Low perfusion pressure during CPB may induce cerebral metabolic and ultrastructural changes

Background. Recently we reported on cerebral metabolic changes suggesting ischemia in piglets during nitroprusside-induced low-pressure CPB. We here investigated whether a mean arterial pressure (MAP) of 40–45 mmHg could provoke similar changes by a NO-independent intervention. Methods. Piglets underwent 60 minutes normothermic followed by 90 minutes hypothermic CPB. The LP-group (n=8) had MAP of 40–45 mmHg by phentolamine while the HP-group (n=8) had MAP of 60–80 mmHg by norepinephrine. Cerebral glucose, lactate, pyruvate and glycerol were determined. In the last two animals of each group, cerebral tissue was examined by electron microscopy. Results. Cerebral lactate was higher in the LP-group than the HP-group during normothermic CPB. Compared with baseline, cerebral glucose of the LP-group decreased whereas lactate/pyruvate-ratio, lactate and glycerol-concentrations increased during normothermic CPB. In the HP-group these parameters remained unchanged. Electron microscopy showed 31.2% and 8.3% altered mitochondria in the cortical micrographs taken from the LP- and the HP-group, respectively (p<0.001). Conclusion. MAP below 45 mmHg during CPB was associated with cerebral biochemical and morphological changes consistent with anaerobic metabolism and subcellular injury.

Keywords: Cardiopulmonary bypass; hypotension; cerebral ischemia

Neurological complications after cardiac surgery with cardiopulmonary bypass (CPB) may be caused by hypoperfusion due to inadequate cerebral perfusion pressure [1].

Recently we reported on cerebral metabolic changes compatible with anaerobic metabolism in a group of piglets with low MAP during CPB by infusion of nitroprusside [2]. These changes were absent in a control group with MAP values above 60 mm Hg by norepinephrine suggesting that these findings were related to hypoperfusion. However, we were not able to find significant differences in cerebral blood flow between the groups as measured with colored microspheres. The previous study was in addition limited by the use of nitroprusside, possibly confounding the results through the specific effects of nitric oxide (NO). Studies in cell cultures and in experimental models, have indicated that NO is capable of inhibiting cellular respiration but also that it may exert cerebral protection [3].

In order to exclude exogenous nitric oxide as the cause of the metabolic changes recently described, we conducted a follow up study, using the α-adrenergic blocker, phentolamine, as vasodilatator. Again we monitored cerebral metabolic markers at different levels of MAP in pigs on CPB.

Cerebral ischemia may induce mitochondrial permeability transition, a condition that is morphologically characterized by swelling of the mitochondria in neurons and supportive cells [4], [5]. Such changes have been demonstrated in dendrites from hippocampus after only 5 min of ischemia in an experimental model [6]. In the present study, we performed transmission electron microscopy of brain tissue collected from cortex and thalamus in the last two animals of each study group after 150 min of CPB. The cortical samples were harvested from the watershed areas since these regions may be particularly susceptible to ischemic damage during conditions of inadequate perfusion pressure [7]. All samples were assessed with focus upon mitochondrial integrity.

Our hypothesis was that a MAP below 45 mmHg during CPB would lead to cerebral anaerobic metabolism, with biochemical and structural signs of ischemic injury.

Immature domestic pigs (Norwegian Landrace-Yorkshire hybrid) were used. Animal handling was in accordance with recommendations by The Norwegian Animal Research Authority. Premedication and induction of anesthesia was performed as previously described [2]. Maintenance of anesthesia was achieved by an infusion of midazolam 0.5 mg/kg/h, fentanyl 7.5 µg/kg/h and pancuronium 45 µg/kg/h and inhalation of isoflurane 0.5–2% in 50% oxygen in air [2], [8] (Cato Anesthesia Workstation, Dräger A.G., Lübeck, Germany). After sternotomy, heparin was given (6 mg/kg plus 3 mg/kg after 1 h) and preparations for extracorporeal circulation performed. All animals were stabilized for 60 min before CPB. During CPB isoflurane was added to the CPB oxygenator. At the end of each experiment all animals were killed with an intravenous injection of 20 ml of saturated KCl solution.

Sixteen animals were allocated to a low pressure group (LP-group, n = 8) or a high pressure group (HP-group, n = 8). During CPB the HP-group was given an infusion of norepinephrine to keep MAP at 60–80 mmHg while the LP-group received phentolamine to a MAP of 40–45 mmHg. Six animals of the HP-group were used in a recently published study [2] with identical protocol for these animals.

