|
|
||||||||
J Thorac Cardiovasc Surg 2008;136:962-967
© 2008 The American Association for Thoracic Surgery
Surgery for Congenital Heart Disease |
a Department of Anesthesia, Deutsches Herzzentrum Berlin, Berlin, Germany
b Department of Cardiothoracic and Vascular Surgery, Deutsches Herzzentrum Berlin, Berlin, Germany
c Department of Perfusion, Deutsches Herzzentrum Berlin, Berlin, Germany
d Department of Congenital Heart Disease, Deutsches Herzzentrum Berlin, Berlin, Germany
e Institute of Physiology, Charité–Universitaetsmedizin Berlin, Berlin, Germany
Received for publication October 8, 2007; accepted for publication December 17, 2007. * Address for reprints: Andreas Koster, MD, Department of Anesthesia, Deutsches Herzzentrum Berlin, Augustenburger Platz 1, D-13353 Berlin, Germany. (Email: koster{at}dhzb.de).
| Abstract |
|---|
|
|
|---|
Methods: Twenty patients (body weight < 10 kg) undergoing open cardiac procedures with cardiopulmonary bypass were enrolled. Blood samples were obtained in parallel from inferior and superior caval vein cannulas and mixed venous line and assessed for venous oxygen saturation and lactate levels. Data were compared to simultaneously measured tissue oxygenation indices obtained by near-infrared spectroscopy from brain and lower limb.
Results: Venous oxygen saturation was lower and lactate concentration higher in blood from superior relative to inferior venous line. Mixed venous oxygen saturation correlated with venous oxygen saturation from inferior venous line and tissue oxygenation index of lower limb. No correlation was found between mixed venous oxygen saturation and venous oxygen saturation from superior venous line or cerebral tissue oxygenation index.
Conclusion: In neonates and small infants undergoing cardiac surgery with cardiopulmonary bypass, considerable regional differences exist in venous oxygen saturation. Mixed venous oxygen saturation primarily represents lower-torso oxygen status but poorly reflects and systematically overestimates upper-body oxygenation. Near-infrared spectroscopy yields additional information on regional oxygenation and may be valuable in early and sensitive detection of regional malperfusion in critical organs such as the brain.
| Introduction |
|---|
|
|
|---|
To date, the only commonly used online-monitored parameter for the quality of perfusion and oxygenation during CPB in neonates and small infants is the mixed venous oxygen saturation (S
O
2), as optically measured in the common venous line of the CPB system.8
The S
O
2 is, however, a poor predictor of regional oxygenation levels in different organs and compartments of the body.9,10
Physiologically, regional inhomogeneities in tissue oxygenation can be attributed to organ-specific differences in tissue perfusion, in the regulation of vasomotor tone, in local diffusion distances, or in oxygen consumption. This situation is further complicated during surgical correction of CHD, because preexisting malformations of the vascular system—including hypoplastic vessels, shunts, or collateralization—as well as impaired arterial inflow or venous drainage through the CPB cannulas may cause additional inhomogeneities of perfusion.
Recently, near-infrared spectroscopy (NIRS), a noninvasive method for transcutaneous measurement of oxyhemoglobin and deoxyhemoglobin concentrations in tissue, has increasingly been used to monitor cerebral oxygenation during CPB.11
The use of multiple sensor–detector systems allows in addition the comparison of regional tissue oxygenations, such as between the head and the flank12
or between different cerebral areas.13
In a recent case report, we documented regional malperfusion in a patient with complex aortic anatomy by means of simultaneous transcranial and lower torso NIRS monitoring.14
This investigation was performed to assess systematically the utility of regional oxygenation monitoring by NIRS relative to that of S
O
2 measurements during CPB in neonates and small infants undergoing surgical correction of CHD.
| Materials and Methods |
|---|
|
|
|---|
Near-infrared Spectroscopy
A near-infrared spectrometer (NIRO-200; Hamamatsu Photonics KK, Hamamatsu City, Japan) equipped with two independent emitter-sensor pairs was used for simultaneous measurement of regional oxygenations in the brain and in skeletal muscle of the lower torso. After induction of anesthesia, the first pair of NIRS optodes was positioned noninvasively on the patient's forehead with a spatial separation of 4 cm and an estimated optical path length of 15.4 cm between the emission and detection probes.15
A second pair of optodes was placed on the right thigh with an estimated optical path length of 14.4 cm. Light was generated by three pulsed laser diodes at wavelengths of 775, 810, and 850 nm, respectively, and emitted into the tissue through the source optode. Backscattered light from the tissue was collected by the detection probe equipped with a dual-segmented photodiode chip. With the spatially resolved spectroscopy method, the tissue oxygenation index (TOI) was calculated from measured changes of light attenuation along the intersegmental distance of the chip.16,17
The TOI is the ratio of oxygenated to total hemoglobin and thus reflects mean hemoglobin oxygen saturation within the scanned tissue section.
