VASCULAR SPASM IN CAT CORTEX AFTER ISCHEMIA/tfa/7 et al. References 1. Ames A, Wright RL, Kowada M, et al: Cerebral ischemia II. The noreflow phenomenon. Am J Pathol 52: 437^147, 1968 2. Ginsberg MD, Myers RE: The topography of impaired microvascular perfusion in the primate brain following total circulatory arrest. Neurology (Minneap)*22: 998-1011, 1972 3. Chiang J, Kowada M, Ames A, et al: Cerebral ischemia III. Vascular changes. Am J Pathol 52: 455-476, 1968 4. Little JR, Kerr FWL, Sundt TM Jr: Microcirculatory obstruction in focal cerebral ischemia: An electron microscopic investigation in monkeys. Stroke 7: 25-30, 1976 5. Arsenio-Nunes ML, Hossmann KA, Farkas-Bargeton E: Ultrastructural and histochemical investigation of the cerebral cortex of cat during and after complete ischemia. Acta Neuropathol (Berl), 26: 329-344, 1973 6. Stullken EH, Sokoll MD: The effects of heparin on recovery from ischemic brain injuries in cats. Anesth Analg (Cleve) 55: 683-687, 1976 7. Hallenbeck JM: Prevention of post ischemic impairment of microvascular perfusion. Neurology (Minneap) 27: 3-10, Jan 1977 8. Hart MN, Galloway GM, Dunn MJ: Perivascular anoxia-ischemia lesions in the human brain. Neurology (Minneap) 25: 477-482, 1975 9. Padykula HA, Herman E: Factors affecting the activity of adenosine triphosphatase and other phosphatases as measured by histochemical techniques. J Histochem Cytochem 3: 161-167, 1955 10. Kowada M, Ames A, Majno G, et al: Cerebral ischemia I. An improved experimental method for study; cardiovascular effects; and demonstration of an early vascular lesion in the rabbit. J Neurosurg 28: 150-156, 1968 57 11. Fischer EG: Impaired perfusion following cerebrovascular stasis — A review. Arch Neurol 29: 361-366, 1973 12. Hossmann KA, Kleihues P: Reversibility of ischemic brain damage. Arch Neurol 29: 375-384, 1973 13. Wade JG, Amtorp O, Sorenson SC: No-reflow state following cerebral ischemia. Arch Neurol 32: 381-384, 1975 14. Keatinge WR: Electrical and mechanical responses of vascular smooth muscle to vasodilator agents and vasoactive polypeptides. Circ Res 18: 641-649, 1966 15. Zervas NT, Lavyne MH, Negoro M: Neurotransmitters and the normal and ischemic cerebral circulations. N Engl J Med 293: 812-816, 1975 16. Goodman LS, Gilman A (eds): The Pharmacological Basis of Therapeutics, p 558. New York, MacMillan 1970 17. Shepherd JT, Vanhoutte PM: Veins and Their Control, p 246. London, W. B. Saunders 1975 18. Duckies SP, Bevan RD, Bevan JA: An in vitro study of prolonged vasospasm of a rabbit cerebral artery. Stroke 7: 174-178, 1976 19. Heistad DD, Marcus ML, Ehrhardt JC, et al: Effect of stimulation of carotid chemoreceptors on total and regional cerebral blood flow. Circ Res 38: 20-25, 1976 20. Raper AJ, Kontos HA, Patterson JL Jr: Effect of hypoxia on pial precapillary vessels (abst) Fed Proc 3 1 : 401, 1972 21. Harper AM, Deskmukh VD, Rowan JO, Jennett WB: The influence of sympathetic nervous activity on cerebral blood flow. Arch Neurol 27: 1-6, 1972 22. Emerson TE, Parker JL: Effects of local increases of venous pressure on canine cerebral hemodynamics, pp 10-13. In Langfitt TW et al (eds) Cerebral Circulation and Metabolism, New York, Springer-Verlag, 1975 133 Xenon Inhalation Method: Significance of Indicator Maldistribution for Distinguishing Brain Areas with Impaired Perfusion An Index for Total Flow URS W. BLAUENSTEIN, M.D., JAMES H. HALSEY, JR., M.D., EDWIN M. WILSON, D . S C , AND EDWARD L. WILLS, P H . D . S U M M A R Y This paper introduces a new index for the assessment of regional cerebral blood flow. The index is proportional to total flow, and is obtained from the ratio of regional count rate to arterial indicator input to a region. This index is a more sensitive indicator of impaired perfusion than the traditional flow rate indices which ex- press flow per unit mass of tissue per minute. It accounts for brain tissue partly or totally deprived of its blood supply. Examples of clinical application are reported. A good correlation with the findings of computer-assisted tomography has been found. THE CLINICAL APPLICATION of the 133Xe inhalation method for the assessment of regional cerebral blood flow (rCBF) in cerebrovascular disease is still limited by methodological difficulties. In addressing the problem of indicator maldistribution in ischemic brain regions1"6 — commonly referred to as "look through" phenomenon or artifact3"6 — 86Kr has been cited as a better alternative to 133Xe if the surface of the brain is directly accessible. 4 ' 5 However, for non-invasive measurements of rCBF the converse appears to be true. The primary characteristic emission of 85Kr is predominantly fi particle (99.6% (3; 0.4% 7). Thus, measurements with 85Kr are either limited to a depth of 2.5mm below the brain surface in the open skull,6'7"9 or, when detecting the 7 emission through the intact skull, require doses several times greater than with 133 Xe. For whichever 7 is detected, indicator maldistribution is equally possible whether using the inhalation or the injection method. The localization and quantitation of impaired indicator delivery to brain regions partly or totally deprived of their blood supply would clearly be of clinical diagnostic value. They may be achieved by appropriate analysis of the indicator clearance curves which seem to contain the necessary information about indicator maldistribution. A parameter for total flow may offer a solution for this methodological problem. Our preliminary experience with such a parameter is reported here. From the Stroke Research Center and Department of Neurology, University of Alabama Medical Center, Birmingham, AL 35294. Supported in part by NINCDS Grants NS 08802 and NS 11109. Dr. Blauenstein was supported in part by the Swiss Biologico-Medical Scholarship Foundation. Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 STROKE 58 Calculation Of rCBF Clearance curve analysis was operationally carried out in accordance with the two-compartment model developed by Obrist and co-workers.10"12 By means of the resulting set of parameter values Pi, p2, k,, and k2, reconstruction of the curve fitted through the data points originally recorded was then accomplished to obtain the fitted value of the total count rate N-f(t) at any given time t from start-fit to end-fit time. This was done using the following equations (1, 2): = p,-/C A (u)-e -k,(t-u) du (1) where N',(t) is the count rate accounted for by the i:th compartment at time t, not corrected for recirculation; t = a given time after start of 133Xe inhalation; pt = weighting coefficient of the i: th compartment according to Obrist's assumptions; CA(t) = isotope concentration in the endtidal expired air (« arterial blood10-12) at time t; k, = decay constant of the i:th compartment. (2) =1 In the concept of the two-compartment model the rCBFparameters f, (ml/lOOg/min) and w, (%) are derived by means of the following relations (3-5): = A, • k, (3) where Xi represents the blood-brain tissue partition coefficient for gray and white matter, adjusted for hemoglobin concentration (e.g. Xgray = 0.8 and Xwhlte = 1.5 for a hemoglobin concentration of 15g%13). The parameter f, is the normalized flow rate of the i :th compartment. The relative weight of the first compartment, w,, is determined by p./f. + l where w2 = 100 - w, . 100, (4) (5) The values for fi and Wj are readily obtained from equations (3) to (5) by using the computer solutions for pt and k1; and the hemoglobin adjusted X,. To the extent that the basic assumption of the two-compartment model is valid — i.e. that a faster clearing compartment, considered to represent mainly cerebral gray matter, is separated from a slower clearing compartment, considered to represent white matter and extracerebral tissue — the two-compartmentally derived rCBF-parameters will provide meaningful results. Risberg et al." and Obrist et al.12 recognized, however, that in certain pathological states, where compartmental boundaries shift, the concept may not hold true. From the equations given above it can be easily seen that the outcome of f! and Wi depends essentially upon the proper substitute for the X, related to the decay constants k, for each compartment i. Under pathological circumstances, where reduced perfusion of gray matter may show up in the second compartment and increased perfusion of white matter — e.g. due to reactive hyperemia — may reveal itself in the first compartment according to the order of magnitude of their clearance rates, the association of X! to k, and A2 to k2 is not suitable. Such situations are well known under the term VOL 9, No 1, JANUARY-FEBRUARY 1978 "slippage" which describes the manifestation of clearance rates of tissue fractions of gray or white matter in the wrong compartment of the two-compartment model. This phenomenon introduces a source of considerable uncertainty in the analysis of rCBF results. In a series of about 1,200 measurements in 280 stroke patients we found it generally difficult to achieve conclusive interpretation of impaired rCBF based solely upon f, and w,. Nevertheless the two-compartment model provides an excellent parameterized fit to the clearance curves from which a useful new set of parameters can be derived to describe rCBF. By rearranging we obtain from equations (1) and (2): p, =l /C A (u)-e -k,(t-u)'du (6) The quotient on the right-hand side of equation (6) is a constant and independent of indicator dose. Its physiological meaning is more readily elucidated by considering the parameters represented by the coefficients p,. Referring to Obrist,12 Pi = a - w, • f, (7) where w, is the relative weight and f, the normalized flow rate of the i: th compartment. The proportionality constant a implies a whole product of different coefficients, i.e. A, B, C, D, . . . W, etc., most of which cannot be readily assessed in vivo. For example, the coefficient A relates units of count rate to concentration, B accounts for differences in counting geometry, C for differences in detector sensitivity, D for regional differences of skull thickness, etc., whereas W accounts for the absolute weight of brain tissue in thefieldof view of a detector. Most of these coefficients, although partly subjected to biological variations, are generally assumed to be reasonably constant and reproducible from subject to subject and from measurement to measurement in the same subject. They are commonly lumped together into a single coefficient that is in the following denoted by /?. The coefficient W, however, certainly represents a biological variable which contributes essentially to the size of p(. Its product with the relative weight w( is the absolute weight W, of the tissue under observation. Equation (7) can thus be rewritten as f, . Pi = P (8) The normalized flow rate f, (ml/min/lOOg) is given by the ratio of totalflowFt (ml/min) to total tissue weight perfused W,(g), thus p, = P • W, • F, W, (9) In a two-compartment system, i. Fi (10) The sum of the two-compartmental pi is thus proportional to total flow (ml/min) in the tissue under