After stabilization, both study groups underwent 60 min of normothermic (38°C) followed by 90 min of hypothermic (28°C) CPB. Domestic pigs normally have a core temperature of 38–39°C. Consequently, body temperature during normothermic CPB was kept at this level.

Cannulation of the right atrium and aorta was performed as recently described [2].

The CPB-circuit consisted of a membrane oxygenator with reservoir and integrated heat exchanger (Quadrox, VHK 4200, Jostra AG, Hirrlingen, Germany). 1 115 ml of acetated Ringer's solution was used as prime filling the machine reservoir to a level of 300 ml. The left ventricle was vented by a 17 F vent catheter (E061, Edwards Lifesciences, Irvine, California, USA). CPB pump flow was 110 ml/kg/min and flow pattern nonpulsatile. During CPB the height difference between the machine reservoir and the right atrium was fixed (73±3 cm). Continuous free venous drainage was ensured by visual inspection and monitoring of the right atrial pressure (CVP). Alpha-stat pH management was used. Nasopharyngeal and rectal temperatures were measured continuously.

MAP and CVP were monitored by fluid-filled catheters (Secalone®T, 18 G, BD Medical, Singapore) introduced into the right femoral artery and right atrium and connected to transducers (TranspacR IV, Abbot Critical Care Systems, Sligo, Ireland).

A cranial burr hole with a diameter of 0.5 cm was made in the frontal bone, 0.5 cm anterior to the coronal suture and 1.0 cm lateral to the midline. Dura mater was incised, and a pressure transducer (Codman MicroSensor ICP Transducer, Johnson & Johnson professional Inc., Raynham, Massachusetts, USA) introduced 2 cm into the brain parenchyma and connected to a Codman ICP express ™ monitor.

Maintenance fluid was given by infusion of acetated Ringer's solution 5 ml/kg/h. All blood losses prior to bypass were substituted by Ringer's solution in volumes three times the blood loss. During CPB, blood loss to the open chest was returned by suction to circulation via the extracorporeal machine reservoir.

The fluid level in the reservoir was used as guide for volume supplementation. Whenever the level fell below 300 ml, Ringer's solution was added to restore the level.

Urine output was measured every 30 min via a suprapubic catheter. Accumulated fluid balance was calculated for each experiment.

A CMA 70 microdialysis catheter, cut off 20 000 Da, membrane length 20 mm (CMA Microdialysis AB, Stockholm, Sweden) was inserted through the cranial burr hole and placed with the tip 25 mm into the brain parenchyma. The microdialysis catheter was perfused with CNS perfusion fluid (CMA) 0.3 µl/min by a microinfusion pump (CMA 107). Samples of dialysate were analyzed by a CMA 600 Microanalyser, every 30 min, with respect to glucose, glycerol, lactate and pyruvate.

Blood samples were collected from the arterial line every 30 min for measurements of hematocrit (Hct), serum-glucose, serum-osmolality as well as acid-base parameters. The analysis were performed as previously described [2].

Before killing the animal at the end of the experiment, the scalp was removed from the forehead. Following circulatory arrest, the scull was opened, the brain removed and pieces of cortical brain tissue collected from the area between the frontal and parietal lobe bilaterally, the area between the parietal and the occipital lobe bilaterally and from thalamus. The samples were submerged in 2.5% glutaraldehyde in sodium cacodylate buffer within 3 min from circulatory arrest and, while immersed, they were further divided into biopsies, sized about 1 mm3. Postfixation was performed with 1% aqueous osmium (OsO4) before they were embedded in plastic resin (Agar 100 Resin). They were then sliced into 60 nm thick sections, stained with 2% uranyl sulphate and lead citrate and examined in a transmission electron microscope (JEM 1230, Jeol Ltd. Tokyo, Japan).

Fifteen micrographs with neurons or supportive cells were randomly photographed from each of the cortical sections and from the thalamus section. The mitochondria were blindly assessed with regards to integrity by two investigators. To consider a mitochondrion changed, two of the following criteria were required: swelling, clearing of matrix and destruction of cristae [5]. The fractions of changed mitochondria in the cortical and the thalamic sections were determined.

