|
|
||||||||
Ann Thorac Surg 2002;74:109-114
© 2002 The Society of Thoracic Surgeons
a Department of Cardiothoracic Surgery, Würzburg, Germany
b Department of Neurology, University Hospital, Würzburg, Germany
Accepted for publication March 19, 2002.
* Address reprint requests to Dr Reents, Klinik für Herz- und Thoraxchirurgie, Josef-Schneider-Str. 6, 97080 Würzburg, Germany
e-mail: wilko.reents{at}mail.uni-wuerzburg.de
| Abstract |
|---|
|
|
|---|
Methods. Forty-seven patients undergoing elective coronary artery bypass grafting with cardiopulmonary bypass underwent preoperative and postoperative neuropsychological evaluation. Patients were classified according to the presence or absence of postoperative cognitive dysfunction. Cognitive dysfunction was defined as an individual test score decrease of more than one standard deviation in two or more of the five tests. During operation ScO2 was continuously measured using an INVOS 4100 device. Cerebral oxygen saturation values were analyzed with reference to two cutoff points, which should reflect low cerebral oxygenation: an ScO2 less than 40% and a drop of more than 25% from individual baseline values. The duration and extent of ScO2 values below these two cutoff points was compared between the patients with and without cognitive dysfunction.
Results. Sixteen patients (34%) showed postoperative cognitive dysfunction. Cerebral oxygen saturation values less than 40% occurred in 17 patients for a mean (± standard error of the mean) of 17.2 ± 6.5 minutes, whereas a decrease of more than 25% from baseline values occurred in 37 patients for 52.7 ± 7.8 minutes. The duration and extent below the two cutoff ScO2 values was similar in patients with and without cognitive dysfunction.
Conclusions. Intraoperative regional ScO2 as assessed by near-infrared spectroscopy with the INVOS 4100 device is not predictive for postoperative cognitive performance in patients undergoing coronary artery bypass grafting with cardiopulmonary bypass.
| Introduction |
|---|
|
|
|---|
Transcranial near-infrared spectroscopy (NIRS) permits continuous noninvasive measurement of oxygenated and total hemoglobin within a small compartment of the frontal cortical region. The resulting regional cerebral oxygen saturation (ScO2) represents the oxygen saturation derived from several vascular components (eg, arterial, capillary, and venous blood), thus being a relative measure of the actual oxygenation status in a restricted area. It is influenced by the actual oxygen supply (depending on the arterial cerebral blood flow, the arterial PO2, and the hemoglobin concentration) and the cerebral oxygen consumption (depending on the depth of anesthesia and temperature) [7]. A correlation between jugular venous bulb oxygen saturation, which assesses the balance between global cerebral oxygen supply and demand, and ScO2 as assessed by NIRS has been demonstrated [8, 9]. Several studies with the application of NIRS during pediatric cardiac operations [10, 11] and during procedures with deep hypothermic circulatory arrest [12, 13] suggested that brain oxygenation can be assessed by NIRS and that NIRS monitoring may optimize neurologic outcome. Because of only limited available data [14] regarding intraoperative cerebral oxygenation as assessed by NIRS and their impact on neuropsychological outcome, the aim of this study was to evaluate the value of NIRS for predicting postoperative cognitive performance. We hypothesized that patients with postoperative cognitive decline will experience longer or more profound periods of intraoperative cerebral hypoxia-ischemia as assessed by NIRS.
| Material and methods |
|---|
|
|
|---|
|
-stat method. During CPB the threshold for transfusion of packed red blood cells was a hematocrit less than 21%. Cardiac arrest was achieved by intermittent cold blood cardioplegia. Anticoagulation was achieved by administration of heparin (500 IU/kg) before the onset of CPB and was checked by means of the activated clotting time, which had to be in excess of 400 seconds during CPB. At the end of CPB anticoagulation was reversed by protamine. All patients received aprotinin (Antagosan; Hoechst GmbH, Frankfurt, Germany), 2 x 106 IU at the beginning and 0.5 x 106 IU/h during CPB.
