| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL ARTICLE |
Unit of Behavioural Medicine, UCL, London, UK
1 Cardiothoracic Surgery Unit, The Heart Hospital, UCLH NHS Foundation Trust, London, UK
Stanton P Newman, DPhil. Tel: +44 207 6799468, Fax: +44 207 6799426; Email: s.newman{at}ucl.ac.uk, Unit of Behavioural Medicine, UCL, Charles Bell House, 67-73 Riding House Street, London W1W 7EJ, United Kingdom.
ABSTRACT
This study compared the occurrence of intraoperative microemboli and postoperative changes in neuropsychological performance in 195 patients undergoing coronary artery bypass grafting who were randomized to intermittent crossclamp fibrillation or cardioplegic arrest. Cerebral microemboli were recorded from cannulation to 15 min after decannulation, using transcranial Doppler in 166 patients. Microemboli in relation to 9 surgical events were also noted. Neuropsychological change scores were obtained by comparing cognitive performance preoperatively with that at 6–8 weeks after surgery. The median number of microemboli detected was 105 (range, 9–1,757) in the fibrillation group, and 110 (range, 1–1,306) in the cardioplegia group, with no significant difference between groups. There was also no significant difference between groups in the generation of microemboli during any of the surgical events. Neuropsychological tests were completed postoperatively by 177 participants, with no significant differences in performance found between the 2 groups. Given the equivalence of the effect of intermittent crossclamp fibrillation and cardioplegic arrest on microemboli and neuropsychology, consideration of which form of myocardial protection to employ should perhaps focus more on which method affords most protection to the heart.
Key Words: Coronary Artery Bypass • Myocardial Protection • Neuropsychology • Microemboli
INTRODUCTION
Since the mid 1970s, cardioplegic arrest has been the most widely practiced technique for myocardial protection during coronary artery bypass grafting (CABG). Nevertheless, intermittent crossclamp fibrillation has continued to be preferred by a minority of surgeons for both high- and low-risk patients.1 Several studies have shown fibrillation to be as safe or safer than cardioplegic arrest, through preconditioning.2–6 Exponents of fibrillation claim that the technique is simpler, allows the myocardium to be reperfused during the procedure, and potentially shortens ischemic and cardiopulmonary bypass (CPB) times.2 However, fibrillation is associated with more handling and repeated clamping of the aorta. This may cause atheromatous plaque to release debris (microemboli) into the systemic circulation, which could find its way into the cerebral circulation.7 The surgical maneuvers most likely to cause this are aortic cannulation/decannulation, crossclamp application/ removal, and construction of proximal anastomoses.8,9 Microemboli in the cerebral circulation can be detected noninvasively by continuous Doppler ultrasonography of the carotid or middle cerebral artery. The number of microemboli determined by transcranial Doppler and also found in post-mortem studies have been associated with CPB time.10 Microemboli are thought to play a major role in neuropsychological changes after CABG.11 Neuropsychological decline occurs in up to 79% of patients within 5 days of surgery, 18.5%–40% at 6–8 weeks, and 7%–57% at 6 months.12 The relationship between neuropsychological performance and micro-emboli is not always clear-cut.9,11,13–15 This is not surprising given the size of the microemboli, the fact that techniques to discriminate between gaseous and particulate microemboli are only recently available, and such emboli may be lodged in an area not covered by the cognitive tests employed. Only one small study compared postoperative neuropsychological performance and microemboli in patients undergoing fibrillation or cardioplegic arrest during CABG, finding no significant differences between the 2 groups.6 However, with only 70 participants followed up at 6 months, the study was underpowered, and it excluded patients with aortic disease, which may have reduced the impact of repeated crossclamping. Also, flow rates and body temperatures differed between groups, leaving the possibility that other factors could have confounded the results. This randomized study set out to compare the generation of microemboli and postoperative neuropsychological changes in a large group undergoing either intermittent fibrillation or cardioplegic arrest for routine CABG.
PATIENTS AND METHODS
Prior to commencement of the study, ethics approval was obtained from UCL/UCLH ethics committees. Between February 2002 and November 2004, 200 patients scheduled for nonemergency CABG were prospectively randomized to intermittent crossclamp fibrillation or a single aortic crossclamp technique. Participants were recruited into the study on the day before surgery. Patients with prior transient ischemic attack or stroke were excluded, as were those with a history of psychiatric illness. Only patients who were fluent in English and had good (corrected or uncorrected) visual and auditory acuity were enrolled. Randomization was achieved with sealed envelopes given to the perfusion department. Participants were not informed of the type of myocardial protection they received until the end of the study. Surgeons selected to participate in this study were proficient in both techniques.
A battery of 9 neuropsychological tests was administered on 2 occasions by the same trained psychologist in the same quiet room. Assessments were usually performed on the day before surgery and 6–8 weeks postoperatively, immediately prior to the scheduled out-patient appointment. The tests assessed cognitive function under 3 broad-based categories of memory, attention and concentration, and perceptuomotor ability. These tests have been widely used to examine neuropsychological change following CABG, and comply with the Statement of Consensus.16,17 Where available, parallel forms were used to limit the effects of learning between assessments. To obtain an estimation of premorbid intelligence quotient, the National Adult Reading Test was administered preoperatively. It has been suggested that mood may impact on neuropsychological performance; therefore, the state subscale of the Spielberger State-Trait Anxiety Inventory and the Center for Epidemiological Studies Depression Scale were administered at each assessment. The assessor was blinded to the type of myocardial protection the participant was to receive. A brief neurological examination was conducted immediately prior to neuropsychological testing, 24 h postoperatively, at discharge, and at the 6–8-week follow-up.
