Transient Middle Cerebral Artery Occlusion by Intraluminal Suture: I. Three-Dimensional Autoradiographic Image-Analysis of Local Cerebral Glucose Metabolism—Blood Flow Interrelationships during Ischemia and Early Recirculation

Abstract

Using autoradiographic image-averaging strategies, we studied the relationship between local glucose utilization (LCMRglc) and blood flow (LCBF) in a highly reproducible model of transient (2-hour) middle cerebral artery occlusion (MCAO) produced in Sprague-Dawley rats by insertion of an intraluminal suture coated with poly-L-lysine. Neurobehavioral examination at 60 minutes after occlusion substantiated a high-grade deficit in all animals. In two subgroups, LCBF was measured with ¹⁴C-iodoantipyrine at either 1.5 hours of MCAO, or at 1 hour of recirculation after suture removal. In two other matched subgroups, LCMRglc was measured with ¹⁴C-2-deoxyglucose at 1.5 to 2.25 hours of MCAO, and at 0.75 to 1.5 hours of recirculation after 2 hours of MCAO. Average image data sets were generated for LCBF, LCMRglc, and the LCMRglc/LCBF ratio for each study time. Middle cerebral artery occlusion for 2 hours induced graded LCBF decrements affecting ipsilateral cortical and basal ganglionic regions. After 1 hour of recirculation, LCBF in previously ischemic neocortical regions increased by 40% to 200% above ischemic levels, but remained depressed, on average, at about 40% of control. By contrast, frank hyperemia was noted in the previously ischemic caudoputamen. Mean cortical LCBF values during MCAO correlated highly with their respective LCBF values after 1 hour of recirculation (R = 0.93), suggesting that postischemic LCBF recovery is related to the depth of ischemia. Despite focal ischemia, LCMRglc during ~2 hours of MCAO was preserved, on average, at near-normal levels; but following ~1 h of recirculation, LCMRglc became markedly depressed (on average, 55% of control in previously densely ischemic cortical regions). Regression analysis indicated that this depressed glucose utilization was determined largely by the intensity of antecedent ischemia. By pixel analysis, the ischemic core (defined as LCBF 0% to 20% of control) comprised 33% of the ischemic hemisphere, and the penumbra (LCBF 20% to 40%) accounted for 26%. The penumbra was concentrated at the coronal poles of the ischemic lesion and formed a thin shell around the central ischemic core. During 2 hours of MCAO, the LCMRglc/LCBF ratio within the ischemic penumbra was increased four-fold above normal (average, 179 umol/100 mL). In marked contrast, after ~1 h recirculation, this uncoupling had almost completely subsided. The companion study (Zhao et al., 1997) further analyzes these findings in relation to patterns of infarctive histopathology.

Keywords

Focal cerebral ischemia Stroke Penumbra 2-deoxyglucose Cerebral blood flow Flow-metabolism couple

Graded decrements in cerebral perfusion in the focally ischemic brain set into motion a complex sequence of pathophysiological events, which are sensitively reflected in local alterations of cerebral glucose utilization (LCMRglc), blood flow (LCBF), and the interrelationship between metabolism and flow—variables that are readily amenable to quantitative analysis by autoradiographic methods (Sakurada et al., 1978; Sokoloff et al., 1977; Zhao et al., 1995). The complete characterization of the pathophysiological processes leading to brain injury in cerebral ischemia is greatly aided by the use of consistent animal models having a high degree of interanimal reproducibility (Ginsberg, 1996; Ginsberg and Busto, in press). This is particularly the case in models of focal ischemia produced by selective arterial occlusion, in which differences in rat strain (e.g., spontaneously hypertensive versus normotensive strains) as well as in the site, extent, and technique of vascular occlusion all appear to exert important influences on lesion size, location, and consistency (Coyle, 1986; Brint et al., 1988; Duverger and MacKenzie, 1988; Markgraf et al., 1993).

In recent years, several groups have developed methods of transient or permanent middle cerebral artery occlusion (MCAO) in rodents by the insertion of an intraluminal suture filament into the artery (Koizumi et al., 1986; Nagasawa and Kogure, 1989; Zea Longa et al., 1989). These models have been widely applied in the rat to study a variety of ischemia-related problems, such as the time-course of neuronal changes (Bartus et al., 1995; Garcia et al., 1995), neurological deficit (Garcia et al., 1995), local cerebral blood flow (Nagasawa and Kogure, 1989; Memezawa et al., 1992a, 1992b), tissue energy state (Folbergrová et al., 1992), gene expression (Kinouchi et al., 1994; Kokaia et al., 1995; Iadecola et al., 1996), apoptosis (Li et al., 1995; Chopp et al., 1996), ischemic depolarizations (Chen et al., 1993; Röther et al., 1996), hypothermic (Chen et al., 1992) and pharmacologic (Belayev et al., 1995) neuroprotection, and magnetic resonance imaging (Mintorovitch et al., 1991; Minematsu et al., 1992, Kohno et al., 1995; Röther et al., 1996). The adaptation of MCA suture-occlusion to the mouse has permitted studies in transgenic and gene-knockout strains (Kamii et al., 1994; Hara et al., 1996; Huang et al., 1996).

In previous studies of MCA suture-occlusion, success in achieving infarction has been variable (93%, Koizumi et al., 1986; and 56%, Zea Longa et al., 1989). Subarachnoid hemorrhage may occur because of perforation of intracranial arteries (Laing et al., 1993); high mortality has been reported in some studies (Nagasawa and Kogure, 1989); and differences in total infarct volume may arise when similarly-sized sutures from different manufacturers are used (Kuge et al., 1995). To deal with these methodological problems, we modified the method of Zea Longa et al. (1989) by coating the suture with poly-L-lysine, a polycationic molecule thought to increase adhesion of the suture to the endothelial surface of the occluded vessel (Belayev et al., 1996a; Zhao et al., 1996). In validation studies, we have reported a low coefficient of variation of infarct size with this technique (Belayev et al., 1996a), and we have applied the method to study blood-brain barrier function (Belayev et al., 1996b) and the effect of pharmacotherapeutic agents (Belayev et al., 1995).

The high reproducibility of this MCA suture-occlusion model lends it ideally to the precise study of the interrelationships between local hemodynamic and metabolic events in the focally ischemic brain by permitting the application of sophisticated image-processing strategies by means of which image data sets derived from LCBF and LCMRglc studies in separate groups of replicate animals may be coaligned, mapped into a common atlas-derived template, averaged to form mean data sets, and coanalyzed via pixel-based methods to permit statistical comparison of multimodal data. The feasibility of this approach has been documented in previous publications (Zhao et al., 1993, 1995; Belayev et al., 1995; Prado et al., 1996; Ginsberg et al., 1997), and has been applied to analyze metabolism-blood flow interrelationships and histopathology in a model of photothrombotic distal MCAO (Back et al., 1995; Alexis et al., 1996).

In the current study, we studied LCBF, LCMRglc, and their interrelationship both during a 2-hour period of MCAO by intraluminal suture as well as after 1 hour of postischemic recirculation. The companion study (Zhao et al., 1997), in turn, reports correlations of these data with histopathology using similar pixel-based image-analysis strategies.

MATERIALS AND METHODS

Studies were performed on 20 adult male Sprague-Dawley rats, weighing 320 to 380g, obtained from Charles River Laboratories, Inc, (Wilmington, MA, U.S.A.). All animals underwent right MCAO by insertion of an intraluminal suture. The rats were assigned in a randomized manner to experiments using the autoradiographic measurement of either LCBF or LCMRglc.

Surgical preparation

Animals were fasted overnight but were allowed free access to water. Atropine sulfate (0.5 mg/kg intraperitoneally) was injected 10 minutes before anesthesia. Anesthesia was induced with 3.5% halothane in a mixture of 70% nitrous oxide and a balance of oxygen. Rats were orally intubated and mechanically ventilated. During ventilation, the animals were immobilized with pancuronium bromide (0.6 mg/kg intravenously). Temperature probes were inserted into the rectum (CMA/150 temperature controller, CMA/Microdialysis AB, Stockholm, Sweden) and left temporalis muscle (Hypo-33-1. TG SMP-M, Omega Engineering, Stanford, CT, U.S.A.), and separate heating lamps were used to maintain rectal and cranial temperatures at 37 to 37.5°C. Polyethylene catheters were introduced into the right femoral artery and vein for blood pressure recording and blood sampling. Mean arterial blood pressure was measured via an indwelling femoral arterial catheter connected to a precalibrated pressure transducer (Model P23XL, Gould instrument Systems, Valley View, OH, U.S.A.) and was recorded continuously (Gould Model RS3400 polygraph). Serial measurements (15 minutes before MCAO, 15 minutes and 90 minutes after onset of MCAO, and 30 minutes after the onset of reperfusion) were made of arterial blood gases and pH (Model ABL50, Radiometer America, Inc., Westlake, OH, U.S.A.) and plasma glucose (Model 2300 Stat, Yellow Springs Instrument Co, Inc., Yellow Springs, OH, U.S.A.) (Table 1).

TABLE 1.

