Abstract
Objective: Neurocognitive dysfunction is a common complication after cardiac surgery with cardiopulmonary bypass (CPB). Studies using magnetic resonance imaging (MRI) have demonstrated that new focal brain lesions can occur after coronary artery bypass grafting (CABG), even in patients without apparent neurological deficits. Diffusion-weighted MRI is superior to conventional MRI and allows for sensitive and early detection of ischemic brain lesions. We prospectively investigated cerebral injury early and 3 months after CABG using diffusion-weighted MRI and related the findings to clinical data and neurocognitive functions. Methods: Twenty-nine patients [67.6±8.6 (52–85) years, 5 females] undergoing elective CABG with CPB were examined before surgery, at discharge and 3 months after surgery. A battery of standardized neuropsychological tests and questionnaires on depression and mood were administered. Conventional and diffusion-weighted MRI of the brain was performed and new lesions were analyzed. Clinical characteristics, neuropsychological test performance and radiographic data were collected and compared. Results: There was no major neurological complication after CABG. Thirteen patients (45%) exhibited 32 new ischemic lesions on postoperative diffusion-weighted MRI. The lesions were small, rounded and equally dispersed in both hemispheres. Eight patients had at least two lesions. At discharge, significant deterioration of neuropsychological performance was observed in 6 of the 13 tests compared to baseline assessment. By 3 months postoperatively, 5 of the 6 tests returned to preoperative levels. Verbal learning ability, however, remained impaired. The presence of new focal brain lesions was not associated with impaired neuropsychological performance. There was also no correlation between clinical variables, intraoperative parameters and postoperative complications and MRI findings. Conclusions: Although neurocognitive decline after CABG is mostly transient, memory impairment can persist for months. New ischemic brain lesions on postoperative diffusion-weighted MRI do not appear to account for the persistent neurocognitive decline.
1 Introduction
Coronary artery bypass grafting (CABG) is the most common major surgical procedure performed worldwide [1]. Advances in surgical techniques, anesthesia management, perfusion methodology and postoperative intensive care treatment have markedly decreased mortality and cardiac morbidity over the past 2 decades. Nevertheless, it is well recognized that cardiac surgery implies a substantial risk for central nervous system complications. Severe focal neurological defects such as strokes and transient ischemic attacks occur in 0.4–5.7% of patients, the incidence of postoperative encephalopathy is reported to be 10–28% [1]. Neuropsychological deficits are considerably more frequent sequelae of cardiac surgery and are observed in up to 80% of patients early postoperatively [2]. These deficits often resolve by 6 weeks to 6 months after surgery where the incidence is 20–50 and 10–30%, respectively. It has been reported that a substantial rate of neurocognitive decline is seen even at 1–5 years postoperatively [3]. In the latter subgroup of patients, however, the cause of cognitive deficits is subject to much debate, and it may well be that the deficits late after CABG are due to the effect of time rather than the influence of operation.
The pathogenesis of cognitive dysfunction after cardiac surgery is still uncertain. Variables that have been postulated to explain the development of postoperative neurocognitive decline include advanced age, concomitant cerebrovascular disease and severity of cardiovascular disease, and intraoperative factors such as embolization, cerebral hypoperfusion or hypoxia, activation of inflammatory processes, aortic cross-clamp (ACC) or cardiopulmonary bypass (CPB) time, and possibly low mean arterial pressure in patients with impaired cerebrovascular autoregulation [4]. There is increasing evidence that multiple microemboli arising from the ascending aorta, the heart chambers or the bypass circuit are the primary pathophysiological mechanisms producing diffuse ischemic cerebral injury [5–8]. Several methods have been used as potential indicators of cerebral injury associated with cardiac operations including comprehensive neurological and neuropsychological examinations, transcranial Doppler ultrasonography, quantitative electroencephalography, P 300 auditory evoked potentials, near infrared spectroscopy, serum studies of S100 protein and neuron-specific enolase, and magnetic resonance techniques. Using conventional magnetic resonance imaging (MRI), some investigators have observed new focal brain lesions after CABG in up to one-third of patients without overt neurological complications [9–11]. Advanced magnetic resonance techniques such as diffusion-weighted imaging (DWI) even allow for sensitive and early detection of ischemic damage within minutes after onset [12] and are therefore superior to conventional imaging. Diffusion-weighted MRI in conjunction with neuropsychological assessment has been scarcely employed in the study of patients after coronary re-vascularization. Also, most studies using diffusion-weighted sequences of MRI lack in a follow-up for more than 10 days or were performed on smaller groups. Finally, the clinical significance of new focal brain lesions for the development of neurocognitive deficits after heart surgery is still uncertain [10,13–15]. The purpose of the present prospective study was to determine the incidence of brain damage in patients undergoing elective CABG using conventional and diffusion-weighted MRI and neuropsychological testing before surgery, at the time of hospital discharge and 3 months after surgery. The relation between radiological data and clinical and surgical variables and neuropsychological test performance was analyzed.
