0148-396X/90/2704-0554802.00/0 NEUROSURGERY Copyright € 1990 by the Congress of Neurological Surgeons Vol. 27. No. 4. 1990 Printed in U.S.A Xenon-enhanced Computed Tomographic Measurement of Cerebral Blood Flow in Patients with Chronic Subdural Hematomas Akira Tanaka, M.D., Shinya Yoshinaga, M.D., and Masato Kimura, M.D. Department of Neurosurgery, Fukuoka University, Chikushi Hospital, Fukuoka, Japan We compared clinical symptoms with extent of brain shift on computed tomographic (CT) scans and quantitative and three-dimensional measurements of cerebral blood flow (CBF) on xenon-enhanced CT scans in 10 patients with chronic subdural hematomas. Five patients had only headache and minimal or no brain shift on a CT scan. The other five had hemiparesis and/or mental disturbance in addition to headache and moderate or severe brain shift on a CT scan. The mean hemispheric CBF decreased about 7% in patients with headache and about 35% in patients with hemiparesis and/or mental disturbance. It decreased also on the side without the hematoma. The CBF reduction was always more pronounced in the putamen and thalamus than in the cortex. On the contrary, the cortex CBF was mostly preserved or even elevated in both groups of patients. We speculate that CBF reduction in patients with a chronic subdural hematoma occurs initially in central cerebral areas like the basal ganglia and thalamus, and then extends to the entire hemisphere including the cortex as brain compression and displacement progress. Central cerebral area involvement might be more responsible for clinical symptoms than the cortex. (Neurosurgery 27:554-561, 1990) Key words: Central cerebral area, Cerebral blood flow. Chronic subdural hematoma, Cortex, Xenon-enhanced computed tomography Chronic subdural hematomas are slowly expanding intra- cranial mass lesions that can cause reversible reduction of cerebral function by compressing and displacing the brain. Compression of the subjacent brain and subsequent devel- opment of focal cerebral edema can lead to cerebral dysfunc- tion and cause clinical symptoms (2, 3). Cerebral blood flow (CBF) studies in patients with chronic subdural hematomas are relatively few (1, 4-8, 10), and studies of quantitative CBF measurements with a direct ana- tomical correlation are rare. We compared clinical symptoms with extent of brain shift on a computed tomographic (CT) scan and quantitative and three-dimensional measurements of CBF on xenon-enhanced CT scans in 10 patients with chronic subdural hematomas. On the basis of the data we obtained, we speculated as to which intracerebral structures are most responsible for clinical symptoms in patients with chronic subdural hematomas. MATERIALS AND METHODS CBF was studied in 10 patients. aged 23 to 79 years, before and after evacuation of a chronic subdural hematoma. The second study was done 3 to 4 weeks postoperatively, when all the patients had improved clinically. CBF studies were performed on a xenon-enhanced CT scan CBF analyzing system adapted to a Siemens SOMATOM DR-III. CBF mapping was created to obtain the arterial build- up curve, by recording the end-tidal build-up curve, and the tissue build-up curve, by analyzing the sequential images of a CT scan taken while a patient inhaled 30% xenon gas for 6 minutes. CBF values were derived by placing regions of interest (ROIs) on the hemisphere, cortex, thalamus, and putamen on both sides to delineate them ona CT scan without contrast enhancement. Three patients had bilateral hematomas, so that the number TABLE 1 Preoperative Clinical Symptoms and Computed Tomographic (CT) Findings in 10 Patients with Chronic Subdural Hematoma* Clinical Symptoms Computed Tomographic Findings Case Age (yr)/Sex Head- Hemi- Mental = é Mass ache paresis Disturbance Ses Bocation Density Effect 1 23/M + = - R F, P High + 2 77/M + = = L. F Low - 3 69/F + - - L FP Isodense + 4 52/M + + + Js F.P Isodense +++ =| T5/F + - = L F.P. Low + 6 46/M + = - B F.P/F High/high + 7 72/F bs = ae B FB. T.P/E.:P High/high ++ 8 79/M + + + I F.P High +++ 79/M Se + + B F. T. P/F Mixed/high +++ 10 53/M + + = Il F.P Mixed ++ “R, right: L. left: B. bilateral: F. frontal. P. parietal: T. temporal. nr a October 1990 of sides with a hematoma was 13 and the number of sides without was seven. The mean values between the sides with the hematoma and those without the hematoma and between patients with hemiparesis and/or mental disturbance and headache pre- and