| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1995;59:558-561
© 1995 The Society of Thoracic Surgeons
Anesthesiology and Neurological Surgery, College of Physicians and Surgeons of Columbia University, New York, New York, and Department of Anesthesiology, Duke Heart Center, Duke University Medical Center, Durham, North Carolina
Disney, of course, has the best casting. If he doesn't like an actor, he rips him up. Alfred Hitchcock
All of us strive to portray what we believe to be the closest approximation of ``truth'' in our research efforts. The difficulties of reconciling our knowledge of things as something absolute or some distortion of our perceptions are as old as science itself. Elsewhere in this issue Cook and co-workers [1] report findings regarding cerebral blood flow (CBF) during cardiopulmonary bypass (CPB). In their work they raise the methodologic question of how to measure CBF and go a step further by concluding that one method is superior to another. Their conclusions deserve relatively careful scrutiny.
A paramount consideration in comparing the multiple methods for determinations of cerebral perfusion is to understand that no method perfectly describes actual blood flow moving through the capillary bed in precise quantitate terms. All measurements, especially in the cardiac operating room, are relative approximations with inherent errors and limitations. If we are interested in the effects of physiologic and pharmacologic interventions on CBF, we may use any reproducible method that covaries with other similarly reproducible and comparable methods; the choice of method then is dictated by the nature and particular limitations of the experiment. The necessary and sufficient condition for being an acceptable method is that, under the experimental conditions in question, the various CBF estimates covary together. Figure 3 in Cook and co-workers' article makes this point: CBF values for both Kety-Schmidt (KS) and xenon-133 clearance covary together. Because the absolute values in large numbers of studies for both methods are relatively close [2], whether KS overestimates or 133Xe clearance underestimates ``true'' CBF is a secondary issue, especially if the question is how CBF changes under the variable physiologic conditions of CPB.
The results of Cook and co-workers' study showing a discrepancy in CBF using 133Xe clearance and KS nitrous oxide saturation under the experimental conditions of normothermic and hypothermic CPB are not surprising because the two methods measure different things with different diffusible ``indicators''. What is surprising about Cook and co-workers' article, however, is the limited interpretation of the observed data that tends to obscure rather than clarify the methodologic issues for the readership. This is in part due to a remarkable failure to cite pertinent published work on the subject [3–7] including an especially important validation study of the two methods (using the same indicator) performed in dogs during CPB [3]. We strongly disagree with the proposition that one method is more ``valid'' than the other. We believe that the usefulness of Cook and co-workers' article is that it serves to illustrate that there are important differences in the CBF methodologies that are used for clinical purposes (Table 1
).
|
shows that the method of CBF produces different values. Of the several methods used clinically, the 133Xe clearance and Kety-Schmidt with argon or N2O are the two that have been used in the operating room during CPB. Both methods generate CBF data that are discrepant under relatively similar temperature, hemodilution, acid-base management, and hemodynamics. Of interest is the wide variability in reported flows with each method: during hypothermia (27° to 29°C) there is a range of 19 to 34 mL • min-1 • 100 g-1 with KS and 12 to 21 mL • min-1 • 100 g-1 with 133Xe clearance.
|
In attempting to understand why KS and 133Xe clearance produce different values the theoretical assumptions upon which both are based (the Fick principle) need to be understood: (1) tracer does not affect CBF, (2) concentration of tracer is the same in tissue and venous blood, (3) partition coefficients for tracer are the same in normal and abnormal states (eg, low flow), and (4) partition coefficients for tracer are the same in different physiologic states (eg, hypothermia). None of these four assumptions has been validated in the setting of CPB, and there are very real concerns especially regarding KS. With regard to the first and third assumptions, N2O increases CBF [15, 16], and with lower-flow states encountered during CPB, tissue and venous blood may not be in equilibrium and partition coefficients are not precisely known. With regard to the fourth assumption, temperature affects the solubility of N2O and xenon differently, making comparison of the two problematic and complicating the question of ``validity.'' This is the reason that in our validity study [3] we used the same tracer for both the KS and 133Xe clearance.
The KS technique uniquely assumes the following: (1) jugular bulb blood is free of extracranial contamination and (2) uptake and clearance of tracer is proportional to flow. Neither of these assumptions has been proven during CPB as Prough and Rogers [2] pointed out, and one reason for this is that it is technically very difficult to administer, collect, and measure nitrous oxide during CPB (Prough DS, personal communication). Cook and co-workers could have strengthened their article if they had addressed these methodologic issues in their report.
The 133Xe clearance technique uniquely assumes the following: (1) contamination by extracranial scalp and muscle blood flow is negligible and (2) all tracer arrives in the brain before any of it begins to leave. The first assumption that extracranial circulation does not significantly affect CBF has been demonstrated repeatedly to be valid [5, 7, 14]. If the second assumption is not correct, then there will be a distributed input function, one ``smeared'' over time (Fig 1). This is the case with intravenous and inhalational administration of tracer and may be dealt with mathematically by deconvolution [5, 7]. The effect of having such a smeared input function is underestimation of the flow calculated. There is little doubt that injection of 133Xe tracer in the pump circuit (and perhaps in the great vessels proximal to the internal carotid artery) could result in smearing of the input function. It could explain the relatively low CBF values reported by some in Table 2
.
|
In our validation report [3] we demonstrated the importance of this choice of compartmental modeling, which significantly alters CBF calculation. We always have used the initial slope model, believing it representative of gray matter brain flow of interest because of relevance to cerebral protection. This issue of modeling brings up another distinction between KS and 133Xe clearance: there is a theoretical sampling issue when comparing the two methods. The ratios of subcortical to cortical tissue are different. With KS, one sees a preponderance of subcortical gray and white matter, whereas 133Xe washout is heavily weighted toward cortical gray matter.