observation. Since f) is considered to be a constant for a given recording system, Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 59 NEW INDEX FOR rCBF/Blauenstein et al. the sum of the Pi can be used as an index for total flow hereinafter referred to as $ for convenience. Note that the calculation of does not require the knowledge of the blood-brain tissue partition coefficients X|. It is recognized that contamination by radiation from the contralateral hemisphere ("cross-talk") and from ipsilateral regions adjacent to the tissue cone under observation (scatter radiation) interferes with an experimentally ideal measurement of $ . The constant /? depends largely upon the technical characteristics of the monitoring devices of a given recording system. Absolute values of $ are thus not immediately reproducible in different laboratories using the 133Xe inhalation method as long as /? is not specified. Inter-laboratory communication, however, can be obtained by use of dimensionless ratios where the analyzed parameters are expressed as a percentage of reference values represented, e.g., by homologous regions of the contralateral hemisphere, by the ipsilateral hemispheric mean value, or by analogous regions of preceding measurements in the same subject. In the figures of this paper the regional values for * are expressed as a percentage of the total sum of $ for all 14 regions considered. The index for total flow, $, is determined by both the size of the compartments and their flow rates. A large tissue fraction with low perfusion rates and a small one with high perfusion rates may yield the same total flow. In order to distinguish such differences in perfusion rates, a parameter for flow rate is needed. This can be derived from the p, as ordinate values and the ^ as decay constants, expressing the first-minute decrease of the p ( as a percentage of the sum of the initial ordinate values — denoted by S — according to the equation S = p,(l - e klt ) + p2(l - e 100, (11) Pi + P2 where t = 1 minute. The proportionality constant (8 implied in Pi according to equation (10) cancels out on the righthand side of equation (11). It follows that: S = -e klt ) -e kgt ) 100, (12) where F, represents the absolute flow for each compartment. Referring to Obrist's definition of the fractional flow, FF,12, equation (12) can be rewritten as S = FF^l-e"1^1) -100 + F F 2 ( l - e ~ k j t ) -100. Methods 1S3 The clearance of Xe was measured by the Obrist inhalation method.10"12 The technical details of the procedure and the specifications of our system are described elsewhere.20 In brief, two lead shielded blocks enclosing seven detectors each in a fixed parallel array were placed perpendicularly to the two lateral planes of the subject's head. This provides an optimal alignment of corresponding probes to cover homologous regions in both hemispheres in the approximate locations shown in the sketch of figure 3. The use of a Plexiglas grid system assured reproducibility of positioning of the subject's head relative to the detectors in sequential measurements to within ±0.5 cm. The scintillation detectors consisted of Nal(Tl) crystals with %" diameter and %" length, being recessed behind 1" long cylindrical lead collimators with an inner diameter of %". The electronic windows of the pulse height analyzers were set to encompass both gamma and x-ray radiation, i.e. from 20 to 100 kev. The systemic arterial carbon dioxide tension (Paco 2 ) was analyzed from arterial blood samples in patients and was estimated from the endtidal points of the capnographic recordings (Beckman LB-2) of the expired air sample in normal subjects. The mean arterial blood pressure (MABP) was calculated as diastolic pressure plus one-third of the pulse pressure. The computerized axial transverse tomography (CT-scan) was performed in standard fashion (EMIscanner). Material Normal reference data for $ and S were obtained from 46 right-handed healthy volunteers with a mean age of 24.5 ± 5.5 (SD) years, ranging from 18 to 43 years (1 woman, 45 men). On 32 of the subjects two rest measurements were performed on consecutive days; 14 volunteers underwent a single measurement at rest. Normal reference values for f, and W[ are described elsewhere.21 To illustrate the significance of a parameter for total flow in the assessment of abnormal indicator delivery, three examples were arbitrarily chosen from more than 1,200 rCBF measurements on 280 patients with cerebrovascular disease: (1) cerebral hemiatrophy; (2) ischemic cerebral infarction in association with transient ischemic attacks; (3) unilateral ischemic cerebral infarction with progression, followed by a second infarction in the contralateral hemisphere. The related case histories are mentioned together with the rCBF data in the following section. (13) Results Note that like the index for total flow <£, this flow rate S does not require the knowledge of the blood-brain tissue partition coefficients X,. The parameter S is similar to other initial slope indices reported in the literature.14"18 Historically the concept of an initial slope has been well established for many years for the 133Xe injection method. 