2 Materials and methods
2.1 Patients and study protocol
A total of 35 patients with coronary heart disease undergoing elective CABG gave their informed consent and were enrolled in the study. Patients who had a history of hemodynamically relevant cerebrovascular disease, psychiatric illness, previous cardiac operation or concomitant valvular disease, renal insufficiency (i.e. serum creatinine >2.0 mg/dl), active liver disease, alcohol abuse, who were unable to read and speak the German language, who had contraindications for MRI (e.g. pacemaker, shrapnels) or who were supposed to be not compliant were excluded. All patients were non-emergent and awaited operation on the regular surgery ward. Patients were studied at three points in time: before operation (base line, examination 1 [E1]), early before discharge (E2), and 3 months after operation (E3). A physical examination, neurological examination, neuropsychological testing and a MRI scan of the brain were performed preoperatively and postoperatively in all 29 patients who completed the 3-month follow-up investigation.
2.2 Neuropsychological assessment
Cognitive brain function was assessed using a well-validated battery of 13 standardized psychometric tests including two questionnaires. All tests were performed by the same investigator who was an experienced neuropsychologist. To minimize learning effects the tests existed in parallel forms and were randomly assigned to the respective examination. The testing comprised the assessment of performance in five major cognitive domains known to be vulnerable to organic injury: attention, rate of information processing, memory and verbal learning, ability of logical thinking, and visual-spatial perception. The affective status of the patients was assessed with the use of standardized questionnaires on depression and mood. To measure the patient's attention and psychomotor speed Reitan's Trail making test version B (TMB) and Zimmermann's divided attention test were administered. Reitan's Trail making test version A was used to measure the speed of information processing. The digit span subtest of the Wechsler Memory Scale—Revised was further used to test short-term (digit span forward) and working memory for verbal material (digit span backward). The verbal learning test was used to test the ability of learning verbal material and to recall it immediately (word list immediate) and after a 30-min delay (word list delayed). The Corsi block tapping test investigates the memory of spatial structures and exists in two forms (Corsi block tapping forward and backward). The Horn's performance test 55+ subtest 9 tests for visual-spatial abilities and the subtest 3 tests for recognition of categories and regularities of geometrical objects. Except for the TMA, TMB and the divided attention test, a higher score indicates a better function. In addition to the psychometric tests, the patients were asked to fill in questionnaires on mood (von Zerssen's Bf-S', 28 items) and depression (general depression scale—short version, 15 items, modified from hospital anxiety and depression scale-short version). In both questionnaires a lower score indicates a lower level of depression and discontent.
2.3 Clinical evaluation
Clinical examination was carried out by an experienced neurologist before surgery, early before discharge, on the day of 3-month follow-up evaluation and in between, when necessary. The examination consisted of a medical history, physical examination and detailed neurological examination. By 3-month evaluation, neurological examination, neuropsychological testing and brain MRI were obtained on the same day of visit.
