated before therapeutic embolization by superselective injection of anesthetics into individual arterial branches so as to determine whether permanent occlusion would affect eloquent function. In Experiment 1, we used this adaptation of the Wada procedure to study three right-handed adult patients with left frontal cerebral AVMs by injecting vessels in Wernicke's and Broca's areas, respectively, and assessing language functions. The results showed that superselective testing in the inferior division of the left MCA in all three patients produced a dense Wernicke's aphasia. Injections into the left frontal regions, however, resulted in right paresis in all patients, but no language deficits including no loss of fluency. In Experiment 2, Patient 2 underwent fMRI activation for spontaneous word-list generation using multi-slice echo planar BOLD techniques at 1.5 Tesla. A voxel-by-voxel comparison of rest vs activation for each task was performed with a Z-score threshold of 2.5 SD for activated voxels. There was activation in the right hemisphere in the insula, frontal operculum pars opercularis, and inferior frontal gyrus, an area homologous to Broca's area in the left hemisphere. There was also activation in the left hemisphere in the Rolandic region, but language function was unaffected during Wada testing in this area. These data suggested that features of expressive language were no longer controlled by the left frontal lobe where the AVM was located, and provided new evidence for interhemispheric re-organization under conditions of chronic neurovascular disease.

Evolution of cortical activation during recovery from corticospinal tract infarction.

BACKGROUND AND PURPOSE: Recovery from hemiparesis due to corticospinal tract infarction is well documented, but the mechanism of recovery is unknown. Functional MRI (fMRI) provides a means of identifying focal brain activity related to movement of a paretic hand. Although prior studies have suggested that supplementary motor regions in the ipsilesional and contralesional hemisphere play a role in recovery, little is known about the time course of cortical activation in these regions as recovery proceeds. METHODS: Eight patients with first-ever corticospinal tract lacunes causing hemiparesis had serial fMRIs within the first few days after stroke and at 3 to 6 months. Six healthy subjects were used as controls. Statistically significant voxels during a finger-thumb opposition task were identified with an automated image processing program. An index of ipsilateral versus contralateral activity was used to compare relative contributions of the 2 hemispheres to motor function in the acute and chronic phases after stroke. RESULTS: Controls showed expected activation in the contralateral sensorimotor cortex (SMC), premotor, and supplementary motor areas. Stroke patients differed from control patients in showing greater activation in the ipsilateral SMC, ipsilateral posterior parietal, and bilateral prefrontal regions. Compared with the nonparetic hand, the ratio of contralateral to ipsilateral SMC activity during movement of the paretic hand increased significantly over time as the paretic hand regained function. CONCLUSIONS: The evolution of activation in the SMC from early contralesional activity to late ipsilesional activity suggests that a dynamic bihemispheric reorganization of motor networks occurs during recovery from hemiparesis.

Imaging physiologic dysfunction of individual hippocampal subregions in humans and genetically modified mice.

We have developed a variant of functional magnetic resonance imaging (fMRI) designed to be sensitive to static neuronal function. This method is based on resting instead of dynamic changes in oxygen-dependent signal and therefore allows for a spatial resolution that can detect signal from different hippocampal subregions in human subjects as well as in mice. We found that hippocampal signal was significantly diminished in elderly subjects with memory decline compared to age-matched controls, and different subjects showed dysfunction in different subregions. Among healthy elders, signal intensity from the subiculum was correlated selectively with memory performance. This method does not require an activation task; it can be used in anesthetized normal and in genetically modified and cognitively impaired mice. In mice the signal was found to be sufficiently sensitive to detect functional changes in the absence of underlying anatomical changes.

Differential regional dysfunction of the hippocampal formation among elderly with memory decline and Alzheimer's disease.