15 ' 19 To date the attempts to apply it to the inhalation method,"' "• 18 however, have been limited by the necessity to correct for recirculation of indicator. This in turn requires a monitoring time of much more than two minutes after start of clearance to obtain a proper estimate of k2 and in turn a suitable parameter set pi and k^19 Normal Subjects The results of the 78 rCBF measurements performed under rest conditions on 46 healthy right-handed subjects are summarized in table 1 and figure 1. Since two rCBF measurements were performed on each of 32 subjects, they were weighted by a factor of 0.5; together with the single measurements on the remaining 14 subjects a sample size of N = 46 was obtained. $ is bilaterally highest in the pre-central (superior-frontal), inferior-frontal, and centro-temporal regions. The interhemispheric asymmetries pre-frontally, centro-temporally, and temporo-occipitally are statistically Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 STROKE 60 NORMAL SUBJECTS (N = 4 6 ) VOL 9, No 1, JANUARY-FEBRUARY TABLE 1 Regional Cerebral Blood Flow Indices for Total Flaw (*) and Flow Rale (S) in 46 Right-handed Healthy Subjects s * L Mean S.E. *% A B C D E G H 1978 B C 0 E G H FIGURE I. Regional cerebral blood flow indices for total flow ($) and flow rate (Sj in 46 right-handed healthy subjects. In this and the following figures, the regional values for $ are expressed as a percentage, <&%, of the total sum of $ of all 14 regions. * = Statistically significant interhemispheric differences. The solid line refers to the left hemisphere, the broken line to the right. The letters A through H refer to the brain regions shown in the sketch of figure 3: (A) pre-frontal; (B) pre-central, superior-frontal; (C) inferior-frontal; (D) Rolandic; (E) centro-temporal; (G) parietal; (H) temporo-occipital. A B C D E G H HM *358 446 475 375 *453 399 *384 *413 17 18 16 15 14 12 10 14 R Mean S.E. L Mean S.E. 332 439 474 377 469 398 347 405 50.9 49.2 •47.8 16 15 14 15 16 14 11 13 46.0 46.4 45.9 43.5 47.1 0.8 0.7 0.6 0.6 0.6 0.6 0.6 0.6 R Mean S.E. 50.9 49.3 48.7 46.0 46.9 45.8 43.5 47.3 0.7 0.7 0.7 0.7 0.6 0.7 0.6 0.6 Mean Pacoj 40.4 ± 2.3 (SD) mm Hg; mean MABP 95.0 ± 12.4 (SD) mm Hg. •Statistically significant interhemispheric differences. S.E. => standard error. The letters A through H refer to the brain regions shown in the sketch of figure 3. HM = hemispheric mean; L/R = left/right hemisphere. The interhemispheric correlation coefficients of the seven homologous regions range from 0.859 to 0.917, that of the hemispheric mean values is 0.982. Patients significant at the level of p < 0.0001, 0.047, and < 0.0001, respectively. The hemispheric mean of $ is significantly higher on the left side {p = 0.034). The interhemispheric correlation coefficients of six of the seven pairs of homologous regions range from 0.859 to 0.962, that of the temporo-occipital regions is 0.672. The hemispheric means of 4> show an almost complete interhemispheric correlation (r = 0.967). S is bilaterally highest in the frontal lobe, gradually declining in size from the pre-frontal to the temporo-occipital regions. The interhemispheric asymmetry inferior-frontally is statistically significant (p = 0.001), that of the hemispheric means of S reaches a level of p = 0.076. Case 1: A trophy of a Cerebral Hemisphere A 22-year-old woman with mental retardation and chronic left hemiparesis was referred for evaluation because of progressing mental deterioration and worsening of the seizure disorder from which she had been suffering since age 8. The pneumoencephalogram (fig. 2) and carotid angiogram revealed atrophy of the right hemisphere. The results of the rCBF measurement are shown in table 2 and figure 3. The parameter $ is unequivocally lower in all the regions of the right hemisphere compared with the homologous counterparts on the left side. The right hemispheric mean of FIGURE 2. Pneumoencephalogram of case I (atrophy of the right cerebral hemisphere). Left: anleroposterior view, supine position. The frontal horn and the body of the right lateral ventricle is markedly enlarged. Right: postero-anterior view, prone position. The occipital horn of the right lateral ventricle is considerably enlarged. Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 61 NEW INDEX FOR rCBF/ Blauenstein et al. TABLE 2 rCBF in Case 1: Atrophy of the Right Cerebral Hemisphere * L A B C D E G H HM 45 10 I i i 35 B c o E G B C D E G H FIGURE 3. Regional cerebral bloodflowin case I (atrophy of the right cerebral hemisphere). Labels as in figure 1. <£ is 20.4% lower than the left. The right hemispheric mean of S, however, is only 2.6% lower than the left. S indicates a comparable size of the clearance rates in both hemispheres, while $ reveals a considerable interhemispheric asymmetry. Since $ is proportional to total flow, the larger left hemisphere of this patient reveals the bigger total flow than the atrophic right hemisphere, as would be expected from the radiological findings (fig. 2). The flow per unit of time and of tissue mass, as implied in S, is, however, approximately similar for both hemispheres. The main rCBF finding in this case is, therefore, the interhemispheric asymmetry in total perfusion of either hemisphere and not an essential difference in the perfusion rate per unit of tissue mass. To illustrate the salient features of $ and S, their behavior in three homologous brain regions is described in detail: Pre-frontally (A) and inferior-frontally (Cj. The interhemispheric differences of S are in the same direction as those of $, but of much smaller magnitude. S of the rightsided regions (A) and (C) is only 1.1% and 1.9%, respectively, lower than that of the left-sided homologous regions. $ in contrast is 23.7% and 22.4% smaller, respectively. Because of lack of a weighting factor, S does not show the pronounced interhemispheric asymmetry