2.4 Magnetic resonance imaging of the brain
MRI scans of the brain were obtained by means of a standard 1,5 Tesla whole body system (Magnetom, Siemens AG, Erlangen, Germany). The MR-protocol included the following sequences: (a) transaxial T2-weighted (time to repeat [TR], 5120 ms, echo time [TE], 104 ms, average 1, matrix 256×256); (b) coronal T2-weighted (TR 4810 ms, TE 113 ms, averages 2, matrix 512×512); (c) transaxial fluid-attenuated inversion recovery [FLAIR]-weighted (TR 9000 ms, TE 115 ms, average 1, matrix 256×256); (d) conventional transaxial T1-weighted (TR 500 ms, TE 14 ms, averages 2, matrix 256×256); and (e) transaxial diffusion-weighted sequence (TR 4600 ms, TE 137 ms, averages 2, matrix 128×128, b-values 0, 500 and 1000). Field of view was 230 mm for all sequences. Slice thickness was 6 mm for turbo–spin–echo sequences and 5 mm for diffusion-weighted sequence. Images were evaluated by consensus of two experienced neuroradiologists using standardized evaluation forms. The presence of previous brain abnormalities, localization and volume of new lesions were recorded. On looking for the presence of any new brain lesion on postoperative MRI scans, the investigators were blind to the patient's individual postoperative course and the baseline status. For volumetry, the images were magnified fourfold, the area of lesion was delineated manually in each image slice and the volume was calculated using standard scanner software.
2.5 Anesthesia and surgery
CABG was performed with standard anesthesia techniques and surgical procedures. Ethmidate, sufentanyl and rocuronium were administered intravenously as needed to induce general anesthesia and isofluran to maintain it. All patients underwent a median sternotomy and CPB in mild hypothermia (32 °C). CPB was established by aortic and single venous cannulation. Aortic side clamping was performed for fashioning proximal anastomoses. Cardiac arrest was achieved by applying cold Bretschneider cardioplegia solution through a cardioplegia cannula. Arterial partial pressure of carbon dioxide (paCO2) was maintained at 35–40 mmHg and the arterial tension of oxygen was maintained at 200–250 mmHg during CPB. Non-pulsatile blood flow during CPB was maintained at 2.4 l/minm2 body surface area at normothermia and adjusted at hypothermia. Mean arterial pressure was kept higher than 50 mmHg throughout the operation with vasoactive agents when necessary. Packed red cells were administered when necessary to keep the hematocrit above 21% during CPB. The bypass circuit consisted of a roller pump (Stöckert SIII, Stöckert GmbH, Munich, Germany) and a 40-μm arterial blood filter. Heparin was added to the heart-lung machine to achieve an activated clotting time above 400 s during CPB. The acid–base status was managed using the α-stat protocol.
2.6 Statistical analysis
All analyses were performed using SPSS software for windows (SPSS 11.0, SPSS Inc., Chicago, IL, USA). A P-value below 0.05 was considered to be statistically significant. All variables were tested for normal distribution with the Kolmogorov Smirnov-1 sample test. Differences of neuropsychological variables between the three times of testing were analyzed with repeated measures univariate analysis of variance and the Friedman test, respectively. For comparison of two group means, Student's t-test and Wilcoxon test were used, respectively. The relation between demographic and clinical variables (e.g. preoperative stroke risk factors, operative parameters and postoperative adverse events) and cognitive function (as determined by the difference between the postoperative and preoperative neuropsychological test score) or MRI findings was analyzed using repeated measures multiple linear regression analysis. In order to analyze whether cognitive dysfunction at the 3-month follow-up examination was related to new ischemic MRI lesions, a two-step analysis was performed: (1) the unpaired Student's t-test was calculated between patients with and without new MRI lesions for the respective neuropsychological variable, (2) in case of a significant difference between the two subgroups, a regression analysis was administered.