Near-infrared spectroscopy
Cerebral oxygen saturation was continuously monitored using the INVOS 4100 (Somanetics, Troy, MI) with bifrontal placement of two sensors after induction of anesthesia and arrival at the operating theater until skin closure. The device measured ScO2 every 10 seconds, and ScO2 values were stored automatically on floppy disk for both hemispheres. Surgeons, anesthesia staff members, and perfusionists were blinded to the measurement of ScO2 so that no interventions were made according to ScO2 values. Cerebral oxygen saturation values were evaluated with regard to two cutoff points that should reflect critically low oxygenation [1517]: the duration patients experienced an ScO2 less than 40% and the duration they experienced a relative ScO2 decrement of more than 25% below their individual baseline value. Individual baseline values were estimated as the mean of the initial 10 measurements. To estimate the extent of ScO2 deterioration, the corresponding area under the curve was calculated.
Neurologic and neuropsychological assessment
A complete neurologic examination for the assessment of stroke signs and of overt encephalopathy (defined as the presence of delirium according to the Diagnostic and Statistical Manual of Psychiatric Diseases, DSM IV) was performed preoperatively and on the first, third, and sixth postoperative day. Each patient underwent a comprehensive neuropsychological examination preoperatively and on the sixth postoperative day. The applied tests were selected according to the recommendations for neuropsychological examination of patients undergoing cardiac surgery [18, 19]. A comparable test setup has been evaluated in a large cohort of patients undergoing CABG [2]. Specific tests used and the investigated cognitive domain were as follows: (1) d2-letter cancellation test (sustained concentration, attention) [20]; (2) Halstead-Reitan trail-making test B (attention, psychomotor speed, hand-eye coordination) [21]; (3) Bentons visual retention test (visual short-term memory, visuomotor abilities) [22]; (4) block design test from the Wechsler adult intelligence test (clumsiness, visuospatial and constructive abilities) [23]; and (5) digit span test from the Wechsler adult intelligence test (verbal immediate memory, attention span) [23]. To control for clinically significant depression, the depression inventory of Beck and colleagues [24] was performed additionally.
Data analysis
For each individual patient, the difference between preoperative and postoperative test scores in every neuropsychological test was calculated. Cognitive dysfunction was defined as a test score decrease of one standard deviation or more in two or more of the five tests. The standard deviation for each test was calculated from preoperative test scores of the study population. Patients were divided into two groups according to the presence or absence of cognitive dysfunction.
For ScO2 data analysis the time below the two cutoff points as well as the corresponding area under the curve was assessed for each hemisphere, and the mean of both measures was calculated for an individual patient. The area under the curve was calculated according to the following formula:
[ScO2(n) + ScO2(n + 1)]/2 x 10 seconds with ScO2(n) and ScO2(n + 1) representing the difference between the measured value (if it was less than the cutoff value) and the cutoff value at the respective measurement points; calculation was performed with Microsoft Excel 7.0 (Microsoft Corp, Redmond, WA). For the patient groups with or without cognitive dysfunction the corresponding time and the area under the curve below the cutoff ScO2 values was tested for difference using the Mann-Whitney U test.
Furthermore, two subgroups of patients with a diametral neuropsychological outcome were identified. These were the patients with postoperative delirium and, on the contrary, the patients with constant or improved test scores in all applied neuropsychological tests. Again, the corresponding time and the area under the curve below the cutoff ScO2 values were compared in these subgroups. The demographic, clinical and intraoperative characteristics of these patient groups were analyzed using Students t test for continuous and
2 test or Fishers exact test for categorical variables with a probability value less than 0.05 indicating significance. Power calculation was performed with nQuery Advisor 4.0 (Statistical Solution Ltd., Los Angeles, CA).
| Results |
|---|
|
|
|---|
|
Neurologic and neuropsychological outcome
Postoperatively none of the patients showed clinical signs of stroke. Six of the 50 patients had symptoms of overt postoperative encephalopathy with delirious state, reduced level of consciousness, and disorientation, which in each case resolved during the hospital stay. These 6 patients all had a test score decrease of one standard deviation or more in two or more tests, thus having cognitive dysfunction as measured by the applied neuropsychological tests.