Physiological variables in rats subjected to MCAO

	Groups
	1A*LCMRgl(n = 5)	1B*LCBF(n = 5)	2A†LCMRglc(n = 5)	2B†LCBF(n = 5)
Cranial temperature (°C)	37.2 ± 0.1	37.2 ± 0.1	37.1 ± 0.1	37.1 ± 0.1
Rectal temperature (°C)	37.4 ± 0.2	37.4 ± 0.1	37.3 ± 0.1	37.3 ± 0.1
pH	7.40 ± 0.05	7.41 ± 0.03	7.41 ± 0.02	7.41 ± 0.03
Po₂, mm Hg	120 ± 16	120 ± 25	123 ± 18	110 ± 8
Pco₂, mm Hg	39.3 ± 1.8	39.1 ± 1.8	38.2 ± 2.2	37.9 ± 1.8
MABP, mm Hg	114 ± 15	111 ± 12	119 ± 13	116 ± 13
Plasma glucose, mg/dL	129 ± 18	125 ± 26	126 ± 13	116 ± 18

Values are mean ± SD. LCMRglc, local cerebral glucose utilization; LCBF, local cerebral blood flow; MABP, mean cerebral blood pressure; MCAO, middle cerebral artery occlusion.

Average of measurements at 15 minutes pre-MCAO and 90 minutes of MCAO.

†

Average of measurements at 15 minutes pre-MCAO, 90 minutes of MCAO, and 30 minutes of reperfusion.

Middle cerebral artery occlusion

Middle cerebral artery occlusion was induced as described by Zea Longa et al. (1989). Under an operating microscope, the right common carotid artery was exposed through a midline neck incision and was carefully dissected free from surrounding nerves and fascia from its bifurcation to the base of the skull. The occipital artery branches of the external carotid artery were then isolated, and these branches were dissected and coagulated. The internal carotid artery was isolated and carefully separated from the adjacent vagus nerve, and the pterygopalatine artery was ligated. Next, a 4-cm length of 3-0 monofilament nylon suture (Harvard Apparatus, Inc., South Natick, MA, U.S.A.) was inserted via the proximal external carotid artery into the internal carotid artery and then into the circle of Willis, effectively occluding the MCA.

To produce MCAO, we used a poly-L-lysine coated suture, as previously described (Belayev et al., 1996a). Before use, the tip of the suture was blunted by heating near a flame. A 20-mm distal segment of the suture was then coated with poly-L-lysine solution (0.1% [wt/vol] in deionized water, [Sigma, St. Louis, MO, U.S.A.]) and dried in a 60°C oven for 1 hour. The suture was inserted 19 to 20 mm from the bifurcation of the common carotid artery, according to the animal's body weight. After the intraluminal suture was placed, the neck incision was closed with a silk suture. The animals were then allowed to awaken from anesthesia and were returned to their cages.

Two animal groups, each consisting of LCBF and LCMRglc subgroups, were studied. Rats in subgroups 1A (LCBF) and 1B (LCMRglc) were subjected to a 2-hour period of MCAO and were studied without reperfusion. In subgroups 2A (LCBF) and 2B (LCMRglc), rats were reanesthetized after 2 hours of MCAO with the same anesthetic combination as used initially. Temperature probes were reinserted, and the intraluminal suture was carefully removed. Autoradiographic studies were performed after ~1 hour of recirculation. Each of these 4 subgroups consisted of n = 5 rats.

Neurobehavioral testing

Behavioral testing based on our previously published neurological function scale (Belayev et al., 1996a) was performed in all 20 rats before MCAO and at 60 minutes of MCAO by an investigator (L.B.) who was blinded to the experimental groups; for this testing, rats were allowed to awaken from anesthesia at 15 minutes after suture insertion. Any rats showing convulsions or sustained disturbances of consciousness were excluded from the study. In our experience, most of the latter have proven to have subarachnoid hemorrhage secondary to suture-induced rupture of the internal carotid artery. In the current study, only 1 animal (from group 2A) needed to be excluded for this reason.

Autoradiographic studies of local cerebral glucose utilization and blood flow

In both Group 1 (~2 hours of MCAO) and Group 2 (2 hours of MCAO plus ~1 hour of recirculation), rats were randomly assigned to either LCBF or LCMRglc subgroups (n = 5 each). To measure LCBF in group 1 A, ¹⁴C-labeled iodoantipyrine was injected at 1.75 h after MCAo. In group 2A, ¹⁴C-labeled iodoantipyrine was injected at 1 hour of postischemic recirculation. In each subgroup, 20 μCi of the radioactive tracer (specific activity 40 μCi/mmol, New England Nuclear, Boston, MA, U.S.A.) dissolved in 1 mL of isotonic saline was infused intravenously at a constant rate over 45 seconds via a Harvard infusion pump. Arterial blood samples were taken at approximately 2-second intervals from a freely flowing femoral arterial line. In two other subgroups, LCMRglc was measured with ¹⁴C-2-deoxyglucose (2-DG). In subgroup 1B, an intravenous bolus of 20 μCi of 2-DG (specific activity 40 to 45 μCi/mmol, New England Nuclear) dissolved in 0.1 mL saline was administered at 1.5 hours after the onset of MCAO. In subgroup 2B, the MCA suture was removed after 2 hours of MCAO, and 2-DG was administered 45 minutes later. Arterial blood samples were taken every 10 seconds during the first minute, at 1 minute intervals for 5 minutes, and at 2.5- to 5-minute intervals for the remainder of the 45-minute study period. Plasma samples were assayed for 2-DG by liquid scintillation counting and for plasma glucose by means of a glucose analyzer (Yellow Springs Instrument Co.). As the temporal centroid of 2-DG uptake into the brain occurs approximately 15 to 20 minutes after its bolus administration, these time-points (i.e., 1.75 hours after MCAO, and 1 hour of recirculation, respectively) corresponded to the times of the LCBF study in subgroups 1A and 2A, respectively, thus permitting more valid comparisons between the LCMRglc and LCBF subgroups.

Autoradiographic studies were terminated by decapitation. Brains were quickly removed (<1.5 minutes) and frozen over liquid nitrogen. They were subsequently embedded and sectioned subserially in a cryostat (20-μm thickness, 100-μm intervals) as previously described (Zhao et al., 1995). These sections, together with calibrated ¹⁴C-methylmethacrylate standards, were exposed to Kodak Hyperfilm Beta-max film for 10 days. Films were digitized by means of a Xillix CCD camera (70 μm/pixel resolution) interfaced to an MCID image analysis system (Imaging Research, Inc., St. Catherines, Ontario, Canada). Image files were transferred to a clustered DEC-Alpha minicomputer system (Digital Equipment Corp., Nashua, NH, U.S.A.) for processing. Operational equations of the iodoantipyrine method (Sakurada et al., 1978) and the 2-DG method modified for variable plasma glucose (Sokoloff et al., 1977; Savaki et al., 1980) were used to compute LCBF and LCMRglc, respectively. We assumed normal values for the kinetic constants and lumped constant of the 2-DG model (Sokoloff et al., 1977), as we have previously shown, using the ¹⁴C-methylglucose method (Dienel et al., 1991; Greenberg et al., 1992), that even in the acute focal-ischemic penumbra, in which LCBF is reduced to 20% to 40% of control, the brain glucose space, and hence the lumped constant, are unaltered (Yao et al., 1995). Because this is not the case at LCBF values <20% of control, we have excluded the ischemic core from our analysis of the LCMRglc/LCBF ratio (see Results).

Generation of averaged three-dimensional autoradiographic image data sets

Three-dimensional autoradiographic image reconstruction was based on a novel image alignment algorithm which we have previously published in detail (Zhao et al., 1993, 1995) and have implemented in studies of brain ischemia (Back et al., 1995; Prado et al., 1996) and traumatic brain injury (Ginsberg et al., 1997). In this method, a linear affine transformation was first used to register sequential coronal sections of each brain, and transformation parameters were calculated by point-to-point disparity analysis. Each coronal section was aligned to its adjacent neighbor by applying a reverse transformation with the estimated translation and rotation parameters. After alignment of LCBF or LCMRglc image data sets in individual rats, corresponding coronal sections of all brains of each LCBF series or of each LCMRglc series, respectively, were placed in register with one another at a common coronal reference level (bregma + 0.7 mm, [Paxinos and Watson, 1982]). One brain was designated as the template, and all other sections were mapped into its contours at each coronal level by means of an averaging procedure similar to that used for image alignment. These procedures resulted in quantitative three-dimensional image data sets of average LCBF and LCMRglc for groups 1 and 2, which could each be displayed in pseudo-color on a video terminal. Three-dimensional standard-deviation data sets of LCBF and LCMRglc were generated in a similar manner.

Next an LCMRglc/LCBF ratio image data set was calculated by the following equation:

R_{k} = \frac{1}{N \times M} \sum_{i}^{N} \sum_{j}^{M} \frac{P_{i k}}{Q_{j k}},

(1)

where R_k denotes the mean ratio value at pixel k; P_lk is the LCMRglc reading at the kth pixel from the ith animal in the LCMRglc study (group with a total of N animals; and Q_jk is the LCBF reading at the kth pixel, geometrically corresponding to that of the LCMRglc reading, from the jth animal in the LCBF study group with a total of M animals (Back et al., 1995). Image analysis programs developed in our laboratory allowed every coronal section of each three-dimensional autoradiographic map to be displayed separately, both on the average images and on the individual images from which they were derived.