evidenced by $. Temporo-occipitally (H). In contrast to the findings in regions (A) and (C), the parameter S is temporo-occipitally 23% higher on the right side, whereas $ is 5% smaller on the right side compared with the left. Apparently, the parameterized estimate of total flow, $>, is in the right temporo-occipital region lower than in the left despite the distinctly higher perfusion rate on the right side evidenced by S. The parameters $ and S combined suggest diagnostically relative hyperperfusion of a small tissue mass in the right temporo-occipital region. The two-compartmentally derived parameters f, and wt are also given in table 2. Since they are supposed to be highly dominated by gray matter flow represented in the fast clearing first compartment,12 a comparison with * and S is of interest. The latter parameters are not predicated on separation of fast and slow compartments and are, therefore, not subjected to the phenomenon of slippage. The flow rate f, shows pre-frontally and inferior-frontally the higher values in the right atrophic hemisphere, in contrast to S. The relative weight W! is consistently lower on the right side than on the left. Its interhemispheric asymmetries for 291 320 353 225 305 191 180 266 S R 222 243 274 179 220 176 171 212 f L R 45.7 41.9 41.9 36.7 39.7 35.5 33.0 39.2 45.2 38.1 41.1 33.1 35.5 34.0 40.6 38.2 L Wl R L R 43 37 73 76 40 35 67 63 40 33 69 73 34 28 59 55 61 58 45 36 32 28 58 58 56 79 26 25 63.3 66.0 37.1 31.7 Paooi 43.5 mm Hg; MABP 102 mm Hg. Labels as in table 1. homologous regions, however, do not correlate in size with those revealed by $. Care 2: Small Cerebral Ischemic Infarction in Association with Transient Ischemic Attacks A 60-year-old right-handed man with a history of transient loss of speech two years prior to admission was referred for weakness of the right arm lasting half a minute, and incoherent speech lasting a few hours. During his hospital course he developed intermittent dysphasia, dysarthria, and numbness in the right hand, but no motor weakness. The CT-scan performed one day after occurrence of the first attack revealed an area of diminished density in the left temporo-parietal region; the cerebral ventricles and sulci appeared mildly to moderately enlarged (fig. 4). A left carotid angiogram carried out eight days after the first episode showed a 30% stenosis of the left internal carotid artery without evidence of ulceration; there were marked arteriosclerotic irregularities in the cavernous and distal portion of the internal carotid artery, but no intracranial abnormalities were otherwise found. The results of the rCBF measurement performed on the same day as the CT-scan are shown in table 3 and figure 5. Both $ and S indicate in this patient severely impaired rCBF mainly in the distribution of the left middle cerebral artery (regions B, C, D, E, and G). Based upon the CT-scan an area of indicator maldistribution may be expected in the left temporo-parietal region. $ reveals the most pronounced interhemispheric asymmetries in the regions E, G, and B with 30%, 22%, and 20%, respectively, lower values on the left side than on the right. The corresponding values for S are 17%, 13%, and 12%. In the centro-temporal region E the TABLE 3 rCBF in Case 2: Cerebral Ischemic Infarction in the Left Hemisphere in Association with Transient Ischemic Attacks s * A B C D E G H HM L 179 213 256 163 191 177 225 201 R 178 267 272 201 272 227 253 239 L 34.6 33.1 33.4 31.2 28.5 27.9 30.9 31.4 fi R 35.5 37.5 34.8 33.4 34.3 31.9 32.3 34.2 Pacoi 45.3 mm Hg; MABP 128 mm Hg. Labels as in table 1. Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 L Wl R L R 34 53 33 50 36 34 46 57 33 36 51 51 49 50 27 31 34 37 41 50 26 35 45 45 35 35 44 46 46.6 50.3 32.1 34.4 62 STROKE VOL 9, No 1, JANUARY-FEBRUARY 1978 FIGURE 4. CT-scans, slices 2B, 3A, 3B, and 4 A, of case 2 (ischemic infarction in the left hemisphere in association with transient ischemic attacks). An area of diminished density is seen in the left temporoparietal area. The ventricles and cerebral sulci are mildly to moderately enlarged. 30% reduction of <£ compared with the contralateral side is due to not only the 17% reduction in the perfusion rate, but also to an apparent loss of tissue mass perfused. The same may hold true for the left parietal (G) and precentral (B) regions. Comparable extensive interhemispheric differences are not disclosed by f, and w,. Case 3: Progression of a Unilateral Ischemic Cerebral Infarction Followed by a Second Infarction in the Contralateral Hemisphere A severe left hemiparesis developed in a 72-year-old diabetic woman 12 hours prior to admission and progressed to a persistent hemiplegia 58 hours after onset. On the i—: 10 35 I 30 A B C 0 E G H B C 0 E FIGURE 5. Regional cerebral blood flow in case 2 (ischemic infarction in the left hemisphere in association with transient ischemic attacks}. Labels as in figure I. Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 63 NEW INDEX FOR rCBF/Blauenstein et al. TABLE 4 rCBF in Case 8: Progression of an Ischemic Infarction in the Right Cerebral Hemisphere, Followed by a Second Infarction in the Contralateral Hemisphere Which Occurred Between Measurements #5 and #6? #i A B C * D E G H HM A B C S D E G H HM #3 #2 #6 #5 #4 L R L R L R L R L R L R 169 197 272 154 248 174 234 155 212 261 144 216 150 237 179 198 258 147 234 176 240 162 204 255 138 227 156 223 164 190 254 137 233 159 234 158 202 240 143 217 167 231 214 238 335 156 304 203 331 194 225 308 131 249 176 337 176 205 283 144 242 173 247 176 201 272 156 249 179 246 97 94 156 75 112 156 120 157 160 120 179 127 145 207 196 205 195 196 194 254 231 210 211 114 144 36.5 33.8 34.2 32.3 37.1 37.3 32.6 29.3 36.5 31.4 29.9 29.2 30.2 32.4 32.5 32.7 32.4 31.8 32.2 28.0 31.9 31.0 29.1 35.0 25.8 30.9 25.9 28.0 31.1 30.7 30.9 28.1 32.2 31.9 26.2 26.1 33.7 28.5 25.6 30.5 31.0 31.7 49.2 48.1 43.9 42.8 45.7 43.7 44.0 47.1 40.4 43.1 38.8 38.9 37.6 44.0 35.9 35.0 39.1 31.1 35.3 32.3 36.9 38.7 32.8 38.6 33.9 34.2 31.7 34.3 25.9 25.9 31.4 21.8 26.6 23.1 29.0 26.7 24.2 26.6 23.4 27.4 25.2 28.0 32.3 31.6 29.4 30.1 45.3 41.4 35.1 34.9 26.2 25.9 33.8 29.6 HI Pacos and MABP of each measurement are depicted in figure 9. Labels as in table 1. seventh day of the course the patient became aphasic and hemiplegic on the right side as well. The CT-scan obtained on the third and eleventh day after onset are shown in figure 6. The first scan coincided temporally with the progression of the left hemiparesis to hemiplegia. It showed a circumscript low density area in the right hemisphere. The second scan was performed four days after the occurrence of the right hemiplegia and aphasia. It showed an extensive low density area in the left hemisphere in addition to the initially found small low density area in the right hemisphere. Serial rCBF measurements correlated well with the stepwise progression in the clinical course which ultimately proved fatal. The results for * and S are shown in table 4 and presented graphically for the regional $ and S in figures 7 and 8, respectively, and for the hemispheric means of $ in figure 9. The initial rCBF measurement (#1) performed 18 hours after onset showed impaired flow in the distribution of the right middle cerebral artery characterized by lower values of either * or S or both mainly in the right-sided Rolandic (D), FIGURE 6. CT-scans of case 3 (progression of a right-sided ischemic cerebral infarction, followed by a second infarction in the contralateral hemisphere). Left: low density area in the right hemisphere on the third day after onset. Right: extensive low density area in the left hemisphere on the eleventh day after onset, in addition to the initially found circumscript low density area in the right hemisphere. Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 STROKE 64 VOL tn #3 #4 18 26 46 70 M\ 50 9, No 1, JANUARY-FEBRUARY 1978 40 30 50 20 40 30 45 20 C 35 45 25 D 35 45 25 E 35 25 45 35 G 45 25 H 35 25 L 45 r TIME AFTER ONSET 18 26 46 AFTER I PROGRESSION AFTER 2. PROGRESSION 70 12 113 HOURS 55 HOURS 16 14 10 DAYS DAYS DAYS FIGURE 7. Regional cerebral blood flow in case 3. Time course of * through six consecutive measurements (§1 through §6). The regional values for 4> are expressed as a percentage, *%, of the total sum of 4> of all 14 regions. HS = hemispheric sum of $%. The letters A through H refer to the brain regions shown in the sketch of figure 3. The solid lines refer to the left hemisphere, the broken lines to the right. centro-temporal (E), and parietal (G) regions, compared with the left. In the second (#2) and third (#3) measurements 26 and 46 hours after onset, the interhemispheric asymmetries of $ and S between homologous regions varied in size and orientation throughout the hemispheres, revealing either moderate aggravation, attenuation, or even inversion of the asymmetries found in the initial (#1) measurement. The progression of the neurological deficit from left hemiparesis to hemiplegia occurred between the third (#3) and fourth (#4) measurements. It was followed by a distinct aggravation and spatial extension of the interhemispheric asymmetries of both * and S in measurement #4, with the symptomatic hemisphere showing the lower values. Remarkably, the small asymmetry of the temporo-occipital regions, which predominantly represent vertebro-basilar/ posterior cerebral circulation, remained relatively constant. Table 4 and figures 8 and 9 indicate that the clinical worsening was also accompanied by a distinct general increase of both <£ and S. This is not obvious infigure7 because it expresses regional values as a percentage of the sum of these 14 values. (It may be speculated that in this instance the general flow increase contributed to the focal and perifocal HM 35 25 ii3 HOURS 16 DAYS FIGURE 8. Regional cerebral blood flow in case 3. Time course of the parameter S through six consecutive measurements (§1 through §6). HM = hemispheric mean ofS. Other labels as in figure 7. edema in the lesion area thus worsening the functional deficit.) The fifth (#5) rCBF measurement, performed 55 hours after the first progression of the neurological deficit, HEMISPHERIC MEAN OF * #4 #5 #6 250 200 150 100 18 26 46 70 113 HOURS 16 DAYS MABP (mmHg) 152 IS3 160 97 120 83 PoCO2 (mmHQ> 400 4 0 0 395 38 0 385 31.5 FIGURE 9. Cerebral blood flow in case 3. Time course of the hemispheric mean of 4> through six consecutive measurements (§1 through §6). The solid line refers to the left hemisphere, the broken line to the right. Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 NEW INDEX FOR rCBF/Blauenstein revealed a peculiar attenuation or even reversal of the interhemispheric asymmetries, accompanied by a marked reduction of * and S in all regions. Two days later the patient developed an extensive infarction in the left hemisphere, as documented by the CT-scan obtained four days later (fig. 6). The last rCBF measurement (#6), performed 10 days after the disastrous cerebrovascular event in the left hemisphere, disclosed a decrease of the mean hemispheric $ of 45% and 32% on the left and right side, respectively, compared with the preceding rCBF measurement #5 (fig. 9). The interhemispheric asymmetries were most pronounced in the regions A, B, D, E, and G. The parameter S declined simultaneously in both hemispheres to a similar degree, except in the inferior-frontal region C where the right side decreased more than the left. The reduction of the mean hemispheric S from the fifth to the sixth measurement was bilaterally 25%. In this instance the greater impairment of the blood supply to the left hemisphere was disclosed by 4> and not by S. Discussion Proportionalized Total Flow,

