Ischemic edema in stroke. A parallel study with computed tomography and cerebrospinal fluid markers of disturbed brain cell metabolism

Mass Effect and Atrophy after Stroke

Cerebral edema, resolution of edema, and subsequent development of cerebral atrophy were studied prospectively in 83 patients with ischemic stroke with CT 3 days, 2 weeks, and 6 months post-stroke. Nineteen patients had large (diameter > 3 cm), 25 medium sized (diameter ≥ 1.5 ≤ 3 cm), and 15 lacunar infarcts (diameter < 1.5 cm). In 24 patients no infarcts were seen. Changes of Evans' ratio (ER), septum-caudate distance (S/C), and width of widest cortical sulci (SuW) were taken as markers of mass effect/atrophy. These parameters were within normal limits in most cases. However, when all CT scans performed in each patient were compared, changes of ER, S/C, and SuW became apparent as evidence of mass effect and subsequent atrophy development. Mass effect occurred in 81 percent and atrophy in 58 percent of patients with large infarcts. In patients with medium sized infarcts, mass effect occurred in 38 percent and atrophy in 45 percent.

Pilot results of in vivo brain glutathione measurements in stroke patients

Measurement of glutathione concentration for the study of redox status in subjects with neurological disease has been limited to peripheral markers. We recruited 19 subjects with large strokes. Using magnetic resonance spectroscopy we measured brain glutathione concentration in the stroke region and in healthy tissue to calculate a glutathione-ratio. Elevated glutathione-ratio was observed in subacute (<72 hours) subjects without hemorrhagic transformation (mean=1.19, P=0.03, n=6). No trend was seen when all subjects were considered (n=19, 3 to 754 hours, range=0.45 to 1.41). This technique can detect glutathione changes because of disease, and may be valuable in clinical trials of stroke and other neurological diseases.

Hyperacute therapies for childhood stroke: a case report and review of the literature

Objective. The optimal management of pediatric patients with arterial ischemic stroke (AIS) is not known. Despite this, goal-oriented, time-sensitive therapies geared to rapid reestablishment of arterial blood flow are occasionally applied with beneficial effects. The inconsistent approach to AIS is in part due to a lack of knowledge and preparedness. Methods. Case report of a 12-year-old male with right middle cerebral artery (MCA) occlusion resulting in dense left hemiplegia and mutism and review of the literature. Intervention(s). Mechanical thrombectomy, intra-arterial administration of rt-PA, vasodilators, and platelet inhibitors, and systemic anticoagulation and subsequent critical care support. Results. Restoration of right MCA blood flow and complete resolution of neurologic deficits. Conclusion. We report the gratifying outcome of treatment of a case of AIS in a pediatric patient treated with hyperacute therapies geared to arterial recanalization and subsequent neurologic critical care and review the pertinent literature. Guidelines for the emergency room management of pediatric AIS from prospective, randomized trials are needed.

Effect of endovascular hypothermia on acute ischemic edema: morphometric analysis of the ICTuS trial

INTRODUCTION

Pilot studies of hypothermia for stroke suggest a potential benefit in humans. We sought to test whether hypothermia decreases post-ischemic edema using CT scans from a pilot trial of endovascular hypothermia for stroke.

METHODS

Eighteen patients with acute ischemic stroke underwent therapeutic hypothermia (target = 33 degrees C) for 12 or 24 h followed by a 12-h controlled re-warm using an endovascular system. CT scans obtained at baseline, 36-48 h (right after cooling and re-warming) and 30 days were digitized, intracranial compartment volumes measured using a validated stereological technique, and the calculated change in CSF volume between the three time-points were used as an estimate of edema formation in each patient. Patients were grouped retrospectively for analysis based on whether they cooled effectively (i.e., to a temperature nadir of less than 34.5 degrees C within 8 h) or not.

RESULTS

Eleven patients were cooled partially or not at all, and seven were effectively cooled. Baseline demographics and compartment volumes and densities were similar in both groups. At 36-48 h, the total CSF volume had significantly decreased in the not-cooled group compared to the cooled group (P < 0.05), with no significant difference in mean volume of ischemia between them (73 +/- 73 ml vs. 54 +/- 59 ml, respectively), suggesting an ameliorative effect of hypothermia on acute edema formation. At 30 days, the difference in CSF volumes had resolved, and infarct volumes (73 +/- 71 ml vs. 84 +/- 102 ml, respectively) and functional outcomes were comparable.