Autoradiographic region-of-interest analysis was performed at 8 representative coronal levels (ranging from bregma +1.2 to −7.8 mm) by superimposing a digitized atlas based on the atlas of Zilles et al. (1985) (Fig. 1). A polygon tool was used to demarcate multiple cortical and subcortical regions of interest on each digitized atlas section. The atlas templates were then fitted by disparity analysis to the corresponding autoradiographic sections of individual rats of each respective subgroup; automated region-of-interest readings were taken; and averages and standard deviations were computed for each region of interest of the two LCBF or LCMRglc subgroups. Mean values for the LCMRglc/LCBF ratio were obtained from region-of-interest measurements on the averaged ratio data set. (Tables 2 through 4 present the anatomic levels and atlas levels used in this analysis. Some structures were measured at multiple coronal levels [as noted in the “Atlas Level” column of Tables 2 through 4] and an average structure value computed for each rat.)

FIG. 1.

Digitized templates at eight coronal levels (bregma +1.2, −0.3, −1.3, −1.8, −3.8, −4.8, −6.3, and −7.8 mm) based on the atlas of Zilles et al. (1985), used for autoradiographic region-of-interest analysis. These correspond to the atlas levels 0 to 7 referred to in Tables 2 to 4

TABLE 2.

Local cerebral blood flow (mL/g/min)

	During 2 MCAO					1-H Postischemic RecirculationRight

		Right		Left		Right		Left		% diff. (Late-Early)

Structure	Atlas Level	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Right	Left
Cingulate cortex (Cg)	0,1,2	0.963	0.599	1.349	0.842	0.802*	0.130	1.133	0.277	−17	−16
Frontal cortex (Fr1, Fr2)	0,1,2,3	0.625*	0.216	1.236	0.424	0.628	0.264	0.904	0.177	1	−27
Hindlimb area (HL)	2,3	0.492*	0.132	1.242	0.449	0.473*	0.207	0.788	0.100	−4	−37
Forelimb area (FL)	0,1,2	0.338*	0.131	1.278	0.444	0.479*	0.221	0.858	0.132	42	−33
Primary somatosensory cortex (Par1)	0,1,2,3	0.221*	0.156	1.376	0.435	0.433*	0.150	0.870	0.126	96	−37†
Primary somatosensory cortex (Par1)	4	0.222*	0.178	1.163	0.277	0.425*	0.200	0.857	0.174	91	−26
Supplementary somatosensory cortex (Par2)	1,2,3	0.169*	0.188	1.329	0.402	0.412*	0.184	0.858	0.163	143	–35†
Temporal cortex (Te1, Te2, Te3)	4,5	0.220*	0.146	1.228	0.318	0.440*	0.168	0.984	0.405	100	–20
Inferolateral cortex (Gu, AID, AIV, Pir)	0	0.227*	0.108	1.423	0.484	0.448*	0.123	0.929	0.107	97†	–35
Inferolateral cortex (Gu, AID, AIV, AID, Pir)	1,2,3,4	0.131*	0.097	1.063	0.304	0.406*	0.182	0.708	0.140	211†	–33†
Medial occipital cortex (Oc2MM, Oc2ML, Oc1M)	5,6	0.365*	0.239	1.015	0.434	0.620	0.304	0.859	0.200	70	–15
Lateral occipital cortex (Oc1B, Oc2L)	5,6	0.257*	0.170	1.056	0.366	0.473*	0.248	0.957	0.251	84	–9
Posterior occipital cortex (Oc1B, Oc2L)	7	0.267*	0.164	0.934	0.402	0.559	0.319	0.882	0.207	109	–6
Entorhinal cortex (Ent)	5,6	0.312*	0.125	0.933	0.311	0.419*	0.190	0.689	0.114	34	–26
Entorhinal cortex (Ent)	7	0.430*	0.130	0.927	0.361	0.592	0.260	0.852	0.222	38	–8
Dorsolateral caudoputamen (CPu)	0	0.327*	0.065	1.170	0.290	0.514*	0.189	0.812	0.202	57	–31
Dorsolateral caudoputamen (CPu)	1	0.142*	0.118	1.253	0.350	1.254	0.898	0.974	0.330	782†	–22
Ventromedial caudoputamen (CPu)	0	0.668*	0.269	1.543	0.473	0.567*	0.135	0.910	0.209	–15	–41†
Ventromedial caudoputamen (CPu)	1	0.172*	0.068	1.398	0.417	1.039	0.874	0.978	0.248	503	–30
Posterior caudoputamen (CPu)	2,3	0.139*	0.119	1.136	0.309	1.112	0.671	0.809	0.093	698†	–29
Globus pallidus (GP)	2,3	0.115*	0.111	1.152	0.340	1.037	0.727	0.808	0.136	804†	–30
Dorsal hippocampus (CA1, CA2, CA3, CA4, DG)	4	0.544	0.239	0.709	0.225	0.505	0.155	0.641	0.134	–7	–10
Ventral hippocampus (CA1-CA4, DG, S)	5,6	0.498	0.229	0.851	0.302	0.522*	0.171	0.814	0.175	5	–4
Dorsal thalamus	4,5	1.192	0.481	1.270	0.310	0.712	0.159	0.920	0.210	–40	–28
Ventral thalamus	4,5	1.165	0.308	1.114	0.289	0.678	0.186	0.714	0.143	–42	–36†

Different from L-sided value (P < 0.05, repeated-measures ANOVA followed by Bonferroni comparisons).

†

Difference between early and late values (P < 0.05, repeated-measures ANOVA followed by Bonferroni comparisons).

Atlas levels 0-7 correspond to coronal levels of bregma +1.2, −0.3, −1.8, −3.8, −4.8, −6.3, and −7.8 mm, respectively.

TABLE 3.

Local cerebral glucose utilization (μmol/100g/min)

	During 2-H MCAO					After 1-H Postischemic Recirculation

		Right		Left		Right		Left		% diff. (Late-Early)

Structure	Atlas Level	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Right	Left
Cingulate cortex (Cg)	0,1,2	41.1	5.5	47.2	3.1	37.8	17.7	49.4	11.4	–8	5
Frontal cortex (Fr1, Fr2)	0,1,2,3	36.4	10.0	37.8	4.7	26.5*	8.3	41.2	5.8	–27	9
Hindlimb area (HL)	2,3	43.6	15.7	36.6	8.6	22.9*	3.3	38.6	4.5	–48†	5
Forelimb area (FL)	0,1,2	52.3	18.2	36.4	4.8	23.1*	4.2	38.8	6.7	–56†	7
Primary somatosensory cortex (Par1)	0,1,2,3	52.7	14.7	38.1	7.0	22.7*	3.0	39.8	5.6	–57†	4
Primary somatosensory cortex (Par1)	4	67.9	28.9	41.0	11.2	23.3*	2.4	38.8	6.8	–66†	–5
Supplementary somatosensory cortex (Par2)	1,2,3	47.3	24.6	42.3	7.4	19.6*	5.1	41.7	6.8	–59†	–2
Temporal cortex (Te1, Te2, Te3)	4,5	57.7	18.1	48.2	14.1	23.2*	2.1	48.6	9.7	–60†	1
Inferolateral cortex (Gu, AID, AIV, Pir)	0	54.3	10.6	48.8	6.3	26.7*	2.9	47.9	7.0	–51†	–2
Inferolateral cortex (Gu, AID, AIV, AID, Pir)	1,2,3,4	41.4	19.8	41.0	5.3	21.5*	4.2	40.5	6.0	–48	–1
Medial occipital cortex (Oc2MM, Oc2ML, Oc1M)	5,6	32.7	9.5	41.4	10.8	30.0*	5.7	44.6	3.8	–8	8
Lateral occipital cortex (Oc1B, Oc2L)	5,6	46.7	12.8	41.6	12.1	26.1*	6.8	44.2	4.5	–44†	6
Posterior occipital cortex (Oc1B, Oc2L)	7	44.8	12.0	48.5	14.3	29.5*	5.1	45.5	7.5	–34†	–6
Entorhinal cortex (Ent)	5,6	40.5	15.5	36.8	7.8	29.2	10.5	40.2	5.3	–28	9
Entorhinal cortex (Ent)	7	34.3	10.1	47.8	11.2	41.6	11.2	45.5	11.1	21	–5
Dorsolateral caudoputamen (CPu)	0	56.4	8.8	45.4	7.8	26.1*	0.9	46.4	13.1	–54†	2
Dorsolateral caudoputamen (CPu)	1	45.5	16.4	50.7	4.1	27.8*	4.9	54.4	8.3	–39†	7
Ventromedial caudoputamen (CPu)	0	51.0	12.5	47.8	6.0	31.6*	4.2	46.6	8.0	–38†	–3
Ventromedial caudoputamen (CPu)	1	45.1	14.8	50.3	5.1	28.6*	6.4	55.9	8.5	–36	11
Posterior caudoputamen (CPu)	2,3	41.5	18.0	46.9	6.1	24.1*	7.1	51.5	9.0	–42	10
Globus pallidus (GP)	2,3	32.2	26.1	42.9	6.9	21.2*	9.0	50.5	9.0	–34	18
Dorsal hippocampus (CA1, CA2, CA3, CA4, DG)	4	38.5	12.0	37.9	8.7	34.0	3.8	39.6	8.2	–12	4
Ventral hippocampus (CA1-CA4, DG, S)	5,6	35.9	10.7	38.0	10.4	34.5	3.7	39.4	4.5	–4	4
Dorsal thalamus	4,5	38.8	15.9	48.1	11.6	37.2*	4.5	51.6	5.0	–4	7
Ventral thalamus	4,5	33.1	12.2	37.8	10.7	33.2	3.7	36.5	3.8	0	–4

Different from L-sided value (P < 0.05, repeated-measures ANOVA followed by Bonferroni comparisons).