recorded from the bulkiest parts of the brain will yield the largest values. In the normal subjects the highest values of * occur in the pre-central, inferior-frontal, and centro-temporal regions of both hemispheres which is in accordance with their anatomical size and blood supply (fig. 1). A shrunken vascular bed in an atrophic hemisphere in turn will carry less total flow as exemplified in case 1 (fig. 3). Any reduction in size of the vascular bed necessarily diminishes total flow, e.g. ischemic sequestration of brain tissue by thrombo-embolic vascular occlusion reduces the original capacity of the vascular bed and thus 4> as illustrated by cases 2 and 3 (figs. 5, and 7-9). Any space-demanding process, particularly brain edema, may reduce the potential size of the vascular bed and thus $. This may hold true as well for severely increased intra- et al. 65 cranial pressure of any cause which decreases the capacity of the vascular bed by its extrinsic compressive force. An empirical approach may be helpful in the assessment of the informational value of <£. If one postulates that the sensitivity of a parameter to detect rCBF differences in the brain can serve as a rating scale for its usefulness, then $ appears to be superior to the other parameters available at present. As reported elsewhere21 the rCBF indices f,, w,, FF,, and ISI reveal in normal subjects significant interhemispheric differences only in the inferior-frontal regions. $ in the present series of normals (table 1, fig. 1), however, indicates statistically significant interhemispheric asymmetries in the pre-frontal, centro-temporal, and temporooccipital regions. Although these asymmetries are small in comparison to the standard error of the corresponding regional mean values, they are nonetheless statistically significant due to the strong interhemispheric correlation of $. Remarkably, $ in the inferior-frontal regions is not different between the two hemispheres despite the statistically significant difference of the flow rate S. Hence, it appears that $ contains different information about rCBF than S, implying conceptually total flow and not flow per unit of tissue mass. Influence of the Blood-Brain Tissue Partition Coefficient X, At present there is no way to assess in vivo the appropriate set of X, in a given subject at a given time for all tissue fractions of a given brain region. Moreover, experimentally determined X, for gray and white matter cannot reliably be related to the clearance constants derived, as mentioned earlier in connection with slippage. Thus, the substitution of X, into the equations to calculate rCBF would not contribute to the clarification of the rCBF data obtained. Rather it may result in unrecognized distortions which can be very misleading in the interpretation of rCBF and its correlation to clinical findings. The unavailability of individually appropriate sets of X, is, however, no handicap to arrive at useful data. The original and basic information about rCBF is contained in the clearance curves themselves. As shown above the computer-analysis of the clearance curves by Obrist's model 10 ' 2 provides directly the coefficients p, from which the new parameter for total flow, , can be derived without any knowledge of the A, involved. The proportionality constant /? implied in the p, is independent of the X|. A breakdown of the p! into their physiological components represented by the absolute tissue weights W| and the flow rates f( that would require the substitution of X, is not needed to obtain the parameter for total flow. As exemplified by case 3, $ does not behave erratically in successive rCBF measurements despite progressive cerebrovascular pathology rendering ample cause for shifts in compartmental boundaries and alterations of \t due to ischemic infarction. The six rCBF measurements appear to correlate well with the clinical course of the patient. Influence of "Cross-Talk" The contamination of unilateral regional measurements of 133Xe clearance by radiation from the opposite hemisphere, as implied in the term "cross-talk," influences Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 STROKE 66 the outcome of $ to a substantial degree. The higher the count rate in the contralateral regions the more undesired counts will be picked up by the monitoring device in the ipsilateral region. Thus, $ in its uncorrected form will always reveal a value higher than is expected without cross-talk. Recent studies carried out in our laboratory indicate that the rCBF parameters can be partially corrected for cross-talk so that they are more useful for clinical purposes. Among the various parameters $ is the most sensitive to cross-talk. It is, therefore, remarkable that it nonetheless discloses interhemispheric differences of rCBF not found with the other parameters.21 Flow Rate, S Several initial slope indices have proven useful in