Elevated brain lactate accumulation and increased neurologic deficit are associated with modest hyperglycemia in global brain ischemia

Elevated brain lactate accumulation and increased neurologic deficit are associated with modest hyperglycemia in global brain ischemia

Resuscitation, 19 (1990) 271-289 Elsevier Scientific Publishers Ireland Ltd. 271 Elevated brain lactate accumulation and increased neurologic defici...

1MB Sizes 0 Downloads 15 Views

Resuscitation, 19 (1990) 271-289 Elsevier Scientific Publishers Ireland Ltd.

271

Elevated brain lactate accumulation and increased neurologic deficit are associated with modest hyperglycemia in global brain ischemia* JoAnne E. Natale, Susan M. Stantea and Louis G. D’Alecyb Departments of Physiology” and Surge@, The University of Michigan Medical School, Ann Arbor, MI (U.S.A.) (Received March 27th, 1989; revision received December 26th, 1989; accepted December 28th, 1989)

This study determined if hyperglycemia: (1) augments ischemic cerebral cortical lactate accumulation during complete cerebral ischemia; and (2) exacerbates subsequent neurologic morbidity and mortality. Dextrose (DSW, n = 8) or normal saline (n = 6) was administered i.v. prior to 10 min of global cerebral ischemia induced by normothermic cardiac arrest in dogs. Before arrest plasma glucose was significantly higher in the DSW-treated group than saline-infused (407 * 3 1 vs. 11 9 f 20 mg/dl, P < 0.05). By 6 h post-arrest, seven of eight DSW-infused dogs died, compared to one of six saline-infused dogs (P = 0.002). DSW-infused dogs showed significantly greater neurologic deficit at 2, 6, and 12 h post-arrest. In a complementary protocol, dogs were pretreated in the same manner, however, six cerebral cortical brain biopsies were taken before, during, and immediately after cardiac arrest. Plasma glucose was 320 + 17 mg/dl in the DSW-infused dogs and lower (P < O.OOl), 140 + 5 mg/dl, in the saline-infused group. Cerebral cortical lactate accumulation was slightly but significantly greater during ischemia and early reperfusion in animals receiving dextrose. Neither plasma nor cerebrospinal fluid (CSF) creatine kinase isoenzymes nor plasma or CSF lactate concentrations, measured during and for 25 min after cardiac arrest, served as a good prognostic indicator of 24 h neurologic morbidity or mortality. Therefore, induction of complete cerebral ischemia in the presence of moderate hyperglycemia is associated with profound neurologic dysfunction and striking mortality. A qualitative but not quantitative increase in brain lactate accumulation is consistent with the hypothesis that lactate may contribute to the increased severity of neurologic dysfunction with hyperglycemia. Cardiac arrest and resuscitation glycolysis - Brain biopsy

Lactate -

Neurologic function -

Cerebral ischemia -

Anaerobic

INTRODUCTION

Successful resuscitation from cardiac arrest is often accompanied by permanent neurologic dysfunction. If mild hyperglycemia is associated with a cardiac arrest or

*Support for this study was provided in part by PHS grant ROl NS25171-01. This work was presented in part at the 38th Fall Meeting of the American Physiological Society, San Diego, California, 1987. California, 1987. Address correspondence and reprint requests to: Louis G. D’Alecy, Department of Physiology, The University of Michigan Medical School, M7799 Medical Science II, 1301 Catherine Street, Ann Arbor, MI, U.S.A. 0300-9572/90/$03.50 0 1990 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland

272

a surgically-induced or spontaneous CNS ischemic episode, the attendent neurologic injury is exacerbated [l--8]. Although the mechanisms of hyperglycemia-exacerbated ischemic injury are not precisely understood, the most widely proposed mechanism is that during ischemia, anaerobic glycolysis is accelerated and its endproduct, lactate, accumulates [9-IS]. In the CNS, cellular injury has been associated with tissue lactate accumulation [ 13,141. However, neither laboratory nor clinical studies have assessed, in the same experimental setting, the effect of clinically-relevant levels of hyperglycemia on both neurologic dysfunction and ischemic lactate accumulation produced by cardiac arrest-induced complete global cerebral ischemia. The two major objectives of this study are: (1) To determine if mild hyperglycemia before and continuing after complete cerebral ischemia is associated with exacerbated neurologic deficit, and increased mortality; and (2) To determine whether such exacerbated neurologic deficit is quantitatively as well as qualitatively associated with an augmentation of cerebral ischemic glucose metabolism to lactate and the release of creatine kinase isoenzymes, a marker for cellular injury, into the plasma and cerebrospinal fluid (CSF). Two separate and complementary protocols using controlled cerebral ischemia, induced by cardiac arrest, have allowed both neurologic function and brain lactate accumulation to be quantified in the same animal model. MATERIALS

AND METHODS

This study consists of two complementary protocols, involing two separate groups of dogs. The function protocol determined whether infusion of 5% dextrose in sterile water (D5W) would exacerbate neurologic dysfunction and increase mortality subsequent to complete cerebral ischemia induced by 10 min of cardiac arrest. The complementary biopsy protocol determined whether a similar infusion of DSW would lead to the accumulation of higher concentrations of cerebral cortical lactate during the same ischemic interval and for the following 25 min. Surgical preparation, cardiac arrest and resuscitation procedures were similar for each protocol and have been previously described [ 16,171. Brief summaries of these procedures are provided below. Surgical preparation Fasted adult male mongrel dogs were premeditated with morphine sulfate and anesthetized with halothane for implantation of arterial and venous catheters and subdermal ECG electrodes. Body temperature was monitored with an esophageal temperature probe at the level of the heart and maintained at 39.0 & l°C with a heating pad and proportional controller. End expiratory CO, tension was continously monitored. To expose the heart for fibrillation and direct cardiac compressions during resuscitation, a left thoracotomy and pericardiectomy were performed. CSF pressure and samples were obtained in a subset of animals via a cannula inserted in the cisterna magna through the atlanto-occipital membrane. The animals used in the function protocol were randomly assigned to one of four treatment groups: (1) full surgical sham; (2) DSW-sham; (3) 0.9% NaCl (Saline); or