CONCLUSIONS

Endovascular hypothermia decreases acute post-ischemic cerebral edema. A larger trial is warranted to determine if it affects final infarct volume and outcome in stroke.

Efficacy of multiple intraarterial papaverine infusions for improvement in cerebral circulation time in patients with recurrent cerebral vasospasm

OBJECT

Cerebral vasospasm that is caused by aneurysmal subarachnoid hemorrhage and that is refractory to maximal medical management can be treated with selective intraarterial papaverine infusions. The effects of single papaverine treatments on cerebral circulation time are well known. The purpose of this study was to assess the efficacy of multiple, repeated papaverine infusions on the cerebral circulation time in patients with recurrent vasospasm.

METHODS

A retrospective study was conducted in 17 patients who received multiple intraarterial papaverine infusions in 91 carotid artery (CA) territories for the treatment of cerebral vasospasm. Cerebral circulation times were measured from the first angiographic image, in which peak contrast was seen above the supraclinoid internal CA, to the peak filling of cortical veins. Glasgow Outcome Scale (GOS) scores assessed 12 months after discharge were reviewed. Cerebral circulation times in 16 CA territories were measured in a control group of 11 patients. Seventeen patients received a total of 91 papaverine treatments. Prolonged cerebral circulation times improved after 90 (99%) of 91 papaverine treatments. The prepapaverine mean cerebral circulation time was 6.54 seconds (range 3.35-27 seconds) and the immediate postpapaverine mean cerebral circulation time was 4.19 seconds (range 2.1-12.6 seconds), an overall mean decrease of 2.35 seconds (36%, p < 0.001). Recurrent vasospasm reflected by prolonged cerebral circulation times continued to improve with subsequent papaverine infusions. Repeated infusions were just as successful quantitatively as the primary treatment (mean change 2.06 seconds). The mean cerebral circulation time in the control group was 5.21 seconds (range 4-6.8 seconds). In five patients a dramatic reversal of low-attenuation changes was detected on computerized tomography scans. The mean GOS score at 12 months after discharge was 3.4.

CONCLUSIONS

The preliminary results indicate that multiple intraarterial papaverine treatments consistently improve cerebral circulation times, even with repeated infusions in cases of recurrent vasospasm.

Central nervous system damage during cardiac surgery assessed by 3 different biochemical markers in cerebrospinal fluid

Three cerebral biochemical markers, adenylate kinase (AK), neuron-specific enolase (NSE) and protein S-100, were determined in the cerebrospinal fluid (CSF) of male patients 24 h after coronary artery bypass grafting to investigate the extent of possible center nervous system (CNS) damage and relation to the type of oxygenator and the use of an arterial line filter. The patients were randomized into three groups for extracorporeal circulation (ECC); bubble oxygenator without an arterial line filter (Group I, n = 30), bubble oxygenator with a filter (Group II, n = 29) and a flat-sheet membrane oxygenator without a filter (Group III, n = 33). Pathologically high CSF levels of AK and NSE were found 24 h after ECC in respectively 93% and 95% of the patients. All protein S-100 concentrations were within the normal range. Isolated high CSF concentrations of AK, NSE and protein S-100 were observed in group I. Levels of AK and NSE were the lowest in group III, although there was no statistical difference between the groups. In conclusion, our study suggested that CNS damage caused by ECC involved neurons rather than glial cells. AK and NSE in the CSF seemed to be markers of ischaemic neuronal damage. Postoperative levels of biochemical markers in the CSF tended to be the lowest in the flat-sheet membrane oxygenator group.

Cerebrospinal fluid lactate dehydrogenase levels in early stroke and transient ischemic attacks

We examined the concentrations of lactate dehydrogenase in the cerebrospinal fluid of 25 patients with strokes and 15 patients with transient ischemic attacks less than or equal to 8 hours after the onset of the vascular event and in a control group of 21 patients. We found significantly higher concentrations in the stroke patients (40.9 +/- 14.5 units/l) than in the transient ischemic attack patients (11.8 +/- 2.9 units/l, p less than 0.001) and the controls (11.2 +/- 6.7 units/l, p less than 0.001). Among the stroke patients, we found a significantly higher lactate dehydrogenase concentration in those with cortical strokes (n = 12, 50 +/- 12.3 units/l) than in those with lacunar white matter infarcts (n = 5, 26.4 +/- 6.5 units/l; p less than 0.001) and those with basal ganglia infarcts (n = 8, 36.37 +/- 11.7 units/l; p less than 0.05). Our study offers a supplementary examination for diagnosing cortical or subcortical infarction during the early stage of the event, with the possibility of distinguishing precisely stroke from transient ischemic attack during the first hours after onset of the event.