After killing the animal, pieces of the cerebral hemispheres were collected, divided in three samples and put into preweighed vials to determine wet weight. The samples were then dried at 70°C and weighed repeatedly until stable dry weight. The total water content (TTW) was expressed as g water/g dry weight.

Repeated measures analysis of variance with one within-group factor (time) and one between-group factor (different MAP) was used. Six animals of the HP-group was recently used in a similar study comparing low pressure by nitroprusside (n = 6) with high pressure by norepinephrine (n = 6) [2]. The initial repeated measures ANOVA was therefore performed with the three groups included: the historical LP-group (nitroprusside), the LP-group (phentolamine) (n = 8) and the HP-group (norepinephrine) (n = 8). If a significant between- group difference was found, the one way ANOVA was performed to compare the groups at selected time points and if this was significant, the independent t-test was used to compare the two groups of the present study.

When a significant within-group difference was found, the two groups of the present study were further analyzed by paired t-test to compare baseline values with the values at end of hypothermic CPB. For cerebral metabolic markers and Hct, the relation between baseline and end of normothermic CPB was also evaluated. Bonferroni's correction was applied for multiple comparisons.

Statistical analyses of age, weight and TTW in the two groups were conducted by the independent t-test. Fisher's exact test was used to analyze the relation between the fractions of changed mitochondria in the two study groups.

All results are presented as mean with SD in parentheses. An alpha-level of 0.05 was considered significant. The analyses were done by use of the statistical package SPSS, version 13.0 for Windows.

The animals of the HP-group and the LP-group weighted 31.1 (3.9) and 27.6 (1.9) kg and was 11.2 (1.0) and 9.9 (0.6) weeks old, respectively, the HP-group being slightly older and heavier than the LP-group (p < 0.05). The gender of the HP-group and the LP-group was 3/5 and 6/2 (male/female).

The dose of norepinephrine in the HP-group was 0.34 (0.2) µg/kg/min while the dose of phentolamine in the LP-group was 55.0 (13.4) µg/kg/min.

The results of serum-glucose, serum-osmolality, acid-base parameters and Hct are presented in Table I. No significant difference between the two groups was found. Serum-osmolality showed a slight decrease from pre-bypass throughout the CPB period in the HP-group (p < 0.05) and the LP-group (p < 0.01). Base excess declined in both the HP-group (p = 0.01) and the LP-group (p < 0.05) during the CPB, but remained, together with serum-osmolality, within normal range.

Table I.  Blood chemistry.

325 Serum glucose (S-glu) (mmol/l), serum osmolality (S-osm) (mosmol/kg), hematocrit (Hct) (in per cent), pH, PCO2 and PO2 (kPa), base excess (BE) (mmol/l) of the HP-group (HP) (n = 8) and the LP-group (LP) (n = 8) throughout normothermic and hypothermic cardiopulmonary bypass (CPB). The values are presented as mean with SD in parenthesis. •: P < 0.05; ••: P< 0.01; •••: P< 0.001 (within-group, compared with pre-bypass).

Hct fell markedly in both study groups 60 min after start of CPB (p < 0.001). Cumulative net fluid balance in the HP- and LP-group was 243.1 (84.2) ml/kg and 195.8 (50.7) ml/kg (p > 0.05), respectively.

MAP was 68.9 (3.9) and 42.1 (2.9) mmHg at the end of normothermic CPB and 78.8 (5.3) and 43.5 (2.7) mmHg at the end of hypothermic CPB in the HP-group and the LP-group, respectively (Figure 1A). CVP recorded from the right atrium did not differ between the groups. The values were 2.8 (2.0) and 1.6 (2.3) after 60 min of normothermic CPB and 2.2 (0.8) and 1.8 (2.2) following 90 min of hypothermic CPB in the HP-group and LP-group (Figure 1B). Compared with pre-bypass level, a small reduction was seen in the HP-group (p < 0.01) at end of CPB.

Graph: Figure 1. A: Mean arterial pressure (MAP), B: Central venous pressure (CVP), C: Intracranial pressure (ICP) and D: Cerebral perfusion pressure (CPP) in the low pressure group (LP-group) (open circles) and high pressure group (HP-group) (closed circles) during 150 min of cardiopulmonary bypass. Values are presented as mean with SD. ***: P < 0.001 (between group, same time); : P < 0.01; : P < 0.001 (within group, compared with baseline).