The hippocampal formation is composed of separate anatomical regions interconnected to form a circuit, and investigating abnormal hippocampal function is most revealing at the level of these regions. Until recently, regional analysis of the hippocampal formation could be performed only in animals or in human postmortem tissue. Here, we report a method using functional magnetic resonance imaging that evaluates the hippocampal regions in vivo, and we use this method to study elderly with normal memory, with isolated memory decline, and with probable Alzheimer's disease (AD). Although age-related memory decline occurs commonly, the cause of this decline remains unknown, with disagreement as to whether this decline represents one or more etiologies. Analysis revealed two distinct patterns of regional dysfunction among elderly with isolated memory decline--one pattern similar to that found in elders with AD, involving all hippocampal regions, and a second pattern with dysfunction restricted to only one hippocampal region, the subiculum. These results offer direct evidence of hippocampal dysfunction associated with memory decline in the elderly, and implicate both predementia AD and non-AD processes as possible underlying causes.

Prospective magnetic resonance imaging identification of focal cortical dysplasia, including the non-balloon cell subtype.

The purpose of this study was to determine the role of high-resolution T2-weighted fast multiplanar inversion-recovery (FMPIR) magnetic resonance (MR) imaging in detecting and delineating microscopic focal cortical dysplasia (FCD). We performed MR scans with FMPIR on 42 patients with suspected neocortical epilepsy. Ten MR studies were read prospectively as showing FCD; these case histories, electroencephalographic studies, and neuroimaging data were reviewed. Eight of these patients subsequently underwent focal cortical resection guided by intraoperative electrocorticography. The MR findings were correlated with pathological findings in these 8 patients. For purposes of radiological-pathological correlation, the FCD lesions were divided into two classes. Radiological classification was based on the absence (type A) or presence (type B) of T2 prolongation of the subcortical white matter. Pathological grading as type I or type II was based on a previously described pathological grading system. Specific MR findings associated with FCD included focal blurring of the gray-white matter interface (n = 9), thickening of the cortical ribbon (n = 7), and T2 prolongation of the subcortical white matter (n = 4). In 3 patients, the only MR finding that suggested FCD was localized blurring of the gray-white matter junction. In 2 of these 3 patients, the MR diagnosis of FCD could be made only by FMPIR. FCD was confirmed histologically in 7 of 8 patients, with insufficient tissue for complete histopathological evaluation in 1 case. Radiological classification of FCD agreed with pathological classification in 5 of 7 cases. Correlation of MR findings with intraoperative electrocorticography results indicated that the MR study localized the epileptogenic lesion correctly in 8 of 8 cases. Scalp ictal electroencephalographic studies localized the epileptogenic lesion in 5 of 8 cases; positron emission tomographic scans were focally abnormal in 3 of 3 cases. FMPIR MR imaging permitted accurate diagnosis and localization of FCD in all patients with pathologically proved FCD. MR identification of FCD aided presurgical planning and intraoperative management of these patients.

Early diagnosis and endovascular interventions for ischemic stroke.

In recent years there have been formidable advances in the war against stroke. The understanding and detection of stroke have undergone major progress at a rate previously unseen, partly due to major contributions from neuroradiology. Current routine neuroradiologic evaluation of acute stroke relies mainly on computed tomography scanning, although a number of radiologic modalities are becoming available that are based on various physical and chemical tissue properties, such as magnetic resonance imaging, single photon emission computed tomography, positron emission tomography, and magnetic resonance spectroscopy. All these new techniques allow the study of nervous tissue at the cellular and biochemical levels. A review of current diagnostic techniques for stroke follows in the first part of this article. The current status of endovascular therapy for ischemic stroke is reviewed in the second part of this article.

Sensitivity-enhanced echo-planar MRI at 1.5T using a 5 x 5 mesh dome resonator.

In this work a 5 x 5 mesh dome resonator that has been optimized for functional brain imaging is presented. The resonator was reduced in length and diameter compared with previous versions to reduce sample losses, thus enhancing the signal-to-noise ratio of the acquired data. In addition, a 5 x 5 mesh design was employed, which offered improved axial homogeneity over an earlier 3 x 3 mesh version. The new resonator exhibited high sensitivity and good homogeneity over the brain volume, permitting analysis of functional activation over large areas of the cerebral cortex. In a direct comparison with a standard clinical head-imaging resonator, the high sensitivity of the 5 x 5 mesh dome resonator resulted in greater statistical confidence in functional activation.

