HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Boivie, P. Articles by Engström, K. G. Search for Related Content PubMed PubMed Citation Articles by Boivie, P. Articles by Engström, K. G. Related Collections Cerebral protection Extracorporeal circulation Related Article

J Thorac Cardiovasc Surg 2003;125:1451-1460
© 2003 The American Association for Thoracic Surgery

Cardiopulmonary Support and Physiology

Embolic material generated by multiple aortic crossclamping: A perfusion model with human cadaveric aorta

From the Departments of Surgical and Perioperative Science,^a and Integrative Medical Biology,^b Section for Anatomy and Department of Medical Biosciences, Section for Pathology, Umeå University Hospital, Umeå, Sweden.

This work was supported by Swedish Society for Medical Research and funds of the Medical Faculty, Umeå University Hospital, Swedish Medical Research Council (12X-11204), Swedish Heart and Lung Foundation, and the Heart Foundation of North Sweden.

Received for publication May 10, 2002. Revisions requested July 22, 2002; revisions received Aug 27, 2002. Accepted for publication Sept 11, 2002. Address for reprints: Patrik Boivie, Cardiothoracic Division, Department of Surgical and Perioperative Science, Umeå University Hospital, S-901 85 Umeå, Sweden (E-mail: patrik_boivie{at}hotmail.com).

	Abstract

Top Abstract Introduction Materials and methods Results Discussion References

 
Background: Atherosclerosis of the ascending aorta and use of aortic crossclamping are risk factors for neurologic injury during cardiac surgery.
Objectives: Repeated aortic manipulation is part of the surgical approach to most cardiac operations. The aim of this study was to assess the amount and size of particulate matter that is dislodged from the aortic wall as a function of repeated aortic crossclamping.
Methods: In 10 subjects undergoing autopsy the aorta was dissected and mounted in a perfusion model. The ascending aorta was crossclamped and washed out 10 times, with the perfusate collected in aliquots (1 to 10). The aliquots were examined by computerized image processing, both macroscopically and under the microscope for calcified and cellular material.
Results: Aortic crossclamping produced substantial output of particulate matter. After repeated aortic crossclamping the number of particles decreased (P = .012) and approached the baseline for aliquots 6 to 10. The average particle diameter was 0.63 ± 0.03 mm, with a maximum of 4.74 mm. Similar variability in particle outputs were recorded microscopically, with findings of both calcified and cellular material. Nine of 10 aortas had calcifications seen during simple visual inspection.
Conclusions: The washouts of dislodge material at aortic crossclamping had embolic potential. During the initial aortic crossclamping procedures the amount of particles was substantial, both macroscopically and microscopically. On the microscopic scale noncalcified cellular debris represents a significant pool of embolic material. Repeated aortic crossclamping reduced the amount of particles. These findings question surgical techniques associated with repeated aortic crossclamping.

	Introduction

Top Abstract Introduction Materials and methods Results Discussion References

 
See related editorial on page 1200.

Brain damage is a significant part of morbidity and mortality after cardiac surgery. The scenario of brain damage is complex and has two extremes, major stroke versus diffuse cognitive defects. ¹ Atherosclerosis of the ascending aorta and aortic crossclamping (ACC) has been identified as major risk factors for perioperative stroke during cardiac surgery. ^2,3 Diffuse symptoms of brain damage occurs in the postoperative period and are often discussed in relation to the use of extracorporeal circulation, such as from fat microembolism of retrieved mediastinal blood, ^4,5 or in terms of immunological mechanisms. ⁶ In addition, cerebral injury may be secondary to hypoperfusion and anesthetic management. ⁷

Most types of cardiac surgery involve manipulation of the aorta by clamping. ⁸ A number of techniques have been used to reduce the risk of stroke during cardiac operation, ^8,9 including echo visualization of the aorta. ¹⁰ In coronary artery bypass grafting (CABG) on the nonbeating heart, two different methods are employed to perform distal anastomoses, either cardioplegic arrest or the older method of fibrillating heart. The purpose of this study was to investigate the mechanism of injury caused by ACC. The focus was on repeated clamping, such as with the fibrillating heart method.

