|
|
||||||||
Ann Thorac Surg 2003;76:2037-2042
© 2003 The Society of Thoracic Surgeons
a Department of Anesthesia, University of Pennsylvania, Philadelphia, Pennsylvania, USA
b Department of Neurology, University of Pennsylvania, Philadelphia, Pennsylvania, USA
c Department of and Surgery, University of Pennsylvania, Philadelphia, Pennsylvania, USA
Accepted for publication June 13, 2003.
* Address reprint requests to Dr Floyd, Department of Anesthesia, Hospital of the University of Pennsylvania, 3400 Spruce St, Philadelphia, PA 19104-4283, USA
e-mail: floydt{at}uphs.upenn.edu
| Abstract |
|---|
|
|
|---|
METHODS: Cerebral blood flow was measured using the noninvasive continuous arterial spin labeling perfusion magnetic resonance imaging method. Cerebral blood flow, mean arterial pressure, hemoglobin, hemoglobin oxygen saturation, and cardiopulmonary bypass time were recorded in 12 subjects before and 6 ± 2 days after cardiac surgery.
RESULTS: Cerebral blood flow increased from 44.6 ± 15.6 mL100 g-1min-1 to 64.4 ± 20.1 mL100 g-1min-1 after cardiac surgery, or 49.1% ± 26.7%, (p < 0.0001). The absolute change in CBF (
CBF) was predicted by the following regression model:
CBF = -55 + 0.64(Age) + 0.53(CBFPre) -3.3(
Hgb); R2 = 0.81; p = 0.003, where CBFPre is the baseline preoperative CBF and
Hgb is the change in hemoglobin from preoperative to postoperative periods.
CONCLUSIONS: Cerebral blood flow increases after cardiac surgery, and anemia appears to be an important cause. Age appears also to be an important covariate, with advancing age further increasing the magnitude of this hyperemia. The interrelationship of aging and anemia, in determining perioperative changes in CBF, and potentially cerebral oxygenation, may have important implications for the understanding of perioperative stroke and cognitive dysfunction after cardiac surgery.
| Introduction |
|---|
|
|
|---|
Cerebral blood flow (CBF) during cardiopulmonary bypass is increased in multiple published reports. Cook and associates [7] identified a fall in hematocrit as primarily responsible for this phenomenon. We have identified only three published reports documenting CBF changes after cardiac surgery. Smith and coworkers [8] found that CBF decreased 1 week after cardiac surgery relative to preoperative levels, and Abildstrom and colleagues [9] recently reported that CBF decreased 6 days after cardiac surgery in spite of postoperative anemia. A third report documents an increase in CBF 1 month after cardiac transplantation yet attributes this to an improvement in cardiac output [10] with no assessment of the effect of anemia. Interpreting any change in CBF surrounding cardiac surgery must address the impact of anemia on those changes [11]. These inconsistencies in intraoperative and postoperative findings warrant further investigation.
To begin to address these points we have conducted an observational study to explore changes in CBF after cardiac surgery and their relationship to changes in perioperative hemoglobin. Additionally, because age is associated with cerebrovascular disease [12] and potential limitations in cerebrovascular reserve [13], as well as stroke [1] and cognitive dysfunction [14] after cardiac surgery, we examined the influence of age on any anemia-related increase in CBF.
| Patients and methods |
|---|
|
|
|---|
Subjects underwent continuous arterial spin labeling perfusion magnetic resonance imaging (CASL-P-MRI) measurement of CBF (mL100 g-1min-1) before (during preoperative testing or on hospital admission) and after cardiac surgery (as soon as medically stable and before discharge). Table 1 lists all abbreviations, definitions, and equations. Mean arterial pressure (MAP) and arterial oxygen saturation (SaO2) were measured noninvasively at the time of each magnetic resonance imaging (MRI) study. Steady-state hemoglobin (Hgb) values were obtained from preoperative testing and postoperatively within 24 hours of imaging. In all cases Hgb data reflected steady-state conditions before and after surgery, ie, no ongoing losses. Cardiopulmonary bypass duration (TCPB) was recorded from the perfusion record. An estimated arterial oxygen content (CaO2Est) was arrived at using a modification (Table 1) of the well-described equation CaO2 = 1.36(Hgb)SaO2/100 + (0.0031)PaO2. Assumptions include that the noninvasive measurement of SaO2 accurately reflects arterial SaO2, and that ignoring the contribution of dissolved O2 to CaO2 is tolerable. Within the physiologic range of PaO2, ignoring this term should cause minimal error. Estimated cerebral oxygen delivery (CDO2Est) was calculated according to the formula in Table 1.