†

Difference between early and late values (P < 0.05, repeated-measures ANOVA followed by Bonferroni comparisons).

Atlas levels 0-7 correspond to coronal levels of bregma +1.2, −0.3, −1.8, −3.8, −4.8, −6.3, and −7.8 mm, respectively.

TABLE 4.

Local cerebral glucose utilization/local cerebral blood flow ratio (μmol/100mL)

	During 2-H MCAO*					After 1-H Postischemic Recirculation†
		Right		Left		Right		Left		Early/Late Multiple
Structure	Atlas Level	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Right‡	Left§
Cingulate cortex (Cg)	0,1,2	56.0	12.2	48.8	18.2	41.1	26.0	41.7	17.1	1.4	1.2
Frontal cortex (Fr1, Fr2)	0,1,2,3	74.8	34.6	38.3	16.6	54.0	20.4	46.3	15.3	1.4	0.8
Hindlimb area (HL)	2,3	110.8	39.3	35.4	11.9	72.1	11.4	49.5	8.2	1.5	0.7
Forelimb area (FL)	0,1,2	185.5	56.9	33.7	6.7	67.8	12.2	53.6	13.5	2.7	0.6
Primary somatosensory cortex (Par1)	0,1,2,3	428.7	178.0	33.5	10.8	63.0	14.2	51.6	10.5	6.8	0.6
Primary somatosensory cortex (Par1)	4	569.1	224.7	37.1	8.7	75.8	16.5	50.9	8.6	7.5	0.7
Supplementary somatosensory cortex (Par2)	1,2,3	645.3	203.7	40.9	25.6	50.4	14.3	49.7	9.8	12.8	0.8
Temporal cortex (Te1, Te2, Te3)	4,5	458.7	124.4	46.5	30.1	66.5	16.6	55.4	12.2	6.9	0.8
Inferolateral cortex (Gu, AID, AIV, Pir)	0	336.2	96.4	44.6	29.4	83.0	18.1	50.7	7.5	4.0	0.9
Inferolateral cortex (Gu, AID, AIV, AID, Pir)	1,2,3,4	514.1	186.8	45.1	21.3	70.8	26.6	52.8	19.2	7.3	0.9
Medial occipital cortex (Oc2MM, Oc2ML, Oc1M)	5,6	133.8	60.0	52.5	37.5	65.7	23.8	66.4	55.0	2.0	0.8
Lateral occipital cortex (Oc1B, Oc2L)	5,6	250.6	64.7	46.0	23.4	78.6	15.7	58.3	39.5	3.2	0.8
Posterior occipital cortex (Oc1B, Oc2L)	7	214.4	50.4	65.8	46.7	83.2	21.3	64.4	34.4	2.6	1.0
Entorhinal cortex (Ent)	5,6	214.3	84.8	44.5	17.9	97.0	22.5	64.8	27.6	2.2	0.7
Entorhinal cortex (Ent)	7	93.3	16.8	58.0	10.6	82.6	16.2	56.9	20.7	1.1	1.0
Dorsolateral caudoputamen (CPu)	0	209.0	73.0	43.7	8.6	72.6	15.9	55.4	8.4	2.9	0.8
Dorsolateral caudoputamen (CPu)	1	520.2	149.6	43.9	6.0	36.6	7.7	59.7	6.9	14.2	0.7
Ventromedial caudoputamen (CPu)	0	108.4	71.8	34.5	4.5	69.2	14.7	54.2	4.7	1.6	0.6
Ventromedial caudoputamen (CPu)	1	397.4	149.9	40.2	6.3	62.5	21.4	62.6	13.4	6.4	0.6
Posterior caudoputamen (CPu)	2,3	581.0	129.4	46.2	7.3	37.8	8.5	65.7	6.6	15.4	0.7
Globus pallidus (GP)	2,3	486.8	111.6	41.5	3.6	45.2	11.7	64.4	4.5	10.8	0.6
Dorsal hippocampus (CA1, CA2, CA3, CA4, DG)	4	100.6	51.6	59.8	6.2	72.1	21.5	69.3	16.3	1.4	0.9
Ventral hippocampus (CA1-CA4, DG, S)	5,6	132.7	85.2	53.2	12.0	88.5	28.5	52.5	9.6	1.5	1.0
Dorsal thalamus	4,5	44.1	18.9	41.3	8.0	55.7	11.8	59.8	5.7	0.8	0.7
Ventral thalamus	4,5	32.2	7.9	40.5	17.1	57.1	10.7	55.5	7.6	0.6	0.7

Right- and left-sided values differ (paired t-test on means, t = 5.81, df = 24, P < 0.0001).

†

Right- and left-sided values differ (paired t-test on means, t = 2.82, df = 24, P = 0.009).

‡

Early and late right-sided values differ (paired t-test on means, t = 5.18, df = 24, P < 0.0001).

Early and late left-sided values differ (paired t-test on means, t = −7.34, df = 24, P < 0.0001).

Atlas levels 0-7 correspond to coronal levels of bregma +1.2, −0.3, −1.8, −3.8, −4.8, −6.3, and −7.8 mm, respectively.

To facilitate data analysis, the ischemic core was defined as those image pixels of the 2 hours of MCAO average LCBF data set in which LCBF was reduced to 20% of control levels or below (i.e., below 0.24 mL/g/min). The ischemic penumbra was defined as those image pixels having LCBF between 20% and 40% of control (i.e., 0.24 to 0.47 mL/g/min). These threshold levels are consistent with the previous literature and with our earlier analysis of LCBF and LCMRglc in a model of photothrombotic distal MCAO (Back et al., 1995).

Statistical analysis

Tabular values are expressed as mean ± SD. Repeated-measures analysis of variance with post-hoc Bonferroni tests (to correct for multiple comparisons) were used to compare left- and right-sided LCBF (Table 2) and LCMRglc values (Table 3) in each group, and to compare corresponding structures of each hemisphere in the 2 hours of MCAO and 1 hour of recirculation groups (SAS general linear models routines, SAS Institute, Cary, NC, U.S.A.). Paired t-tests were used to compare right- versus left-sided structure means and to compare corresponding hemispheres within each subgroup. Differences were considered statistically significant at the P < 0.05 level.

RESULTS

Physiological variables

These are presented in Table 1. No intergroup differences were noted for arterial blood pressure, blood gases, or plasma glucose.

Neurological observations

All rats showed a normal neurological score (= 0) before MCAO, and each rat developed a marked neurological deficit when examined at 60 minutes of MCAO (score in each rat = 9 out of possible 10). In previous studies, this uniformity of neurological deficit has correlated with highly consistent histological infarcts after 3-day survival (Belayev et al., 1996a). The neurological deficits resulting from 2 hours of MCAO in this model are presented in greater detail in the companion article (Zhao et al., 1997).

Local blood flow during middle cerebral artery occlusion and after recirculation

These data are presented in Table 2, and shown in Figs. 2 and 3. Right MCAO for 2 hours induced a gradient of ipsilateral LCBF decrements affecting cortical and basal ganglionic regions (right- versus left-sided LCBF, F(1,8) = 22.1, P = 0.0015, repeated-measures analysis of variance). In this model of proximal MCAO, ischemia was most dense (11% to 19% of the contralateral hemisphere control value of 1.2 mL/g/min) in primary somatosensory cortex, supplementary somatosensory cortex, temporal cortex, and inferolateral neocortical areas; and 10% to 14% of control in regions of the caudoputamen and globus pallidus (Table 2). A lesser intensity of ischemia was present in hindlimb and forelimb areas and in occipital and entorhinal (Ent) cortices (Table 2).

FIG. 2.

Average local cerebral blood flow (LCBF) (upper row) and average local glucose utilization (LCMRglc) (lower row) data sets at four representative coronal levels (bregma +1.2, −1.3, −3.8, and −6.3 mm). (A) LCBF at 2 hours of middle cerebral artery occlusion (MCAO); (B) LCBF at 1 hour of recirculation; (C) difference image, highlighting patterns of postischemic recirculation in relation to antecedent ischemia (where A = 1 hour of recirculation and B = 2 hours of MCAO). (D) LCMRglc at ~2 hours of MCAO; (E) LCMRglc at ~1 hour of recirculation; (F) difference image, depicting the pattern of local metabolic deterioration after reperfusion (where A = ~1 hour of recirculation and B = ~2 hours of MCAO).

FIG. 3.

Thresholded pseudocolor images of the ischemic penumbra at the four coronal levels shown in Fig. 2. The penumbra was defined as those pixels of the 2 hours of MCAO image data set having LCBF values between 20% and 40% of contralateral control (i.e., 0.24 to 0.47 mL/g/min). These pixels, thus defined, were analyzed for their LCBF (columns 1 and 2), LCMRglc (columns 3 and 4), and the LCMRglc/LCBF ratio (columns 5 and 6) during ~2 hours of MCAO (columns 1, 3, and 5) and after ~1 hour of postischemic recirculation (columns 2, 4, and 6). The manner in which the ischemic penumbra surrounds the ischemic core (shown in black) and is concentrated at the frontal and parieto-occipital poles of the ischemic lesion is evident. Penumbral pixels show normal levels of LCMRglc during ischemia (column 3) but moderate-to-severe deterioration of glucose metabolism during recirculation (column 4). Correspondingly, the metabolism/flow ratio is markedly increased during 2 hours of MCAO (column 5), but this subsides after ~ 1 hour of recirculation (column 6).