Advances in brain tumor diagnostic imaging.

In this review I will discuss recent developments in diagnostic magnetic resonance imaging of brain neoplasms, including refinements of clinically accepted methods and experimental techniques with as yet unproven clinical applications. The techniques and applications to be discussed are (1) fast magnetic resonance imaging methods to assess tumor perfusion (blood volume), blood-brain barrier and blood-tumor barrier, and diffusion status; (2) a specialized magnetic resonance imaging method, magnetization transfer, which is sensitive to microscopic structural changes in tissue and may also be used to improve paramagnetic contrast enhancement; (3) new magnetic resonance imaging paramagnetic contrast agents including recent sensitivity and efficacy trials; and (4) magnetic resonance spectroscopy and magnetic resonance spectroscopic imaging.

Illusory contours activate specific regions in human visual cortex: evidence from functional magnetic resonance imaging.

The neural basis for perceptual grouping operations in the human visual system, including the processes which generate illusory contours, is fundamental to understanding human vision. We have employed functional magnetic resonance imaging to investigate these processes noninvasively. Images were acquired on a GE Signa 1.5T scanner equipped for echo planar imaging with an in-plane resolution of 1.5 x 1.5 mm and slice thicknesses of 3.0 or 5.0 mm. Visual stimuli included nonaligned inducers (pacmen) that created no perceptual contours, similar inducers at the corners of a Kanizsa square that created illusory contours, and a real square formed by continuous contours. Multiple contiguous axial slices were acquired during baseline, visual stimulation, and poststimulation periods. Activated regions were identified by a multistage statistical analysis of the activation for each volume element sampled and were compared across conditions. Specific brain regions were activated in extrastriate cortex when the illusory contours were perceived but not during conditions when the illusory contours were absent. These unique regions were found primarily in the right hemisphere for all four subjects and demonstrate that specific brain regions are activated during the kind of perceptual grouping operations involved in illusory contour perception.

Echo-planar imaging.

Echo-planar imaging is a fast magnetic resonance (MR) imaging technique that allows acquisition of single images in as little as 20 msec and performance of multiple-image studies in as little as 20 seconds. Echo-planar imaging achieves its speed by obtaining all spatial-encoding information after a single radio-frequency (RF) excitation. Conventional imaging requires multiple-RF excitations, separated by the repetition time (TR), to acquire this information. An "infinite" TR, routine lipid suppression, and sensitivity to magnetic susceptibility are other features of echo-planar imaging. Standard pulse sequences are used to obtain echo-planar images, which have diagnostic utility similar to that of conventional MR images. Echo-planar imaging is less sensitive to motion than is conventional MR imaging and allows imaging of rapidly changing physiologic processes such as blood flow and kinetic activity. Echo-planar imaging is opening new areas of MR imaging research and clinical applications.

31P NMR phospholipid characterization of intracranial tumors.

Phospholipid extracts from 48 intracranial tumors were analyzed using 31P NMR. Phospholipids commonly identified in the tumor spectra included phosphatidylglycerol (PG), phosphatidic acid (PA), diphosphatidylglycerol (DPG), uncharacterized phospholipid (U), ethanolamine plasmalogen (EPLAS), phosphatidylethanolamine (PE), phosphatidylserine (PS), sphingomyelin (SM), lysophosphatidylcholine (LPC), phosphatidylinositol (PI), a choline phospholipid (CPLIP), and phosphatidylcholine (PC). Differences in the mean relative mole-percentage of phosphorus concentrations of individual phospholipids were used to differentiate among tumors. Neural sheath tumors (neurilemmoma, neurofibroma and fibrosarcoma) were noted to contain significantly elevated levels of SM relative to tumors of neural glial origin and individually, glioblastoma multiforme was noted to contain depressed levels of SM relative to neurilemmoma, neurofibroma and meningioma. Significantly decreased levels of PA were noted for glioblastoma relative to neurilemmoma along with significantly decreased levels of PE relative to meningioma. Elevated levels of LPC and CPLIP were seen in glioblastoma multiforme relative to meningioma. Additional findings included elevated levels of PC for glioblastoma multiforme relative to neurofibroma, and neurilemmoma was differentiated from neurofibroma with elevated levels of PA and depressed levels of PI. 31P NMR phospholipid analysis provides supplemental biochemical information which may be used to improve the interpretation of spectra acquired in vivo, and reveals important tumor-specific biochemical information which may further improve the understanding of the biological behavior of intracranial tumors.