3 Results
Thirty-five patients underwent CABG and participated in the prospective study. Six patients dropped out of the study postoperatively because of the following reasons: one patient died within 24 h after surgery, one had CABG without CPB, one sustained a major middle cerebral artery infarction on postoperative day 3 with hemiparesis and aphasia, two refused to comply at the 3-month follow-up investigation and one had postoperative renal failure and pneumonia requiring long-term ventilation. Finally, the study population consisted of 29 patients who completed all preoperative and postoperative examinations including the 3-month follow-up study. There were 5 women and 24 men with a mean age of 67.6±8.7 years (range 52–85 years).
The preoperative neuropsychological testing was performed 3.0±2.0 days (1–7 days) before CABG with MRI scanning obtained within 24 h. The first postoperative neuropsychological testing was performed 5.0±2.0 days (2–20 days) after CABG and the 3-month follow-up study was done 96.8±9.9 days (86–122 days) after operation. Twenty-two of the 29 patients (76%) had at least three risk factors for atherosclerosis. The most common risk factors were hypertension (76%), hyperlipidemia (76%) and diabetes (35%). Duration of extracorporeal circulation was 96.1±27.9 min (38–160 min) to achieve on average 3.1±0.8 grafts (1–5 grafts). Lacunes were seen on preoperative brain MRI in 11 of 29 patients (38%). Table 1 shows the most relevant demographic, clinical and perioperative data of the study group. There was no statistical difference between the two subgroups of patients with and without new lesions on postoperative MRI.
Clinical characteristics and perioperative variables of patients with and without new lesions in brain MRI after CABG
There was no evidence of major focal neurological abnormalities on serial examinations following CABG in the 29 patients. Transient psychotic symptoms requiring medication with haloperidol or benzodiazepines were observed in five patients (17%) and always restored within 1–3 days. These patients underwent neurocognitive testing at the time of full recovery from the episodic symptoms as considered by a neurologist.
Diffusion-weighted MRI of the brain detected new lesions on postoperative scans in 13 of the 29 patients (45%). Five patients had 1 new lesion, 7 patients had 2–4 lesions, and 1 patient had 7 lesions. There were 32 new focal lesions, 18 in the left hemisphere and 14 in the right. The lesions were rounded and their volume was small ranging from 32.5 to 745.5 mm3. Most of the lesions were supratentorial with 10 lesions in the frontal lobe, 9 in the parietal, 7 in the occipital and 1 lesion in the temporal lobe. Three of the four infratentorial lesions were in the cerebellum. The greatest single new lesion was found in a 78-year-old patient without overt neurological abnormalities on serial postoperative examinations (Fig. 1) . Preoperative Doppler sonography showed plaques in the carotid arteries on both sides but no stenosis, and preoperative brain MRI was normal. The patient was discharged from hospital on postoperative day 4 after an uneventful course. In all five patients with more than three lesions, the abnormalities were dispersed in different cerebral artery territories.
Diffusion-weighted MRI of a 78-year-old man with an uneventful postoperative course after CABG. No brain abnormalities were seen on the preoperative scan. Postoperative MRI disclosed a focal discrete area of increased signal in the left lower cerebellar hemisphere. The lesion was interpreted as indicative of a small ischemic brain injury secondary to embolization.
Table 2 summarizes the scores for the neuropsychological tests before and after CABG. Neurocognitive assessment early after surgery disclosed a significant decline in performance in 4 of the 11 psychometric tests compared to the preoperative status (repeated measures univariate analysis of variance, P<0.05). Fig. 2 demonstrates the performance in attention for each patient before and after surgery as measured by the Trail making test B. As compared to baseline performance most of the patients exhibited a deterioration in attention early postoperatively (Fig. 2a). By 3 months postoperatively, the function of attention had recovered (Fig. 2b). This is expressed in the mean TMB scores which increased from 114.6±36.8 s at E1 to 145.2±79.6 s at E2 (P=0.021, Table 2) (indicating a reduction in psychomotor speed), and then dropped to baseline values at E3 (P=0.066). Although the differences of the scores for the divided attention test and the digit span backward subtest were not significant between E1 and E2, they tended to indicate impaired function (divided attention, P=0.052; digit span backward, P=0.070, Table 2). The scores in depression and mood were also increased early postoperatively (Table 2) pointing out that the affective condition of the patients was markedly impaired few days after CABG. At the 3-month follow-up examination, 3 of the 4 tests with cognitive decline early postoperatively and both depression and mood scores returned to their baseline values. However, the test score for verbal learning ability decreased from 59.2±10.3 at E1 to 54.3±10.5 at E2 (P=0.013) and further dropped to 52.7±10.4 at E3 (P=0.012, E1 vs E3) (Table 2). The individual performance in verbal learning ability is shown in Fig. 3a and b .