213

(4) D5W. Animals assigned to either the sham or saline groups received 500 ml normal saline i.v. prior to arrest, those in the sham-DSW and D5W groups received 500 ml DSW. All fluids were infused through a 16-gauge catheter inserted in the left saphenous vein during the surgical period. Cardiac arrest When the surgical preparation was complete, the animal was switched from breathing halothane/oxygen to room air ventilation. After corneal reflexes returned (stage 3, plane 1 of surgical anesthesia), artificial ventilation was halted and the heart directly, electrically fibrillated. Resuscitation After 10 min of ventricular fibrillation, ventilation was restored and direct cardiac compressions begun to maintain mean arterial pressure (MAP) between 100 and 125 mmHg. Forty pg/kg epinephrine, 1 mg/kg lidocaine, 4 meq/kg sodium bicarbonate, and 25 mg/kg calcium chloride were administered via the centrally placed i.v. line. Dopamine, 10 pg/kg/min, was infused for vasopressor support. Cardioversion was attempted after the resuscitation drugs were administered. Subsequent defibrillatory discharges and resuscitation drugs were administered as necessary to assist in the restoration of spontaneous circulation. After closing the chest, the dogs were ventilated until spontaneous ventilation ensued, and were extubated upon the return of the gag reflex. The dopamine infusion rate was adjusted to maintain MAP between 75 and 100 mmHg as long as necessary but no longer than 6 h. Intravenous infusion of test fluid was continued at a rate of 100 ml/h for the first 6 h post-arrest, then all animals received an infusion of normal saline at 60 ml/h until 24h. The above cardiac arrest and resuscitation procedure was modified for animals in the sham and DSW-sham groups. Animals were not placed on the room air ventilator subsequent to completion of surgical preparation, but remained on 0.5% halothane. The heart was fibrillated, and upon confirmation of ventricular fibrillation, direct cardiac compressions were immediately begun. Cardioversion was attempted when MAP exceeded 100 mmHg; resuscitation drugs were administered only as indicated. Halothane/oxygen was replaced with room air ventilation after the chest was closed. In this way these animals received the full extent of the surgical insult while sustaining the shortest period of ventricular fibrillation and resuscitation. The experimental groups (non-sham) were those with the extended (10 min) ischemia in contrast to the sham-groups which had the minimum ischemia but all the other surgical manipulations. The antibiotic spectinomycin (10 mg/kg i.m.) was administered to the recovery dogs and morphine sulfate provided for analgesia if behavior suggested the presence of pain. (No animals exhibited behaviors that required postoperative analgesia.) Dogs surviving to 24 h post-arrest and dogs in the biopsy protocol (at 35 min postarrest) were euthanized with 120 mg/kg sodium pentobarbital i.v. Postmortem examinations of the heart, lungs, and wound sites were conducted to identify iatrogeny. Animals were excluded from this study if their cause of death could be attribtuted to any cause (hemorrhage, fibrillation, etc.) other than neurologic

274

impairment. This experimental procedure conformed to the guidelines established by the American Physiological Society and the NIH (Guide for the Care and Use of Laboratory Animals, NIH Publication No. 85-23, 1985), and was approved by The University of Michigan Unit for Laboratory Animal Medicine’s Vertebrate Animal Use Committee (approval No. DOO1068D). Evaluation of neuroiogic deficit A well-standardized neurologic deficit score was assigned at 1, 2, 6, 12, and 24 h post-arrest by two investigators who were aware of each dog’s treatment [5,8,16,17]. Interobserver variability was resolved through consultation of the detailed description of each functional level [17]. Of the 100 points on the scale ranging from 0 (normal) to 100 (maximal deficit) , 18 are assigned to consciousness, 18 to respiratory function, 16 to cranial nerve function, 20 to spinal nerve function, and 28 to motor function as described previously. Method modifications for biopsy protocol The treatment groups are the same in the function and biopsy protocols except for the absence of a sham-DSW group in the biopsy study. (Because anaerobic metabolism is necessary for lactate accumulation and the sham groups have minimal ischemia time (anaerobic time) we did not feel justified in testing a sham-DSW group for lactate accumulation.) Left parietal, occipital, and temporal brain lobes were exposed for tissue sampling by craniotomy and excision of the dura mater. Six pairs of simultaneous blood and cerebral cortical samples were obtained just prior to arrest, at 5 and 10 min of ischemia, after 5 min of direct cardiac compressions (CPR), and at 25 and 35 min post-arrest when spontaneous circulation has been restored. A cannula was inserted into the cisterna magna to obtain CSF. Cerebral cortical samples were obtained for L-lactate analysis with a pneumatic biopsy drill since this instrument enables tissue samples to be obtained and frozen in less than 0.5 seconds [18]. This biopsy drill operated as follows: a rotating (10 000 rev./min) sharpened hollow bore (3 mm) needle was manually advanced (4-6 mm) vertically into the tissue and a cut plug of tissue aspirated by high vacuum onto a liquid nitrogen cooled strainer where it was dispersed. To assure rapid freezing of all surfaces of the tissue, Freon-22 was immediately aspirated into the freezing compartment. Approximately 50-150 mg of tissue was removed; the size of the sample depending largely upon the time the sampling tip was in the brain parenchyma. Once the sample was frozen, the vacuum was stopped, and the sample removed from the strainer cup and stored in liquid nitrogen prior to grinding into a fine powder at - 20°C. Each ground sample was added to a preweighed test tube containing 1.O ml cold 3 M perchloric acid and mixed. The tubes were then reweighed to calculate actual tissue weight, and 2.0 ml distilled, deionized water was added to the perchloric acid-brain mixture. The tubes were centrifuged at 3200 x g (3750 rev./min) for 20 min at 4OC. CSF was deproteinized immediately upon withdrawal in 4OC 0.6 M perchloric acid (1:2) and centrifuged at high speed for 1 min. Similarly, whole blood was immediately deproteinized in cold 1 M perchloric acid (1:3), vortexed, placed on ice for 15 min, spun at 3000 x g for 20 min at 4OC and stored at - 70°C [19]. All prepared supernatants were neutralized with 5 M K2C03 and stored at

215

- 70°C until fluorometric lactate analysis performed by the Biochemistry Core Facility of the Michigan Diabetes Research and Training Center [20]. Plasma glucose and total creatine kinase (CK) were analyzed spectrophotometrically. CSF and plasma CK isoenzymes (CK-BB and CK-MM) were separated electrophoretically and the magnitude of each determined with soft laser densitometry using a fluorescence source at 366 nm. Data analysis Multiple linear regression analysis was used to determine if the physiologic variables before cardiac arrest (Table I) were associated with neurologic outcome and survival. Comparisons of all outcome parameters were assessed with one way analysis of variance (ANOVA-Scheffe) among each of the three groups (see Results for Sham, Saline, and D5W groups). Neurologic deficit scores were compared parametrically with ANOVA with a Bonferroni correction and non-parametrically with Kruskal-Wallis test. Fisher’s exact analysis and Breslow survival curve analysis (BMDP statistical program) provided significance values for survival data. Cerebral cortical lactate curves for D5W and Saline groups were compared using area analysis. All average data are expressed as mean f 1 standard error of the mean (S.E.M.).