After 1 hour of postischemic recirculation, LCBF in the previously densely ischemic neocortical regions increased by approximately 40% to 200% above ischemic values but remained, on average, at approximately 0.4 to 0.5 mL/g/min, or approximately 40% of control levels. By contrast, frank hyperemia (reaching 500% to 800% above ischemic values) was noted in the previously ischemic caudoputamen and globus pallidus (Table 2). Interestingly, LCBF in the nonischemic left hemisphere, when measured at 1 hour of recirculation, was depressed to 0.86 ± 0.11 mL/g/min, or 26% below the mean hemisphere value of 1.17 ± 0.20 mL/g/min measured during MCAO (paired t-test on average of structure means in Table 2, P < 0.0001). Figure 2 illustrates quantitative average LCBF images at 4 representative coronal levels, together with a difference image comparing LCBF at these levels before and after recirculation.

A striking correlation was noted between respective mean cortical LCBF values measured during MCAO and after 1 hour of recirculation (R = 0.93) (Fig. 4). This finding suggests that the extent to which LCBF within a previously ischemic cortical region recovered toward normal during early recirculation was largely determined by its depth of antecedent ischemia.

FIG. 4.

Mean LCBF values of cortical structures of the ischemic (right) and nonischemic (left) hemisphere, plotted versus their corresponding LCBF values after early recirculation (data taken from cortical structure means presented in Table 2). Corresponding data at the two time points are highly correlated.

Local glucose utilization during middle cerebral artery occlusion and after recirculation

Despite severe ischemia (Table 2), right-sided LCMRglc measured during ~2 hours MCAO showed, on average, no significant depression below contralateral values. The average of hemisphere structure means (Table 3) was identical for the right and left hemispheres (44.5 ± 8.9 and 43.2 ± 4.9 μmol/100g/min, respectively), and neither right-sided cortical nor basal ganglionic LCMRglc were altered, on average, with respect to contralateral values. (However, in autoradiograms of individual rats, heterogeneous loci of focally elevated or depressed LCMRglc could be identified. As these loci differed topographically from animal to animal, they tended to be obscured by the image-averaging procedure.) By contrast, after ~1 hour of postischemic recirculation, glucose metabolic rate became markedly depressed (on average, 55% of contralateral values) in the previously densely ischemic right cortical areas, and a similar trend was noted for the caudoputamen (Table 3). Right hemisphere LCMRglc values after ~1 hour of recirculation were, on average, 63% of corresponding left-hemisphere values [F(1,8) = 51.8; P = 0.0001] (Table 3). Although left-hemisphere LCMRglc mean values were virtually identical during MCAO and at ~1 hour of recirculation (43.2 ± 4.9 versus 44.7 ± 5.5 μmol/100g/min, respectively), paired t-testing of structure means revealed the former to lie slightly but significantly (P = 0.008) below the latter. Quantitative mean LCMRglc images at representative coronal levels, and difference images comparing LCMRglc before and after recirculation, are shown in Fig. 2.

Linear regression analysis revealed that mean LCBF values for cortical structures measured both during MCAO and after ~1 hour of recirculation were, in each instance, highly correlated to LCMRglc of those respective structures after ~1 hour of recirculation (Fig. 5). These relationships indicate that the postischemic depression of local glucose utilization was largely determined by the depth of antecedent ischemia, and that postischemic LCBF of these structures was depressed pari passu with postischemic LCMRglc.

FIG. 5.

Linear regression analysis of mean cortical LCBF during 2 hours of MCAO (A) and after 1 hour of recirculation (B), in each case plotted versus LCMRglc at ~1 hour of recirculation. In each case, a high degree of correlation is apparent. (Data are taken from the cortical structure means presented in Tables 2 and 3.)

Image analysis of the ischemic core and penumbra

Figure 3 shows threshold images of average LCBF, LCMRglc, and the LCMRglc/LCBF ratio in the ischemic penumbra at four representative levels. Pixel measurements (Table 5) revealed that the ischemic penumbra at 2 hours of MCAO (defined as LCBF 20% to 40% of control) amounted, on average, to 26.4% of the entire right hemisphere, virtually equal to the size of the ischemic core (33.4%). As shown in Fig. 3, the ischemic penumbra was concentrated at the frontal and parieto-occipital poles of the ischemic lesion; in the central portions of the ischemic lesion, the penumbra formed a thin shell around the ischemic core.

TABLE 5.

Pixel-based analysis of ischemic core and penumbra*

	Core		Penumbra
	2h MCAO	1h Recirculation	2h MCAO	1h Recirculation
Volume (% of right hemisphere)	33.4%		26.4%
LCBF	0.15 (0.05)	0.53 (0.33)	0.32 (0.07)	0.47 (0.15)
LCMRglc	—†	18.2 (8.9)	36.0 (18.0)	21.9 (9.6)
CMR/CBF Ratio	—†	62.0 (45.1)	179 (132)	71 (44)

Units: LCBF, mL/g/min; LCMRglc, μmol/100g/min; CMR/CBF ratio, μmol/100mL. Values in parentheses are standard deviations of pixel readings on average data sets.

Defined on basis of 2 hours of MCAO LCBF criteria. Ischemic core and penumbra were defined as those pixels of the average 2h-MCAO data set having LCBF of 0% to 20% of control (0 to 0.24 mL/g/min) and 20% to 40% of control (0.24–0.47 mL/g/min), respectively.

†

LCMRglc and CMR/CBF ratio were not quantitated in the early ischemic core because of uncertainties as to the lumped constant and kinetic constants of the 2DG model at severely reduced LCBF.

The mean CBF value of the ischemic penumbra, averaged over all image pixels, was 0.32 mL/g/min (Table 5). After 1 hour of recirculation, LCBF in these same (former penumbral) pixels had increased on average, to 0.47 mL/g/min. Table 5 and Fig. 3 show that LCMRglc within the early ischemic penumbra (at ~2 hours of MCAO) was well preserved, whereas after ~1 hour of recirculation, penumbral LCMRglc had declined severely (to 22 μmol/100g/min; Table 5). The ratio of local glucose utilization to blood flow (LCMRglc/LCBF) is tabulated in Tables 4 and 5 and is depicted for the penumbra in Fig. 3. In the acute ischemic penumbra (at ~2 hours of MCAO), the LCMRglc/LCBF ratio was markedly elevated (on average, 179 μmol/100 mL, or about four-fold above the mean left-hemisphere value). In marked contrast, after ~1 hour of recirculation, the metabolism/flow uncoupling noted in the penumbra at 2 hours MCAO had almost completely subsided (71 μmol/100mL), but comparison of the global mean ratio for right- versus left-sided structures showed a small but significant persistent right-sided elevation (66.0 ± 15.9 versus 56.5 ± 6.9 μmol/100mL, respectively; P = 0.009; data of Table 4). In addition, a small but highly significant difference was noted between the left-hemisphere LCMRglc/LCBF ratio measured at 2 hours of MCAO (44.6 ± 8.2) versus at 1 hour of recirculation (56.5 ± 6.9 μmol/100mL) (P < 0.0001; Table 4).

Figure 6 shows the relationship between right-sided LCBF and LCMRglc values during 2 hours of MCAO (Fig. 6A) and after recirculation (Fig 6B), for those cortical regions having LCBF values in the core or penumbral ranges during ischemia (0% to 40% of control). During MCAO, this relationship was complex, reflecting normal or elevated LCMRglc in the face of moderate ischemia, i.e., uncoupling (Fig. 6A). By contrast, these same right-sided cortical structures showed a vastly different metabolism/flow relationship after recirculation, which now could be well described by a first-order curve fit (Fig. 6B).

FIG. 6.

Relationship between mean LCBF and LCMRglc during 2 hours of MCAO (A) and after 1 hour of recirculation (B) for those 13 cortical structures having mean LCBF within the core or penumbral range at 2 hours of MCAO (data taken from Tables 2 and 3). The 95% confidence intervals are also plotted. During ischemia, the preservation of near-normal glucose utilization at penumbral flow levels is apparent (a third-order polynomial fit is shown). After recirculation, flow and metabolism are well described by a first-order fit.

DISCUSSION

This study was greatly enabled by the opportunity to combine a highly consistent and technically straightforward model of transient focal ischemia with a robust, pixel-based autoradiographic image-averaging approach that facilitated the comparative analysis of averaged multimodal image data sets obtained under differing physiological conditions (ischemia and recirculation) and permitted extensive topographic and statistical analysis.

Model consistency

The hemodynamic consistency of this model of MCA suture-occlusion is evident in the well-demarcated topography of the early ischemic lesion (Fig. 2) and the acceptable interanimal variability of LCBF within ischemic brain structures (Table 2; mean coefficient of variation for LCBF at 2 hours of MCAO, 35%). We attribute the consistency of the hemodynamic lesion, in part, to our modification of suture preparation by blunting its tip and applying a poly-L-lysine coating to enhance adhesion of the suture to the vascular endothelium (Belayev et al., 1996a). The advantages of this modification include a 100% incidence of infarction after 2 hours of MCAO, predictable size and location of the infarct, and a high degree of interanimal reproducibility of infarct size (coefficient of variation 8%), together with the consistent production of high-grade neurological deficits. This interanimal consistency facilitated the design of the present study, in which data derived from matched animal groups of modest size (n = 5 each) were sufficiently robust to permit detailed quantitative intergroup comparisons.