Relationship between CT attenuation changes and post-traumatic CSF-CKBB-activity after severe head injury in man

In order to evaluate if it is practically possible to assess the volume of contused brain tissue from the CT pictures, a comparison has been carried out between the size of the cerebral contusion(s)--as estimated from the CT scans--and the post-traumatic CSF-CKBB activity, in a series of 29 patients with severe head injury. A clearance curve for the elimination of CKBB from the CSF was constructed. The relation between contusion volume and CSF-CKBB-activity was not statistically significant, while the relationships between contusion volume and outcome, and between CSF-CKBB, as estimated at 6 hours after from the clearance curve, and outcome, were.

Adverse effects on the brain in connection with isoflurane-induced hypotensive anaesthesia

As a marker of brain cell injury, adenylate kinase (AK) was measured in cerebrospinal fluid (CSF) in 10 patients given anaesthesia with isoflurane-induced hypotension for corrective surgery of dentofacial deformities. Nine out of 10 patients displayed a marked increase in CSF-AK postoperatively compared with preoperative values. The postoperative mean value displayed a 400% increase compared to the corresponding preoperative value. This difference was statistically significant (P = 0.001). The rise in CSF-AK was most probably the result of an enhanced efflux of AK into CSF subsequent to a presumed hypoxic injury to brain cells.

Adenylate kinase activity in the cerebrospinal fluid of hypoxic newborns

Adenylate kinase (AK) activity in the cerebrospinal fluid (CSF), described as a marker of brain edema and lesions in adults, was studied in 79 newborns with severe respiratory distress within 24 h after admission to the Intensive Care Unit (ICU). The CSF-AK activity was compared with CSF lactate concentration, CSF lactate dehydrogenase activity (LDH), and CSF and serum creatine kinase isoenzyme BB (CK-BB) activity. Newborns were divided into Group I with moderate to severe brain dysfunction and Group II with mild or no detectable brain dysfunction on discharge from the ICU. Mean CSF-AK activity (11.31 U/L) in Group I was significantly (p less than 0.001) higher than in Group II (2.82 U/L). Correlation between CSF-AK and CSF lactate was r = 0.714, p less than 0.01 and between CSF-AK activity and CSF-LDH activity was r = 0.550, p less than 0.01 in Group I. Preliminary data indicate that CSF-AK activity within 24 h after ischaemia is an indicator of hypoxic brain lesions in newborns. Its prognostic value for the infant's development remains to be determined by further study.

Adenylate kinase enzyme activity in cases of brain infarction

The adenylate kinase (AK) enzyme activity in plasma and CSF of acute brain infarctions was examined. The normal values of enzyme activity in plasma reached from 1.7-5.6 U/l, and in CSF from 0.23-0.71 U/l. According to this present classification a significant CSF increase in AK activity was found with semi-severe and severe brain infarctions. With the CCT an increased enzyme activity was shown with infarction in or close to the cortex. In no case was an alteration of AK activity in the serum sample. CSF samples showing blood contamination or pleocytosis led to false pathological results with examination of AK activity.

Glycerol and dextran combined in the therapy of acute stroke. A placebo-controlled, double-blind trial with a planned interim analysis

The results of clinical trials investigating various therapies in acute ischemic stroke have been inconsistent. The effect of glycerol therapy and a combination therapy of glycerol and dextran was evaluated in a double-blind, placebo-controlled study. Repeated neurologic examinations (Day 0, Weeks 1, 6, 12, and 24) according to a modified Mathew score were performed on 62 patients. Statistical analysis showed no superiority of either treatment compared with placebo in acute ischemic stroke. A retrospective estimation of the Type II error of the study yielded approximately p = 0.25. A major side effect was hemolysis in 98% of patients treated with glycerol.