Average physiologic and resuscitation procedure variables for the function protocol (mean -+ Table I. 1 S.E.M.). Groups were compared by pairwise comparisons using ANOVA (Scheffe). Significant difference at the 95% level are indicated by: *for the sham vs. saline-infused groups, 5 for the saline vs. D5Winfused groups, and **sham vs. DSW-infused groups. MAP: mean arterial pressure; resuscitation time: time until MAP > 75 mmHg without mechanical support. Saline (n = 6)

Sham (?I = 7) Pre-arrest Operative time (minutes) Body weight (kg) Mean arterial pressure mmHg) Heart rate (beats/minute) End expiratory CO, (%) Arterial pH Deep body temperature (“C) Resuscitation Resuscitation time (minutes) Number of countershocks Ventilation time (minutes) Epinephrine inf. dose (Ccg/kg) Lidocaine dose (mg/kg) Sodium bicarbonate (meq/kg) Calcium chloride (mg/kg) Dopamine infusion time (minutes)

36 + 17.9 + 99 + 90*6 2.6 rt 7.37 f 38.5 f

1.8 + 1.7 f 12 f 2.9 f 1.1 i: 0.5 + 0’ 222

3 0.7 8 0.3 0.2 0.2

0.3 0.4 2 2.9+ 0.4 0.2;

D5W (?I = 8)

47 f 18.8 f 99*8 110 + 2.4 f 7.39 f 38.4 f

10 0.3 0.1 0.2

40 + 18.8 f 121 f 110 + 2.1 f 1.35 + 39.1 f

2.2 1.7 15 46.6 1.3 4.3 25 74

0.2 0.3 29 4.2 0.3 0.39 0.4 2.0

2.6 zk 0.5 1.2 + 0.2 29 + 3** 44.4 * 5.8** 1.7 f 0.2 7.0 + 0.6** 25” 164 f 68

f f f f + + + 2

5 0.5

5 0.4 9 14 0.2 0.2 0.2

276

0

4

8 12 18 20 SURVIVAL TIME (hours)

24

Twenty-four hours survival curves for Sham (n = 7) , Saline (a = 6) , and DSW (n = 8) Fig. 1. groups following 10 min of cardiac arrest and resuscitation. Comparison between Saline and D5W groups by Breslow profile analysis indicate significantly greater survival in animals receiving dextrose-free i.v. fhrids.

RESULTS

Function protocol Since sham-saline and sham-DSW groups did not differ in any of the pre-arrest or outcome variables (except plasma glucose), these animals were combined into one group designated as “Sham” in Table I and in Figs. 1 and 2. The two groups subjected to the 10 min of ischemia are designated as the “Saline” and the “DSW” groups. The average values for physiologic variables before cardiac arrest are provided at the top of Table I. Using multiple linear regression analytic techniques,

1

2 8 HOURS POST-ARREST

12

Mean ( f 1 S.E.M.) postresuscitation neurologic deficit (0 = no deficit; 100 = maximum de& Fig. 2. tit) for all live dogs. Sample sizes are indicated in or near each bar. The mean of the Sham group is represented by the open bars: the Saline group, the cross-hatched bars; and the DSW group, the shaded bars. The mean Sham neurologic scores are consistently significantly lower than the means of the other two groups at all times. The + indicates times when mean neurologic scores from the Saline group are significantly lower than the D5W group by Kruskel-Wallis test and ANOVA with Bonferroni correction.

217

there was no evidence of an association of the pre-arrest variables (operative time, body weight, MAP, heart rate, arterial pH, end expiratory CO,, and body temperature) with survival time, nor was there any significant or consistent relationship between these variables and neurologic deficit scores. No statistically significant differences among the groups were detected in physiologic variables before arrest. However, MAP and heart rate tended to be lower in the Sham group, compared to the other groups presumably due to the effects of continued halothane anesthesia. All 21 dogs were successfully resuscitated, defined as return of spontaneous circulation capable of maintaining MAP > 100 mmHg, and no dogs died from non-neurologic complications. Resuscitation procedure variables are presented in the lower half of Table I. Analysis of variance with pairwise comparison of the groups showed that on the average, D5W group required more sodium bicarbonate to combat arterial acidosis, and spent almost twice as long on the ventilator than Saline group. However, the D5W group required an average of only 24 s longer to resuscitate, and received the same amounts of epinephrine and lidocaine as Saline group. Dogs in the sham groups required smaller doses of all resuscitation drugs than the groups exposed to the full 10 min of ventricular fibrillation. Urine output in the first 6 h post-arrest by D5W group was 78 + 17 ml/h, not different from Saline group (149 by Student’s t-test). f 39 ml/h P = 0.102 Pre-arrest infusion of 500 ml of D5W during surgical instrumentation significantly elevated arterial plasma glucose (Table II). In Sham and Saline groups plasma glucose levels rose immediately post-arrest, yet remained significantly lower than in the D5W groups at 0.25, 1, and 2 h post-arrest. Comparison of these data by profile analysis indicates no detectable difference between the Sham and Saline group curves, but significantly higher plasma glucose (P< 0.01) in the D5W group than in Saline or Sham groups. In Table II, glucose concentrations in CSF are reported for sham-DSW, saline-, and DSW-infused dogs. D5W infusion sham-saline, significantly elevated and maintained higher CSF glucose level in spite of a rapidly diminishing blood-to-CSF glucose concentration gradient. Twenty-four hours survival curves are illustrated in Fig. 1. Only one of 8 dogs receiving glucose lived longer than 8 h, and all were dead by 16 h post-arrest, yet 83% of dogs (five of six) in the Saline group were alive at the same time (P = 0.003, Fisher’s exact test at 16 h). Breslow survival curve analysis indicates that the Saline group had increased survival in the first 24 h post-arrest compared to the D5W group (P = 0.002). All dogs of the Sham group (saline and DSW-infused) lived 24 h. The Sham dogs showed significantly less mortality than Saline group (P = 0.013 by Breslow test, P = 0.021 by Fisher’s exact test at 24 h) and D5W group (P = 0.000 by both tests). The effect of dextrose administration on neurologic deficit is similarly striking. Mean neurologic deficits at 1,2,6, and 12 h post-arrest are illustrated in Fig. 2. During the first 12 h post-arrest, Saline animals tended to improve neurologically, and had consistently and significantly less deficit than the D5W group of dogs at all times past 1 h. Only D5W group of dogs exhibited convulsive-like activity (7 of 8 dogs) in the first 12 h post-arrest (P = 0.023). All Sham dogs (saline and D5Winfused) had zero neurologic deficit at 24 h. The mild neurologic deficit in Sham animals at 1 and 2 h post-arrest can be attributed to residual sedative effects of