ICP increased to 26.6 (7.5) mmHg in the HP-group (p < 0.001) and to 22.8 (4.5) mmHg in the LP-group (p = 0.001) during CPB. No between-group difference was found (Figure 1C). The corresponding CPP was 47.8 (8.0) and 23.5(5.7) mmHg at the end of normothermic CPB and 52.3(9.5) and 20.8(4.0) mmHg at the end of hypothermic CPB in the HP-group and the LP-group, respectively. The between-group differences in CPP were significant during normothermic- (p < 0.001) as well as hypothermic CPB (p < 0.001) (Figure 1D).

Results are presented in Table II. Cerebral glucose fell significantly after 90 min of CPB compared with pre-bypass values in the LP-group (p < 0.05) while it remained stable in the HP-group (p > 0.05). In the LP-group cerebral lactate increased after 90 min CPB compared with pre-bypass (p < 0.001). The values of the HP-group were significantly lower than the corresponding value of the LP-group (p < 0.05). Lactate/pyruvate ratio remained essentially unchanged in the HP-group during CPB (p > 0.05) while the corresponding values increased significantly in the LP-group after 90 min of CPB (p < 0.05). Cerebral glycerol in the LP-group was elevated after 90 min of CPB compared with baseline (p < 0.05) while glycerol of the HP-group remained within normal range (p > 0.05).

Table II.  Cerebral microdialysis.

326 Cerebral concentrations of glucose (C-glucose) (mmol/l), lactate (C-lactate) (mmol/l), lactate-pyruvate ratio (C-LP-ratio) and glycerol (C-glycerol) (µmol/l) in fluid collected by microdialysis. Values are presented as mean (SD). •:P < 0.05; •••: P < 0.001 (within group, compared with pre-bypass); *:P < 0.05 (between group differences, same time).

TTW of the brain was similar in the two study groups with 4.50 (0.3) g/g dry weight in the LP-group and 4.55 (0.49) g/g dry weight in the HP-group (p > 0.05).

The fractions of changed mitochondria in cortical micrographs were 8.3% and 31.2% in the HP-group and the LP-group, respectively (p < 0.001). Typical findings are depicted in Figure 2. In the thalamic micrographs 0.6% and 6.7% of the mitochondria were changed in the HP-group and the LP-group, respectively (p < 0.001).

Graph: Figure 2. Micrographs of brain tissue from cortex of the high pressure group (HP-group) (B and D) and the low pressure group (LP-group) (A and C), respectively. Arrows show normal mitochondria (B and D) and pathological mitochondria (A and C). Magnification X 30 000.

The principal finding of the present study is that animals with a MAP below 45 mmHg during CPB developed changes in metabolic markers consistent with cerebral anaerobic metabolism and membrane degradation. Cerebral lactate concentration was higher in the LP-group compared to the HP-group after 90 min of CPB. Although the between-group differences of the remaining cerebral metabolites did not quite reach statistical significance, the LP-group exhibited the same pattern of metabolic changes as the low-pressure group of the recent investigation [2]. Furthermore, the examined animals of the LP-group had higher frequency of mitochondrial alterations, assessed by electron microscopy, as compared to the animals of the HP-group, implicating that these animals may have been subjected to ischemia.

Most data indicate that the cerebral autoregulation, is preserved during CPB with α-stat pH management, although large individual variations may exist [9], [10]. The vasoactive drugs used in this study, are α-adrenergic agents and their possible impact on cerebral autoregulation should be taken into account. Phentolamine is a reversible blocker of adrenergic α1 and α2 receptors with a t1/2 of 19 min [11]. Few data exist on the relation between phentolamine and the cerebral autoregulation. Vlahov et al., found no effect on regional cerebral autoregulation after administration of phentolamine 1 mg/kg in cats [12]. Exogenous norpinephrine is supposed to be restricted from passing the intact blood brain barrier although some transport may take place across endothelial cells of the choroid plexus [13]. A study of healthy volunteers given an infusion with norepinephrine did not result in altered cerebral blood flow velocity as measured with transcranial Doppler ultrasonography [14]. However, infusion of phentolamine resulted in higher flow velocity despite the return to baseline values of MAP. Hence, in the present study, the metabolic consequences of a MAP below the lower limit of autoregulation may have been attenuated by vasodilatation of cerebral arteries due to local effects of phentolamine.