The effects of acute cerebrovascular disease on serum and cerebrospinal fluid parameters

In order to evaluate the effects due to acute cerebrovascular disease on serum and cerebrospinal fluid (CSF) parameters, biochemical data from 312 patients were studied. In serum, CK, LD and alpha 1 protein fraction showed a moderately close relationship to the severity of stroke. In CSF there could not be observed any relation between enzymatic activities (LD, ASAT, CK) and the severity of the disease. The same result was found for CSF protein and the ratio CSF albumin/serum albumin, parameters which are indicative for Blood Brain Barrier (BBB) disturbances. Substances from the intermediate metabolism in the Central Nervous System (CNS) (lactate and pyruvate) were evidently raised in CSF and there was a clear relation between the CSF concentration and the severity of stroke. No indication for IgG immunoglobulin abnormalities in CSF was found. The concentration of neurotransmitter metabolite 5HIAA in CSF was significantly higher in (in)completed stroke than in Transient Ischaemic Attack.

CSF cyclic AMP and CSF adenylate kinase in cerebral ischaemic infarction

The severity of neurological deficits, size of hypodense zone on CT, concentration of cAMP and activity of adenylate kinase in cerebrospinal fluid (CSF) were evaluated at predefined intervals in the acute stage of supratentorial cerebral ischaemic infarction in 52 patients. Patients with cerebral infarction had raised activities of adenylate kinase CSF as compared with normal persons. Patients with marked neurological deficits, only slight improvement of neurological signs and large infarction zones on CT had higher average activities of adenylate kinase and lower concentration of cAMP in CSF. Alterations of CSF adenylate kinase and CSF cAMP values were most distinct on the 3rd day after the stroke. Reasons for the changes may be metabolic disorders following brain ischaemia.

Different patterns of CSF neurotransmitter metabolism in patients with left or right hemispheric stroke

In 30 ischemic stroke patients, divided into 2 groups depending on the side of their hemispheric cerebral lesion, the authors evaluated the levels of CSF homovanillic acid (HVA) and 5-hydroxyindoleacetic acid (5-HIAA). The changes of these metabolites in CSF samples collected 3, 14 and 25 days after stroke have been correlated to the clinical course. In both groups, which were similar in respect to the localization of the infarcted area and to the volume of the lesion, the levels of HVA and 5-HIAA increased in the first 2-3 days and gradually declined to normal values in the following 3 weeks, in parallel with the regression of neurological deficits. The increase of HVA and 5-HIAA was statistically significant only in left hemisphere-injured patients. A linear regression analysis between the clinical score values and the CSF levels of the two metabolites at different time-points of observation revealed a significant correlation only for the HVA in the left-lesioned patients.

CT enhancement after prolonged high-dose contrast infusion in the early stage of cerebral infarction

To demonstrate the BBB break-down on the CT image in the acute stage of cerebral infarction, a 3 hour continuous drip infusion of 200 ml of meglumine amidotrizoate, rather than the conventional bolus injection, was used. In this study, 22 examinations were carried out in 18 patients in whom cerebral infarction due to temporary or permanent obstruction of the cerebral artery was diagnosed by CT and angiography on admission. With each examination, the first CT was obtained prior to contrast infusion, and second immediately after the end of 3 hours of continuous contrast infusion. The EMI number was calculated at 3 regions of interest in the infarction. Within 3 days after stroke episode, 4 out of 5 patients with temporary vascular obstruction demonstrated enhancement, as well as 6 out of 9 patients with permanent vascular obstruction. Between 4 and 14 days after the stroke episode, all of 8 patients showed enhancement. To further clarify the extravasation of the contrast medium during the first 3 days of a cerebral infarction, a third CT scan was performed 3-hrs after finishing the contrast infusion in 4 patients. In these latter patients, blood was sampled at the time of each of the 3 CT series. The EMI number of the blood samples was also measured. In all 4 patients, the Gado's tissue-blood ratio (the EMI number of the CT lesion divided by that of the blood sample) was higher than 17.2% in the second, and higher than 54.7% in the third CT scan. Thus break-down of the BBB which was demonstrated by prolonged contrast infusion is an earlier event in human cerebral infarction than is usually accepted.

CT in hemispheric ischaemic attacks

The authors report the three principal characteristics of hemispheric ischaemic attacks on CT scan and their development over time: hypodensity, mass effect and contrast enhancement. Their pathophysiological mechanism and prognostic importance are discussed. The topography of the territories of the large arteries vascularising the brain and their main branches is described on the basis of twelve sections cut parallel to the cantho-meatal line. Finally, selected special aspects are described (misleading aspects, haemorrhagic infarcts, border zone infarcts).