Sham@ = 7) Saline (n = 6) DSW (n = 8)

Body temperature

Sham (n = 3) Saline (n = 3) DSW (n = 3)

ICP (mmHg)

Sham@ = 7) Saline (n = 6) DSW (n = 8)

(“C)

-

-

-

-

7.34 f 0.044 7.40 + 0.044 7.37 2 0.037

35 + 3 56 f 2 4425 49 f 14

CSFglucose (mg/dl) Sham@ = 2) Sham-DSW (n = 2) Saline (n = 3) D5W (n = 2)

PH

130 f 12 121 f 7 119 +: 9

-

Pretreatment

Sham (n = 4) Sham-DSW (n = 3) Saline (n = 6) DSW (n = 8)

Plasma glucose (mg/dl)

Variable

f 20** Yk145 fj + I$ + 319

38.5 f 0.2 38.4 k 0.2 39.1 + 0.2

9.7 + 1.2 9.3 + 2.7 12.3 + 0.3

7.37 + 0.02 7.39 + 0.01 7.35 + 0.02

< 35 76 f 21 < 35 68 f 33

128 409 119 407

Prearrest 18** 339 5 19$ 213

-

13.0 f 1.5’; 11.3 + 4.4 31.0 + 2.09

7.37 f 0.01 7.38 f 0.02 7.36 + 0.02

< 35** 112 2 5# < 35s 145 + 3fj

139 f 366 + 156 + 324 +

0.25 h

13** 195 5 ll$ 169

38.4 + 0.6’ 39.0 + 03 39.8 + 0.5

15.0 + o** 3.7 f 1.3 22.2 -c 4.85

7.31 -t 0.01 7.39 f 0.01* 7.36 f 0.02

37 + 2*+ 138 + 39 tj < 35$ 148 f 39

129 f 256 f 147 f 272 f

lh

8** 129 $ 18s 179

38.5 f 0.6) 39.0 f 0.23 40.2 + 0.46

3.7 -c 7** 5.7 + 2.4 25.0 + 3.5$

7.31 It 0.01 7.40 + 0.02; 7.39 + 0.017**

< 35** 169 f 39 5 < 35$ 177 + 95

105 f 230 f 137 f 259 +

2h

8** 423 5 19 235

41.9 + 0.5’9

39 .3 -c 0 .45** 38.3 + 0.12

5.3 f 9+* 5.7 2 1.9 27.0 + 7.05

7.34 f 0.01 7.40 f 0.02 7.40 + 0.046

< 35** 183 f 35 5 35 f 3$ 186 2 79

92 -t 168 + 128 k 198 f

3h

10 19 4 16

38.5 + 0.3?** 39.6 + 0.4’ 42.1 f 0.Q

5.3 k 2** 12.7 + 1.9 21.0 + 2.1

7.36 + 0.01 7.38 + 0.01 7.29 -t 0.105

< 35** 161 k 49 5 < 35$ 146 + 115

92 + 150 f 92 + 99 f

6h

, and body temperature in sham, saline, and DSW-infused animals (function protocol). Samples sizes are indicated as a superscript if they differ from the size indicated on the left. Since 35 mg/dl was the lower limit of the glucose assay, values below 35 are indicated as < 35 mg/dl. Significant differences at the 95% level (ANOVA with Scheffe) are indicated by: *for the sham vs. saline-infused groups, 5 for the saline vs. DSW-infused groups, **sham vs. DSW-infushed groups, $ for the sham-DSW vs. saline-infused groups, 3 5 for the sham vs. sham-DSW groups, and # for the DSW-infused vs. sham-DSW groups.

Table 11. Plasma glucose, CSF glucose, arterial pH, intracranial pressure (ICP)

:

219

morphine. Despite the exclusion of dead animals from neurodeficit analysis, the mean score remained significantly higher (less function) in the D5W group than either the Saline or Sham groups, even though the sample size of the D5W group declines at each time point after 1 h (Fig. 2). Glucose was consistently detected (> 35 mg/dl) in CSF post-arrest only in the D5W groups (Table II). The sham-D5W group with normal recovery and the D5W group subjected to 10 min of cardiac arrest had similar CSF glucose concentrations. intracranial pressure (ICP) , as indicated by cisternal pressure, tripled post-arrest in DSW-infused dogs (Table II), and remained high for 6 h post-arrest. In salineinfused animals, ICP fell at 1 h post-arrest and remained steady before rising at 6 h. Since pre-arrest body temperature is an important determinant of neurologic recovery following cerebral ischemia, cardiac arrest was not induced until esophageal temperature was between 38.0°C and 40.0°C. Body temperature was not controlled post-arrest. At 3 and 6 h post-arrest (Table II) , esophageal temperatures in the D5W group were significantly higher than in other dogs (42OC vs. 39OC). Biopsy protocol Comparison of physiologic variables before arrest among the Sham, Saline, and D5W groups in the biopsy protocol revealed no differences except a lower heart rate in the Sham group, presumably due to halothane anesthesia during the cardiac arrest (Table III). Saline and DSW-infused animals were similar with respect to ease of resuscitation as indicated by doses of resuscitation drugs and resuscitation time.

Average values f 1 S.E.M. for physiologic and resuscitation procedure variables for the Table III. biopsy protocol. Groups were compared using pairwise comparisons using ANOVA with Scheffe. Significant differences at the 95% level are indicated by: *for the sham vs. saline-infused groups, 5 for the saline vs. DSW-infused groups, and **sham vs. DSW-infused groups. MAP: mean arterial pressure; resuscitation time: time until MAP > 75 mmHg without mechanical support. Sham (n = 4) Pre-arrest Operative time (min) Body weight (kg) MAP (mmHg) Heart rate (beats/min) End expiratory CO, (%) Arterial pH Deep body temperature (OC) Resuscitation Resuscitation time (min) Number of countershocks Epinephrine dose @g/kg) Lidocaine dose (mg/kg) Bicarbonate dose (meq/kg) CaCl, dose (mg/kg)

44f3 18.5 f 108 + 46 + 2.9 + 7.37 f 38.6 f

3.5 1.7 2 1.2 1.0 0.0

1.2 7 4* 0.2 0.4 0.2

f 0.2* f 0.5 f 2* +: 0.2 f 1.0* & o.o*

Saline (n = 7)

DSW (n = 7)

44&5 17.8 f 0.8 122 + 12 95 * 8 3.3 f 0.2 7.40 f 0.3 38.5 f 0.3

48 f 18.0 f 126 f 93 f 3.4 f 7.36 + 38.6 f

5 0.6 11 7** 0.3 0.2 0.2

8.6 1.6 74 2.3 5.4 28.5

6.6 f 2.7 f 104 f 3.0 + 6.2 + 32.0 2

1.0 0.9 20** 0.6** 0.6** 4.6**

+ 2 f + ? 2

2.1 0.4 8 0.4 0.6 3.6

280

w

300-

8 2 w

200-

D5W (nd)

Saline (n=7)

.