Neuropsychological test scores at baseline (E1), at discharge (E2) and 3 months (E3) after CABG
Trail making test B measuring attention and psychomotor speed. Arrows indicate the difference in individual test score between the different testing points. The score at baseline examination was set to zero. Downward arrows mean a decrease in function and vice versa. (a) Comparison between examination before surgery (E1) and at discharge (E2), and (b) between E1 and examination at 3-month follow-up (E3).
Word list test measuring individual verbal learning ability. Arrows indicate the difference in individual test score between the different testing points. Downward arrows mean a decrease in function and vice versa. (a) Comparison between E1 and E2, and (b) between E1 and E3.
There was no statistical correlation between the presence of new brain lesions on MRI and persistent verbal learning deficit at the 3-month follow-up examination (Fig. 4) . The mean test score difference for verbal learning between E1 and E3 was not significantly different between the 13 patients with new lesions and the 16 patients without new lesions [score difference: −8.38±11.72 for 13 patients with new lesions (Fig. 4a) vs. −4.93±12.40 for 16 patients without new lesions (Fig. 4b), P=0.45]. None of the clinical variables (e.g. age, hypertension, preoperative brain abnormalities), surgical parameters (e.g. intraoperative minimal Hb concentration, ACC time, CPB time, minimal pO2) or postoperative major adverse events (e.g. atrial fibrillation, hypotension, hypertension) were predictive for the perioperative change of neuropsychological variables. The CPB time was slightly longer in patients with new lesions compared to those without lesions (106±25.4 vs. 87.3±27.3, P=0.055). There was also no association between clinical factors and new postoperative MRI lesions (multivariate analysis using a multiple linear regression analysis, Table 1).
Individual performance in the word list test in patients with and without new lesions on postoperative MRI. Arrows indicate the difference in individual test score between the different testing points. Downward arrows mean a decrease in function and vice versa. (a) Patients without lesions, and (b) patients with lesions.
4 Discussion
In the present prospective study on 29 patients undergoing elective CABG, neurocognitive function was markedly impaired postoperatively. The deficits pertained to major cognitive domains such as attention, memory and executive functions, but usually resolved within weeks. However, verbal learning ability remained deteriorated even after 3 months. New small brain lesions were found in 45% of patients on postoperative diffusion-weighted MRI, but this did not cause clinically overt neurological deficits. There was no significant correlation between the presence of new brain lesions and the persistent neurocognitive dysfunction 3 months after operation.
Advances in modern cardiac surgery over the past 2 decades have resulted in a relatively low mortality and cardiac morbidity. Nevertheless, cardiac surgery may be associated with adverse neurological and neuropsychological outcome. Clinically recognized cerebral infarction occurs in 0.8–6% of patients [16]. Moderate complications such as obvious deterioration in intellectual function and seizures have been detected in an additional 3% of cases. Postoperative encephalopathy or delirium occurs in 10–28% of patients and is often transient [1,2]. More subtle postoperative changes in cognitive function as detected by a decline in neuropsychological testing scores from baseline are considerably more common. The incidence of severe cognitive decline and behavioral change related to open heart surgery is highest at the time of discharge ranging from 50 to 80% [3]. In the following weeks, many symptoms restore, but even 6 months after operation neurocognitive decline is found in approximately 10–30% of patients [3]. In some patients, longer-term cognitive dysfunction is reported to be present 1–5 years after operation [1,3]. The relatively high prevalence and persistence of postoperative cognitive decline is of particular importance since it may result in decrease in quality of life and work performance [2]. The frequency of postoperative neuropsychological impairment and the pattern of early decline and late recovery in this study is in line with numerous studies [1,3,17]. Significant cognitive decline had occurred early postoperatively in three cognitive domains. By 3 months, most of these deficits had recovered to baseline in most patients, but verbal learning ability was still severely deteriorated. Longitudinal assessment of cognitive function needs to be awaited to determine the significance of this impairment on long-term outcome of patients. The cause of postoperative cognitive dysfunction is yet unknown, and particularly cognitive decline years after operation is subject to much debate. It may represent the effect of age in patients with underlying neurological disease, or it may be the result of subtle brain injury at the time of coronary surgery [15]. Further long-term studies with patients undergoing CABG compared to patients with coronary disease without CABG as a control group (treated either medically or by percutaneous coronary intervention) are required to provide answers to these questions.