An experimental model was designed using human cadaveric aorta. The spectrum of dislodged material following ACC was evaluated using computerized image processing, both of macroscopic and microscopic particulate matter. The macro- versus microscopic approach was done with reference to possible mechanisms behind stroke and diffuse brain damage, respectively.

	Materials and methods

Top Abstract Introduction Materials and methods Results Discussion References

 
Subjects
The study was based on 10 subjects undergoing autopsy. The subjects all belonged to the high-risk group of atherosclerosis (Table 1). The subjects were selected at random from the daily program of autopsy during a 2-month study period. The experiments were performed within 72 hours after death. Prior to the investigation 2 subjects were tested for method evaluation. During the study an additional 2 subjects were excluded due to findings of aortic dissection and technical circumstances during the autopsy, respectively. The study was approved by the Umeå University ethical committee (Dnr 01-142).

Retrograde perfusion was implemented (Figure 1). A catheter (8/10.6 32, Polystan A/S, Värlöse, Denmark) was positioned in the descending aorta and fixed with ligatures so that the perfusion tip was localized at the subclavian artery. The cannula was connected to a constant pressure device consisting of a 10-L infusion bag positioned at a hydrostatic level corresponding to 60 mm Hg pressure. The perfusion medium consisted of 9 g/L NaCl. The aorta was flushed with about 400 mL perfusion medium to remove debris, followed by a 50-mL sample of the perfusate representing the baseline prior to ACC (denoted as ACC 0). A standard 70-mm aortic crossclamp (Pilling Co, Fort Washington, Pa) was applied to the ascending aorta at a location that resembles a standard surgical procedure. This was typically about 2 cm proximal to the brachiocephalic trunk. To precisely repeat the crossclamping at the same location, the crossclamp jaws were fixed to the aortic adventitia with stitches. To gain the same occluding force of repeated clamping, the clamp was closed to half of its full range in all experiments.

View larger version (48K): [in this window] [in a new window]   Fig. 1. The ACC perfusion model is schematically illustrated. The surgical situation at aortic declamping was simulated by a pressurized human cadaveric aorta and retrograde washouts. The crossclamp was released to fill a 50-mL test tube from the ascending aorta, followed by reclamping. The procedure was repeated 10 times. Large side branches were occluded by ligatures and small arteries by clamps. The perfusion cannula was connected to a constant pressure of 60 mm Hg.

Sample processing
Each washout was centrifuged at 1500g for 10 minutes at 22°C. The supernatant was carefully aspirated, leaving the deposit in the test tube. Distilled water (50 mL) was added to the sample to lyse remaining erythrocytes and for washing purpose. The sample was recentrifuged and the supernatant was again aspirated. The collected material was fixed for 10 minutes at 22°C, by adding 5 mL of 10% formalin in phosphate buffer. The fixation was stopped by adding distilled water, followed by washing and recentrifugation. The material was stained for 10 minutes at 22°C with 20 µL of cresyl violet for basophilic material, ¹¹ followed by 2 washing and centrifugation cycles. The sample was collected in a plastic Pasteur pipette and deposited on an uncoated microscopic slide. The droplet was spread out and left to dry at room temperature. When a sample contained high amounts of particulate matter, it was spread out on 2 separate microscopic slides. This was done to avoid severely condensed material that would interfere with the image analysis. With duplicate slides, which occurred in 20% of all samples, the particulate matter from both slides was added together during image analysis. A typical deposit is shown in Figure 2.

View larger version (83K): [in this window] [in a new window]   Fig. 2. Typical deposits collected on a microscopic slide are shown. (a) Macroscopic view of a calcified particle at x30 magnification. A section is indicated for close-up microscopy in panel b. (b) Microscopic view of particle deposits at x 240 magnification. The images were subject to gray-scale contrast enhancement.

For microscopic evaluation the samples were viewed in an inverted microscope (Olympus CK40-F200, Olympus Optical Company Ltd, Tokyo, Japan) equipped with a CDPlan x10-FPL objective lens, and recorded using a black and white camera (C5405-01 Hamamatsu Photonics, Hamamatsu City, Japan). In this configuration 2 microscopic views were sampled at random from each slide. The microscopic data therefore represent a qualitative measure of the deposit.