Sham (n=4)

K

2 lootn

9

n

ama

C_

Spontareos circuItion

CPA

0 &&,O

5

10 15

25

35

MINUTES POST-ARREST Fig. 3.

Arterial plasma glucose in mg/dl f 1 S.E.M. D5W group have statistically higher plasma glucose levels at all times post-treatment (ANOVA and profile analysis). No difference was found between Sham and Saline groups at any time point.

Sham dogs received significantly less resuscitation drugs since these drugs were administered only as necessary. Administration of DSW prior to cardiac arrest significantly elevated plasma glucose to 320 mg/dl, compared to 140 mg/dl in saline-infused groups (Fig. 3). Plasma glucose concentrations did not differ between these groups before treatment, and Sham animals showed little change in plasma glucose levels post-arrest and resuscitation. All 18 dogs in thebiopsy protocol were successfully resuscitated. Cerebral cortical lactate concentration rose sharply during ischemia produced by cardiac arrest in both DSW and Saline groups (Fig. 4). The profile of cerebral cortical lactate levels in animals receiving dextrose indicated significantly greater (P < 0.05, area analysis)

L

-

s

&Sham

(n=4)

cadiacanes(

spomansous ci,culatbn

CPR

1

0

5

10

15

25

35

MINUTES POST-ARREST

Fig. 4. Cerebral cortical lactate concentrations in pmol/g wet tissue mass + 1 SEM. Cerebral cortical lactate concentrations in Sham dogs were significantly less than in the other two groups (ANOVA, alpha level = 0.05) at all times. Cerebral cortical lactate profile was significantly greater in animals receiving D5W compared to the other groups.

281 12 3 L E

10

2

6

i!i

4

b

2

8

0

7

I 0

cmsaomnl cm SPomnsow cimblbn , , , , 5 10 15, 25 35 MINUTES POST-ARREST

Fig. 5. Average CSF (top figure) and blood lactate concentrations in mM f 1 S.E.M. CSF lactate concentration in Sham dogs were significantly lower (ANOVA) than levels in Saline group at lo”, and both Salineand DSW dogs at 15.25, and 35 min. Similarily, blood lactate levels in Sham dogs were significantly lower than values from Saline and DSW groups at 5, 15,25, and 35 min (ANOVA). At 25 min, blood lactate concentrations were significantly greater in the DSW dogs than in Saline dogs.

lactate accumulation during ischemia and early reperfusion. Following reperfusion, cerebral cortical lactate in DSW-infused dogs fell toward levels in Saline group. In striking contrast, cortical lactate in the Sham group of dogs was not significantly increased by cardiac arrest and immediate resuscitation. CSF and arterial blood lactate profiles were distinct from the brain lactate (Fig. 5). CSF lactate rose gradually during the CPR period, reaching levels similar to cerebral cortical concentrations by 35 min post-arrest. D5W treatment did not alter the pattern of CSF lactate concentration in ischemia. Both blood and CSF lactate concentrations were virtually unchanged from preischemic levels in the Sham group. Similarly in ischemic groups, arterial lactate concentration did not rise until the reperfusion period. Unlike CSF lactate, blood lactate levels were higher at 25 min post-arrest in DSW group compared to Saline group. The potential relationship between blood lactate and cerebral cortical lactate concentration can be examined by plotting values from simultaneous samples of blood and cerebral cortical lactate at six times during cardiac arrest and resuscitation (Fig. 6). A similar comparison can be made between CSF lactate and cerebral cortical lactate concentrations (Fig. 6).

1u

0

BLOOD

0

20 30 LACTATE (mM)

10 CSF

20 LACTATE

30

(mM)

Fig. 6. Effect of blood lactate (mM, top panel) or CSF lactate (mM, bottom panel) on cerebral cortical lactate concentration @mol/g). The bold line is the line of identity. Symbols represent simultaneous tissue and blood samples at six times during cardiac arrest and resuscitation for all dogs (treatment groups are not distinguished).

Table IV gives the mean plasma CK-MM and -BB and CSF CK-BB activity for three treatment groups. Plasma CK-MM activity in all groups increased following cardiac arrest; little to no increase was noted in CSF or plasma CK-BB. Neither CSF nor plasma CK-BB levels were correlated with neurologic deficit in any of the three groups. DISCUSSION

The new information obtained from these studies concerns four separate but related points. First, it has not been previously demonstrated that clinically-relevant doses of i.v. dextrose exacerbates the severe neurologic deficit induced by 10 min of normothermic cardiac arrest. In previous studies from this laboratory, cardiac arrest was for 6 min [8] or 8 min [5]. Second, ischemic cerebral cortical lactate accumula-

DSW (n = 5)

CSF CK-BB (WA) Sham (n = 4) Saline(n = 7)

Saline@ = 7) DSW (n = 5)

Plasma CK-BB (N/l) sham (n = 4)

Sham (n = 4) Saline@ = 7) DSW (n = 5)

Plasma CK-MM (Iv/r)

Variable

undet. undet. 45 f 10

undet. 19 f 5 0

undet. 115 f 16 62 f 9

Pretreatment

undet. 6.5 f 5.0 undet.

2.5 f 2.561 undet.I

28 f 14 40 + 10 45 f 12

367 f 81 443 2 206 308 -c 110

5 min

undet.

28 f 15 49 f 21 51 f 8

321 f 661 434 f 187 299 f 109l

Prearrest

10 f 66 2.2 f 2.2

undet.