Many different methods have been used as potential sensitive indicators of cerebral injury associated with heart surgery. Because of its superiority to other imaging methods, MRI has been increasingly employed in the study protocols before and after open heart surgery. Using T2 MRI sequences, Toner et al. (1994) found new cerebral lesions in 30% (6/20) of patients undergoing CABG, mainly deep white matter lesions. In a larger series of Vanninen et al. (1998), postoperative T2 MRI brain scans revealed new focal lesions in 21% (8/38) of patients. No patient experienced major neurological complications and there was no significant deterioration in mean cognitive performance 3 months after surgery. For the detection of ischemia, new MRI technology with diffusion-weighted sequences is more sensitive than T2-weighted images [18], but studies using DWI are yet scanty. Bendszus et al. (2002) prospectively evaluated brain damage before and after CABG using diffusion-weighted MRI. They found 17 new ischemic DWI lesions in 26% of patients. Similarly, Restrepo et al. (2002) demonstrated that focal brain diffusion abnormalities can occur after CABG, even in patients without clinically identified neurological deficit. The correlation of the presence of new MRI lesions with postoperative neurocognitive decline is variable in the few studies that have assessed cognition. Some investigations report that the new MRI lesions were associated with neurocognitive decline [10,14,15]. In the series of Restrepo et al. (2002), from 13 patients studied, all four patients with postoperative DWI defects had a larger neurocognitive decline than their counterparts with stable MRI. Other clinical characteristics were similar between patients with and without new DWI lesions, including stroke risk factors [15]. In contrast, other studies, including our own, failed to find an association between the presence of new postoperative lesions and neuropsychological decline [11,13]. In the present study, this was even true when cognitive decline was persistently impaired 3 months after operation. The occurrence of new focal brain lesions in our series is more frequent than in others [11,13], but less than in a valve replacement surgery study where it was 58% [19]. Despite a good interobserver agreement for the detection of new diffusion lesions on postoperative scans, we cannot preclude an overestimation of subtle areas of hyperintensity particularly in patients with preexisting cerebral vasculopathy. Since the new lesions are also visible on T2-weighted images, there is evidence of structural brain damage. The discrepancy between the presence of new MRI lesions and the absence of clinically overt neurological deficits prompts the proposal that the severity of brain damage is limited. It may be hypothesized, however, that new postoperative lesions result in subtle brain injury that will leave patients more susceptible to cognitive decline from degenerative or atherosclerotic disease years later. Larger series are required to determine the significance of postoperative brain abnormalities with regard to neurological and neuropsychological long-term outcome of patients.