postoperatively were compared. RESULTS The preoperative clinical symptoms and CT findings in 10 patients with chronic subdural hematomas are presented in Table 1. Headache was accompanied by hemiparesis and/or mental disturbance in the 5 patients in whom the hematoma TABLE 2 Pre- and Postoperative Measurements of Cerebral Blood Flow Preoperative Postoperative Case MABP Paco, MABP Paco, (mmHg) (mmHg) (mmHg) (mm Hg) 1 67 38 71 40 2 90 37 83 39 3 % 40 80 42 4 102 36 92 37 5 97 41 80 43 6 93 35 85 41 aq 98 38 89 41 8 | 40 90 37 9 120 41 100 37 10 95 38 83 37 Mean 96.9 + 38.4 + 85.3 + 39.4 + +SD 13.8 2,1 7.9 2.3. “MABP, mean arterial blood pressure; Paco,. arterial carbon diox- ide pressure. XENON-ENHANCED CT MEASUREMENT 555 had moderate or severe mass effect (degree of brain shift) on a CT scan. The patients with minimal or no mass effect had only headache. The degree of brain shift did not always correlate with hematoma volume. Mean arterial blood pressure (MABP) and arterial carbon dioxide pressure (Paco,) at the time of CBF measurements are shown in Table 2. MABP fell by 11.6 mm Hg and Paco, rose by 1.0 mm Hg on average postoperatively. The former is statistically significant (P < 0.05), but the latter is not. The CBF readings in the hemisphere. cortex, thalamus, and pu- tamen on both sides in each patient before and after operation for a chronic subdural hematoma are shown in Table 3. The average CBF values are compared on the side with the hematoma, the side without, and the normal control in Figure 1. The value in each hemisphere was reduced on average to 35.1 ml/100 g/min on the side with the hematoma and to 37.2 ml/100 g/min on the side without the hematoma. The reduction rate on each side was 22% and 17.3% of 45.0 ml/ 100 g/min seen in the normal control. The value in the cortex was reduced on average to 47.0 ml/100 g/min on the side with the hematoma and to 45.2 ml/100 g/min on the side without the hematoma. The reduction rate on each side was 7.1% and 10.7% of 50.6 ml/100 g/min seen in the normal control. The value in the thalamus was reduced on average to 54.2 ml/100 g/min on the side with the hematoma and 59.7 ml/100 g/min on the side without the hematoma. The reduction rate on each side was 26.2% and 18.7% of 73.4 ml/ 100 g/min seen in the normal control. The value in the putamen was reduced on average to 36.8 ml/100 g/min on the side with the hematoma and 45,3 ml/100 g/min on the side without the hematoma. The reduction rate on each side was 31.4% and 15.5% of 53.6 ml/100 g/min seen in the normal control. Thus. CBF was reduced on both sides. The reduction in the cortex was. however. more marked on the TABLE 3 Cerebral Blood Flow in the Hemisphere. Cortex, Thalamus, and Putamen before and afier Operation for Chronic Subdural Hematoma® Preoperative Cerebral Blood Flow Postoperative Cerebral Blood Flow Case Side"/ (m1/100 g/min) (m1/100 g/min) Volume (ml) = = = = H Cc TO P H oma I P 1 50.1 76.6 87.3 62.2 50.4 43.1 723 60.2 49.1 74.9 72.6 PA 47.9 40.5 68.7 58.9 2 L/20 Ot 53.2 22.2 31.2 34.9 42.4 34.3 3 1/70 34.3 47.4 51.9 42.5 45.9 59.7 54.8 43.4 62.0 73.3 62.3 43.9 56.7 51.4 4 L/58 26.8 31.6 28.9 17.5 51.4 65.1 38.9 25.4 25.9 23.0 28.4 47.8 71.0 43.3 5 L/20 35.3 39.4 55.0 37.1 47.3 64.6 39.1 36.7 373 61.7 43.9 44.1 64.9 41.4 6 B/45 49.0 74.3 64.6 53.7 47.8 46.2 35u1 5 47.9 61.6 50.3 40.4 51.6 43.3 7 B/35 34.4 46.3 44.0 31.2 45.9 711 39.4 47 40.6 42.3 39.2 48.0 47.6 61.5 41.3 8 L/55 10.5 14.9 20.8 11.2 30.4 $0.2 40.2 20.1 20.2 37.7 29.5 32.2 55.2 37.8 9 B/100 212 29.0; 62.4 12.8 39.1 62.2 30.4 10 23.9 39.3 60.6 29.8 40.1 63.8 41.0 10 L/93 46.0 62.6 68.2 31.6 48.3 49.9 51.8 74.9 33.8 44.6 ’R. right: L. left: B. bilateral. “H, hemisphere: C. cortex: T, thalamus: P. putamen. Upper row of values. reading on the left side: lower row. reading on the right side. 556 TANAKA et al. side without the hematoma. unlike the CBF in the hemi- sphere. thalamus. and putamen. The reduction was greatest in the putamen. followed by the thalamus. hemisphere. and cortex. with a statistically significant reduction only in the thalamus (P < 0.05). The average CBF values are compared in the headache group. the neurology group (in which headache is accom- panied with hemiparesis and/or mental disturbance). and normal controls (Fig. 2). In the neurology group, the reduction rate was 35.1% in the hemisphere, 24.9% in the cortex. 