34 + 17 71 + 42 29 f 12

376 f 68 452 + 218 241 f 88

10 min

33 f 20 44 f 22’ 7.5 f 7.5’

32 f 12 76 f 38 55 f 12

403 f 60 451 + 198 247 ZJZ52

15 min

40 f 143 282 f 916 102 f 42

42 f 20 145 + 66 87 f 9

580 f 75 734 k 348 491 + 68

25 min

124 f 56* 825 + 2199 220 f 56

659 + 37 186 f 96” 96 * 31

833 f 86 1216 f 6836 1350 + 352

35 min

Plasma and CSF CK isoenzymes (NJ/I) during cardiac arrest and resuscitation. Samples sizes are indicated as a superscript if they differ from the size indicated on the left. undet.: undetectable levels. Significant differences at the 95% level are indicated by: *for the sham vs. saline-infused groups, 5 for the saline vs. DSW-infused groups; significant differences (paired f-test) between pre-arrest and 35 min points are indicated by q.

Table IV.

284

tion was significantly greater in hyperglycemic animals. Taken together these results are consistent with the hypothesized injurious role of tissue lactate in post-ischemic neuronal dysfunction. Third, animals subjected to the same surgical manipulation but resuscitated immediately (sham group), rather than after 10 min of ischemia, show no morbidity or mortality, regardless of blood glucose concentration. A fourth finding concerns the potential weaknesses of biochemical assessment of neurologic injury from early CSF sampling. Each of these points is discussed further below. Hyperglycemia, ischemic neurologic dysfunction and cerebral lactate accumulation In-hospital cardiac arrests often occur during a period when the patient has received intravenous fluids which frequently contain 5% dextrose. The stress of the hospitalization and the use of dextrose containing intravenous fluids contribute to moderate (200-400 mg/dl glucose) hyperglycemia in this setting. This study tested for an association among induced moderate hyperglycemia, augmented ischemic cerebral lactate accumulation, and subsequent neurologic dysfunction. The slight but statistically significant increase in the cerebral lactate profile we observed does not seem to be sufficient to account for the dramatic functional changes between DSW and Saline groups. That is, the association of ischemic lactate accumulation and subsequent function appears qualitative rather than quantitative. It should be noted, however, that Myers and others have shown that a threshold exists for cerebral necrosis when tissue lactate exceeds 16-20 pmol/g tissue [ 10,151 and such a ‘threshold effect’ could explain this lack of a stronger correlation. Further evidence suggests that lactate concentrations below this threshold are not neurotoxic [16]. In addition, the small increases in brain lactate between the 5 and 10 min sampling time does not correlate with the marked increase in neurologic deficit associated with increasing cardiac arrest time from 5 or 6 min [5] to 10 min. Likewise the addition of dextrose produces a minimal increase in cortical lactate at the 5 min sampling time but in recovery studies after 6 min of arrest [5], the addition of dextrose produced a marked and significantly worse neurologic outcome. Thus while these data generally support the theory that hyperglycemia exacerbates post-ischemic CNS injury, these data do not conclusively establish that damage is quantitatively linked to tissue lactate accumulation. It should be noted that in each case cardiac arrest is assumed to produce zero cerebral blood flow such that variations in “incomplete” ischemia should not be a factor in this model. Cortical brain lactate has relevance to potential deficits of cortical origin, however, deeper brain structures may have markedly different patterns of lactate accumulation. Thus, these data cannot exclude other, perhaps deeper regional changes in lactate correlating more closely with neurologic outcome. Not only was tissue sampling limited to specific cortical regions, but all samples were obtained during the ischemic and immediate reperfusion periods (35 min post-arrest). Since neurologic deficit is dramatically different between dextrose and saline-treated groups, tissue sampling during the prolonged recovery period (12-24 h) may have revealed a correspondingly dramatic dichotomy in lactate accumulation whioh developed long after the resuscitation, but the literature refutes this proposition [21,21]. In rat and canine models of complete cerebral ischemia, Siesjo [21] and Michenfelder [22] have

285

demonstrated that cerebral cortical lactate concentrations return to pre-ischemic levels by 60 min of reperfusion. In a partial ischemia model Welsh et al. [12] reported sustained lactate levels after 90 min of recirculation following 15 min of ischemia in cats. Interpretation of these later studies is further complicated by the marked levels of hyperglycemia (greater than 700 mg/dl) in both glucose-treated and control animals [ 121. Therefore in the complete ischemic situation, we initially hypothesized that lactate accumulation during and immeditately following ischemia would have been predictive of the ultimate neurologic outcome. While a qualitative association was established, a quantitative link between brain lactate accumulation and exacerbated injury was not apparent in these data. Cerebral cortical lactate concentrations exceeded blood lactate levels in almost every instance, suggesting that cerebral cortical lactate accumulation is due to brain ischemic glycolytic metabolism, not to passive lactate flux from blood to brain. Similarly, cerebral cortical lactate levels exceeded CSF lactate concentrations, thus providing additional evidence that cerebral cortical lactate accumulation is not due to passive lactate flux. In addition, it is apparent that measurement of CSF lactate does not accurately reflect cerebral cortical lactate levels, and therefore should be interpreted cautiously as an index of brain lactate accumulation. A relationship between poor neurologic outcome and persistent ICP above 1520 mmHg has been demonstrated in other studies of brain injury and stroke [2326]. ICP was markedly elevated in DSW-infused dogs; whether the elevated ICP contributed to or was the result of the higher incidence of seizure-like activity in these same animals is unknown. It is, however, provocative that both Myers [ 1l] and Siemkowicz [27] observed similar seizure-like myoclonic activity and grossly demonstrable brain edema (and Evans blue extravascularization) in monkeys pretreated with glucose prior to 14 min of circulatory arrest. It is recognized that in a variety of tissues [28-311, temperature contributes to ischemic outcome. DSW-infused dogs became hyperthermic by 3 h after resuscitation. The elevation in body temperature may be due to increase in glycolytic metabolism (due to increased substrate availability), less heat loss, increase in heat generation by seizure-like myotonic activity, or an elevation in circulating cytokines, such as tumor necrosis factor or interleukins, causing fever [32,33]. Lastly, the combination of ischemia and hyperglycemia may cause hypothalamic damage and destruction of the thermoregulatory neurons. Regardless of the initiating cause, it is likely that the delayed postresuscitation hyperthermia further exacerbated the neurologic dysfunction in the DSW-treated group [3 11. Sham protocol Animals in the Sham groups provide important controls in these studies and highlight the utility of the model. Since all sham animals survived the entire recovery period with minimal to no neurologic deficit, it is reasonable to conclude that the open chest surgical preparation, the ventricular fibrillation/defibrillation, and the resuscitation procedures do not significantly impact on 24-h morbidity or mortality. The ischemic morbidity and mortality is correlated with the intentional, controlled ischemia of longer duration including 6 [5], 8 [8], or 10 min. Similarly, the sham group demonstrates that the serial brain biopsies alone have minimal impact on tis-

