ORIGINAL 
Annals of Nuclear Medicine Vol. 7, No. 3, 167-171, 1993 

Evaluation of critically perfused area in acute ischemic stroke for therapeutic reperfusion : A clinical PET study 

Shuichi HIGANO,* Kazuo UEMURA,** Fumio SHISHIDO,*** Iwao KANNO ** Noriaki TOMURA**** and Kiyohiko SAKAMOTO* 

*Department of Radiology, Tohoku University School of Medicine, Sendai, Japan * * Department of Radiology and Nuclear Medicine, Research Institute for Brain & Blood Vessels-AKITA. Akita, Japan * * * Division of Clinical research, National Institute of Radiological Science. Chiba. Japan ****Department of Radiology, Akita University School of Medicine. Akita. Japan 

To evaluate critically perfused areas in the acute ischemic brain, 9 patients were studied by positron emission tomography (PET) within 7-32 hours after the onset. The cerebral blood flow (CBF) and oxygen metabolic rate (CMR02) were evaluated and compared with se-quentia] change in CT findings. In all the regions developing subsequent necrosis on CT, CBF dropped below 17 ml/100 g/min. But in some of these lesions, CMR02 remained above the minimum value for regions in which infarction did not develop, and the tissue density on CT obviously remained normal for several hours after PET scan. The mean CBF in these lesions (14.0 ml/100 g/min, range : 9.9-17,3 ml/100 g/min)_ was significantly higher than that in ischemic areas with low density on CT before or just after PET study (.-10 ml/100 g/min, range: 7.7-14.1 m//100 g/min). These findings suggest that a part of the tissue with CBF between l0-17 m//100 g/min is still viable at least 7 hours after the onset of ischemia, but becomes non-viable in a longer period of ischemia. These lesions should respond to effective treatment, including therapeutic reperfusion. 

Key words : cerebral blood fiow, ischemic threshold, cerebral infarction, positron emrsbron tomography

INTRODUCTION 

WITH THE REC.ENT DEVELOPMENT Of therapeutic inter-vention and thrombolytic agents such as a tissue-type plasminogen activator, thrombo-embolysis has ag. ain seemed to be a feasible therapy in acute ischemic stroke.1-3 Indeed, there were some reports that spontaneous recanalization or thrombolytic therapy in the acute period of ischenvia produced complete recovery or an improved neurological state.2,4 How-ever, inappropriate reperfusion after ischemia may exacerbate brain edema or carry an increased risk of hemorrhagic infarction.5,6 Patients who are good candidates for the therapy should be selected to prevent worsening of the patient's condition as much as possible. Although many studies of cerebral blood fiow (CBF) subjected to ischemic threshold for infarction in animals and humans have demon-strated that the threshold varies depending on the magnitude of CBF reduction and the duration of ischemic insult,7-12 there are only a few clinical PET data on ischemic stroke in the early stages. For this reason, we have reevaluated critically perfused areas in patients with ischemic stroke in its acute phase from the standpoint of CBF and the cerebral oxygen metabolic rate (CMR02) by means of positron emission tomography (PET). 

SUBJECTS AND METHODS 
Patient selection 
Nine patients with acute ischemic stroke were re-trospectively selected criteria. according to the ollowing 
a) Patients presented with neurological deficit of sudden onset which corresponded to the carotid territory. 
b) Patients have no history of cerebral infarction before this insult. 
c)_ PET measurements of CBF and CMR02 were completed within 48 hours after the onset of sym ptoms. 
d) Final distribution of infarction on CT was re-latively large (.more than 2 cm in longer dia-meter) with involvement of cerebral cortices. 
e) No thrombolytic therapy was performed. 

The clinical and neuroradiological data for the selected cases are summarized in Table I . The patients were 7 males and 2 females, aged from 59 to 76 years old (.mean+-SD : 68+-5.3 years old). PET studies were performed from 7 to 32 hours after the onset (Seven of the 9 patients were studied within 24 hours). Cerebral angiography was performed in all the patients between O and 10 days after the onset. In 2 patients (cases nos. 6 & 7) with no occlusive lesion on angiogram, spontaneous recanalization must have occurred before the examination. Informed consent for a PET study was obtained either from each patient or his/her relative before the study in all the cases. 