Neuroprotection following focal cerebral ischaemia with the NMDA antagonist dextromethorphan, has a favourable dose response profile.

Although N-methyl-D-Aspartate (NMDA) antagonists protect against focal cerebral ischaemia, there is concern that the high doses necessary for neuroprotection may cause unacceptable adverse effects. We studied the dose response characteristics of the clinically available NMDA antagonist dextromethorphan in a rabbit model of transient focal ischaemia. Thirty-three anaesthetized rabbits underwent occlusion of the left internal carotid and anterior cerebral arteries for 1 h followed by 4.5 h of reperfusion. One hour after the onset of ischaemia, they were treated with an i.v. infusion of varying doses of dextromethorphan or normal saline. Seventeen additional unanaesthetized, nonischaemic rabbits received similar infusions of dextromethorphan to correlate brain with blood levels and to evaluate adverse effects. Rabbits with plasma dextromethorphan levels 500-1500 ng ml-1 had a 64% reduction in ischaemic neuronal damage (p < 0.05); those with levels > 1500 ng ml-1 showed 92% attenuation of neuronal damage and 65% decrease in ischaemic oedema (p < 0.01). Drug levels suggest that dextromethorphan's neuroprotection is not mediated by its active metabolite dextrorphan. Unanaesthetized rabbits with plasma levels > 2500 ng ml-1 demonstrated severe gait ataxia. These results demonstrate that systemic treatment with dextromethorphan after 1 h of focal ischaemia can significantly protect against cerebral damage if adequate plasma and brain levels are achieved. Dextromethorphan was concentrated 7-30 x in brain compared with plasma, and brain levels were highly correlated with plasma levels (r = 0.89). Neuroprotective doses of dextromethorphan were tolerated with only transient side effects.

High-dose single-fraction brain irradiation: MRI, cerebral blood flow, electrophysiological, and histological studies.

Radiation-induced alterations in cerebrovascular and metabolic function form the basis for the radiosurgical treatment of selected intracranial vascular malformations and tumors in human patients. However, the underlying mechanisms, temporal progression, and modifying factors involved in the radiosurgical obliteration of these intracranial lesions as well as the risks of delayed radiation injury to surrounding normal brain remain poorly understood. In this report, the rabbit brain was used as an animal model to examine the effects of high-dose single-fraction X-irradiation on magnetic resonance imaging (MRI) appearance, neurophysiologic function, and histological integrity. At approximately 10 weeks following left-hemisphere irradiation with 60 Gy (225 kVp) X rays, MRI studies showed radiation-induced changes including blood-brain barrier (BBB) perturbations in the white matter regions and the hippocampus. Significant reductions in regional cerebral blood flow (rCBF) ratios were found in the hippocampus and certain regions of the cortex in irradiated animals. However, no changes in somatosensory evoked potentials (SEP) were observed. Histological studies demonstrated telangiectatic vessels, spreading edema in the white matter, and focal regions of necrosis and hemorrhage in the irradiated cortices and hippocampi. These results demonstrate that the irradiated rabbit brain may be used as an experimental model to correlate the spatiotemporal pattern of functional changes with radiologic and histological changes in delayed radiation injury.

Dextromethorphan alters cerebral blood flow and protects against cerebral injury following focal ischemia.