33.1% in the thalamus, and 53.2% in the putamen. The reduction was statistically significant in the hemisphere. thalamus, and putamen (P < 0.05), but not in the cortex. In the headache group, it was 6.9% in the hemisphere, 16.1% in the thalamus, and 4.9% in the putamen. The reduction was statistically significant only in the thalamus (P < 0.05). The CBF was adversely elevated 13.4% in the cortex of this group, although it was not statistically significant. CBF values were more reduced in the neurology group than in the headache group except in the cortex, and this represented the greatest reduc- tion. The pre- and postoperative CBF changes in the hemisphere. cortex, thalamus, and putamen are presented in Figure 3. The headache group, in which the preoperative reduction was not too obvious, showed further reduction postoperatively. Values were restored to 70 to 80% of normal postoperatively in the neurology group in which preoperative reduction had been obvious. The restoration rate was greater in the putamen and CBF(mI/100g/min) — | 80 60 40 Y N N 20 Hematoma Side Non-Hematoma Side Neurosurgery, Vol. 27, No. 4 the thalamus than in the hemisphere and the cortex, although it was statistically significant only in the putamen (P < 0.05). Three patients are presented with both CT scans without contrast enhancement and xenon-enhanced CT scans. Case 10 A 53-year-old man was admitted because of headache and aright hemiparesis. A CT scan without contrast enhancement disclosed a chronic subdural hematoma of mixed density in the left frontoparietal region with a moderate degree of brain shift (Fig. 4). A xenon-enhanced CT scan showed a slight reduction of CBF in the putamen and thalamus on the left side, although CBF in the cortex under the hematoma was well preserved. Postoperatively, the patient was free of symp- toms, and the hematoma disappeared completely. CBF was homogeneous, but lower than preoperative values. Case 8 A 79-year-old man was admitted with headache, a right hemiparesis, and mental disturbance. A CT scan without contrast enhancement disclosed a chronic subdural hema- toma of high density in the left frontoparietal region with a severe degree of brain shift (Fig. 5). A xenon-enhanced CT scan showed a diffuse reduction of CBF in the left hemisphere including not only the putamen and thalamus, but also the cortex. Postoperatively, the patient was free of symptoms, and p<0.05 OT Normal Control Fic. 1. Mean CBF values in the hemisphere. cortex. thalamus. and putamen on both sides (i.¢.. side with the hematoma and side without the hematoma) in patients with a chronic subdural hematoma and in normal controls. October 1990 CBF(ml/100g/min) 80 60 40 WS 20 Neurology Group the hematoma disappeared completely. CBF values returned to normal in both hemispheres. Case 4 A 52-year-old man was admitted with headache, a right hemiparesis, and mental disturbance. A CT scan without contrast enhancement disclosed a chronic subdural hema- toma of isodensity in the left frontoparietal region with a severe degree of brain shift (Fig. 6). A xenon-enhanced CT scan showed a diffuse and severe reduction of CBF in both hemispheres. Postoperatively, the patient was free of symp- toms, and the hematoma disappeared completely. CBF values normalized on both sides. DISCUSSION Chronic subdural hematomas are slowly expanding intra- cranial mass lesions that can cause reversible reduction of cerebral function by compressing and displacing the brain. The exact mechanism remains unclear. Browder and Rabiner (2) obtained autopsy evidence of regional cerebral swelling in patients who died as a result of a chronic subdural hematoma. They suggested that the focal cerebral edema directly beneath the compressed surface of the brain is more responsible for clinical symptoms than is the compression itself. Further- more. distortion of the brain caused by swelling of the cen- trum semiovale has been identified radiologically in some Headache Group Fic. 2. Mean CBF values in the hemisphere, cortex, thalamus, and putamen in neurology (hemiparesis and/or mental disturbance) and headache groups patients with a chronic subdural hematoma and in normal controls. XENON-ENHANCED CT MEASUREMENT p< 0.05 557 Normal Control patients who failed to recover or continued to deteriorate after evacuation of the hematoma (3). Conversely, Tyson et al. (9) stated that symptomatic swelling of the white matter subjacent to a chronic subdural hematoma is undoubtedly rare and was recognized in only 6% of their