Memezawa et al. (1992a), using the MCA suture-occlusion method of Koizumi et al. (1986), studied LCBF autoradiographically at 1 and 3 hours of MCAO and reported a gradient of LCBF decrements affecting dorsolateral, lateral, and inferolateral neocortices (approximately 0.6 to 0.7, 0.2 to 0.3, and 0.1 to 0.2 mL/g/min, respectively) and caudoputamen. The highest degrees of ischemia (at 60 minutes of MCAO) were in the somatosensory cortex (7% to 8% of contralateral control) and lateral caudoputamen (4% to 12% of control) (Memezawa et al., 1992a). Nagasawa and Kogure (1989) also reported severe LCBF reductions in lateral caudoputamen (2% of control) and frontoparietal somatosensory cortex (10% of control) with 1 to 3 hours of MCAO. Our LCBF findings at 2 hours MCAO are similar.

Topography of the ischemic lesion: core versus penumbra

A unique feature of this study is that pixel-based analytical methods could be applied to averaged data sets encompassing virtually the entire hemisphere so as to yield precise estimates, both of the LCBF and LCMRglc alterations within multiple atlas-defined anatomic regions (Tables 2 to 4), as well as of overall affected tissue volumes (Table 5). In agreement with our previous observations in a rat model of photothrombotic distal MCAO (Back et al., 1995), the total early hemodynamic lesion (defined as pixels with LCBF < 40% of contralateral values at 2 hours of MCAO) was very sizeable, amounting to nearly 60% of hemispheric pixels at the 8 coronal levels analyzed (Table 5). Remarkably, the ischemic penumbra alone comprised fully 44% of the total hemodynamic lesion. (In our earlier MCAo study (Back et al., 1995), this figure was 54%.) As neuroprotectant pharmacotherapy is thought chiefly to target the ischemic penumbra (Ginsberg, 1995), the substantial size of the early penumbra in this model commends it for use in such studies.

Anatomic areas having ischemic-core CBF values in this study included primary and supplementary somatosensory cortices, temporal and inferolateral cortex, and widespread portions of caudoputamen. Penumbral-level CBF was present in fore- and hindlimb areas, occipital and entorhinal cortex, and portions of caudoputamen, while hippocampus was only marginally ischemic, and thalamic CBF values were in the supraischemic range (Table 2). This CBF distribution is consonant with the consistent forelimb placing deficits observed in this model (more completely analyzed in the comparison article [Zhao et al., 1997]). Thus, this model shows gradients of LCBF within the ischemic range extensively involving both cortical and subcortical sites, in keeping with other models of proximal MCAO (Tamura et al., 1981; Tyson et al., 1984; Takagi et al., 1995).

Physiology of the ischemic penumbra

It is useful to conceptualize the penumbra as a hemodynamically altered zone suspended precariously below the LCBF threshold needed, on the one hand, to support synaptic transmission but above that required, on the other hand, for maintenance of cellular ionic homeostasis (Astrup et al., 1981; Strong et al., 1983; Heiss, 1992; Hossmann, 1994; Obrenovitch, 1995; Back et al., in press). The experimental literature suggests that although the absolute CBF thresholds for failure of synaptic transmission and ionic homeostasis may vary with animal species, anesthesia, and ischemia model, nonetheless when these critical flow levels are expressed as percentages of control, consistent threshold levels emerge— ~40%, and ~12% to 15%, respectively (Morawetz et al., 1978; Tamura et al., 1981; Heiss and Rosner, 1983; Strong et al., 1983; Hossmann, 1994; Back et al., in press).

Although controversy continues as to the time-frame within which penumbral deterioration occurs (Heiss et al., 1992; Ginsberg and Pulsinelli, 1994; Fisher and Garcia, 1996; Furlan et al., 1996; Ginsberg, 1997), it is generally appreciated that the penumbra is temporally and spatially dynamic and metastable (Heiss et al., 1994). As careful correlation of local cortical electrical activity with LCBF during ischemia has shown that both the severity and duration of ischemia determine pathological outcome (Heiss and Rosner, 1983), it is likely that subtle differences in LCBF within the various subregions of the penumbra may eventuate in major differences in histopathological outcome.

Dissociation of glucose metabolism and blood flow in the early penumbra

The present results provide further evidence that a marked dissociation of LCMRglc and LCBF characterizes the early ischemic penumbra, attributable to a preservation, on average, of near-normal glucose metabolic rate (Table 3; Fig. 6) in the face of reduced flow (Table 2), resulting in dramatic elevations of the LCMRglc/LCBF ratio in penumbral pixels (Tables 4 and 5; Figs. 3 and 6). This amounted on average to 179 μmol/100 mL—thus representing an almost four-fold elevation of the coupling ratio compared with normal brain. The extent of metabolism-flow dissociation in this study was similar in degree to that observed in our previous study of photothrombotic distal MCAO (Back et al., 1995). In other studies of MCAO, deoxyglucose autoradiography has yielded the consistent observation that zones of normal or increased glucose metabolism are present within an acute ischemic focus, often just peripheral to core-regions of markedly suppressed metabolism (Ginsberg et al., 1977; Welsh et al., 1980; Hossmann et al., 1985; Tanaka et al., 1985; Nedergaard and Astrup, 1986; Back et al., 1995; Yao et al., 1995; Ginsberg et al., 1996; Back et al., in press).

The similarity of this phenomenon across diverse focal ischemia models thus dispels any suspicion that it is a model-dependent epiphenomenon. Furthermore, the near-disappearance of metabolism-flow uncoupling after ~1 hour of recirculation in the current study (Table 4; Fig. 3) substantiates that it is indeed dependent on critical degrees of blood flow reduction.

Evidence from recent studies suggests that recurrent ischemic depolarizations arising in the acute penumbra constitute the probable stimulus driving the inappropriately increased metabolic activity within the penumbra (Hossmann, 1994; Back et al., in press). Recurrent ischemic depolarizations were observed in our previous study of photothrombotic distal MCAO (Back et al., 1995), and the induction of additional depolarization enhanced both histopathological injury and neurobehavioral deficits in that model (Alexis et al., 1996; Back et al., 1996). In other studies, the suppression of ischemic depolarizations by pharmacological antagonism of N-methyl-D-aspartate receptors (Gill et al., 1992; Iijima et al., 1992) or non-N-methyl-D-aspartate receptors (Mies et al., 1994), or by moderate hypothermia (Chen et al., 1993), substantially reduced the volume of ischemic infarction. Ischemic depolarizations seem to be injurious in that, in contrast to the situation with normal spreading depression, LCBF in ischemic tissue is unable to increase in response to heightened metabolic demand, so that tissue oxygen tension falls episodically (Back et al., 1994) and high-energy stores may become depleted (Takeda et al., 1993).

Blood flow and glucose metabolism during early recirculation

The recovery of LCBF after 2 hours of MCAO was remarkable for its regional heterogeneity. While frank hyperemia was evident in basal ganglia (Fig. 2; Table 2), neocortical areas that had been equally ischemic (e.g., somatosensory, temporal, and inferolateral cortices-Table 2) remained hypoperfused at 1 hour of recirculation. Although the reasons for this remain unclear, it is nonetheless apparent that the postischemic LCBF recovery of the various neocortical regions was highly correlated to their respective levels of LCBF during the antecedent ischemic period (R = 0.093, Fig. 4). This is consistent with the possibility that higher degrees of neocortical ischemia in this model may enhance intravascular sedimentation and sludging of blood elements, offering a rheologic impediment to subsequent microvascular reperfusion (Hossmann, 1993).

Reports of LCBF during recirculation in MCA suture-occlusion models are scanty. In one such study in which the suture method of Koizumi et al. (1986) was applied, LCBF increased to approximately 30% of contralateral levels after 15 minutes of reperfusion, with a high degree of interanimal variability (Memezawa et al., 1992b). Heterogeneous reperfusion was also noted in the study of Nagasawa and Kogure (1989) although their differing study-time points prevent precise comparisons with the current findings.

Local glucose utilization, although heterogeneous, was preserved at near-normal levels during MCAO (Table 3; Figs. 2, 3 and 6A) but tended to decrease by about 40% to 60% in previously ischemic structures at ~1 hour of recirculation, to levels which in previous studies from our laboratory (Yao et al., 1995) have signalled irreversible injury. The high degree of correlation noted between LCBF in cortical structures measured during MCAO and the LCMRglc of those same structures 1 hour later (R = 0.93, Fig. 5B) strongly supports the hypothesis that this early metabolic deterioration is primarily determined by the severity of antecedent ischemia. The companion article (Zhao et al., 1997) presents a detailed pixel-based correlative analysis of LCMRglc and histopathology in this model to support the view these low LCMRglc levels indeed imply irreversible tissue injury. Thus, ischemia of severe degree in this model, if lasting for 2 hours, triggers injury processes that cannot be reversed by partial spontaneous reperfusion alone.