Dementia--and adenylate kinase activity in cerebrospinal fluid

Adenylate kinase (AK) was determined in a lumbo-cisternal cerebrospinal fluid (CSF) gradient and in CSF from patients with mental deterioration. AK activity was unchanged in the gradient and different from albumin and albumin ratio (alb CSF/albumin serum), a blood-brain barrier marker in the same gradient. No differences between normal pressure hydrocephalus, multiinfarct dementia and dementia of the Alzheimer type could be found with regard to adenylate kinase activity in CSF.

Pathology and pathophysiology of brain ischaemia

The current views on the pathophysiology of cerebral ischaemia have recently been challenged by the new experimental results and animal models devised in recent years. These data are briefly reviewed. Ischaemic macroscopical and microscopical lesions in the human brain are outlined. Ischaemic changes can be either regional or total. The types, topography and aetiology of cerebral infarcts, lacunes and venous infarcts are described. The consequences of marked or total ischaemia affecting the whole cerebral blood supply are mentioned.

X-ray computed tomography (CT) study of small, deep and recent infarcts (SDRIs) of the cerebral hemispheres in adults. Preliminary and critical report

The evolution of CT signs of small, deep infarcts of the cerebral hemispheres in thirty adults, in the first five weeks, has been retrospectively studied. The relevant literature has been reviewed and an attempt has been made to present a synthesis, accompanied by a commentary. It is impossible now to give the frequency of each type of evolution, but the main data are as follows: The shortest delay of visibility of an hypodense area is about 17 to 19 h, but at 27 h the densities may still be normal. The evolution of the hypodense area is also variable: after a minimum attenuation is reached--at approximately 72 h--there is a risk of "fogging effect", which reduces the visibility of ischemic lesions; it could be seen from the end of the 1st week to the beginning of the 4th, but its frequency and its duration have yet to be better determined. In our series, contrast enhancement has been found in the gray matter of the basal ganglia between the 8th and the 22nd days--but according to some observations recorded in the literature, it may be found from the second to the twenty sixth day--and there was no obvious contrast enhancement in the white matter. The significance of the evolving CT signs is discussed in connection with the clinical applications, principally in the management of these patients, and with the attempts to correlate the clinical and CT findings.

Early mortality following stroke: a prospective review

Analysis of early deaths after stroke is important, since some deaths may be preventable. Previous studies have relied on retrospective and often incomplete clinical data, for comparison with pathological findings. The present study is based on 1073 consecutive stroke patients admitted to an intensive care stroke unit from a well-defined population. There were 212 deaths within the first 30 days, yielding a mortality rate of 20%. Clinical, radiological, and laboratory data were collected prospectively according to a standardized protocol. Autopsies were performed on 90 of the 212 patients, and CT scanning on a further 27. Early mortality after stroke exhibits a bimodal distribution. One peak occurs during the first week, and a second during the second and third weeks. The majority of deaths in the first week are due to transtentorial herniation. Of these, deaths due to hemorrhage usually occur within the first three days, whilst deaths due to infarction peak between the third and sixth day post ictus. After the first week, deaths due to relative immobility (pneumonia, pulmonary embolism and sepsis) predominate, peaking towards the end of the second week. Cardiac deaths occur throughout the first month, and unfortunately account for many deaths in patients with small functional deficits.

CSF lactate and CT findings in middle cerebral artery infarction. A comparative study

The extent of edema related to infarction assessed by computed tomography was compared with the CSF-lactate concentration in patients with middle cerebral artery (MCA) infarction on the first, third and seventh day following the stroke. A linear correlation between the extent of infarction edema and CSF-lactate level was most distinct on the third day. CSF-lactate concentration on the third day can be considered as a measure of the extent of the accompanying edema which in our study reached its maximum at this time in comparison to the first and seventh day.

Secondary brain stem hemorrhage in stroke

The occurrence of secondary brain stem hemorrhage was studied in 435 autopsies from patients with recent cerebral hemorrhage, infarction or ruptured cerebral aneurysms. The frequency of secondary brain stem hemorrhage was found to be 45% in cerebral hemorrhage, 15% in cerebral infarction, and 36% in ruptured aneurysms. In the majority of cases the secondary brain stem hemorrhage occurred a few days after the onset of cerebral hemorrhage or infarction. Ruptured aneurysms showed a more widespread temporal distribution of secondary brain stem hemorrhage. The median survival time was 2 days in cases of cerebral hemorrhage, 4 days in ruptured aneurysm and 4 days in cerebral infarction. The frequency of secondary brain stem hemorrhage was significantly lower in patients younger than 20 years. No significant difference was found in its distribution between the sexes. Secondary occipital lobe infarction was present in 3.5% of the patients. It is concluded that secondary brain stem hemorrhage is a common major contribution to the cause of death in stroke.