|
Perfusion data were saved as raw echo amplitudes and transferred to a workstation for processing. Custom software, written in the Interactive Data Language (IDL Research Systems, Boulder, CO), was used for image reconstruction. The 45 pairs of labeled and control images were corrected for motion and then averaged to produce a single set of perfusion-sensitive images. Perfusion was quantified using a cerebrospinal fluid reference as previously described [20]. Global CBF was determined by averaging perfusion values across all brain voxels.
Mean arterial pressure and SaO2 were measured noninvasively using a Medrad 9500, MRI-compatible monitor (Medrad, Inc, Indianola, PA). Both measurements were made with the subject recumbent at the end of MRI scanning.
Statistical analysis was performed using SAS JMP Professional, v.5 (SAS Institute, Cary, NC). The paired Student's t test was used to estimate differences in mean CBF from the preoperative to postoperative periods. Multiple linear regression methods were used to model the contribution of the variables tested to the outcome
CBF. The regression coefficient, R2, the F ratio, and p values are presented for analysis of variance modeling of the response. Additionally, t ratio, p values, power (ß), and least significant number for each partial regression coefficient are presented. The final model was selected from significance of partial coefficients. A probability of p
0.05 was chosen to test significance for the whole model and for the partial regression coefficients.
Absolute CBF change from the preoperative to postoperative periods (
CBF) was analyzed as a function of
Hgb,
SaO2, age, preoperative CBF (CBFPre),
MAP, and TCPB. Alternately,
CaO2Est was substituted in model development for
Hgb and
SaO2 as it reflects variations in both of these variables.
| Results |
|---|
|
|
|---|
Data are presented for the 12 subjects who completed the study. Subjects were imaged preoperatively and 6 ± 2 days postoperatively (range, 3 to 12 days). Cerebral blood flow for the 12 subjects increased from preoperative levels of 44.6 ± 15.6 to 64.4 ± 20.1 mL100 g-1min-1 postoperatively, for an average increase of 19.8 ± 9.2 mL100 g-1min-1 or 49.1% ± 26.7%, p < 0.0001. Representative colorized and scaled CASL-P-MRI CBF images for both periods from subject 5 are presented in Figure 1.
|
CBF was predicted best by the variables age, CBFPre, and
Hgb, in the following model:
CBF = -55 + 0.64 (Age) + 0.53 (CBFPre) -3.3 (
Hgb); R2 = 0.81, F ratio = 11.0, and p = 0.003 for the model. A statistical summary for each term, to include power and least significant numbers at
= 0.05, is presented in Table 2.
A plot of the actual
CBF versus the model predicted
CBF with 95% confidence intervals is presented in Figure 2.
|
|
Hgb,
SaO2, and
CaO2. Age was associated with a trend toward smaller postoperative decreases in Hgb (R2 = 0.09, p = 0.36), slightly greater decreases in SaO2 (R2 = 0.11, p = 0.31), and smaller decreases in CaO2Est (R2 = 0.15, p = 0.23), none of which were significant.
Figure 3
demonstrates a rather tight relationship between age and
CDO2Est (R2 = 0.95, p < 0.0001). It can be seen that individuals younger than the age of approximately 50 years experienced a fall in CDO2Est in the face of anemia, in spite of an increasing CBF. It can also be seen that individuals older than the age of approximately 50 years responded in an opposite manner, and experienced an increase in CDO2Est under similar conditions.
|
| Comment |
|---|
|
|
|---|
A significant change in CBF from the preoperative to postoperative periods after cardiac surgery with cardiopulmonary bypass was observed. Cerebral blood flow increased from baseline levels of 44.6 ± 15.6 to 64.4 ± 20.2 mL100 g-1min-1, for an average increase of nearly 50%, even at 6 days postoperatively (p < 0.0001). In preliminary univariate analysis before multiple regression modeling,
Hgb did not reach significance as a predictor of change in CBF. In multiple linear regression models, however, after controlling for the effects of other variables such as age,
Hgb emerged as a significant predictor of
CBF (Table 2). Change in CaO2Est, an estimate of oxygen content, was also found to be a significant (p = 0.04) predictor of
CBF when substituted in our regression models for
Hgb and
SaO2, yet was not more predictive than
Hgb.