To summarize, by studying LCBF and LCMRglc in matched series at both 2 hours of MCAO and 1 hour of postischemic recirculation in a commonly used model of temporary MCAO, we have established the following: (1) the presence of consistent graded LCBF decrements within cortical and basal ganglionic regions during ischemia; (2) a topographically well-demarcated ischemic penumbra (Fig. 3), characterized by preserved LCMRglc and uncoupling of glucose utilization and blood flow during ~2 hours of MCAO; (3) regionally heterogeneous restoration of LCBF on recirculation, comprising both hyperemic foci within basal ganglia and persistent moderate hypoperfusion of cortical regions; and (4) substantially depressed LCMRglc within the previous ischemic core and penumbra by ~1 hour of recirculation (suggesting nonviability), associated with near-normalization of the LCMRglc/LCBF ratio. The companion study which follows (Zhao et al., 1997) places these findings in pathophysiological perspective by correlating them to patterns of infarctive histopathology in this model.

Footnotes

Acknowledgments

The authors thank Isabel Saúl and Ofelia Alonso for their superb assistance in the autoradiographic studies, and Judith Y. Loor-Estades for her skillful autoradiographic preparation.

Abbreviations used

References

Alexis

Back

Zhao

Dietrich

Watson

Ginsberg

(1996) Neurobehavioral consequences of induced spreading depression following photothrombotic middle cerebral artery occlusion. Brain Res 706:273–282

Astrup

Siesjö

Symon

(1981) Thresholds in cerebral ischemia—the ischemic penumbra. Stroke 12:723–725

Back

Ginsberg

Dietrich

Watson

(1996) Induction of spreading depression in the ischemic hemisphere following experimental middle cerebral artery occlusion: effect on infarct morphology. J Cereb Blood Flow Metab 16:202–213

Back

Kohno

Hossmann

K-A

(1994) Cortical negative DC deflections following middle cerebral artery occlusion and KCI-induced spreading depression: effect on blood flow, tissue oxygenation, and electroencephalogram. J Cereb Blood Flow Metab 14:12–19

Back

Nedergaard

Ginsberg

: The ischemic penumbra: pathophysiology, and relevance of spreading depression-like phenomena. In: Cerebrovascular Disease: Pathophysiology, Diagnosis and Treatment ( Ginsberg

Bougousslavsky

, eds), Blackwell Scientific, Cambridge (in press)

Back

Zhao

Ginsberg

(1995) Three-dimensional image-analysis of brain glucose metabolism/blood flow uncoupling and its electrophysiological correlates in the acute ischemic penumbra following middle cerebral artery occlusion. J Cereb Blood Flow Metab 15:566–577

Bartus

Dean

Cavanaugh

Eveleth

Carriero

Lyeth

(1995) Time-related neuronal changes following middle cerebral artery occlusion: implications for therapeutic intervention and the role of calpain. J Cereb Blood Flow Metab 15:969–979

Belayev

Alonso

Busto

Zhao

Ginsberg

(1996a) Middle cerebral artery occlusion in the rat by intraluminal suture: neurological and pathological evaluation of an improved model. Stroke 27:1616–1623

Belayev

Busto

Zhao

Ginsberg

(1995) HU-211, a novel noncompetitive methyl-D-aspartate antagonist, improves neurological deficit and reduces infarct volume after reversible focal cerebral ischemia in the rat. Stroke 26:2313–2320

10.

Belayev

Busto

Zhao

Ginsberg

(1996b) Quantitative evaluation of blood-brain barrier permeability following middle cerebral artery occlusion in rats. Brain Res 739:88–96

11.

Brint

Jacewicz

Kiessling

Tanabe

Pulsinelli

(1988) Focal brain ischemia in the rat: methods for reproducible neocortical infarction using tandem occlusion of the distal middle cerebral and ipsilateral common carotid arteries. J Cereb Blood flow Metab 8:474–485

12.

Chen

Chopp

Zhang

Garcia

(1992) he effect of hypothermia on transient middle cerebral artery occlusion in the rat. J Cereb Blood Flow Metab 12:621–628

13.

Chen

Chopp

Bodzin

Chen

(1993) Temperature modulation of cerebral depolarization during focal cerebral ischemia in rats: correlation with ischemic injury. J Cereb Blood Flow Metab 13:389–394

14.

Chopp

Jiang

Zhang

Prostak

(1996) Antibodies against adhesion molecules reduce apoptosis after transient middle cerebral artery occlusion in rat brain. J Cereb Blood Flow Metab 16:578–584

15.

Coyle

(1986) Different susceptibilities to cerebral infarction in spontaneously hypertensive (SHR) and normotensive Sprague-Dawley rats. Stroke 17:520–525

16.

Dienel

Cruz

Mori

Holden

Sokoloff

(1991) Direct measurement of the lambda of the lumped constant of the deoxyglucose method in rat brain: determination of lambda and lumped constant from tissue glucose concentration or equilibrium brain/plasma distribution ratio for methylglucose. J Cereb Blood Flow Metab 11:25–34

17.

Duverger

MacKenzie

(1988) The quantification of cerebral infarction following focal ischemia in the rat: influence of strain, arterial pressure, blood glucose concentration, and age. J Cereb Blood Flow Metab 8:449–461

18.

Fisher

Garcia

(1996) Evolving stroke and the ischemic penumbra. Neurology 47:884–888

19.

Folbergrová

Memezawa

Smith

M-L

Siesjö

(1992) Focal and perifocal changes in tissue energy state during middle cerebral artery occlusion in normo- and hyperglycemic rats. J Cereb Blood Flow Metab 12:25–33

20.

Furlan

Marchal

Viader

Derlon

J-M

Baron

J-C

(1996) Spontaneous neurological recovery after stroke and the fate of the ischemic penumbra. Ann Neurol 40:216–226

21.

Garcia

Wagner

Liu

K-F

X-j

(1995) Neurological deficit and extent of neuronal necrosis attributable to middle cerebral artery occlusion in rats: statistical validation. Stroke 26:627–635

22.

Gill

Andine

Hillered

Persson

Hagberg

(1992) The effect of MK-801 on cortical spreading depression in the penumbral zone following focal ischaemia in the rat. J Cereb Blood Flow Metab 12:371–379

23.

Ginsberg

(1995) Neuroprotection in brain ischemia—an update—Parts I and II. Neuroscientist 1:95–103; 164–175

24.

Ginsberg

(1996) The validity of rodent brain-ischemia models is self-evident. Arch Neurol 53:1065–1067

25.

Ginsberg

(1997) Delayed OEF elevation is an unreliable index of tissue at risk [Letter]. Ann Neurol 41:411–412

26.

Ginsberg

Belayev

Zhao

Back

Dietrich

Busto

(1996) The pathophysiological significance of metabolism/blood flow uncoupling in acute ischemic and traumatic brain injury. In: Pharmacology of Cerebral Ischemia—1996 ( Krieglstein

, ed), Medpharm Scientific Publishers, Stuttgart, pp 245–258

27.

Ginsberg

Busto

: Small-animal models of global and focal cerebral ischemia. In: Cerebrovascular Disease: Pathophysiology, Diagnosis and Treatment ( Ginsberg

Bougousslavsky

, eds), Cambridge, Blackwell Scientific Publications, (in press)

28.

Ginsberg

Pulsinelli

(1994) The ischemic penumbra, injury thresholds, and the therapeutic window for acute stroke [Editorial]. Ann Neurol 36:553–554

29.

Ginsberg

Reivich

Giandomenico

Greenberg

(1977) Local glucose utilization in acute focal cerebral ischemia: local glucose utilization in acute focal cerebral ischemia: local dys-metabolism and diaschisis. Neurology 27:1042–1048

30.

Ginsberg

Zhao

Alonso

Loor-Estades

Dietrich

Busto

(1997) Uncoupling of local cerebral glucose metabolism and blood flow following acute fluid-percussion injury in rats. Am J Physiol 272:H2859–H2868

31.

Greenberg

Hamar

Welsh

Harris

Reivich

(1992) Effect of ischemia and reperfusion on lambda of the lumped constant of the [¹⁴C]deoxyglucose technique. J Cereb Blood Flow Metab 12:70–77

32.

Hara

Huang

Panahian

Fishman

Moskowitz

(1996) Reduced brain edema and infarction volume in mice lacking the neuronal isoform of nitric oxide synthase after transient MCA occlusion. J Cereb Blood Flow Metab 16:605–611

33.

Heiss

W-D

(1992) Experimental evidence of ischemic thresholds and functional recovery. Stroke 23:1668–1672

34.

Heiss

W-D

Graf

Wienhard

Lottgen

Saito

Fujita

Rosner

Wagner

(1994) Dynamic penumbra demonstrated by sequential multitracer PET after middle cerebral artery occlusion in cats. J Cereb Blood Flow Metab 14:892–902

35.

Heiss

W-D

Huber

Fink

Herholz

Pietrzyk

Wagner

Wienhard

(1992) Progressive derangement of periinfarct viable tissue in ischemic stroke. J Cereb Blood Flow Metab 12:193–203

36.

Heiss

W-D

Rosner

(1983) Functional recovery of cortical neurons as related to degree and duration of ischemia. Ann Neurol 14:294–301

37.

Hossmann

K-A

(1993) Ischemia-related neuronal injury (review). Resuscitation 26:225–235

38.