For the whole patient cohort, test scores of all tests showed a slight, but nonsignificant decrease on the sixth postoperative day as compared with baseline (Table 2). According to individual test score differences, there were 20 patients without a test score decrease of one standard deviation or more in any of the five tests, and 11 patients with a test score decrease of one standard deviation or more in one of the five tests. Sixteen patients had a test score decrease of one standard deviation or more in two or more of the five tests. Thus according to the above-mentioned definition, in our study population 16 of 47 patients (34%) showed early postoperative cognitive dysfunction.
|
|
| Comment |
|---|
|
|
|---|
Two points of our study require special consideration: the neuropsychological test procedure and the cutoff points to define low cerebral oxygenation as assessed by NIRS. There is an ongoing discussion about methodological issues regarding the assessment of cognitive function in patients undergoing cardiac operations. One major topic is the arbitrary definition of global cognitive dysfunction [4, 25]. We conservatively defined a change of cognitive function as a test score decrease of one standard deviation in two or more of five tests to discriminate patients according to their early postoperative cognitive performance and to allow for comparison with other published studies. According to this definition, 34% of our patients showed a cognitive decline 6 days after CABG, which is at the lower range of the reported incidence using the same definition [4, 25]. This might be related to the selected patient population with a low prevalence of risk factors for an adverse neurologic outcome [26]; age more than 75 years, cerebrovascular disease, stenosis of the internal carotid artery, and reoperation were all exclusion criteria in the present study. Six patients showed signs of postoperative delirium. According to the neuropsychological test results, all these patients were classified as having cognitive dysfunction. Although the number of patients was small, this observation confirms the validity of the test procedure.
The cutoff points to define low cerebral oxygenation were taken from the available literature. Cerebral oxygen saturation values less than 41% during CABG were associated with a higher number of patients with impairment of rapid antisaccadic eye movements, which was interpreted as frontal lobe dysfunction [16]. During carotid endarterectomy a decrease of ScO2 to less than 50% or a relative decrease of more than 20% from baseline values was associated with transient changes in the amplitude of somatosensory evoked potentials, indicative of focal cerebral ischemia [17]. In another study on patients undergoing awake carotid endarterectomy, two patients with intraoperative hypotension and symptoms of global cerebral ischemia had a concomitant ScO2 decrease of more than 25% of their baseline values. Smaller decreases during cross-clamping of the common and internal carotid arteries were not associated with stroke symptoms [27]. In patients undergoing awake cardioverter-defibrillator testing, signs of global cerebral ischemia, eg, syncope during ventricular fibrillation, were associated with an ScO2 decline of more than 20% of baseline values [15]. We therefore chose the most pronounced cutoff points to define critically low oxygenation. One might argue that critically low cerebral oxygenation could occur below other threshold values. Yet if relevant dysoxygenation of the brain only occurs at lower levels, this should be detected with the calculated area under the curve. A higher absolute threshold value (eg, <50%) seems to be not useful given the wide range of normal baseline ScO2 values (49% to 86% in our study population). An ScO2 decrease of more than 25% relative to individual baseline values occurred in 80% of the patients for an average of 53 minutes. According to that definition, the majority of our patients experienced a prolonged period of low cerebral oxygenation. We additionally calculated the time and corresponding area under the curve with higher threshold values to define low oxygenation (eg, a decrease of ScO2 greater than 20%, 15%, and 10% relative to baseline values) of the 4 patients with constant or improved test scores and the 6 patients with clinically overt postoperative encephalopathy. Again, for all these threshold values the time and the amount of dysoxygenation did not differ between these patient groups with a markedly distinct neuropsychological outcome (data not shown). Taken together, we were unable to detect a critical threshold with respect to postoperative cognitive performance. The above cited studies [1517, 27], which proposed critically low ScO2 values during temporary cerebral ischemia, did not include detailed neuropsychological testing.