The pathogenesis of new ischemic brain injury after CABG is unknown. Possible underlying mechanisms are embolization and hypoperfusion [5,8,20–22]. From the pattern of distribution, most of the MRI lesions in the present study are considered embolic even though hypoperfusion-induced border zone edema and small infarcts may not be precluded in a few cases. Emboli are made of gas, thrombi, fragments of atheromatous plaques from the ascending aorta, cell aggregates, fat or inorganic debris from the bypass circuit tubing [6]. The sensitivity of diffusion-weighted MRI is limited by the spatial resolution. Lesions smaller than the pixel size are hardly to detect. Microbubble emboli would produce diffuse cerebral hypoxemia with cytotoxic edema resulting in a reduced attenuated diffusion coefficient on magnetic resonance spectroscopy [13]. As elevated diffusion coefficient values were found on brains of patients after CABG [13] and with respect to the absence of apparent neurological events, we assume that the lesions are probably due to thromboemboli or macroscopic air emboli rather than microbubbles or big plaque fragments, the latter being probably more harmful.
There are some limitations to our study. One limitation is the restricted number of patients, even though only two studies comprised more patients [11,13]. There is also a possible selection bias, because only patients who seemed compliant were enrolled. The follow-up of patients was 3 months, therefore the significance of the new postoperative MRI lesions on the neuropsychological long-term outcome of patients remains uncertain.
In summary, the present study is, to our knowledge, the first to systematically employ conventional and diffusion-weighted MRI before and after CABG and to follow-up the patients for a mean of 3 months after operation. We found no association between the appearance of new small brain lesions on postoperative MRI and short-term and mid-term neurocognitive outcome. We conclude that the new focal ischemic brain damage detected by MRI does not seem to account for the deterioration of neurocognitive function three months after coronary surgery. Further studies on larger series including control groups are required to corroborate these initial findings. These studies might also help to evaluate the influence of non-surgery related factors on the long-term course of cognitive function in patients with coronary artery disease.
The authors are grateful to Dr Matthias Thielmann and Thorsten Liesebach for technical assistance during the study.
Dr T. Aberg (Umea, Sweden): The brain component after any open heart surgery is the Achilles heel of open heart surgery. We have come a long way the last 20 years in reducing these injuries, and this work is really very important in further elucidating the mechanisms and also the extent of these injuries.
However, as almost always, there is a lack of controls, of what happens without surgery and what happens with the competing technology, the PCI. Are you planning to repeat the same very interesting investigation in patients who were treated without surgery, without the pump and with PCI?
Dr Knipp: We are actually planning to control these patients and to compare them with OPCAB patients without use of pump and even with patients who have major surgical procedures on the vascular system without the use of cardiopulmonary bypass and with no touch on the ascending aorta, which is certainly one of the main sources of embolization causing these MRI lesions.
Mr V. Zamvar (Edinburgh, UK): Did you find any difference in the MRI findings between immediate postoperative period and the three-month postoperative period? Did the lesions disappear?
Dr Knipp: Also early after surgery we didn't find any association between the appearance of new lesions and neurocognitive dysfunction.
Mr Zamvar: Sorry. My question is, what happened to the MRI lesions? Did the lesions disappear at the three-month scan?
Dr Knipp: I see. The lesions were seen on the diffusion-weighted sequences but also on T2-weighted images, indicating structural damage to the brain. Most of the lesions did persist.
Dr R. Frater (Bronxville, NY, USA): You mentioned that the cardiopulmonary bypass was standard, but there surely are a large number of possible variables in the way you conduct bypass which can influence at least cerebral flow and perhaps cerebral function. Can you be sure that the cognitive effects you are talking about are related to the way you conducted the bypass rather than to any embolic phenomena?
Dr Knipp: You mean that there was a difference in laboratory data between the two situations?
Dr Frater: No. What I mean is, if you do cardiopulmonary bypass by itself you are potentially affecting cerebral function through flow, through oxygenation, through carbon dioxide concentrations and so on. Now, you just say it is standard, but in fact you need to define that at the same time as you are looking for emboli, don't you think, in order to be able to separate embolic lesions versus other mechanisms for cerebral dysfunctions?
Dr Knipp: I didn't comment on the details of the cardiopulmonary bypass, but all of the patients had a median sternotomy, all of them were managed by alpha-stat protocol, all patients had a mean bypass flow of 2.4 liters per minute and square meter body surface area, and so on. So this is a standard protocol that we follow. That is why I said that the bypass itself was not changed in this prospective study.