The effects of the N-methyl-D-aspartate (NMDA) antagonist dextromethorphan (DM) on regional cerebral blood flow (rCBF) and cerebral injury were studied in a rabbit model of transient focal ischemia. Anesthetized rabbits underwent 2 h occlusion of the left internal carotid, middle cerebral and anterior cerebral artery, followed by 4 h of reperfusion. Ten minutes after the onset of ischemia they were treated with either i.v. DM 20 mg/kg followed by 10 mg/kg/h (n = 6) or normal saline (NS, n = 5). Control rabbits received DM (n = 3) or NS (n = 2) infusion without arterial occlusion. DM attenuated the sharp, post-ischemic rise in rCBF seen during reperfusion within the ischemic core of NS controls (DM 31% pre-ischemic value, NS 92%). DM also improved the delayed post-ischemic hypoperfusion compared with controls. DM infusion without arterial occlusion did not change rCBF values. Compared with NS controls, DM treated animals demonstrated recovery of the somatosensory evoked potential (DM 96% pre-ischemic values, NS 24%), 76% reduction in cortical edema and 92% decrease in cortical ischemic neuronal damage. We conclude that DM's effect on CBF may contribute to its neuroprotective action.

Protection after transient focal cerebral ischemia by the N-methyl-D-aspartate antagonist dextrorphan is dependent upon plasma and brain levels.

Dextrorphan is a dextrorotatory morphinan and a noncompetitive N-methyl-D-aspartate (NMDA) receptor antagonist. We studied the dose response characteristics of dextrorphan's neuroprotective efficacy and side effects, correlating these beneficial and adverse responses with plasma and brain levels in a rabbit model of transient focal cerebral ischemia. Thirty-three rabbits, anesthetized with halothane, underwent occlusion of the left internal carotid and anterior cerebral arteries for 1 h, followed by 4.5 h of reperfusion. One hour after the onset of ischemia, they were treated with an i.v. infusion of varying dextrorphan doses or normal saline. After killing, the brains were analyzed for ischemic high signal intensity using magnetic resonance imaging (MRI) and for ischemic neuronal damage with histopathology. A separate group of 12 anesthetized ischemic rabbits received similar doses of dextrorphan, correlating plasma with brain dextrorphan levels. Twenty-six additional dextrorphan unanesthetized, nonischemic rabbits received infusions of dextrorphan to correlate behavioral side effects with dextrorphan dose and levels. Compared with controls, dextrorphan 15 mg/kg group had significantly less cortical ischemic neuronal damage (5.3 versus 33.2%, p = 0.01) and a reduction in cortical MRI high signal area (9.1 versus 41.2%, p = 0.02). The dextrorphan 10 mg/kg rabbits showed less cortical ischemic neuronal damage (27.2%) and less MRI high signal (34.8%) but this was not statistically significant (p = 0.6). Dextrorphan 5 mg/kg had no benefit on either neocortical ischemic neuronal damage (35.8%) or MRI high signal (42.9%). The protective effect of dextrorphan was correlated with plasma free dextrorphan levels (r = -0.50, p less than 0.02 for ischemic neuronal damage; r = -0.66, p less than 0.001 for ischemic MRI high signal). All the rabbits with plasma levels greater than 2,000 ng/ml had less than 12% cortical ischemic neuronal damage and less than 34% MRI high signal. All rabbits with plasma levels greater than 3,000 ng/ml showed less than 7% ischemic neuronal damage and less than 11% MRI high signal. Plasma levels of approximately 2,500 ng/ml correlated with brain dextrorphan levels of approximately 6,000 ng/g. Unanesthetized rabbits with plasma levels of approximately 2,500 ng/ml demonstrated loss of the righting reflex. These results demonstrate that systemic treatment with dextrorphan after 1 h focal ischemia can significantly protect against cerebral damage if adequate plasma and brain levels of dextrorphan are achieved. The brain levels necessary to obtain in vivo protection are similar to concentrations that prevent glutamate or NMDA-induced injury in neuronal culture.