Data Acquisition 
The regional CBF, regional oxygen extraction frac-tion (OEF) and regional CMR02 were measured by PET, using HEADTOME 11113 and the 150 labeled gas steady state inhalation method.14 Regional cerebral blood volume (CBV) was obtained by one-minute inhalation of C150 gas. Overestimation of OEF and CMRO 2 due to intravascular non-extracted 150 radioactivity was corrected for CBV, based on the method by Lammertsma.15 Before measuring these physiological variables, a transmission scan was performed for attenuation correction of positron annihilation gamma-rays. Arterial blood sampling for measurement of the 150 concentration in whole blood and plasma was obtained from the radial artery. 
The effective transaxial resolution of HEAD-tome 111 is 10 mm of full width at half maximum (FWHM), and the axial resolution is 12mm of FWHM.13 Five tomographic slices were simul-taneously obtained with 15 mm spacing from 20 mm above the orbito-meatal plane. To accurately adjust the anatomical structures of the PET images to those of the CT images, and to immobi]ize the head during scanning, we used a thermoplastic face mask with landmarks molded for each atient.16,17 
Serial CT scans were carried out by a GE 9800 scanner (General Electric, Mi]waukee, WI, USA) to evaluate the sequential change in findings. An initial emergent CT scan was carried out at 10 mm thickness with 10mm intervals about 15 degrees cranial to the orbito-meatal plane according to our routine CT study . I_n all the patients, CT with planes cor-responding to PET images were also obtained using the thermoplastic fac,e mas k before or after the PET measurement. Intervals between the last CT before the PET and the first one after the PET for cases nos. 7, 8 and 9 were relatively long (Table l), but in these cases, distribution of low density areas noted before PET studies was similar to the final distribu-tion of infarcts. 

Data analysis 
Four physiological values, i.e., CBF, CMR02, OEF and CBV were obtained simultaneously by means of the image analysis system (VAX-II/750 and DeAnza IP 8500) by setting regions of interest (ROIs) in the reference PET images. Based on the hypo-thesis that ischemic vulnerability of the middle cerebral artery (MCA) territory was uniform, about 8 round or elliptic ROIs (12xl2mm-14x40 mm in diameter) were selected in cortical zones of the MCA territory in an affected hemisphere in every patient. These ROIs were located several millimeters deep to the brain surface to minimize partial volume averaging of cerebrospinal fluid spaces. So each region include subcortical white matter as well as gray. On the basis of the sequential changes in serial CT findings, the regions were divided into following four groups ; 

In judging the CT findings, we took into account not only the decrease in tissue density as 'low density' but also effacement of the cortical sulci and ob-scuration of the corticomedullary junction.18 To decrease the partial volume effect from surrounding tissues. we discarded the data from the ROI defined in the border of these categories. Statistical difference was examined by one-way analysis of variance with Scheffe's ranging test for multiple comparison, or by Kruskall-Wallis non-parametric analysis. An a of 0.05 was used to determine statistical significance. 

RESULTS 

Figure I shows the relationship between CBF and CMR02 for every ROI in the groups formed on the basis of the CT findings. In nine LD-B regions in the two patients without an occlusion on the angiogram (cases nos. 6 & 7), CMR02 was lower against the higher blood fiow. The OEF values for these regions were all below 0,30 (mean+-SD : 0.21 +-0.06). As the mean normal OEF value obtained in our previous PET study was 0,42+-0.05 (mean+-SD). ,17 these LD-B regions are considered to be in a state of luxury perfusion. 19 The regions with luxury perfusion did not include the LD-S and LD-L groups. The oxygen metabolic rate for all the regions in which normodensity was obviously maintained on CT during PET measurement (ND & LD-L) was above 1 .6 ml/100 g/min. Except for LD-B areas with luxury perfusion, CBF in the regions of low density or becoming low in density (LD-B. LD-S and LD-L) was below 17 ml/100 g/min. 
Table 2 shows the mean values and ranges of CBF, CMR02 and OEF for 3 groups becoming low in density on CT (LD-B, LDS and LD-L). The data for the regions with luxury perfusion were excluded from this analysis. The mean CBF and CMR02 for the LD-L group were significantly higher than those for another two groups (LD-B & LD-S). There was no difference amon*" these three groups in the mean OEF values (p>0.05). Very few overlaps were found between group LD-L and groups LD-B & LD-S in CMR02, but there were relatively large overlaps in CBF ranges. The CBF ^for LD-L regions was above 10 ml/100 g/min in almost all cases, but the CBF of most LD-B regions was below this level. Table 3 shows mean intervals from the onset to the PET study. The interval for the LD-B group was significantly longer than those for the other two groups, but there was no significant difference be-tween groups of LD-S and LD-L in the intervals. 

DISCUSSION 

In the case of ischemic stroke, the 'lw density' area on CT defined in our study is thought to be attribu-able to vasogenic edema and/or possibly to cytotoxic edema.rs So the regions of LD-B reflect an established infarcted core. The 'normodensity' on CT cannot always represent undamaged tissue, but the ND areas should be morphologically intact tissue, which also may include non-functioning ones, because no definite morphological change had occured through-out the observation. On the other hand, it is difficult to evaluate the viability of the LD-S or LD-L regions only by their CT-normodensity at the PET study. In addition, there was no definite evidence of main-taining normodensity during a PET scan in the LD-S region, because they had already showed hypodensity on CT just after the PET. As the energy production depends on oxidative phosphorylation, CMR02 is considered to reflect the activity of tissues somewhat. Powers et al. used PET to study the minimum CBF and CMR02 required by the human brain to main-tain viability for more than a few hours in 50 subjects with varying degrees of cerebral ischemia.20 They found that CMR02 was an accurate indicator for distinguishing viable from infarcted tissue. Accord-ing to their results, CMR02 Iess than 1.3 ml/100 gl min was inadequate to sustain tissue viability for a prolonged period. In our results, CMR02 of the LD-L region was above 1.6ml/lOO g/min, which was also the minimum CMR02 value for ND regions. This suggests that the LD-L regions should include viable tissue. 
We found no definite boundary in CBF between viable and non-viable tissues even with the exclusion of the areas with luxury perfusion. But regions with infarction or subsequent necrosis (LD-B, LD-S, LD-L) were distinguished from those escaping ir-reversible damage (ND) at 17 ml/100 g/min in blood flow. Considering the interval from the onset to the measurement, the LD-B group was studied signi-ficantly later than both the LD-S and LD-L groups, but there were no significant difference between latter two groups (LD-S, LD-L) in ischemic duration. The difference in outcome of these two groups (LD-S, LD-L) should therefore be attributed to the magnitude of the initial CBF reduction. In fact, the mean CBF for the LD-L group (14.0 ml/100 g/ min, range : 9.9-17.3 ml/100 g/min). was significantly higher than that for the LD-B and LD-S groups (-10ml/100 g/mm range: 7.7-14.1 ml/IOO g/min). These findings suggest that a part of the tissue with CBF between 10 and 17 ml/100 g/min is still viable at least 7 hours after onset of ischemia, but becomes non-viable in a longer period of ischemia. 
As we did not measure subsequent changes in CBF, however, we discuss here the possibility that the CBF of the LD-L regions further decreased after the PET study and this additional reduction in flow produced infarction, First of all, we can refer to many studies about ischemic threshold reported by various investigators. Many experimental studies have demonstrated that membranous failure or ir-reversible damage occurs when CBF is reduced to around 10-12ml,/100 g/min9,12 for 2-3 hours, and lower flow was tolerated for a shorter length of time. Infarction also ensued by reducing CBF to the level of 17--18 ml/100 gfmin for permanent MCA occlu-sion.9,11 In clinical studies in humans, there are some reports about the relationship between CBF and electroencephalographic (EEG) changes in patients undergoing carotid arterectomy.8,21,22 The operation was per^formed without clinical deficit in spite of marked EEG change and CBF reduction to 12-15 m//100 g/min. But the direct cortical responses may be abolished in CBF of 8-10 m//100 gfmin for 50 minutes before infarction occurs. With PET, Baron et al. studied patients with cerebral infarction of 2-38 days after onset, and showed that the CBF threshold for infarction was around 1 1 ml/100 gl min.7 Powers et al. measured the minimum CBF level for the viable tissue (-15 m//100 g/min) and the normal functioning one. (19 ml/100 g/min).20 These results were obtained from subacute to chronic cases more than 48 hours after onset. There is a report of evidence of penumbra in a PET study. Powers et al. studied 4 patients with vasospasm due to subarach-noid hemorrhage.23 Two of them recovered and had CBF of 15.0 and 16.2 m//100 g/min, respectively, in comparison to 12.0 and 1 1.7 ml/100 g/min in the two with no recovery. The CMR02 va]ues were 1 .34 and 2.60 against 0.72 and 1.66 ml/100g/min. These reports were fairly consistent with our results. 
We can also refer to the PF.T study by Heiss et al of subsequent changes in CBF and CMR02 in patients with ischemic stroke.24 They studied 16 patients within 6-48 hours of onset and again 1 3-25 days later. They found that CBF and CMR02 m the core of infarction did not change during the study period. In the peri-infarct tissue, they demonstrated that the progressive derangement of CMR02 oc-curred without a concomitant change in CBF or even with some improvement in flow. These results sug-gest that CBF should not change greatly in ischemic lesions in and around the core of infarction and their metabolism and tissue integrity should pro-gressively deteriorate in the natural course. 
In conclusion, our data demonstrate, in the clinical PET study of humans, that some regions with a residual flow as low as 10-17 ml/100 g/min could maintain viability for at least 7 hours after onset, and that these ischemic lesions demonstrate normo-density on CT in the acute phase although they should subsequently develop irreversible morpho-logical changes. These lesions should exhibit the possibility of effective treatment of ischemia in-cluding therapeutic reperfusion. In addition, we were able to evaluate the viability of ischemic area by normodensity on CT to some degree without measur-ing the blood fiow or metaboiism. 