patients, all of whom were comatose at the time the diagnosis was made. Swelling of the white matter found only in autopsy and severely ill cases might be the final outcome of the disease process of chronic subdural hematomas rather than the actual pathogenesis of clinical symptoms. It is noteworthy that the role of the white matter in chronic subdural hematomas has been recognized. CBF studies in patients with chronic subdural hematomas are relatively few (1, 4-8, 10), and studies of quantitative CBF measurements with direct anatomical correlation are rare. Ueda et al. (10) used a single photon emission CT scan for this purpose, but the xenon-enhanced CT scan used in our study and that of Segawa et al. (8) is more accurate in its anatomical correlation. Ikeda et al. (6) found that CBF decreased on both the side with the hematoma and the side without the hematoma, and the reduction was more pronounced in patients with a hemi- paresis and/or mental disturbance than in patients with head- ache. They also found that patients with hemiparesis showed a local CBF reduction in the frontal and parietal lobes on the side with the hematoma, suggesting that hematoma compres- sion of the subjacent brain might be responsible for the development of hemiparesis and mental disturbance. Fukuda etal. (4) made the same conclusion from their study: however. 558 TANAKA et al. Neurosurgery, Vol. 27, No. 4 CBF(ml/100g/min) o—o Headache Group e—e Neurology Group 80 * P<0.05 60 \ 40 | i ls a oe aS es Pre-Op Post-Op Pre-Op Post-Op Pre-Op Post-Op Pre-Op Post-Op Hemisphere Cortex Thalamus Putamen Fic. 3. Pre- and postoperative CBF values in the hemisphere, cortex, thalamus, and putamen in each patient with a chronic subdural hematoma in the neurology (hemiparesis and/or mental disturbance) and headache groups. A vertical line indicates a mean CBF value with standard deviation in each group. Fic. 4. Case 10. 4, CT scan without contrast enhancement demonstrates a chronic subdural hematoma of mixed density in the left frontoparietal region with a moderate degree of brain shift. B, xenon-enhanced CT scan discloses a slight reduction of CBF in the putamen and thalamus on the side with the hematoma. although CBF under the hematoma is well preserved. C, postoperatively. CBF is homogeneous, but lower than preoperative values. they used a two-dimensional measurement of the '**Xe in- using their method of three-dimensional measurement. There halation method and could not define an individual CBF was no difference in local CBF reduction among the frontal, value for each intracerebral structure. temporal. parietal. and occipital lobes. CBF reduction was Ueda et al. (10) found that the mean hemispheric CBF most pronounced in the central cerebral areas in both groups decreased about 20% in patients with headache and about 20 of patients. The authors concluded that psychosis was due to to 40% in patients with psychosis and/or pyramidal signs a total reduction and pyramidal signs were due to a local CBF October 1990 reduction not in the cortex, but in the central cerebral areas at the basal ganglia level. They stated that headache is more likely due to dural traction than to a CBF change. Segawa et al. (8) found that CBF decreased minimally in the cortex, but 36 to 58% in the white matter, based on xenon-enhanced CT scans, and concluded that CBF reduction in the white matter is more responsible for neurological dysfunction. They also speculated that the white matter is more susceptible to he- matoma pressure and resulting CBF reduction than the gray matter, since the former has much lower perfusion pressure than the latter. Nukui et al. (7) found that CBF decreased to 34 to 43 ml/ 100 g/min in elderly patients with focal symptoms and re- mained normal in young patients without a focal deficit. The reduction was not focal, but generalized over the entire hem- isphere wherever the hematoma was located. They ascribed the CBF reduction to intracranial hypertension, which in their series was below 300 mm H:0O. Brodersen and Gjerris (1) and Gjerris and Brodersen (5) found that CBF decreased to 24 to 38 ml/100 g/min and returned to 34 to 43 ml/100 g/min. which was still below normal values (45 to 55 ml/100 g/min). after surgery when all the patients had improved. The reduc- tion was uniform over the entire hemisphere. and cerebrospi- nal fluid pressure in four study patients was normal. They concluded that it is unlikely that CBF reduction