The average volume of the washout samples was 35.0 ± 0.9 mL (mean ± SEM). Due to technical circumstances a few washouts deviated in volume with a range from 15.0 to 51.5 mL. Due to this variation the amount of deposits was corrected for the sample volume and expressed as particles per milliliter. For the microscopic evaluation, the data were expressed per milliliter perfusate and per microscopic view.

Image processing and geometric measurements
The images were processed by computerized technique to measure particle count, area, and shape factor. The image handling included: (1) contrast enhancement to full gray scale, (2) manual editing for technical artefacts, (3) 1-step pixel dilation followed by pixel erosion, and (4) geometric measurement from preset gray threshold levels.

On the macroscopic level the particulate matter was dense in contrast and evaluated by a single threshold setting for calcified particles. On the microscopic level, at which both calcified dense and cellular semitransparent particulate matter were detected, the image analysis was performed using 2 windows of gray scale attenuation, dense and soft spectrum, respectively. All threshold settings were constant throughout the study. The results were transferred to an Excel spreadsheet.

Statistical analysis
The statistical evaluation contained ordinal, discrete, and numerical data. The geometric data (eg, area and shape factor measurements) showed normality. Mean values ± SEM are given throughout. However, because of the different nature of data only nonparametric statistics were used. For within-group differences Wilcoxon signed rank test was applied. Between-groups differences were evaluated by Kruskal-Wallis test and post hoc comparison by Mann-Whitney U test. Correlation matrices were tested using Spearman's rank coefficient. Slope values refer to linear regression versus clamp opening. A P value above .05 was considered nonsignificant.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

 
Demographics of subjects
Demographic and autopsy data are summarized in Table 1. From a semiquantitive evaluation of calcification, 1 subject was without overt signs of calcification in the aorta. In the remaining 9 cases the aorta had atherosclerotic changes of variable magnitude and ulcerations. None of the subjects had porcelain characteristics (Table 1). There was no significant correlation between subject age and magnitude of aortic calcification.

Method evaluation
We developed and tested a new method of studying the injuries caused by repeated ACC. The method gave reproducible results by which the accumulated deposits varied in number and characteristics with repeated ACC. There was a significant correlation between number of deposits and the magnitude of aortic wall calcification (P = .014). There was no correlation between subject age and number of deposits.

Macroscopic particle evaluation (dense spectrum)
The first-time clamping procedure resulted in a marked increase in observed calcified particles versus the baseline recording (Figure 3, a). With repeated ACC the number of deposits decreased. ACC 1 to 5 produced significantly larger amounts of particles compared to ACC 6 to 10 (Table 2, column A). This finding was also confirmed by a negative slope that was significantly different from zero (Table 2). There were no significant differences in the observed average area or shape factor of the calcified particles between ACC 1 to 5 compared with ACC 6 to 10 (Figures 3, b and c, Table 2). When the embolic particle load was calculated, by multiplying the area with particle numbers, there was again a significant difference between ACC 1 to 5 versus ACC 6 to 10 (Figure 3, d, Table 2). This variability produced a negative slope significantly different from zero (Table 2). The average diameter of recorded particles of all measurements was 0.63 ± 0.03 mm, with a spread from 0.22 to 4.74 mm (0.22 mm represents the technical resolution for the macroscopic measurements).

View larger version (20K): [in this window] [in a new window]   Fig. 3. Geometric characteristics of calcified particles on the macroscopic scale are demonstrated. Deposits were collected from the ascending aorta following repeated ACC. A baseline recording prior to clamping is denoted by zero on the horizontal axis. The curves represent deposits produced at ACC 1 to 10. (a) Particle number, (b) particle average area, (c) particle shape factor, and (d) embolic load from multiplying the number of particles with particle area. Mean values ± SEM. Reference is made to Table 2 for statistical results.

View larger version (21K): [in this window] [in a new window]   Fig. 4. Geometric characteristics of particles on the microscopic scale are shown, with the attenuation window set for dense calcified material. Deposits were collected from the ascending aorta following repeated ACC. A baseline recording prior to clamping is denoted by zero on the horizontal axis. The curves represent deposits produced at ACC 1 to 10. (a) Particle number, (b) particle average area, (c) particle shape factor, and (d) embolic load from multiplying the number of particles with particle area. Mean values ± SEM. Reference is made to Table 2 for statistical results.