Oxypurines in cerebrospinal fluid as indices of disturbed brain metabolism. A clinical study of ischemic brain diseases

Using a HPLC method the concentrations of oxypurines were simultaneously measured in CSF of patients with acute cerebrovascular lesions (CVL) and global cerebral ischemia (GCI) in an attempt to study disturbed brain metabolism during cerebral oxygen deprivation. In cerebral infarction both hypoxanthine and xanthine gradually increased from normal levels at admission to pathologically increased on the fourth day from onset of symptoms. There was no correlation between these substances and the clinical score but the maximum CSF-hypoxanthine concentration was significantly correlated to the maximum lesion volume determined by computerized tomography. In GCI the hypoxanthine-xanthine concentrations were considerably increased less than 20 hours from onset of unconsciousness but the initial levels did not predict the final outcome. These findings suggest that the end products of nucleotide degradation accumulate rapidly in acute cerebral hypoxia but more gradually in CVL probably due to growing local edema with subsequent local hypoxia. In controls and patients with CVL the CSF-urate concentrations were positively correlated to those of CSF-albumin. However, in CVL the increase of urate was relatively much more pronounced than the increase of albumin indicating that urate is a sensitive marker of dysfunction of blood-brain barrier.

Pentobarbital protection from cerebral infarction without suppression of edema

We studied the mechanism of barbiturate protection from focal cerebral infarction in cats by examining in detail edema formation 72 hours after acute, permanent occlusion of the left middle cerebral artery (LMCA). Neurological function, gas exchange, vital signs, and intracranial pressure (ICP) were observed during the post-occlusion period, and infarct size and cerebral edema were measured after sacrifice. Infarct size was reduced only when pentobarbital was given before occlusion and continued for 24 hours. Edema formation was not suppressed even though the extent of infarction was. Clinical evidence of stroke developed and ICP rose in most cats after occlusion despite the presence of pentobarbital sufficient to reduce infarct size. Elevated ICP accounted for most premature deaths despite intensive cardiopulmonary support. Water and electrolyte changes in the ischemic hemisphere continued to develop throughout the 72 hour post-occlusion period in pentobarbital-treated cats, suggesting that resolution of edema was delayed by the drug. We conclude that pentobarbital reduces infarct size and attenuates the expected time course of ischemic edema in cats, but that the drug has little effect on the severity of edema that develops after arterial occlusion.

Brain hypoxia: the turning-point in the genesis of the migraine attack?

The hypothesis postulates that a brief episode of focal cerebral hypoxia occurs in every attack of migraine. Clinical biochemical and technical (EEG and CT scans) evidence is summarized suggesting that cerebral hypoxia is seen as the turning-point in the pathogenesis of the attack. It may be provoked by different mechanisms in different patients; the potential role of decreased oxygen supply and of increased oxygen need are reviewed and excess sympathetic drive is considered a potential key mechanism in a majority of patients. Whether or not focal hypoxia leads to a genuine migraine attack, depends largely upon the quality of the whirlpool of biochemical, vascular and hematological changes that follow the hypoxic episode. These changes are discussed and it is concluded that those which have been reported to occur during migraine attacks could be due to a preceding hypoxic event. Finally, the hypoxia viewpoint is confronted with some popular theories about the pathogenesis of migraine. It is found that the other points of view are compatible with the hypoxia hypothesis.

Release of adenylate kinase into cerebrospinal fluid during open-heart surgery and its relation to postoperative intellectual function

In 33 of 36 patients having open-heart surgery, the levels of adenylate kinase in cerebrospinal fluid (CSF-AK) were raised. No such increase was seen in 8 patients who had had lung operations without cardiopulmonary bypass. A significant increase in CSF-AK was recorded in 18 patients whose CSF was examined both preoperatively and postoperatively, and this increase was correlated with change in an index of intellectual function. Since there was no evidence of damage to the blood-brain barrier, these findings point to a causal relation between brain-cell injury during cardiopulmonary bypass and release of AK into CSF. Measurement of CSF-AK may therefore prove useful in research to improve the quality of open-heart surgery.