is secondary to intracranial hypertension. taking into account the fact that intracranial pressure levels below 400 to 500 mm of water have no influence on CBF. On the other hand. the CBF reduction was roughly proportional to the reduction of cere- XENON-ENHANCED CT MEASUREMENT 559 Fic. 5. Case 8. A, CT scan without contrast enhancement shows a chronic subdural hematoma of high density in the left frontoparietal region with a severe degree of brain shift. B, xenon-enhanced CT scan discloses a diffuse reduction of CBF in the left hemisphere including not only the central cere- bral areas, but also the cortex. CBF in the right hemisphere is well preserved. C and D, postopera- tively, the hematoma disappeared totally (C) and CBF values were restored to normal (D). bral function. They also measured the cerebral metabolic rate of oxygen in one patient and found that it was reduced more than CBF. It was speculated that the CBF reduction might be secondary to a reduced metabolic demand rather than cerebral ischemia and suggested that a concomitant CBF reduction in the opposite hemisphere might be the result of transneural depression of cerebral function. In our study, we used a three-dimensional measurement with xenon-enhanced CT scans and found that a mean CBF was decreased 35.1% in the hemisphere, 33.1% in the thala- mus, and 53.2% in the putamen in patients with hemiparesis and/or mental disturbance, and 6.9% in the hemisphere, 16.1% in the thalamus, and 4.9% in the putamen in patients with headache. The CBF reduction was statistically significant (P < 0.05) in all three areas in the former group. but it was statistically significant (P < 0.05) only in the thalamus in the latter group. A mean CBF in the cortex was decreased 24.9% in patients with hemiparesis and/or mental disturbance. but it was adversely elevated 13.4% in patients with headache. although both of these readings were not statistically signifi- cant. CBF was decreased as well on the side without the hematoma. These results are compatible with those of Ikeda et al. (6). Ueda et al. (10), Segawa et al. (8), and Nukui et al. (7). A mass effect on a CT scan was moderate or severe in patients with hemiparesis and/or mental disturbance, and minimal or absent in patients with headache. On the basis of these findings. we speculate that CBF reduction in patients with chronic subdural hematoma occurs initially in the central cerebral areas like the basal ganglia and thalamus. and then. 560 TANAKA et al. Fi. 6. Case 4. .4, CT scan without contrast enha’ iscloses a chronic subdural hematoma of iso- density in the left frontoparietal region with a severe degree of brain shift. B, xenon-enhanced CT scan dis- closes a diffuse and severe reduction of CBF in both hemispheres. C and D, postoperatively. the hematoma disappeared totally (C) and CBF values normalized bilaterally (D). as brain compression and displacement progress, extends to the whole hemisphere, including the cortex. The CBF eleva- tion in the cortex seen mostly in patients with headache might reflect a local hyperemia induced by direct hematoma compression. The CBF reduction was restored to 70 to 80% of normal values postoperatively in patients with hemiparesis and/or mental disturbance in whom preoperative reduction was obvious. These facts seem to suggest that the central cerebral areas might be more susceptible to clinical symptoms than the cortex. Postoperatively, the mean CBF was, however, further reduced significantly in the cortex and minimally in the hemisphere, thalamus. and putamen in patients with headache. The postoperative change in the cortex might re- flect a disappearance of local hyperemia and a return to normal values. On the other hand, it is very difficult to explain the postoperative change in the hemisphere. thalamus, and putamen. Postoperatively, MABP dropped significantly in nine of 10 patients, although Paco, remained relatively stable. The CBF that had been relatively well preserved in these areas might have deteriorated as a result of this fall in MABP. In this sense, this MABP decrease might be responsible, in part. for an incomplete recovery in patients with headache and/or mental disturbance. Postoperative relief of headache and other symptoms might be one of the causes of the MABP decrease. Serial imaging with xenon-enhanced CT scans and time- by-time variations of CBF values in each normal patient are crucial. In our studies. the variations