View larger version (20K): [in this window] [in a new window]   Fig. 5. Geometric characteristics of particles on the microscopic scale are shown, with the attenuation window at soft spectrum to include cellular material. Deposits were collected from the ascending aorta following repeated ACC. A baseline recording prior to clamping is denoted by zero on the horizontal axis. The curves represent deposits produced at ACC 1 to 10. (a) Particle number, (b) particle average area, (c) particle shape factor, and (d) embolic load from multiplying the number of particles with particle area. Mean values ± SEM. Reference is made to Table 2 for statistical results.

Because of the different recording methods between macroscopic and microscopic methods it was not meaningful to compare particle numbers between these groups. However, the two microscopic evaluation methods were possible to compare. Of particular interest was the embolic load from the product of particle-x-area. There was a significantly larger amount of embolic output of cellular material, comparing columns C versus B (Table 2). From a technical image-analysis perspective this was expected as the cellular measuring window also contained calcified material. If the calcified material was subtracted from the data, leaving only cellular debris, it was found that the perfusate contained 169 ± 41 (n x µm²/mL and microscopic view) of cellular embolic material. This was twice the amount of calcified particles, 85.7 ± 43.1 (n x µm²/mL and microscopic view) (Table 2, columns B and C). However, this apparent difference was not statistically significant (P = .070).

Variability between subjects
The data showed great variability between the cases. Although 9 of the 10 subjects had atherosclerotic changes in their aorta, only half of the cases produced major bursts of dislodged output (Figure 6). This was evaluated by calculation of linear regression. In 9 subjects the slope was negative, although in 4 of these only minor negativity was found (Figure 6). The patient without visual aortic calcifications belonged to the group that had a slope close to zero.

View larger version (16K): [in this window] [in a new window]   Fig. 6. Characteristics of individual subjects on the macroscopic scale are demonstrated, using linear regression analysis of particle numbers produced at ACC 1 to 10. Deposits were collected from the ascending aorta following repeated ACC. Five of the 10 subjects produced substantial output of particles, as shown. In only 1 of the 10 subjects was the regression analysis slope positive.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

Before the era of cardioplegic arrest, the method of choice for CABG was by transient fibrillation of the heart. ¹⁷ With this technique the aorta was clamped repeatedly, for each anastomosis. Even though this method may be considered old-fashioned, it is still in use at several clinics around the world. ¹⁸ A parallel scenario is the manipulation from side-biting aortic clamps. In this study we simulated the surgical methods of fibrillating heart or other forms of repeated clamping. The crossclamp was applied repeatedly and in the same location on the aorta, and the dislodged material was collected and geometrically analyzed. The hypothesis was that the magnitude of dislodged particles would diminish with repeated clamping. This assumption was here confirmed, although the baseline was reached after several clamp maneuvers.

Aortic particles that cause stroke are thought to be geometrically large, such as detached calcified plaques. These particles are best analyzed on the macroscopic level. We found an average particle area of macroscopically detected material of about 0.2 mm². The average diameter was 0.6 mm, with maximum of 4.7 mm. Moreover, it was found that the macroscopic material was noncircular with a shape factor of about 0.6 (circular shape equals 1 and linear shape 0). These data can be compared with known diameters of embolic material in occluded cerebral arteries (Willis circle) of about 5 mm. ¹⁹ Our data corroborate histological observations from ongoing studies using intra-aortic filters, in which the diameter of captured particles ranged from 0.1 to 6.0 mm. ¹³

Although the macroscopic material seems to be the most interesting with respect to stroke, cardiac surgery is also known to cause diffuse defects in neuropsychology and cognition. ¹ This may in part reflect small-artery brain emboli. ²⁰ It has been speculated that pericardial suction blood contains liquid fat that occludes brain capillaries' blood flow function. ^4,5 However, this study suggests an additional source of microscopic emboli from ACC. We investigated microscopic particles from two perspectives, both calcified and cellular material. The distinction between these particles was created by gray-scale attenuation at image analysis. Dense opaque material was calcified microparticles. Cellular material was measured using a softer threshold spectrum to expose semitransparent material. Red blood cells that contaminated the samples were lysed by water prior to the analysis, and the measured material was noncalcified cellular debris from the aortic wall. The average area of these particles was approximately 100 to 200 µm², with measured diameters of about 10 to 15 µm for calcified and cellular material, respectively. The particles were more circular in shape than the large-size macroscopic deposits. The capillary diameter is known to be about 5 µm, ²¹ which suggests embolic potential.