Cerebral blood flow is known to vary inversely and linearly with hematocrit [26]. This linearity in CBF reserve in response to anemia seems to hold true even during cardiopulmonary bypass with hemodilution to remarkably low levels [27]. Cerebral blood flow increases in response to a decrease in CaO2 regardless of whether the decrease in CaO2 is secondary to falling PO2 or falling hematocrit [28]. Normal brain is remarkably adaptive to changes in oxygen delivery, increasing CBF and finally oxygen extraction in response to decreasing supply [29]. Hemodilution is associated with a decrease in viscosity in addition to a fall in CaO2, and decreased viscosity may also increase CBF [30]. The acute anemia, which is nearly instantaneous with the commencement of cardiopulmonary bypass and which persists into the postoperative recovery phase, is certainly an important contributor to the increased CBF identified in this study.
Because of the association of age with cerebrovascular disease and its potential impact on cerebrovascular reserve, as well as stroke and cognitive dysfunction after cardiac surgery, we examined the influence of age on changes in CBF after cardiac surgery. Age was the most important coefficient in a multiple linear regression model of
CBF (Table 2). In our study, aging was associated with a robust increase in CBF after cardiac surgery. Although no previous study has looked at an age-related cerebrovascular response to anemia, previous studies testing age-related cerebrovascular reactivity to CO2 have yielded inconsistent results [13, 31]. It is surprising, at the very least, to find that the aged brain has the ability to generate such a marked cerebrovascular response to an anemic challenge.
As the cerebral circulation is controlled to a great extent by CaO2, one could explain the age-related increase in CBF if aging was associated with more severe anemia or greater lung injury, resulting in a greater decrease in SaO2, and or CaO2Est. Our results demonstrate that although there might have been a trend toward a greater fall in SaO2 from the preoperative to postoperative periods in the more aged, the fall in CaO2 was actually less pronounced in the more aged, as the decreasing SaO2 was offset by higher postoperative Hgb values.
Figure 3 demonstrates that although younger subjects tended to experience a decrease in CDO2Est, the more aged tended to experience an increase in CDO2Est. Previous work has found that during hemodilution, even as CBF increases, CDO2 actually falls, and cerebral oxygenation is maintained by increasing oxygen extraction [28, 32]. This pattern is consistent with our results in younger cohorts. These findings contrast markedly, however, from our results in the more aged in whom CDO2 increased after hemodilution.
Signaling and control over the cerebral vasculature in response to oxygen demand and supply is complex and may involve both CaO2 and capillary O2 tension, as well as important parenchymal feedback mechanisms [33]. Is the relative hyperemic response seen in the aged in our study, which is accompanied by an increased CDO2Est, an indication that there is some degree of loss of control over the circulation resulting in truly luxury perfusion (oxygen delivery in excess of a requirement)? Or is it potentially and more ominously an indicator that in spite of the increased CBF and CDO2, the aged brain is not able to efficiently extract oxygen under these conditions, resulting in mild hypoxemia and an accelerating vasodilatory response to that hypoxemia (a physiologic shunt)?
Baseline CBF (CBFPre) did appear to limit the change in CBF in response to anemia. In spite of the relationship between aging and diminished baseline CBF [34], albeit weak, each variable, age and CBFPre, seemed to exert important independent and opposite effects on
CBF (Table 2). Preoperative CBF (CBFPre) or baseline CBF may be an indicator of cerebrovascular disease, and subjects with lower values may be limited in their ability to respond to challenges such as acute anemia.