CPB and Surgery
In all cases, CPB was performed with a S3 mast-mounted roller pump console (Stöckert Instrumente GmbH, Munich, Germany), which provided short tubing connections in our circuit consisting of tubing with 4.76 mm (16 inch) inner diameter in the entire system, with the exception of the roller pump segment of the arterial pump, which consisted of silicone rubber tubing with 6.35 mm (0,25 inch) inner diameter. The entire extracorporeal circuit, incorporating the Capiox RX05 hollow-fiber membrane oxygenator (Terumo Deutschland GmbH, Eschborn, Germany) and an arterial line filter (Dideco D736; Dideco SpA, Mirandola, Italy), had a total priming volume of 200 mL.18
The patients' CHD diagnoses and the surgical procedures are summarized in
Table 1. During CPB, a systemic hemoglobin concentration less than 8 g/L, corresponding to a hematocrit of approximately 24%, was considered the critical trigger for the transfusion of red blood cells.
|
Group data are given as mean ± SEM. Differences between dependent variables were analyzed by Wilcoxon signed rank test (SigmaStat; Systat Software Inc, San Jose, Calif). The Spearman coefficient of correlation (r_s) was calculated to test correlations between parameters, and linear regression analysis was performed (SigmaPlot; Systat).
| Results |
|---|
|
|
|---|
Direct comparison of SvO
2 values revealed a significantly lower SvO
2 in blood samples retrieved from the superior relative to the inferior venous line (
Figure 1, A). The notion that average blood oxygenation is lower in the upper than in the lower body compartment was confirmed by NIRS measurements yielding higher TOI values at the thigh than at the forehead (Figure 1, B). This regional heterogeneity in oxygenation is directly reflected in the metabolic level by higher lactate concentrations in venous blood from the upper than in the lower torso (Figure 1, C). Thus regional oxygen extraction during CPB is higher in the upper body than in the lower body, and direct measurements of SvO
2 or lactate in blood sampled from the mixed venous line of the CPB may overestimate the critical oxygenation status of the cerebral compartment.
|
|
|
|
| Discussion |
|---|
|
|
|---|
Our data demonstrate that in neonates and small infants undergoing surgery with CPB for the correction of CHD, S
O
2 represents predominantly the oxygenation status in the lower body compartment drained by the caval vein inferior line of the CPB circuit, whereas the oxygenation status of the upper body, and particularly the brain, is poorly reflected. This deficiency in monitoring is even more detrimental because relative perfusion is on average worse in the upper body during CPB, as reflected by lower SvO
2 and higher lactate concentration. S
O
2 thus systematically overestimates the adequacy of oxygenation and perfusion in the most critical organ, the brain, and serious cerebral malperfusion may remain unrecognized with the standard monitoring techniques.
Previous clinical studies evaluating the value of postoperative S
O
2 monitoring in neonates and infants have shown conflicting results. In several independent studies, regional cerebral oxygenation assessed by NIRS was correlated significantly with SvO
2 measured through a central venous or pulmonary arterial catheter.12,20,21
Surprisingly, the correlation between cerebral TOI and S
O
2 was particularly high when the central venous catheter was placed with the tip in the inferior as opposed to the superior caval vein or the right atrium.21
Moreover, a Bland–Altman comparison between S
O
2 and regional oxygen saturation revealed wide limits of agreement, suggesting that despite a significant correlation S
O
2 was of poor predictive value for cerebral oxygenation.12
During CPB, the adequacy of S
O
2 measurements to yield relevant information on the cerebral oxygenation status appears to be even further reduced. In our study, S
O
2 was correlated neither with cerebral TOI values determined by NIRS nor with SvO
2 values from the superior venous line. This finding is in agreement with previous studies reporting a lack of correlation between S
O
2 and jugular SvO
2 in both pediatric19
and adult22
patients undergoing CPB. The perception that S
O
2 provides only an unsatisfactory estimate of the cerebral oxygenation status during CPB is also supported by data from an experimental CPB model in swine, in which oxygenation in the sagittal sinus decreased from 66% ± 3% to 33% ± 2%, whereas S
O
2 remained largely unchanged between 75% and 80%.10
Our finding that S
O
2 represents primarily the lower body compartment during CPB is, furthermore, in agreement with a previous study by Lindholm and colleagues,10
who showed that S
O
2 correlates well with hepatic but not jugular vein oxygen saturation in adults undergoing CPB. This study therefore shows, in line with previous reports, the existence of considerable regional heterogeneities in oxygenation in neonates and infants during CPB, thus stressing the need for direct monitoring of critical organs at a local level.