Isoflurane is a potent vasodilatator capable of impairing the cerebrovascular autoregulation in a dose-dependent manner. Clinical studies indicate that a dose of 1 MAC (minimal alveolar concentration) corresponding to an end-tidal concentration of about 2% in pigs [15], do not interfere with the autoregulation. In the present study isoflurane in the range of 0.5–2% was added to the CPB oxygenator [16], [17]. Occasionally the dose was increased further to keep MAP at target level, but only as a transient measure. We therefore may assume that the experimental animals had isoflurane below or equal to 1 MAC most of the time despite the fact that expiratory levels of isoflurane were not measured.

Cerebral microdialysis is an established technique for determining relative concentrations of substances in interstitial fluid [18]. Normal values in piglet were previously reported by Reinstrup et al. [18]. By measuring the levels of cellular energy metabolites such as glucose, lactate and pyruvate, a state of energy depletion may be noted. Due to the low flow rate of dialysate from the microdialysis catheter to the collecting vial, the samples are assumed to reflect local conditions roughly half an hour earlier. The lactate/pyruvate ratio appears to be particularly useful as it reflects the local ratio of [NADH][H+]/[NAD+] [19]. The elevated lactate/pyruvate ratio in the LP-group of our study may implicate local ischemia after 60 min of normothermic CPB. Concomitantly an increase in the levels of cerebral glycerol was found, indicating a possible degeneration of cellular membranes.

During hypothermic CPB, a decreasing trend in cerebral concentrations of lactate, glycerol and lactate-pyruvate ratio was seen. Cerebral blood flow is reduced during hypothermia in parallel with decrease in metabolism. Pressure autoregulation, however, is still preserved, provided that alpha-stat pH-management is used [1]. The lower limit of cerebral pressure autoregulation is not affected, at least down to a temperature level of 33°C [20]. In the present study hypothermia may have diminished a possible mismatch between oxygen demand and supply by reducing cerebral metabolic rate.

Investigations by electron microscopy after brief episodes of ischemia in experimental animals, have demonstrated characteristic morphological alterations in cerebral mitochondria with swelling, clearing and disintegration of cristae [5], [21]. Similar changes were frequently seen in the examined animals of the LP-group. Such a morphological pattern is associated with a biochemical process denoted as the mitochondrial permeability transition (MPT), characterized by opening of a multiprotein channel, the mitochondrial permeability transition pore that bridges the inner and outer mitochondrial membrane. The result is swelling of the mitochondria, inner membrane depolarization, uncoupling of oxidative phosphorylation, increased superoxide production, outflow of Ca2 +  to cytoplasm and release of intermembrane space proteins such as cytochrome c [4]. The MPT has recently been identified as a possible common pathway for necrotic and apoptotic cell death [22].

The present investigation follows up a recently conducted study testing the hypothesis that MAP of about 40 mmHg may lead to cerebral ischemia. Six animals of the HP-group were also used as high pressure group in the previous study. The statistical analyses have been adjusted accordingly by performing the initial repeated measures ANOVA with the two groups of the present study and the low pressure group of the previous study included. Although not randomized, all experiments were performed within a short time period and the all high pressure animals were subjected to identical protocols.

The electron microscopic data is limited by a low number of animals subjected to examination. These data, although revealing highly significant differences between the groups, has to be confirmed in future studies focusing upon the relation between perfusion pressure, microcirculation and cellular metabolism.

In this experimental model, a MAP below 45 mmHg during CPB was associated with changes in cerebral metabolic markers consistent with anaerobic metabolism and degradation of membranes. Furthermore, electron microscopy of cerebral tissue from a small number of the animals revealed changes in the LP-group that might indicate subcellular injury.

The Board of the Faculty of Medicine, University of Bergen, has authorized the "Locus for Circulatory Research" as an officially recognized research group within the faculty. We greatly acknowledge this support. Oddbjørn Haugen is a research fellow of Western Norway Regional Health Authority, Stavanger, Norway. The technical support and creative ideas by Arve Mongstad RP and Else Nygreen RP are highly appreciated. This study was financially supported by the Western Norway Regional Health Authority, Stavanger, Norway and The Frank Mohn Foundation, Bergen, Norway. Preliminary results were presented at the 19th Annual Congress of the European Society of Intensive Care Medicine, Barcelona, 2006.