have ranged to <10°. It is not clear whether CBF reduction is a cause of brain dysfunction or merely a result of a reduced metabolic demand in dysfunctional brain. Simultaneous measurements of CBF and metabolic rate should resolve this question. A second question is why the central cerebral areas are more susceptible to hematoma pressure than the cortex. Might the extent of Neurosurgery, Vol. 27, No. 4 brain displacement and distortion be more important than hematoma pressure in the damage to the central cerebral areas? These questions need to be resolved. ACKNOWLEDGMENT The authors are grateful to Dr. A. Stollman from the Department of Neuroradiology. Mount Sinai Hospital. New York. for editorial assistance. Received for publication. March &. 1990: accepted. final form. April 20, 1990. Reprint requests: Akira Tanaka, M.D.. Department of Neurosur- gery, Fukuoka University. Chikushi Hospital. 377-1, Ohaza-Zoku- myoin. Chikushino-shi, Fukuoka, 818. Japan. REFERENCES 1, Brodersen P. Gjerris F: Regional cerebral blood flow in patients with chronic subdural hematoma. Acta Neurol Scand 51:233- 239. 1975 2. Browder J. Rabiner AM: Regional swelling of the brain in sub- dural hematoma. Ann Surg 134:369-375. 1951. Cook AW. Browder EJ. Carter WB: Cerebral swelling and ven- tricular alterations following evacuation of intracranial extracer- ebral hematoma. J Neurosurg 19:419-423, 1962. 4. Fukuda T. Ikeda Y. Nagai K. Azuma S. Ito H. Miwa T: Exami- nation of senile chronic subdural hematoma—lts relationship to CT findings. intracranial pressure. mean hemispheric cerebral blood flow. and clinical features {in Japanese. English abstr]. Neurotraumatology (Tokyo) $ -61. 1986 3. Gjerris F. Brodersen P: Regional cerebral blood flow in patients with chronic subdural haematoma before and after operation. Acta Neurochir (Wien) 31:267-268. 1975. 6. Ikeda K. Kano A. Hayase H. Yamashima T. Ito H. Yamamoto S: Relationship between symptoms of chronic subdural hema- October 1990 toma and hematoma volume or regional cerebral blood flow [in Japanese. English abstr]. Neurol Med Chir (Tokyo) 24:869-875, 1984. 7. Nukui H, Aiba T: Regional cerebral blood flow in patients with chronic subdural hematoma [in Japanese. English abstr]. No Shinkei Gaisho 4:371-382. 1972. 8. Segawa H. Fujie K. Fujimaki T. Morimoto T. Aritake Y. Sano K: Effect of hematoma pressure on cerebral blood flow and function in chronic subdural hematoma [in Japanese]. Proc Jpn Neurosurg 350, 1986 (abstr). 9. Tyson G, Strachan WE. Newman P. Winn HR, Butler A, Jane J: The role of craniectomy in the treatment of chronic subdural hematomas. J Neurosurg 52:776-781. 1980. 10. Ueda M. Takahashi Y, Ohmiya N. Mikami J, Ito K, Sato H. Matsuoka T. Takeda S. Ohkawara S: Single-photon emission CT- findings in chronic subdural hematoma [in Japanese]. Prog Com- put Tomogr (Tokyo) 7:623-630. 1985. COMMENTS This study measures cerebral blood flow (CBF) in 10 pa- tients with subacute and chronic subdural hematomas before and after surgical evacuation. The CBF was measured pri- marily in gray matter structures with the xenon-enhanced computed tomographic (CT) scan technique. This method is not as accurate as positron emission tomography, but is less costly, hence more readily available. The clinical subgroups of headache and hemiparesis and/or mental disturbance are analyzed with regard to CBF in various structures. The authors find reduction in CBF in the headache group ranging from 16% in the thalamus to 5% in the putamen with a paradoxical increase of 13% in the cortex. Given the lack of statistical significance and up to 10% variability inherent in this technique. these may only be considered trends. In the group with the neurological deficit, however, there is a signif- icant reduction of CBF in the putamen and thalamus, which presumably accounts for some of the symptoms, These data support the hypothesis that chronic subdural hematomas have their greatest effect on deep rather than cortical structures. CBF is shown to be lowered in both hemispheres with the greatest effect being on the ipsilateral putamen. This is con- firmed by the postoperative CBF measurements that display a significant increase in blood flow in the putamen of neuro- logically impaired patients. It is of interest to speculate as to why the deep gray structures show the greatest impairment of CBF from the chronic subdural hematomas. Perhaps the acute take-off or the lack of collateralization