Noninvasive monitoring of regional tissue oxygenation by NIRS may provide such crucial information. In our study, cerebral TOI measured by NIRS was correlated significantly with SvO
2 determined simultaneously in the superior venous line, with a correlation coefficient of 0.70. This is in agreement with previous studies demonstrating significant correlations between jugular bulb oxygen saturation and cerebral oxygenation determined by transcranial NIRS in children.23,24
The validity of the transcranial NIRS approach in yielding adequate and reliable information on cerebral oxygenation is further supported by data from Al-Rawi and coworkers,25
who demonstrated an 87.5% sensitivity and a 100% specificity of the TOI measurement with respect to changes in intracranial perfusion in the adult head.
The interpretation of NIRS data is limited, however, by the technical constraints of the technique and the lack of reliable reference values or critical thresholds in defined patient populations. Approximately 70% of the obtained NIRS signal is derived from the venous compartment, with capillaries and arterioles contributing 20% and 10%, respectively.26
TOI thus reflects SvO
2 in the cerebral compartment only partially, varying not only with changes in tissue perfusion or oxygen extraction but also with changes in the relative distribution of blood volume among the arteriolar, capillary, and venous compartments within the scanned tissue section. Taking this into account, variations in TOI monitored by transcranial NIRS can still yield critical information on changes in the cerebral oxygenation status. The clinical relevance of NIRS monitoring is, however, further complicated by the lack of accepted and validated critical thresholds for cerebral TOI. This applies in particular to the patient population of neonates and small infants monitored in our study. In a recent clinical investigation of 143 infants and children undergoing CHD surgery, Fenton and coworkers27
showed an association of baseline cerebral oxygenation values of less than 50% with perioperative death, suggesting that postponement of the surgical intervention or additional interventions such as red blood cell transfusions or hyperoxic ventilation might be warranted in cases with preoperative TOI values below this threshold. Under perioperative conditions including CPB, regional perfusion, and hypothermia, however, critical oxygenation thresholds in the brain remain largely unknown. This applies in particular to the patient population undergoing correction of CHD, in which individual susceptibility to tissue hypoxia may vary according to the underlying malformation and especially differs between cyanotic and noncyanotic malformations.
The high incidence and critical outcome of neurologic sequelae warrant optimization of cerebral monitoring during CPB in neonates and small infants.2-4
The results of our study demonstrate that the measurement of S
O
2 is inadequate for this purpose and should be supplemented with additional monitoring systems with higher regional resolution. Currently, NIRS provides the only versatile noninvasive technique for point-of-care monitoring during CPB. The extent to which improved regional monitoring of oxygenation in neonates and small infants with CHD undergoing CPB can serve as a perioperative guidance tool to optimize cerebral perfusion and improve clinical outcome remains to be determined in future studies.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Redlin, M. Huebler, W. Boettcher, M. Kukucka, H. Schoenfeld, R. Hetzer, and H. Habazettl Minimizing intraoperative hemodilution by use of a very low priming volume cardiopulmonary bypass in neonates with transposition of the great arteries J. Thorac. Cardiovasc. Surg., October 1, 2011; 142(4): 875 - 881. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Redlin, M. Huebler, W. Boettcher, H. Kuppe, R. Hetzer, and H. Habazettl How Near-Infrared Spectroscopy Differentiates Between Lower Body Ischemia Due to Arterial Occlusion Versus Venous Outflow Obstruction Ann. Thorac. Surg., April 1, 2011; 91(4): 1274 - 1276. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. M. Murkin and M. Arango Near-infrared spectroscopy as an index of brain and tissue oxygenation Br. J. Anaesth., December 1, 2009; 103(suppl_1): i3 - i13. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Koster, M. Huebler, W. Boettcher, M. Redlin, F. Berger, and R. Hetzer A new miniaturized cardiopulmonary bypass system reduces transfusion requirements during neonatal cardiac surgery: Initial experience in 13 consecutive patients. J. Thorac. Cardiovasc. Surg., June 1, 2009; 137(6): 1565 - 1568. [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 |