REFERENCES 

1 . Del Zoppo GJ, Zeumer H, Harker LA : Thrombolytic therapy in stroke : possibilities and hazards. Stroke 17: 595-607, 1986 
2. Mori E, Tabuchi M, Yoshida T, et al : Intracarotid urokinase with thromboembolic occlusion of the middle cerebral artery. Stroke 19: 802-812, 1988 
3. Phillips DA, Fisher M, Smith TW, et al : The safety and angiographic efficacy of tissue plasrninogen activator in a cerebral embolization model. Ann Neurol 23 : 391-394, 1988 
4. Taneda M, Shimada N, Tsuchida T : Transient neuro-logical deficits due to embolic occlusion and im-mediate reopening of the cerebral arteries. Stroke 16: 522-524, 1982 
5. Bell BA, Symon L, Branston NM: CBF and time threshold for the formation of ischernic cerebral edema, and effect of reperfusion in baboons. J Neurosurg 62: 31~tl, 1985 
6. Slivka A, Pulsinelli W : Hemorrhagic complication of thrombolytic therapy in experimental stroke. Stroke 10: 1148-1156, 1987 
7. Baron JC, Rougemout D. Bousser MG, et al : Local CBF, oxygen extraction fraction (OEF), and CMRO 2 : Prognostic value in recent supratentorial infarction in humans. J Cereb Blood Flow Metab 3 (Suppl 1): S1-S2, 1983 
8. Jafar JJ, Crowell RM: Focal ischemic thresholds. In Cerebral blood flow-physiologic and clinical aspects, Wood J (ed.), New York. McGrawHill, pp 449~,57, 1987 
9. Jones TH, Morawetz RB, Crowell RM, et al : Thre-sholds of focal cerebral ischemia in awake monkeys. J Neurosurg 54: 773-782, 1981 
10. Lenzi GL, Frackowiak RSJ, Jones T : Cerebral oxygen metabolism and blood fiow in human cerebral ischemic infarction. J Cereb Blood Flow Metab 2 : 321-335, 1982 
11 . Morawetz RB, Crowell RM, DeGirolami U, et al : Regional cerebral blood fiow threshold during cere-bral ischemia. Fed Proc 38: 2493-2494, 1979 
12. Heiss W-D : Flow threshold damage of brain tissue. Stroke and morphological 14: 329-331, 1983 
13. Kanno I, Miura S, Yamamoto S, et al : evaluation of a HEADTOME-III. 931-938, 1985 positron emission J Comput Assist Design and tomograph : Tomogr 9: 
14. Frackowiak RSJ, Lenzi GL, Jones T, et al : Quantita-tive measurement of regional cerebral blood flow and oxygen metabolism in man, using 150 and positron emission tomography : theory, procedure and normal values. J Comput Assist Tomogr 4 : 727-736, 1980 
15. Lammertsma AA. Jones T : Correction for the pre-sence of intravascular oxygen- 1 5 in the steady-state technique for measuring regional oxygen extraction ratio in the brain : I . Description of the method. J Cereb Blood Flow Metab 3 : 416-424, 1983 
16. Miura S, Kanno I, Iida H, et al : Anatomical adjust-ment in brain positron emission tomography using X-ray CT images. J Comput Assist Tomogr 12: 363-367, 1988 
17. Yamaguchi T. Kanno, I. Uemura K, et al : Reduction in regional cerebral metabolic rate of oxygen during human aging. Stroke 17: 1220-1228, 1986 
18. Tomura N, Uemura K, Inugami A, et al : Ear.ly CT finding in cerebral infarction : Obscuration of the lentiform nucleus. Radiology 168: 463-467, 1988 
19. Lassen NA : The luxury perfusion syndrome and its possible relation to acute metabolic acidosis localized within the brain. Lancet 2: 1113-1115, 1966 
20. Powers WJ, Grubb LR, Darriet D, et al : Cerebral blood fiow and cerebral metabolic rate of oxygen requirement for cerebral function and viability in humans. J Cereb Blood Flow Metab 5 : 600-608, 1985 
21. Sundt TM Jr, Sharbrough FS, Anderson RE, et al : Cerebral blood fiow measurements and electro-encephalograms during carotid endarterectomy. J Neurosurg 41 : 310-320, 1974 
22. Trojaberg W, Boysen G : Relation between EEG, regional cerebral blood fiow and internal carotid artery pressure during carotid endarterectomy. Electroencephalogr Clin Neurophysiol 34: 61-69, 1973 
23. Powers WJ. Grubb RL, Baker RP, et al : Regional cerebral blood fiow and metabolism in reversible ischemia due to vasospasm : Determination by posi-tron ernission tomography. J Neurosurg 62 : 539-546, 1985 
24. Heiss W-D, Huber M, Fink GR, et al : Progressive derangement of periinfarct viable tissue in ischemic stroke. J Cereb blood Flow Metab 12: 193-203, 1992