An additional consideration is Cook and co-workers' method of determining the slope of the washout curve from only two points. This is certainly not the standard method and is only reasonable if each curve is carefully inspected for nonlinearity.
Cook and co-workers' data on cerebral metabolic rate and CBF at hypothermia also deserve careful scrutiny. The results obtained simultaneously by KS and 133Xe clearance appear to have markedly different cerebral arteriovenous oxygen differences. Cerebral arteriovenous oxygen difference is the difference in oxygen content of arterial and jugular venous blood and should be identical for measurements made simultaneously because it is independent of CBF technique. Cerebral arteriovenous oxygen difference from Cook and co-workers' data can be determined by dividing cerebral metabolic rate by the CBF. Accomplishing this for the hypothermic KS and 133Xe clearance measurements yields values of 2.8 (KS) and 3.9 (133Xe clearance) mL/dL. Although there are limitations in using mean data to make assumptions, the results should at least be similar. The marked difference for the same measurement has several troubling explanations: different samples were used to determine cerebral arteriovenous oxygen difference, multiple measurements were averaged for one group without stable brain temperature being reached, jugular bulb samples were drawn at a rapid rate contaminating the jugular venous blood producing differences in jugular blood saturation, or an indicator used to measure CBF significantly affected oxygen extraction or CBF. Regardless of the mechanism, use of different values for cerebral arteriovenous oxygen difference produces questions about the validity of the cerebral metabolic rate results and if cerebral arteriovenous oxygen difference is unstable, the accuracy of the KS measurements.
Much has been learned about cerebral blood flow and brain metabolism during cardiopulmonary bypass [6]. Although the story is far from complete, and not every portrayal is without criticism, we do not have the luxury nor the scientific prerogative to discredit the major player in this story, CBF measurement by xenon clearance. We believe the bulk of scientific evidence supports use of this methodology. From our perspective KS is fraught with more problems for use during CPB, but neither should be abandoned. They are similar but different tools to help us better understand the physiology of cardiopulmonary bypass.
Footnotes
Address reprint requests to Dr Reves, Division of Cardiac Anesthesia, Department of Anesthesiology, Heart Center of Duke University Hospital, Box 3094, Durham, NC 27710.
References
Related Article
This article has been cited by other articles:
|
|
 |
|
  K. Bendjelid, B. Poblete, O. Baenziger, and J.-A. Romand The effects of hypothermic cardiopulmonary bypass on Doppler cerebral blood flow during the first 24 postoperative hours Interactive CardioVascular and Thoracic Surgery, March 1, 2003; 2(1): 46 - 52. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Bonser, C. H. Wong, D. Harrington, D. Pagano, M. Wilkes, T. Clutton-Brock, and M. Faroqui Failure of retrograde cerebral perfusion to attenuate metabolic changes associated with hypothermic circulatory arrest J. Thorac. Cardiovasc. Surg., May 1, 2002; 123(5): 943 - 950. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Grubhofer, A. M. Lassnigg, B. Schneider, M. A. Rajek, T. Pernerstorfer, and M. J. Hiesmayr Jugular Venous Bulb Oxygen Saturation Depends on Blood Pressure During Cardiopulmonary Bypass Ann. Thorac. Surg., March 1, 1998; 65(3): 653 - 657. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. A. Nuttall, D. J. Cook, U. H. Trivedi, R. L. Patel, M. R. J. Turtle, G. E. Venn, and D. J. Chambers Transcranial Doppler Sonography and Cerebral Blood Flow During Cardiopulmonary Bypass Ann. Thorac. Surg., September 1, 1997; 64(3): 891 - 891. [Full Text]
|
|
|
|
|
|
I. MacVeigh, D. J. Cook, T. A. Orszulak, R. C. Daly, and D. E. Munnikhuysen Nitrous Oxide Method of Measuring Cerebral Blood Flow During Hypothermic Cardiopulmonary Bypass Ann. Thorac. Surg., March 1, 1997; 63(3): 736 - 740. [Abstract] [Full Text]
|
|
|
|
|
|
U. H. Trivedi, R. L. Patel, M. R. J. Turtle, G. E. Venn, and D. J. Chambers Relative Changes in Cerebral Blood Flow During Cardiac Operations Using Xenon-133 Clearance Versus Transcranial Doppler Sonography Ann. Thorac. Surg., January 1, 1997; 63(1): 167 - 174. [Abstract] [Full Text]
|
|
|
|
|
|
D. J. Cook, R. E. Anderson, J. D. Michenfelder, W. C. Oliver Jr, T. A. Orszulak, R. C. Daly, and R. D. Bryce Who's Afraid of Kety-Schmidt? Ann. Thorac. Surg., October 1, 1995; 60 (4): 1156 - 1157. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