A similar study to ours found a correlation between low cerebral oxygenation and poorer cognitive outcome in patients undergoing cardiac operation with CPB [14]. However, the neuropsychological examination was performed with the Mini-Mental State Questionnaire, which is a quite crude test instrument for the detection of obvious dementia [28], and NIRS was performed with another device using a different algorithm for measuring cerebral tissue oxygenation. However, although noninvasive measurement of cerebral tissue oxygenation status with NIRS would be ideal, the accuracy of available devices with respect to this aim appears not to be sufficient [29, 30].
We were unable to provide a threshold of regional ScO2 that is critical with respect to postoperative brain dysfunction. The incidence or amount of low ScO2 was quite similar in patients with or without postoperative cognitive decline. A power calculation, which was performed with the incidence of an ScO2 value less than 40% in both patient groups (12 of the 31 patients,
1 = 0.387, without cognitive decline and 5 of 16 patients,
2 = 0.313, with cognitive decline), revealed that the study with this number of patients would only have a power of 4% to detect a difference between the groups. Vice versa, with this incidence one would need 1,500 patients, 1,000 without cognitive decline and 500 with cognitive decline, to achieve a power of 80% for detecting a difference between the groups. If the incidence of a low ScO2 value would have been more discernible, for example, 20% of the patients without cognitive decline experience an ScO2 value below 40% as opposed to 80% of the patients with cognitive decline, the power of the study would be 97% with the studied number of patients. The key question therefore is, which level of specificity is clinically useful. We believe that a device that measures an event in approximately one third of patients with or without cognitive decline is not specific enough to predict cognitive decline. Even if a study with a large number of patients would reveal a significant difference, the clinical value would be questionable owing to the low specificity. Therefore at the present time regional ScO2 as assessed by NIRS with the INVOS 4100 device is not useful for predicting postoperative neuropsychological outcome in patients undergoing CABG.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. M. Hemmerling, M. C. Bluteau, R. Kazan, and D. Bracco Significant decrease of cerebral oxygen saturation during single-lung ventilation measured using absolute oximetry Br. J. Anaesth., December 1, 2008; 101(6): 870 - 875. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. W. Hong, J. K. Shim, Y. S. Choi, D. H. Kim, B. C. Chang, and Y. L. Kwak Prediction of cognitive dysfunction and patients' outcome following valvular heart surgery and the role of cerebral oximetry Eur. J. Cardiothorac. Surg., April 1, 2008; 33(4): 560 - 565. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. W. Fischer Recent Advances in Application of Cerebral Oximetry in Adult Cardiovascular Surgery Seminars in Cardiothoracic and Vascular Anesthesia, March 1, 2008; 12(1): 60 - 69. [Abstract] [PDF] |
||||
![]() |
C. Olsson and S. Thelin Regional cerebral saturation monitoring with near-infrared spectroscopy during selective antegrade cerebral perfusion: Diagnostic performance and relationship to postoperative stroke J. Thorac. Cardiovasc. Surg., February 1, 2006; 131(2): 371 - 379. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Kunihara, T. Grun, D. Aicher, F. Langer, O. Adam, O. Wendler, Y. Saijo, and H.-J. Schafers Hypothermic circulatory arrest is not a risk factor for neurologic morbidity in aortic surgery: A propensity score analysis J. Thorac. Cardiovasc. Surg., September 1, 2005; 130(3): 712 - 718. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Plachky, S. Hofer, M. Volkmann, E. Martin, H. J. Bardenheuer, and M. A. Weigand Regional Cerebral Oxygen Saturation Is a Sensitive Marker of Cerebral Hypoperfusion During Orthotopic Liver Transplantation Anesth. Analg., August 1, 2004; 99(2): 344 - 349. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. P. Alston Cerebral perfusion and cognitive function Ann. Thorac. Surg., April 1, 2003; 75(4): 1361 - 1361. [Full Text] [PDF] |
||||
![]() |
W. Reents, J. Babin-Ebell, and W. Muellges Cerebral perfusion and cognitive function: Reply Ann. Thorac. Surg., April 1, 2003; 75(4): 1361 - 1362. [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 |