Hossmann

K-A

(1994) Neurological progress: Viability thresholds and the penumbra of focal ischemia. Ann Neurol 36:557–565

39.

Hossmann

K-A

Mies

Paschen

Csiba

Bodsch

Rapin

Le Poncin-Lafitte

Takahashi

(1985) Multiparametric imaging of blood flow and metabolism after middle cerebral artery occlusion in cats. J Cereb Blood Flow Metab 5:97–107

40.

Huang

Meng

Ayata

Fishman

Moskowitz

(1996) Enlarged infarcts in endothelial nitric oxide synthase knockout mice are attenuated by nitro-L-arginine. J Cereb Blood Flow Metab 16:981–987

41.

Iadecola

Zhang

Casey

Clark

Ross

(1996) inducible nitric oxide synthase gene expression in vascular cells after transient focal cerebral ischemia. Stroke 27:1373–1380

42.

Iijima

Mies

Hossmann

K-A

(1992) Repeated negative DC deflections in rat cortex following middle cerebral artery occlusion are abolished by MK-801: effect on volume of ischemic injury. J Cereb Blood Flow Metab 12:727–733

43.

Kamii

Kinouchi

Sharp

Koistinaho

Epstein

Chan

(1994) Prolonged expression of hsp70 mRNA following transient focal cerebral ischemia in transgenic mice overexpressing CuZn-superoxide dismutase. J Cereb Blood Flow Metab 14:478–486

44.

Kinouchi

Sharp

Chan

Koistinaho

Sagar

Yoshimoto

(1994) Induction of c-fos, junB, c-jun and hsp70 mRNA in cortex, thalamus, basal ganglia and hippocampus following middle cerebral artery occlusion. J Cereb Blood Flow Metab 14:808–817

45.

Kohno

Back

Hoehn-Berlage

Hossman

K-A

(1995) Modified rat model of middle cerebral artery thread occlusion under electrophysiological control for magnetic resonance investigations: description of technique. Magn Reson Imag 13:65–71

46.

Koizumi

Yoshida

Nakazawa

Ooneda

(1986) Experimental studies of ischemic brain edema. 1. A new experimental model of cerebral embolism in rats in which recirculation can be introduced in the ischemic area. Jpn J Stroke 8:1–8

47.

Kokaia

Zhao

Kokaia

Elmer

Metsis

Smith

M-L

Siesjö

Lindwall

(1995) Regulation of brain-derived neurotrophic factor gene expression after transient middle cerebral artery occlusion with and without brain damage. Exp Neurol 136:73–88

48.

Kuge

Minematsu

Yamaguchi

Miyake

(1995) Nylon monofilament for intraluminal middle cerebral artery occlusion in rats. Stroke 26:1655–1658

49.

Laing

Jakubowski

Laing

(1993) Middle cerebral artery occlusion without craniectomy in rats: which method works best? Stroke 24:294–298

50.

Chopp

Jiang

Yao

Zaloga

(1995) Temporal profile of in situ DNA fragmentation after transient middle cerebral artery occlusion in the rat. J Cereb Blood Flow Metab 15:389–397

51.

Markgraf

Kraydieh

Prado

Watson

Dietrich

Ginsberg

(1993) Comparative histopathologic consequences of photothrombotic occlusion of the distal middle cerebral artery in Sprague-Dawley and Wistar rats. Stroke 24:286–292

52.

Memezawa

Minamisawa

Smith

M-L

Siesjö

(1992a) Ischemic penumbra in a model of reversible middle cerebral artery occlusion in the rat. Exp Brain Res 89:67–78

53.

Memezawa

Smith

M-L

Siesjö

(1992b) Penumbral tissues salvaged by reperfusion following middle cerebral artery occlusion in rats. Stroke 23:552–559

54.

Mies

Kohno

Hossmann

K-A

(1994) Prevention of periinfarct direct current shifts with glutamate antagonist NBQX following occlusion of the middle cerebral artery in the rat. J Cereb Blood Flow Metab 14:802–807

55.

Minematsu

Fisher

Sotak

Davis

Fiandaca

(1992) Diffusion-weighted magnetic resonance imaging: Rapid and quantitative detection of focal brain ischemia. Neurology 42:235–240

56.

Mintorovitch

Moseley

Chileuitt

Shimizu

Cohen

Weinstein

(1991) Comparison of diffusion- and T2-weighted MRI for the early detection of cerebral ischemia and reperfusion in rat. Magn Res Med 18:39–50

57.

Morawetz

DeGirolami

Ojemann

Marcoux

Crowell

(1978) Cerebral blood flow determined by hydrogen clearance during middle cerebral artery occlusion in unanesthetized monkeys. Stroke 9:143–149

58.

Nagasawa

Kogure

(1989) Correlation between cerebral blood flow and histologic changes in a new rat model of middle cerebral artery occlusion. Stroke 20:1037–1043

59.

Nedergaard

Astrup

(1986) Infarct rim: effect of hyperglycemia on direct current potential and (¹⁴C)2-deoxyglucose phosphorylation. J Cereb Blood Flow Metab 6:607–615

60.

Obrenovitch

(1995) The ischaemic penumbra: twenty years on. Cerebrovasc Brain Metab Rev 7:297–323

61.

Paxinos

Watson

(1982) The Rat Brain in Stereotaxic Coordinates. Academic Press, Sydney

62.

Prado

Watson

Zhao

Yao

Busto

Dietrich

Ginsberg

(1996) L-Arginine does not improve cortical perfusion or histopathological outcome in spontaneously hypertensive rats subjected to distal middle cerebral artery photothrombotic occlusion. J Cereb Blood Flow Metab 16:612–622

63.

Röther

deCrespigny

D'Arceuil

Moseley

(1996) MR detection of cortical spreading depression immediately after focal ischemia in the rat. J Cereb Blood Flow Metab 16:214–220

64.

Sakurada

Kennedy

Jehie

Brown

Carbin

Sokoloff

(1978) Measurement of local cerebral blood flow with iodo(¹⁴C)antipyrine. Am J Physiol 234:H59–H66

65.

Savaki

Davidsen

Smith

Sokoloff

(1980) Measurement of free glucose turnover in brain. J Neurochem 35:495–502

66.

Sokoloff

Relvich

Kennedy

DesRosiers

Patlak

Pettigrew

Sakurada

Shinohara

(1977) The (¹⁴C)deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 28:897–916

67.

Strong

Venables

Gibson

(1983) The cortical ischaemic penumbra associated with occlusion of the middle cerebral artery in the cat: 1. Topography of changes in blood flow, potassium ion activity, and EEG. J Cereb Blood Flow Metab 3:86–96

68.

Takagi

Zhao

Busto

Ginsberg

(1995) Local hemodynamic changes during transient middle cerebral artery occlusion and recirculation in the rat: a (¹⁴C)iodoantipyrine autoradiographic study. Brain Res 691:160–168

69.

Takeda

Jacewicz

Takeda

Nowak

Jr Pulsinelli

(1993) DC-potential and energy metabolites in the focal ischemia. J Cereb Blood Flow Metab 13 Suppl 1:S450

70.

Tamura

Graham

McCulloch

Teasdale

(1981) Focal cerebral ischemia in the rat: 2. Regional cerebral blood flow determined by (¹⁴C)iodoantipyrine autoradiography following middle cerebral artery occlusion. J Cereb Blood Flow Metab 1:61–69

71.

Tanaka

Greenberg

Gonatas

Reivich

(1985) Regional flow-metabolism couple following middle cerebral artery occlusion in cats. J Cereb Blood Flow Metab 5:241–252

72.

Tyson

Teasdale

Graham

McCulloch

(1984) Focal cerebral ischemia in the rat: topography of hemodynamic and histopathological changes. Ann Neurol 15:559–567

73.

Welsh

Greenberg

Jones

Ginsberg

Reivich

(1980) Correlation between glucose utilization and metabolite levels during focal ischemia in cat brain. Stroke 11:79–84

74.

Yao

Ginsberg

Eveleth

LaManna

Watson

Alonso

Loor

Busto

(1995) Local cerebral glucose utilization and cytoskeletal proteolysis as indices of evolving focal ischemic injury in core and penumbra. J Cereb Blood Flow Metab 15:398–408

75.

Zea Longa

Weinstein

Carlson

Cummins

(1989) Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke 20:84–91

76.

Zhao

Belayev

Ginsberg

(1997) Transient middle cerebral artery occlusion by intraluminal suture: II. Neurological deficits, and pixel-based correlation of histopathology with local blood flow and glucose utilization. J Cereb Blood Flow Metab 17:1281–1290

77.

Zhao

Ginsberg

Prado

Belayev

(1996) Depiction of infarct frequency distribution by computer-assisted image-mapping in rat brains with middle cerebral artery occlusion: comparison of photothrombotic and intraluminal suture models. Stroke 27:1112–1117

78.

Zhao

Ginsberg

Smith

(1995) Three-dimensional quantitative autoradiography by disparity analysis: theory and application to image-averaging of local cerebral glucose utilization. J Cereb Blood Flow Metab 15:552–565

79.

Zhao

Young

Ginsberg

(1993) Registration and three-dimensional reconstruction of autoradiographic images by the disparity analysis method. IEEE Trans Med Imag 12:782–791

80.

Zilles

(1985) The Cortex of the Rat. New York, Springer-Verlag