286

sue lactate levels until possibly the final (sixth) sample. Furthermore, blood and CSF lactate concentrations remain generally unchanged in Sham animals. Neurologic deficit scores in Sham animals indicate that 1 and 2 h scoring are likely confounded by the neurologic effects of the morphine (administered as a premedication) and possibly by the residual effects of anesthesia. By 6 h post-arrest, however, Sham animals are generally neurologically normal, and are able to stand and drink freely with little or no motor deficit. More interestingly with regard to the sham group given 5% dextrose, these animals had plasma glucose levels comparable to the DSW-infused group exposed to 10 min of ventricular fibrillation, yet their neurologic deficit scores were not different from normoglycemic sham animals. Thus pre-ischemic plasma glucose, alone, does not predict post-arrest neurologic function in the absence of significant ischemia time. Assessment of neurologic injury In this study, neurologic injury was primarily assessed using the lOO-point neurologic deficit score. In addition, we examined CK and lactate in CSF as potential early indices of neuronal injury. In accord with previous studies we found virtually exclusively the brain specific CK-BB dimer in spinal fluid. The high levels of CK-MM in the plasma render the contribution of enzyme from brain to. the CSF insignificant [34-361. All animals had CK-BB present in pre- and post-arrest plasma samples. Our interpretation is that this initial CK-BB does not originate from the brain because: it did not correlate with adverse neurologic outcomes, and 28-51 IU/L of CK-BB appeared in plasma prior to arrest while undetectable levels of CK-BB were detected in the CSF at the same time. The source of this plasma CK-BB is unknown but may be related more to the thoracic trauma of resuscitation than to brain damage [37]. Previous studies have suggested a correlation between CK-BB in CSF and neurologic damage at 48-72 h after injury, with little to no CK-BB activity prior to 48 h [34,38-401. We were likewise unable to detect CK-BB in CSF early after injury further supporting the position that CK-BB has limited early prognostic value. CSF lactate concentration sampled during the ischemic episode or immediately after had little relationship to either cerebral cortical lactate levels, or subsequent neurologic recovery. A kinetic model for cerebral lactate proposed by Kuhr et al. confirms the net transport of lactate out of CNS cells during ischemia is rate limiting [41]. This study further correlates augmented neurologic dysfunction with exacerbated ischemic cerebral cortical lactate accumulation in animals receiving dextrose-containing intravenous fluids. Additionally, the impact of open chest ventricular fibrillation and immediate resuscitation on CNS function were evaluated in sham animals, and demonstrated minimal to no adverse effects on neurologic function in the absence of intentially prolonged ischemia (> 3 min). These data do not, however, establish causal links between lactate accumulation and neurologic dysfunction and question the early prognostic utility of CSF lactate or CK-BB. Furthermore, the use of glucose administration prior to cardiac arrest in this stud+, may limit the generalization of these data to in-hospital cardiac arrest where administration of dextrose-containing intravenous fluids is common. Therefore, induction of complete cerebral ischemia in the presence of moderate hyperglycemia is associated with

287 Beurologic dysfunction and striking mortality. However, a qualitative but not a quantitative increase in brain lactate accumulation supports the hypothesis that lactate may contribute to the increased severity of neurologic dysfunction with hyperglycemia.

profound

ACKNOWLEDGEMENTS

The authors are grateful to Steven W. Ressler, Robert J. Schott, and Robert Takla for their excellent technical assistance, and to Linda Annesley and Dennis Martin, Ph.D. of the Biochemistry Core Facility of the Michigan Diabetes Research and Training Center for the lactate analyses. The authors acknowledge D.A. Pelligrino for assistance in the development of the brain biopsy technique. Thanks also to Puritan-Bennett Corporation, Westmont, IL for the gifts for the Compact-75 Anesthesia Machine and the Bennett Anesthesia Ventilator, and Ames Division of Miles Laboratories, Inc. for the gift of the Seralyzer Reflectance Photometer. Lastly we acknowledge Physio-Control of Redmond, WA for the Lifepak-3 defibrillator/ monitor. REFERENCES 1 2 3

4 5

6

7

8

R.E. Myers and S. Yamaguchi, Nervous system effects of cardiac arrest in monkeys. Arch. Neurol., 34 (1977) 65-74. E. Siemkowicz and A.J. Hansen, Clinical restitution following cerebral ischemia in hypo- normoand hyperglycemic rats. Acta Neurol. Stand., 58 (1978) l-8. W.A. Pulsinelli, D.E. Levy, B. Bigsbee, P. Scherer and F. Plum, Increased damage after ishemic stroke in patients with hyperglycemia with or without established diabetes mellitus. Am. J. Med., 74 (1983) 540-544. W.T. Longstreth and T.S. Inui, High blood glucose level on hospital admission and poor neurological recovery after cardiac arrest. Ann. Neurol., 15 (1981) 59-63. L.G. D’Alecy, E.F. Lundy, K.J. Barton and G.B. Zelenock, Dextrose containing intravenous fluid impairs outcome and increases death after eight min of cardiac arrest and resuscitation in dogs. Surgery, 100 (1986) 505-511. 6 D.R. LeMay, S. Neal, G.B. Zelenock and L.G. D’Alecy, Paraplegia in the rat induced by aortic cross clamping: Model characteristics and glucose exacerbation of neurologic deficit. J. Vast. Surg., 6 (1987) 383-390. E.F. Lundy, T.D. Ball, M.A. Mandell, G.B. Zelenock and L.C. D’Alecy, Dextrose administration increases sensory/motor impairment and paraplegia following infrarenal aortic occlusion in the rabbit. Surgery, 102 (1987) 737-742. E.F. Lundy, J.E. Kuhn, J.M. Kwon, G.B. Zelenock and L.C. D’Alecy, Infusion of five percent dextrose increases mortality and morbidity following six minutes of cardiac arrest in resuscitated dogs. J. Crit. Care, 2 (1987) 4-14. R.E. Myers and S. Yamaguchi, Effects of serum glucose concentration on brain response to circulatory arrest. J. Neuropathol. Exp. Neurol., 35 (1976) 301. R.C. Myers, A unitary theory of causation of anoxic and hypoxic brain pathology, in: Advances in Neurology, Vol. 26: Cerebral Hypoxia and Its Consequences. Editors: S. Fahn, J.N. Davis and L.P. Rowland, Raven Press, New York, 1979, pp. 195-213. R.C. Myers, Lactic acid accumulation as cause of brain edema and cerebral necrosis resulting from oxygen deprivation, in: Advances in Perinatal Neurology, Vol. 1. Editors: R. Korobkin and C. Guilleminault, SP Medical and Scientific Books, New York, 1979, pp. 88-l 14.