of the lenticulostriate and thala- moperforating arterial systems make them particularly sus- ceptible to the effects. Also. as the authors point out. it is not known whether the reduction of blood flow is due to the pressure of the hematoma or the gross distortion of the vessels. XENON-ENHANCED CT MEASUREMENT 561 Tanaka, et al. have collected some interesting data, which, although difficult to interpret. do in fact have several points that are intriguing: 1. Although not unknown from previous studies on cere- bral blood flow (CBF), it is interesting that there is so little difference between the CBF on the side with the hematoma vs. the side without the hematoma. Where there is a difference, it tends to show up in the deep gray structures. 2. Where increases in CBF seem to have occurred, these have generally been in patients with headache or mini- mal deficit and relatively little shift. It is clinically im- portant that those with neurological deficit tend to dem- onstrate /ow CBF. Although not significant in all com- parisons, this difference is intriguing. The mechanism for headache may represent dural irritation or, more likely, vascular dilatation or even a combination of the two. Nevertheless, the mechanism for neurological def- icit seems to include ischemia. 3. Related to the above is the observation that systemic blood pressure declines by 10 to 20 mm Hg after evac- uation of the mass. If one assumes an intracranial pres- sure of 10 to 20 mm Hg before decompression and estimates pre- and postoperative cerebral perfusion pres- sure. then the figures are nearly identical. We believe the preoperative hypertension is the brain’s “attempt” to maintain adequate cerebral perfusion pressure, a mechanism with which one should not interfere lest ischemia (vide supra) be potentiated; the postoperative reduction in systemic arterial blood pressure is predict- able. 4. The flows are quite heterogeneous among the deep gray matter, white matter, and cortical gray structures. This demonstrates the importance of tomographic blood flow techniques when testing clinical hypotheses. It is the major weakness of '**Xe studies in which external probes were used to monitor CBF. In the column of tissue beneath the probe. a profoundly ischemic area may amount to 10 or 20% of the probe’s field of view and could easily include hyperemia in 30 or 40% of that same volume. The net CBF would be increased, and the ischemia masked. The investigators might conclude that the problem was “hyperemia” rather than “ischemia”— a gross error. This is especially true because. as in the study of Tanaka et al., the increase in flow tends to be more cortical and. therefore, would be closer to external probes and produce a disproportionate amount of signal, further tending to mask the fact that deep tissues may well be ischemic. 5. Because of the lack of intracranial pressure data, it is This particular question has clinical relevance in the common situation of a deep intracranial hemorrhage producing a con- tralateral trapped (hydrocephalic) lateral ventricle. A draining ventricular catheter cou d lower intracranial pressure. but at impossible to calculate loca resistance and make esti- mates as to whether some of the low flows represented passive decreases because perfusion pressures had dropped below lower limits of autoregulation or may the cost of allowing further subfalcine herniation and brain distortion. This study is a significant addition to a surprisingly small body of literature on the effects of chronic subdural hemato- mas on CBF. I particularly like the portion of the methodology that performs serial CBF measurements on the patients. In- ternal controls such as these are informative. | look forward to a larger study that would allow more meaningful compar- isons among subgroups of structures and patients. Thomas Francavilla Washington, District of Coltunbia have been the result of metabolic changes. These data remind us of the complexities of regulation of CBF and the relationships among neurological function, met- abolic rate. and distant changes. They demonstrate some of the pitfalls of traditional external counting techniques for monitoring CBF in what is clearly a heterogeneous system. Lastly. these data remind us of how little we really understand of the pathophysiological processes of one of neurosurgery’s most basic lesions: the chronic subdural hematoma. Michael J. Rosner Birmingham, Alabama