We found that the magnitude of dislodged particles differed substantially between subjects. In this study only 5 of the 10 subjects produced large amounts of debris. There was a positive correlation between the amount of particles and the visual appearance of aortic atherosclerosis. Nine of the 10 cases had visible atherosclerosis, but in only half of these were severe amounts of particles dislodged at clamping.

Our study was limited by the use of cadaveric aorta. It is reasonable to expect that some postmortem changes may affect the results in different ways. Furthermore, the conditions at autopsy may not truly represent the surgical conditions in the operating room. The limited number of observations must also be considered. Our model employed retrograde washout, which is a simplification although it resembles the surgical situation at aortic declamping. However, the data appeared consistent, and image analysis revealed significant trends with repeated ACC, a fact that suggests a useful model in surgical science. The clinical impact of our findings can only be speculated upon as a question that relates to the complexity of flow dynamics and vascular anatomy.

In conclusion, we designed an experimental model with cadaveric aorta to analyze injury caused by crossclamping. Our main focus was on surgical methods using repeated ACC, such as with the fibrillating heart technique. Repeated crossclamping reduced the amount of dislodged particulate matter. However, the output of material was profound during the initial series of clamp maneuvers, a finding that questions repeated ACC in surgery.

	Acknowledgments

 
We thank Miss Ann-Christin Güvencel, who kindly made the drawing used for Figure 1. We would also like to thank Mrs Anne-Marie Österdahl, Mr Dan Nylund, and Dr Karin Ernhill at the Department of Clinical Pathology, Umeå University Hospital for their assistance during autopsy.

	References

Top Abstract Introduction Materials and methods Results Discussion References

 

1. Taylor KM. Brain damage during cardiopulmonary bypass. Ann Thorac Surg. 1998;65:20-6.
   
2. Mills NL, Everson CT. Atherosclerosis of the ascending aorta and coronary artery bypass. J Thorac Cardiovasc Surg. 1991;102:546-53.[Abstract]
   
3. Barbut D, Hinton RB, Szatrowski TP, et al. Cerebral emboli detected during bypass surgery are associated with clamp removal. Stroke. 1994;25:2398-402.[Abstract]
   
4. Appelblad M, Engstrom KG. Fat contamination of pericardial suction blood and its influence on in vitro capillary-pore flow properties in patients undergoing routine CABG. J Thor Cardiovasc Surg. 2002;124:377-86.[Abstract/Free Full Text]
   
5. Brown WR, Moody DM, Challa VR. Cerebral fat embolism from cardiopulmonary bypass. J Neuropathol Exp Neurol. 1999;58:109-19.[Medline]
   
6. Gu YJ, van-Oeveren W, Boonstra PW, de-Haan J, Wildevuur C. Leukocyte activation with increased expression of CR3 receptors during cardiopulmonary bypass. Ann Thorac Surg. 1992;53:839-43.[Abstract]
   
7. Taggart DP, Westaby S. Neurological and cognitive disorders after coronary artery bypass grafting. Curr Opin Cardiol. 2000;16:271-6.
   
8. Harringer W, on behalf of the ICEM Study Group. Capture of particulate emboli during cardiac procedures in which aortic cross-clamp is used. Ann Thorac Surg. 2000;70:1119-23.[Abstract/Free Full Text]
   
9. Hammon JW Jr, Stump DA, Kon ND, et al. Risk factors and solutions for the development of neurobehavioral changes after coronary artery bypass grafting. Ann Thorac Surg. 1997;63:1613-8.[Abstract/Free Full Text]
   
10. Van der Linden J, Hadjinikolaou L, Bergman P, Lindblom D. Postoperative stroke in cardiac surgery is related to the location and extent in the ascending aorta. J Am Coll Cardiol. 2001;38:131-5.[Abstract/Free Full Text]
    