288 12

13 14 15 16 17

18

19 20 21 22 23 24 25 26 27 28 29 30

31

32 33

34

35

F.A. Welsh, M.D. Ginsberg, W. Rieder and W.W. Budd, Deleterious effect of glucose pretreatment on recovery from diffuse cerebral ischemia in the cat: II. Regional metabolite levels. Stroke, 11 (1980) 355-363. S. Rehncrona, I. Rosen and B-K. Siesjii, Brain lactic acidosis and ischemic eel1 damage: 1. Biochemistry and neurophysiology. J_ Cereb. Blood Flow Metab., 1 (1981) 297-311. H. Kalimo, S. Rehncrona, B. Soderfeldt, Y. Olsson and B.K. Siesjo, Brain lactic acidosis and ischemic cel1 damage: 2. Histopathology. J. Cereb. Blood Flow Metab., 1 (1981) 313-327. F. Plum, What causes infarction in ischemic brain?: The Robert Wartenberg Lecture. Neurology, 33 (1983) 222-233. J.E. Natale and L.G. D’Alecy, Protection from cerebral ischemia by brain cooling without reduced lactate accum~ation in dogs. Stroke, 20 (1989) 770-777. J.E. Natale, R.J. Schott, E.D. Hall, J.M. Braughler and L.C. D’Alecy, The 21-aminosteroid U74006F reduces systemic lipid peroxidation, morbidity and 24-h mortality after cardiopulmonary arrest in dogs. Stroke, 19 (1988) 1371-1378. D.A. Pelligrino, D.J. Miletich, R.F. Albrecht, D. Visintine, R. Ripper, G. Dominguez and J. I&se, Evaluation of a system for serial biopsy of cerebral cortical tissue in an awake goat. Am. J. Physiol., 247 (1984) R600-R6OQ. N.W. Tie&, Fundamentals of Clinical Chemistry, W.B. Saunders Co., Philadelphia., PA, 1976,pp. 936-937. J.V. Passonneau, L-Lactate-Fluorimetric Method, in: Methods of Enzymatic Analysis. Editor: H.U. Bergmeyer, Academic Press, Inc., New York, 1974, pp. 1468-1472. B.K. Siesjo, Cerebral circulation and metabolism. J. Neurosurg., 60 (1984) 883~-908. J.D. Michenfelder and J.H. Milde, Nimodipine does not affect cerebral lactate leveis following complete ischemia in dogs. J. Cereb. Blood Flow Metab., 7 (1987) 619-624. L. Marshall and R. Smith, Outcome with aggressive treatment in severe head injuries: Part I: Significance of intracranial pressure monitoring. J. Neurosurg., 50 (1979) M-25. J.E. Miller, D.P. Becker, J.D. Ward and N.G. Sullivan, Significance of intracranial hypertension in severe head injury. J . Neurosurg., 47 (1977) 503-5 16. D.A. Bruce, R.C. Raphaely, A.J. Goldberg and R.A. Simmerman, Pathophysiology, treatment and outcome following severe head injury in chifdren. Child’s Brain, 5 (1979) 174-191. A.H. Ropper and B. Shafran, Brain edema after stroke: Clinical syndrome and intracranial pressure. Arch. Neural., 41(1984)909-915. E. Siemkowicz, The effect of glucose upon restitution after transient cerebral ischemia: A summary. ActaNeurol. Stand., 71(1985)417-427. GM. Bogardus and R.J. Schlosser, The influence of temperature upon ischemic renal damage. Surgery, 39 (1956) 970-974. W.W. AngelI, L. Rickkers, E. Dong and N.E. Shumway, Organ viability with hypothermia. J. Thor. Cardiovas. Surg.. 58 (1969) 619-625. R. Busto, W.D. Dietrich, V.T. Globus, 1. Valdes, P. Schemberg and M.D. Ginsberg, Small differences in intraischemic brain temperature critically determine the extent of neuronal injury. J. Cereb. Blood Flow Metab.. 7 (1987) 729-738. M. Chopp, K.M.A. Welch, C.D. Tidwell, R. Knight and J.A. Helpem, Effect of mild hyperthermia on recovery of metabolic function after gobal cerebral ischemia in cats. Stroke, 19 (1988) 15211525. M.J. Kluger, J.J. Oppenheim and MC. Powanda, The Physiologic, Metabolic and Immunologic Actions of Interleukin-1, A.R. Liss, New York, 1985. C.A Dinarello, J.G. Cannon, S.M. Wolff and H.A Bernheim, Tumor necrosis factor (cachectin) is an endogenous pyrogen and induces production of interleukin-1. J. Exp. Med., 163 (1986) 14331450. J.K. Kjekshus, P. Vaagenes and B. Hetland, Assessment of cerebral injury with spinal creatine kinase (CSF-CK) in patients after cardiac resuscitation. Stand. J. Clin. Lab. Invest., 40 (1980) 437444. W.T. Longstreth, K.J. Clayson, W.L. Chandler and SM. Sumi, Cerebrospinal fluidcreatine kinase activity and neurologic recovery after cardiac arrest. Neurology, 34 (1984) 834-837.

289

36 37 38

39

40 41

W.T. Longstreth. K.J. Clayson and SM. Sumi, Cerebrospinal fluid and serum creatine kinase-BB activity after out-of-hospital cardiac arrest. Neurology, 31(1981) 455-458. W. Lamar, I. Woodard and B.E. Statland, Clinical implications of creatine kinase-BB isoenzyme. N. Engl. J. Med., 299 (1978) 834-835. P. Vaagenes, J. Kjekshus and A. Torvik, The relationship between cerebrospinal fluid creatine kinase and morphologic changes in the brain after transient cardiac arrest. Circulation (Suppl. V) , 61 (1980) 1194-1199. A. Mullie, P. Lust, J. Penninky, L. Vanhove, K. Vandevelde, G. Vanhoonacker and M. Krier, Monitoring of cerebrospinal fluid enzyme levels in post-ischemic encephalopathy after cardiac arrest. Crit. CareMed., 9 (1981) 199-400. P. Vaagenes, R. Cantadore, W. Diven and P. Safar, Prediction of brain damage by brain enzyme levels in cerebrospinal fluid (CSF) of dogs after cardiac arrest. Anesthesiology, 61(1984) 145. W.G. Kuhr, C.J. van den Berg and J. Korf, In vivo identification and quantitative evaluation of carrier-mediated transport of lactate at the cellular level in the striatum of conscious, freely moving rats. J. Cereb. Blood Flow Metab., 8 (1988) 848-856.