A follow-up study of 153 consecutive patients who underwent complete excision of an angiographically visualized intracerebral arteriovenous malformation was conducted. The follow-up period ranged from 0.5 to 10.6 years, with a mean of 3.8 years. The presenting clinical event was hemorrhage in about one-half of the patients and seizure in about one-third. There was a marked tendency for postoperative neurological deficits to improve with time, so that whereas the immediate postoperative rate of serious morbidity was 24.2%, only 7.8% of the patients were found to have serious morbidity at follow-up. An additional 3 patients had died. one of an unrelated carcinoma, making the mortality related to arteriovenous malformation 1.3%. The classification of Spetzler and Martin (43) was used retrospectively; the percentages of Grade I (easiest) through Grade V (most difficult) lesions were 7.8%, 22.9%, 28.8%, 26.8%, and 13.8%, respectively. The early result was well correlated to grade, with good or excellent results in 100%, 94.3%, 88.6%, 61%, and 28.6% of the patients in Grades I through V, respectively. At follow-up, 98.7% of the patients with arteriovenous malformations of Grades I, II, and III were in good or excellent condition. The late morbidity and mortality rates for the patients in Grades IV and V were 12.2% and 38.4%, respectively. Of the patients who did not have seizures before surgery, 8.2% had only one or two seizures during the immediate postoperative period, and 7.1% had late seizures that were well controlled with medication in all. Of the patients who had seizures before surgery, over half were either cured or greatly improved with respect to the seizures. In 32.7% there was no change in the frequency of the seizures, and in 12.7% the seizures were more frequent after surgery. There was no history of either proven or suspected intracranial hemorrhage in any of the patients during the entire follow-up period, which in the aggregate totalled 556.3 years.

Background and Purpose: Hemodilution is known to increase cerebral blood flow, but it is not known whether the increase in flow is a direct result of a decrease in viscosity or whether it may be due to compensatory vasodilatation in response to the decrease in oxygen carrying capacity that results from hemodilution. This study is designed to investigate this question.Methods: Changes in regional cerebral blood flow were studied in normal and ischemic brains of 15 and 18 rabbits, respectively. In one group of rabbits graded hemodilution was used to reduce arterial oxygen content progressively in stages; in the second group the arterial oxygen content was reduced in similar stages by progressively larger reductions in the concentration of inspired oxygen (hypoxic hypoxia). In the ischemic animals focal ischemia was produced by embolic occlusion of the right middle cerebral artery.Results: In the normal rabbits, hypoxic hypoxia and hemodilution resulted in similar progressive increases in cerebral blood flow as arterial oxygen content fell. In the ischemic animals, there was a significant fall in cerebral blood flow in the ischemic region in all groups after arterial occlusion. Hemodilution resulted in a progressive increase in cerebral blood flow in both ischemic and nonischemic regions. With hypoxic hypoxia, however, cerebral blood flow in the ischemic region showed no increase or a slight decrease.Conclusions: Even though hypoxic hypoxia results in a marked increase in cerebral blood flow in normal brain, it does not significantly change cerebral blood flow in ischemic brain. In contrast, hemodilution resulting in a comparable degree of hypoxemia is capable of significantly increasing cerebral blood flow in ischemic brain. Therefore, the mechanism of blood flow augmentation by hemodilution in ischemic brain is probably related to a direct hemorheologic effect rather than to the resulting hypoxemia. (Stroke 1992;23:1487-1493 KEY WORDS • cerebral blood flow • cerebral ischemia • hemodilution • rabbits

No abstract

The ability of hemodilution to lower blood viscosity and increase cerebral blood flow has been proven experimentally; however, the optimum hematocrit for maximum oxygen delivery to ischemic brain tissue is not known, and a study was designed to determine this. Fifty dogs were selected for inclusion in the study using criteria based on changes in somatosensory evoked potentials at the time of arterial occlusion, which were found in a previous study to predict the development of a moderate infarction of relatively constant size. Infarctions were induced by permanent occlusion of the left middle cerebral artery and the azygous anterior cerebral artery. The animals selected for inclusion were divided into five groups of 10 dogs each: 1) a control group; 2) a group with 25% hematocrit; 3) a group with 30% hematocrit; 4) a group with 35% hematocrit; and 5) a group with 40% hematocrit. Isovolemic hemodilution was accomplished 1 hour after occlusion of vessels using dextran infusion and blood withdrawal. The animals were sacrificed after 6 days and infarction volume was determined from fluorescein-stained sections. Statistical analysis was performed using Student's t-test and one-way analysis of variance. Mean infarction volume for each group, expressed as a percentage of total hemispheric volume +/- 1 standard error of the mean, was 28.3% +/- 2.8% for the control group, 33.6% +/- 3.4% for the 25% hematocrit group, 17.1% +/- 2.2% for the 30% hematocrit group, 29.2% +/- 4.3% for the 35% hematocrit group, and 29.9% +/- 2.1% for the 40% hematocrit group. The 30% hematocrit group showed the smallest average infarction size and this size differed significantly (p = 0.02) from the average infarction size in the control animals. These results show that, in this model of focal ischemia, a hematocrit of approximately 30% is optimum for protecting the brain.