11. Kawamura S, Niimi K. Counterstaining of Nata-Gyatz impregnated with Cresyl Violet. Stain Technology. 1972;47:1-6.[Medline]
    
12. Vaage J, Jensen U, Eriksson A. Neurologic injury in cardiac surgery. Scand Cardiovasc J. 2000;34:550-7.[Medline]
    
13. Reichenspurner H, Naiva JA, Berry G, et al. Particulate emboli capture by an intra-aortic filter during cardiac surgery. J Thorac Cardiovasc Surg. 2000;119:233-40.[Abstract/Free Full Text]
    
14. Hirose H, Amano A, Takahashi A, Nagano N. Off-pump total arterial bypass grafting using bilateral mammary, gastroepiploic, and radial artery. Ann Thorac Cardiovasc Surg. 2000;6:405-7.[Medline]
    
15. Stefaniszyn HJ, Novick RJ, Sheldon H, Sniderman AD, Salerno TA. Anatomical observations in cadavers during application of a partial exclusion clamp to the ascending aorta. Curr Surg. 1984;41:184-7.[Medline]
    
16. Schmitz C, Blackstone EH. International Council of Emboli Management (ICEM) Study Group results: risk adjusted outcomes in intraaortic filtration. Eur J Cardiothorac Surg. 2001;20:986-91.[Abstract/Free Full Text]
    
17. Kay HR, Barash PG, Hammond G, Geha A, Kopriva CJ, Laks H. Myocardial recovery: perfusion of the fibrillating heart versus cold chemical cardioplegia. J Surg Res. 1981;30:515-21.[Medline]
    
18. Kirklin JW, Baratt-Boyes BG. Myocardial management during cardiac surgery with cardiopulmonary bypass. In: Kirklin JW, Baratt-Boyes BG, editors. Cardiac Surgery. 2nd ed. New York: Churchill Livingstone, 1993. p. 141-2.
    
19. Vogel FS, Fuller GN, Bouldin TW. The nervous system. In: Rubin E, Farber JL, editors. Pathology. 3rd ed. New York: Lippincott-Raven, 1998. p. 1474.
    
20. Moody DM, Brown WR, Challa VR, Stump DA, Reboussin DM, Legault C. Brain microemboli associated with cardiopulmonary bypass: a histologic and magnetic resonance imaging study. Ann Thorac Surg. 1995;59:1304-7.[Abstract/Free Full Text]
    
21. Potter RF, Groom AC. Capillary diameter and geometry in cardiac and skeletal muscle studied by means of corrosion casts. Microvasc Res. 1983;25:68-84.[Medline]
    

This article has been cited by other articles:

				P. Boivie, M. Hansson, and K. G. Engstrom Intraluminal aortic manipulation by means of intra-aortic filter, cannulation, and external clamp maneuvers evaluated versus dislodged embolic material J. Thorac. Cardiovasc. Surg., February 1, 2006; 131(2): 283 - 289. [Abstract] [Full Text] [PDF]

				S. F. Aranki, M. Nathan, P. Shekar, G. Couper, R. Rizzo, and L. H. Cohn Hypothermic Circulatory Arrest Enables Aortic Valve Replacement in Patients With Unclampable Aorta Ann. Thorac. Surg., November 1, 2005; 80(5): 1679 - 1687. [Abstract] [Full Text] [PDF]

				P. Boivie, C. Edstrom, and K. G. Engstrom Side differences in cerebrovascular accidents after cardiac surgery: A statistical analysis of neurologic symptoms and possible implications for anatomic mechanisms of aortic particle embolization J. Thorac. Cardiovasc. Surg., March 1, 2005; 129(3): 591 - 598. [Abstract] [Full Text] [PDF]

				K. G. Engstrom Contaminating fat in pericardial suction blood: a clinical, technical and scientific challenge Perfusion, January 1, 2004; 19(1_suppl): S21 - S31. [Abstract] [PDF]

				J. W. Hammon Aortic nightmares: Can we sleep better? J. Thorac. Cardiovasc. Surg., June 1, 2003; 125(6): 1200 - 1201. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Boivie, P. Articles by Engström, K. G. Search for Related Content PubMed PubMed Citation Articles by Boivie, P. Articles by Engström, K. G. Related Collections Cerebral protection Extracorporeal circulation Related Article