Our results are in agreement with the intraoperative results published by Cook and colleagues [7], yet they do not agree with the postoperative results of Smith and associates [8] or Abildstrom and coworkers [9], which described a decrease in CBF after cardiac surgery. With the plethora of evidence documenting microembolization during cardiopulmonary bypass, one might expect that CBF would decrease in response to an accumulation of vascular debris, yet this is not consistent with our observations. To the contrary, Sungurtekin and colleagues [35] have reported that microembolization might actually result in an increase in CBF and loss of autoregulation. The observations reported by Smith and coworkers [8] and Abildstrom and associates [9] were not examined as a function of hemoglobin or CaO2, and differences between their results and ours may be secondary to differences in transfusion practices. Significant degrees of perioperative anemia are accepted at ours and other institutions, whereas subjects may be more aggressively transfused elsewhere. Interestingly, Gruhn and colleagues [10] recently reported a significant perioperative increase in CBF after cardiac transplantation but attributed this increase to an improvement in cardiac output and did not make any attempt to relate those changes to postoperative anemia.
Our study would have benefited greatly from direct measurements of arterial O2 and CO2 tensions. Beyond what we have already discussed regarding the importance of oxygen content in contributing to any age-related process, the same could be said for CO2. The effects of narcotics, even at the delayed date of our postoperative examinations when narcotic requirements should have been minimal, may have had greater effects on the elderly population in this study. We did not measure changes in viscosity, and although hematocrit is primarily responsible for decreases in viscosity associated with anemia, it is at least possible that viscosity decreased to a greater extent in the more aged secondary to mechanisms other than falling hematocrit, such as lowering of serum protein content. In spite of the fact that the number of subjects in this study is only 12, the power and least significant numbers listed in Table 2 give additional support to these preliminary yet provocative data, especially concerning the relationship of age to the changes in CBF. No other literature, in an animal model or in the human, has compared the cerebrovascular response of the aged versus younger brain to acute anemia, nor has any study compared the translation of that response into actual parenchymal oxygenation across age ranges.
In conclusion, CBF increases after cardiac surgery with cardiopulmonary bypass, and this response depends significantly on the magnitude of the drop in Hgb. The degree of this response appears to increase with age. Aging may be associated with an impaired CBF response to anemia at baseline or after cardiopulmonary bypass. As recent work suggests that cardiopulmonary bypass may not importantly contribute to postoperative cognitive dysfunction [5] and as it has been shown that this sequela occurs in noncardiac [6] as well as cardiac surgery, a process common to both types of procedures warrants investigation. Because acute anemia has been found to create acute cognitive dysfunction even in healthy subjects [36] and chronic anemia to create the same in chronic disease states [37, 38], and as it is a process common to both cardiac and noncardiac surgery, we suggest its potential contribution to postoperative cognitive dysfunction warrants further investigation. Lastly, the impact of aging on cerebral oxygenation in the face of acute anemia is not known and should be addressed.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. F. Gottesman, J. Sojkova, L. L. Beason-Held, Y. An, D. L. Longo, L. Ferrucci, and S. M. Resnick Patterns of Regional Cerebral Blood Flow Associated With Low Hemoglobin in the Baltimore Longitudinal Study of Aging J Gerontol A Biol Sci Med Sci, September 1, 2012; 67(9): 963 - 969. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Baufreton, F. Pinaud, J. Corbeau, A. Chevailler, D. Jolivot, A. Ter Minassian, D. Henrion, and J. de Brux Increased cerebral blood flow velocities assessed by transcranial Doppler examination is associated with complement activation after cardiopulmonary bypass Perfusion, March 1, 2011; 26(2): 91 - 98. [Abstract] [PDF] |
||||
![]() |
E T Petersen, I Zimine, Y-C L Ho, and X Golay Non-invasive measurement of perfusion: a critical review of arterial spin labelling techniques Br. J. Radiol., August 1, 2006; 79(944): 688 - 701. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Bronden, M. Dencker, M. Allers, I. Plaza, and H. Jonsson Differential Distribution of Lipid Microemboli After Cardiac Surgery Ann. Thorac. Surg., February 1, 2006; 81(2): 643 - 648. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |