| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;83:475-482
© 2007 The Society of Thoracic Surgeons
a Department of Cardiac Surgery, St. George’s Hospital Medical School, London
c Department of Clinical Neuroscience, St. George’s Hospital Medical School, London
d Department of Cardiological Sciences, St. George’s Hospital Medical School, London
b Department of Health Sciences, University of York, York, United Kingdom
Accepted for publication September 6, 2006.
* Address correspondence to Dr Jahangiri, Department of Cardiac Surgery, St. George’s Hospital Medical School, London SW17 0QT, United Kingdom (Email: marjan.jahangiri{at}stgeorges.nhs.uk).
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
METHODS: 212 patients admitted for CABG were randomly assigned to on-pump (n = 104) or off-pump (n = 108) surgery. Embolic signals were detected with bilateral transcranial Doppler ultrasonography of the middle cerebral artery. Neurocognitive tests were administered preoperatively, on discharge from hospital, at 6 weeks, and at 6 months after surgery. Composite neurocognitive scores were derived using principal component analysis and were compared between the two groups, using analysis of covariance to adjust for baseline values.
RESULTS: At discharge from hospital, the adjusted composite neurocognitive score was 0.25 standard deviations greater in the off-pump group compared with the on-pump group (95% confidence interval: 0.05 to 0.45; p = 0.01). There was no significant difference at 6 weeks (0.09 standard deviations, 95% confidence interval: –0.11 to +0.30; p = 0.4) and 6 months (–0.002 standard deviations, 95% confidence interval: –0.23 to +0.23; p = 1.0). Median number of embolic signals was 1,605 (751 to 2,473) during on-pump and 9 (4 to 27) in off-pump CABG (p < 0.001). Age, length of education, and on-pump status were independent predictors of the predischarge neurocognitive score (p = 0.02, 0.03, and 0.006, respectively).
CONCLUSIONS: Cerebral emboli are more prevalent during on-pump CABG. At discharge from hospital, neurocognitive function is better after off-pump surgery, possibly as a result of the lower embolic load. However, the difference in neurocognitive function does not persist at 6 weeks and 6 months.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Traditionally, most neurologic complications after CABG have been attributed to the use of cardiopulmonary bypass (CPB) and manipulation of the aorta. Cerebral emboli [6] and hypoperfusion [7], as well as the systemic inflammatory response to CPB [8], are thought to be the main underlying causes. Most emboli arise from manipulation and instrumentation of the heart and aorta, and from the pump circuit [9]. With off-pump CABG (ie, surgery on the beating heart without the use of CPB), fewer emboli are generated [10].
The aim of this study was to determine whether there is a difference in postoperative neurocognitive function between patients undergoing off-pump and on-pump CABG, and whether that can be explained by the number of intraoperative cerebral emboli.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Patients referred for elective first-time isolated CABG at St. George’s Hospital, London, between August 2002 and March 2004 were candidates for inclusion in this study. Patients with the following criteria were excluded: (1) previous cerebrovascular accident or transient ischemic attack; (2) right or left internal carotid artery stenosis of 50% or greater; (3) previous cardiac surgery; (4) concomitant surgery, for example, valve replacement; (5) previous psychiatric illness, for example, depression, schizophrenia; (6) dialysis-dependent renal failure; (7) Q-wave myocardial infarction in the past 6 weeks; (8) very poor left ventricular function (ejection fraction < 20%); (9) illiteracy or nonfluency in English; and (10) absence of an acoustic window for transcranial Doppler (TCD) ultrasound monitoring. Patients were not excluded because of their coronary anatomy.
The primary outcome measure was a postoperative composite neurocognitive score at 6 months. The neurocognitive score at discharge and at 6 weeks, and the total intraoperative microemboli count formed secondary outcomes. Measurement of primary and secondary outcomes was carried out blinded to patient details and group allocation.
Patients were randomly allocated according to a computer-generated randomization list. Blocking was used to construct the allocation sequence, with block size varying randomly from 4 to 10 patients. Assignments were on cards and enclosed in serially numbered, opaque, sealed envelopes, and each bearing on the outside the name and date of birth of the enrolled patient. Envelopes were opened sequentially and only on the day of surgery for that patient.
Transcranial Doppler Ultrasonography Monitoring
Bilateral TCD ultrasonography of the middle cerebral arteries was performed using a Nicolet/EME Companion II machine (Eden Medizinische Elektronik, Kleinstheim, Germany) with two 2 MHz transducers. A sample volume of 10 mm and sweep speed of 5.1 s was set for all patients. A 128-point fast Fourier transform was used for spectral analysis. Fast Fourier transform time-window overlap was more than 50%. The middle cerebral arterey was identified at a depth of 48 to 58 mm through the transtemporal window, and the transducers were fixed in position using a Welder headset (Nicolet/EME GmbH, Kleinstheim, Germany). The Doppler signals were stored onto digital audiotapes using TCD-D8 recorders (Sony Corp, Tokyo, Japan). Recording of signals would commence after opening the pericardium and terminate upon sternal closure.
Audiotapes were individually coded and played back through the same TCD machine for off-line analysis at a later date. A single observer (R.M.) therefore conducted analysis blinded to subject identity and treatment group. An embolic signal (ES) was identified as a unidirectional short-duration intensity increase, accompanied by a characteristic "click" or "chirp" using Consensus criteria [11]. An intensity threshold of 7 dB was one of the criteria used to identify ES. The total number of ES for the whole procedure was counted, and the number of ES for 1 minute after surgical maneuvers was also documented.
Neurocognitive Assessment
Neuropsychologic test selection was based on the Statement of Consensus on assessing neurocognitive outcomes after cardiac surgery [12]. The examination included a battery of tests designed to test memory, visuomotor skills, attention, cognitive speed, and executive functions. The following tests were administered: Medical College of Georgia Complex Figure Test [13]; Grooved Pegboard Test—dominant and nondominant hand [14]; Rey Auditory Verbal Learning Test [15]; Letter Cancellation Test [14]; Trail Making Test—parts A and B [14]; Symbol Digit Modalities Test [16]; and Verbal Fluency Test [14].
Premorbid estimates of verbal intellectual abilities were obtained by administering the vocabulary subtest, and those of nonverbal intelligence by administering the matrix reasoning subtest of the Wechsler Abbreviated Scale of Intelligence (WASI) [17]. All neurocognitive assessment was conducted blinded to treatment details.
Subjects underwent baseline neurocognitive testing 1 week before surgery, and at three subsequent stages: on discharge from hospital and at 6 weeks and 6 months postoperatively. Testing was performed in a standardized manner. Patients were seated in a private room, and testing would only commence after a clear verbal indication that all instructions were understood.
Surgical Procedures
Premedication was administered with morphine (5 to 10 mg intramuscularly) and hyoscine (200 to 600 µg intramuscularly). Anesthesia was induced with either fentanyl (500 µg), propofol (0.5 to 1.0 mg/kg), and vecuronium (0.1 mg/kg) or alfentanil (1 mg), propofol, and pancuronium (0.5 mg/kg). Maintenance anesthesia was provided with isoflurane and propofol.
On-pump CABG was performed with a roller pump (Stöckert S3, Munich, Germany), membrane oxygenators (Avant Sorin, Mirandola, Italy), and a 40 µm arterial blood filter (Dideco, Mirandola, Italy). Moderate hypothermia (32°C) and 
-stat control of acid-base management was used. Perfusion pressure was kept at 60 mm Hg, and a pump flow of 2 to 2.4 L · min–1 · m–2 was maintained throughout CPB. Blood from cardiotomy suction catheters was separated from the pump circuit and washed with a cell saving device (Dideco). In this way, recirculation of fatty microemboli was minimized [18]. After completing the distal anastomoses, the aortic cross-clamp was removed, and the proximal anastomoses then performed using a single side-clamp on the aorta.
Off-pump CABG was carried out through a median sternotomy using a CTS stabilizer (Cardio Thoracic Systems, Cupertino, California). After all distal anastomoses, proximal anastomoses were fashioned onto the aorta using a single side-clamp. Near normothermia (35°C) was maintained, and systolic blood pressure was kept at 70 mm Hg or greater throughout.
Statistical Analysis
Based on a power of 90% to detect a difference of 0.45 standard deviations (SDs) between the on-pump and off-pump composite neurocognitive scores, a sample size of 212 patients was required.
The prespecified analysis of the cognitive tests was to combine them using a principal components analysis. The first component would form a summary score for all the tests. Comparison of the treatment groups was carried out using two-sample t tests and by analysis of covariance (ANCOVA) to adjust for the preoperative scores. Correlation of postoperative composite scores with the preoperative scores was performed using the Pearson correlation coefficient. Emboli counts were analyzed using rank-based, nonparametric tests. A p value less than 0.05 was considered significant. Statistical analyses were performed on Stata 8.2 software (StataCorp, College Station, Texas).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
|
There were 3 nonfatal strokes in the on-pump group and 1 in the off-pump group within 30 days of surgery. The overall 6-month stroke rate for on-pump and off-pump CABG was 5% (5 of 104) and 1% (1 of 108), respectively (p = 0.11).
Intraoperative Cerebral Emboli
The distribution of ES was highly skewed, and therefore nonparametric methods were used for analysis. The median (and interquartile range [IQR]) total intraoperative ES count was 1,605 (750 to 2,475) for on-pump CABG but only 9 (4 to 28) for off-pump CABG (p < 0.001, Mann-Whitney U test). Of the different surgical maneuvers, removal of the side-clamp and cross-clamp was associated with the highest rate of ES detection (Fig 2A). In the on-pump group, there was no difference in the median ES load in the left and right middle cerebral arteries: 825 (362 to 1,322) and 780 (368 to 1,220) respectively (p = 0.3, Wilcoxon paired test).
|
Neurocognitive Outcome
Principal component analysis was carried out on the baseline scores. The first component, the Complex Figure Test, contained the maximum amount of variation and represented 40% of the variation in test scores. A predictive score was then developed from the first component on the preoperative data and used to calculate a composite score for neurocognitive function. The same coefficients were applied to the remaining time points (predischarge, 6 weeks, and 6 months) to calculate a similar score. High scores represented better neurocognitive function.
The composite neurocognitive scores, for both on-pump and off-pump CABG, were all highly correlated with the baseline neurocognitive score (correlations between baseline and predischarge, baseline and 6 weeks, and baseline and 6 months were 0.88, 0.73, 0.78, respectively, for on-pump patients and 0.84, 0.79, 0.78, respectively, for off-pump patients (all p < 0.001).
The difference in the mean composite neurocognitive score (off-pump minus on-pump) was 1.13 (95% confidence interval [CI]: 0.09 to 2.17, p = 0.03) at discharge from hospital. After adjustment for the baseline score, the difference remained significant, estimate 0.73 (95% CI: 0.16 to 1.30, p = 0.01). To make this estimate more interpretable, we have calculated the differences in terms of SDs of the baseline score, which was 2.90. The difference in mean composite score adjusted for baseline score (off-pump score minus on-pump score) was 0.25 SDs (95% CI: 0.05 to 0.45, p = 0.01). At 6 weeks, the adjusted difference was reduced to 0.09 SDs (95% CI: –0.11 to 0.30, p = 0.4); and at 6 months, the difference was –0.002 SDs (95% CI: –0.23 to 0.23, p = 1.0).
We attempted to explain the difference in predischarge scores by determining the influence of the total ES count. Number of years of education, age at surgery, and on-pump status were independent predictors of the adjusted predischarge score, and so were retained in further analyses (increase in score per year of education = +0.04 SDs, 95% CI: 0.004 to 0.07, p = 0.03; increase in score per year of age = –0.01 SDs, 95% CIL –0.025 to –0.002, p = 0.02; and increase in score for on-pump surgery = –0.27 SDs, 95% CI: –0.47 to –0.08, p = 0.006). If on-pump status is not included in the multifactorial model, the number of ES is a significant predictor of the adjusted predischarge score (fall in score = –0.09 SDs for every factor of 10 increase in ES, p = 0.03). However, on-pump status is no longer a significant predictor of the predischarge score when the total number of ES is included in the regression equation (p = 0.08). The difference in neurocognitive score (predischarge minus baseline) with respect to number of ES was also analyzed (see Fig 3).
|
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This is the only large and randomized study to date that has compared both cerebral emboli and neurocognitive function between on-pump and off-pump CABG. The follow-up rate of 79% at 6 months is comparable to previous randomized studies, and in a large longitudinal study assessing neurocognitive function after on-pump CABG [4, 19]. Previous studies comparing neurocognitive function after on-pump versus off-pump surgery have been limited by either small sample size or follow-up that was very short term. Other studies have not assessed cerebral embolization [19, 20] or only used unilateral TCD [21]. By using bilateral TCD, and manually counting emboli according to well-defined criteria rather than relying on automated software, this study’s measurement of embolic load is more accurate.
An early protective effect of off-pump surgery with respect to neurocognitive dysfunction has been shown in previous reports. Neurocognitive testing was limited to the immediate postoperative period, however, and nonstandard tests were used [22, 23]. Other studies have reported markedly diverse incidences of postoperative neurocognitive dysfunction [19, 24]. This difference is related to differences in methodology such as type and number of neuropsychometric tests, timing of testing, and the definition of neurocognitive decline. There is no agreement as to what degree of change is indicative of neurocognitive impairment. One study showed that, depending on the definition used, the incidence of cognitive deficit after CABG ranged from 1.1% to 34% at 6 weeks and from 3.4% to 19.4% at 6 months postoperatively [25]. A widely used definition of neurocognitive dysfunction is the SD criterion, in which a postoperative deficit is defined as a deterioration of 1 SD in one or more tests compared with the preoperative test results [4]. Other categorical definitions of neurocognitive dysfunction include a 20% decline in 20% of tests performed [19]. These methods have limitations. Firstly, an arbitrary cut-off value without justification has been set to derive an incidence of neurocognitive dysfunction. Secondly, the SD and percentage change methods are influenced by regression to the mean [26]. As an example, the longitudinal study by Newman and colleagues [4] of neurocognitive function after CABG found that a significant predictor of postoperative decline was a high preoperative cognitive test score. Lee and associates [24] found significant improvement in memory after off-pump CABG, but this group had lower baseline scores and hence were susceptible to the effects of regression to the mean.
In our analysis, we derived a composite score from all the individual test scores and avoided categorizing patients as "cognitively impaired or not." It has been argued that using composite scores is counterintuitive, however, in that each test reflects a particular function or specific brain area, and that useful information will be lost [14]. Nevertheless, comparing cognitive domains instead of overall cognitive function will result in multiple comparisons that cannot be adequately corrected for. Hence, the use of composite scores has been recommended for analysis of postoperative cognitive function [27]. In addition, we have compared the postoperative scores having adjusted for preoperative scores by using ANCOVA. This takes regression to the mean into account and is a powerful method of analysing test–retest data [28].
We identified age as a predictor of cognitive function at discharge. Increasing age is associated with progressive atherosclerosis of the aorta and embolic-mediated brain injury, as well as altered cerebral blood flow autoregulation [29]. We also found that number of years of education is correlated with better neurocognitive function, and infer that a higher level of education enables patients to recover from this type of brain injury.
Emboli during on-pump CABG can arise from a variety of sources, for example, from aortic cannulation, through application and removal of aortic clamps releasing debris from atherosclerotic plaques, from the bypass circuit, and during the rewarming phase [9, 30]. Previous work has shown that emboli play an important role in determining postoperative neurocognitive function after CABG [6, 31]. It has been suggested that most emboli during on-pump surgery are gaseous in nature and are unlikely to be as harmful as particulate emboli. Although identification of the type of emboli (solid or gaseous) would be very useful, a recent validation study has shown that it is not yet possible to make such a distinction reliably with the current generation of TCD machines [32]. In our study, owing to the lack of an overlap in the number of emboli between the two treatment groups, we were unable to disentangle the effect of emboli from that of CPB on the predischarge neurocognitive score. We therefore cannot conclude that the difference in predischarge neurocognitive function is due to emboli.
Factors other than emboli could also account for the immediate protective benefit of off-pump surgery on cerebral function. Cardiopulmonary bypass is associated with a greater degree of systemic inflammation, which could have a profound effect on cerebral function [33]. Furthermore, the nature of cerebral blood flow is different between the two groups. Perfusion is nonpulsatile during CPB, and that may account for the appearance of diffuse brain edema, which is not seen after off-pump CABG, in which perfusion is pulsatile and therefore physiologic [34].
We have demonstrated that there is little difference in postoperative neurocognitive function at 6 weeks and 6 months between patients undergoing on-pump or off-pump CABG, and that the number of cerebral emboli during surgery only influences neurocognitive function in the immediate postoperative period. It is possible that the reason we found no persistent difference in neurocognitive function between the two groups is that the various exclusion criteria resulted in a low-risk group of patients being studied. A more pronounced advantage for off-pump surgery might have been obtained if the study had been limited to older patients with comorbidity [35].
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y.-H. Liu, D.-X. Wang, L.-H. Li, X.-M. Wu, G.-J. Shan, Y. Su, J. Li, Q.-J. Yu, C.-X. Shi, Y.-N. Huang, et al. The Effects of Cardiopulmonary Bypass on the Number of Cerebral Microemboli and the Incidence of Cognitive Dysfunction After Coronary Artery Bypass Graft Surgery Anesth. Analg., October 1, 2009; 109(4): 1013 - 1022. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. B. Halbersma, S. C. Arrigoni, G. Mecozzi, J. G. Grandjean, A. P. Kappetein, J. van der Palen, F. Zijlstra, and M. A. Mariani Four-year outcome of OPCAB no-touch with total arterial Y-graft: making the best treatment a daily practice. Ann. Thorac. Surg., September 1, 2009; 88(3): 796 - 801. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Koster, F. G.J. Oosterveld, A. G. Hensens, A. Wijma, and J. van der Palen Reply. Ann. Thorac. Surg., July 1, 2009; 88(1): 350 - 351. [Full Text] [PDF]
|
|
|
|
 |
|
 C. H. Moller, L. Penninga, J. Wetterslev, D. A. Steinbruchel, and C. Gluud Clinical outcomes in randomized trials of off- vs. on-pump coronary artery bypass surgery: systematic review with meta-analyses and trial sequential analyses Eur. Heart J., November 1, 2008; 29(21): 2601 - 2616. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. F. Marasco, L. N. Sharwood, and M. J. Abramson No improvement in neurocognitive outcomes after off-pump versus on-pump coronary revascularisation: a meta-analysis Eur. J. Cardiothorac. Surg., June 1, 2008; 33(6): 961 - 970. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Takagi, N. Kawai, and T. Umemoto Reply to the editor. J. Thorac. Cardiovasc. Surg., June 1, 2008; 135(6): 1401 - 1401. [Full Text] [PDF]
|
|
|
|
|
|
R. Motallebzadeh and M. Jahangiri Meta-analysis of randomized controlled trials of cognitive decline after on-pump versus off-pump coronary artery bypass graft surgery. J. Thorac. Cardiovasc. Surg., June 1, 2008; 135(6): 1400 - 1401. [Full Text] [PDF]
|
|
|
|
|
|
G. Djaiani, M. Ali, M. A. Borger, A. Woo, J. Carroll, C. Feindel, L. Fedorko, J. Karski, and H. Rakowski Epiaortic Scanning Modifies Planned Intraoperative Surgical Management But Not Cerebral Embolic Load During Coronary Artery Bypass Surgery Anesth. Analg., June 1, 2008; 106(6): 1611 - 1618. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Lynch and J. Riley Microemboli detection on extracorporeal bypass circuitsa Perfusion, January 1, 2008; 23(1): 23 - 32. [Abstract] [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 |