the calculation of rCBF. In the 133Xe injection method an initial one- or two-minute slope plotted semilogarithmicdlly against time provides a rapid means of comparing regions in the same hemisphere.15'le In the inhalation method, however, the effects of recirculation of indicator preclude the direct use of this approach. Accordingly, the principle was modified to calculate the initial slope index, ISI, for the inhalation method." While this modification is based on computed k! and k2 which have been corrected for recirculation, it is also based on the total count rate NT (X), i.e. explicitly its two-compartmental components N; (X) and NJ (x) at start-fit time x which may range from 1.3 to 2.9 minutes. The fixation of the computed slope of ISI to the arbitrary time interval between the second and third minute after start of 133Xe inhalation14 produces variation in the relative contribution of the two compartments, thus giving rise to methodological distortion of the rCBF data. In order to avoid such distortion the parameter S was introduced as an index for the normalized flow rate. As a single rCBF parameter it is of limited value as are other normalized flow rate indices. In combination with $, however, it may provide clinically useful information about regional cerebral blood flow. Acknowledgment The authors thank Dr. Walter Obrist for his helpful discussion and review of the manuscript. VOL 9, No 1, JANUARY-FEBRUARY 1978 References 1. Hjjedt-Rasmussen K, Shinh^j E, Paulson O, et al: Regional cerebral blood flow in acute apoplexy: The "luxury perfusion syndrome" of brain tissue. Arch Neurol 17: 271-281, 1967 2. Paulson OB: Regional cerebral blood flow in apoplexy due to occlusion of the middle cerebral artery. Neurology (Minneap) 20: 63-77, 1970 3. Sundt TM Jr, Sharbrough FW, Anderson RE, et al: Cerebral blood flow measurements and electroencephalograms during carotid endarterectomy. J Neurosurg 4 1 : 310-320, 1974 4. Donley RF, Sundt TM Jr, Anderson RE, et al: Blood flow measurements and the "look through" artifact in focal cerebral ischemia. Stroke 6: 121-131, 1975 5. Hanson EJ Jr, Anderson RE, Sundt TM Jr: Comparison of "Krypton and m Xenon cerebral blood flow measurements before, during, and following focal, incomplete ischemia in the squirrel monkey. Circ Res 36: 18-26, 1975 6. Hanson EJ Jr, Anderson RE, Sundt TM Jr: Influence of cerebral vasoconstricting and vasodilating agents on blood flow in regions of focal ischemia. Stroke 6: 642-648, 1975 7. Lassen NA, Ingvar DH: Blood flow of the cerebral cortex determined by radioactive Krypton". Experientia 17: 42-43, 1961 8. Ingvar DH, Lassen NA: Regional blood flow of the cerebral cortex determined by Krypton". Acta Physiol Scand 54: 325-338, 1962 9. Lassen NA: Blood flow of the cerebral cortex calculated from "Kryptonbeta-clearance recorded over the exposed surface: Evidence of inhomogeneity of flow. Acta Neurol Scan 41 (Supp 14): 24-28, 1965 10. Obrist WD, Thompson HK Jr, King CG, et al: Determination of regional cerebral blood flow by inhalation of 133-Xenon. Circ Res 20: 124-135, 1967 11. Obrist WD, Thompson HK Jr, Wang HS, et al: A simplified procedure for determining fast compartment rCBFs by ""Xenon inhalation, pp 11-15. In Russell RWR (ed): Brain and Blood Flow. London, Pitman 1971 12. Obrist WD, Thompson HK Jr, Wang HS, et al: Regional cerebral blood flow estimated by 13sXenon inhalation. Stroke 6: 245-256, 1975 13. H0edt-Rasmussen K, Sveinsdottir E, Lassen NA: Regional cerebral blood flow in man determined by intra-arterial injection of radioactive inert gas. Circ Res 18: 237-247, 1966 14. Risberg J, Ali Z, Wilson EM, et al: Regional cerebral blood flow by '"Xenon inhalation. Stroke 6: 142-148, 1975 15. Paulson OB, Cronqvist S, Risberg J, et al: Regional cerebral blood flow: Comparison of 8-detector and 16-detector instrumentation. J Nucl Med 10: 164-173, 1968 16. Olesen J, Paulson OB, Lassen NA: Regional cerebral blood flow in man determined by the initial slope of the clearance of intra-arterially injected m X e . Stroke 2: 519-540, 1971 17. Wyper DJ, Lennox GA, Rowan JO: Two minute slope inhalation technique for cerebral blood flow measurement in man. 1. Method. J Neurol Neurosurg Psychiatry 39: 141-146, 1976 18. Wyper DJ, Lennox GA, Rowan JO: Two minute slope inhalation technique for cerebral blood flow measurement in man. 2. Clinical appraisal. J Neurol Neurosurg Psychiatry 39: 147-151, 1976 19. Unpublished study. 20. Wilson EM, Wills EL, Risberg J, et al: Measurement of regional cerebral blood flow by the 133Xenon inhalation method with an on-line computer. Comput Biol Med 7: 143-157, 1977 21. Blauenstein UW, Halsey JH Jr, Wilson EM, et al: '"Xenon inhalation method. Analysis of reproducibility: some of its physiological implications. Stroke 8: 92-102, 1977 Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016 133Xenon inhalation method: significance of indicator maldistribution for distinguishing brain areas with impaired perfusion. An index for total flow. U W Blauenstein, J H Halsey, Jr, E M Wilson and E L Wills Stroke. 1978;9:57-66 doi: 10.1161/01.STR.9.1.57 Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 1978 American Heart Association, Inc. All rights reserved. Print ISSN: 0039-2499. Online ISSN: 1524-4628 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://stroke.ahajournals.org/content/9/1/57 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Stroke is online at: http://stroke.ahajournals.org//subscriptions/ Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016