Standardized anesthetic induction and maintenance was used in all patients. Following conventional central venous and arterial line insertion, intravenous induction was commenced using fentanyl 5–10 µg · kg–1, propofol 0.5–1.0 mg ·kg–1, or etomidate 0.05–0.1 mg · kg–1 and pancuronium 0.1 mg · kg–1. Patients were intubated and ventilated with isoflurane (0%–2%) in oxygen and air. Heparin 300 units · kg–1 was given to maintain an activated clotting time>400 sec, and while on CPB, patients were maintained on propofol infusion 1–2 mg · kg–1 ·h–1 along with Actrapid 0–5 units h–1 (adjusted to control blood sugar). Heparin was reversed with protamine 1–1.5 mg · per 100 units of heparin. Inotropics were used as necessary. The CPB circuit consisted of hollow-fiber membrane oxygenators (Quantum HF-6700; Bard Ltd.) and an arterial filtration line (3 µm, Prim-Vu; Bard Ltd.). CPB was established between single right atrial and ascending aortic cannulas. Nonpulsatile flow was maintained with a roller pump. Perfusion pressure was kept at 2.4 L min–1 m–2 and reduced to 75% when core cooling reached 32°C–34°C. Mean arterial pressure was maintained as close as possible to 50 mm Hg while on bypass, using vasoactive drugs. Acid-base regulation during CPB was carried out with the alpha-Stat protocol. In patients randomized to cardioplegia, myocardial protection was achieved using 1 L of antegrade cold blood cardioplegia (based on St. Thomas’ Hospital No. 1 solution). Further doses were infused at a rate of 150–200 mL min–1 at 30-min intervals. After distal anastomoses, the crossclamp was removed, proximal anastomoses were completed using a single side-biting clamp, and the vein grafts were carefully de-aired. In the fibrillation group, distal anastomoses were constructed during brief periods (10–15 min) of aortic crossclamping and ventricular fibrillation. When each anastomosis was completed, the aortic crossclamp was removed. If the heart failed to revert to sinus rhythm spontaneously, cardioversion was carried out with internal paddles (10–20 joules). Proximal anastomosis was completed using a partial side-biting clamp. The sequence was repeated for each graft. To ensure crossclamping did not occur in an area of mobile or thick atheroma, both epiaortic ultrasonography and transesophageal echocardiography were employed. A standard cardiac protocol was used in the intensive care unit to wean and extubate the patients.
Microemboli were monitored by transcranial Doppler (Nicolet EME Pioneer 4040; Uberlingen, Germany). The middle cerebral artery was assessed with a 2-MHz pulsed-wave transducer secured by a headband. All measurements were made off-line, using international consensus criteria, and a second observer reviewed 10% of the tapes to assess interrater reliability.18 Microembolic events were counted at 9 time points: aortic cannulation, initiation of CPB, aortic crossclamp applied, aorta declamped, construction of proximal anastomosis, side-biting clamp released, cessation of CPB, aortic decannulation, and 15 min after decannulation. Microemboli detected within 1 min of a maneuver were rated as being associated with that maneuver; however, for proximal anastomosis, microemboli were counted during the entire procedure.
Statistical analysis was performed with SPSS version 12.0.1 software for Windows (SPSS, Inc., Chicago, IL, USA). Univariate analyses were conducted using chi-squared analysis or Fisher’s exact test, where appropriate. To investigate group differences, independent t tests were performed on normally distributed continuous data, and the Mann-Whitney test on non-normally distributed data; to correct for any ties in ranking, the Z value was reported instead of the U. Due to the number of comparisons made, and to avoid the likelihood of making a type 1 error, a significance level of p = 0.01 was set. Early studies investigating neuropsychological decline after CABG rated the participants as either having or not having a deficit. This binary classification is insensitive and unable to take into account any improvements in performance, as it considers only deterioration. Therefore, to make optimal use of the neuropsychological data, each patient’s pre- and postoperative scores were compared to create a change score. Standardized Z change scores were calculated by subtracting the relevant postoperative test score from the preoperative performance, and dividing that score by the preoperative standard deviation for each individual test. This technique of analysis has been adopted in a number of studies on cognitive change in cardiac surgery.14,19 In both methods of analysis, where improved performance in any test was reflected by a lower score, as in timed tests, the direction of the score was reversed so that all improvements gave rise to a positive score, and deterioration to a negative score. A global Z change score was obtained by summing the change scores of individual tests. When a participant was unable to complete a task at follow-up because of cognitive difficulties, a score of 2 standard deviations below the mean for the whole group was recorded for that test. Sample size calculation was based on the Z change scores in a previous study by Arrowsmith and colleagues.14 The sample size was formulated on a 2-sided hypothesis with an alpha significance set at 0.05 and power set at 0.80. The sample size required was 74 per group (total sample, 148). It was estimated that a 25% loss on follow-up was possible; therefore, it was calculated that 200 participants were required to give an adequately powered study. The normality of the distribution of data was examined using the Kolmogorov-Smirnov goodness-of-fit test at each time point. Data found to be not normally distributed were subjected to either square-root or natural log transformations, depending upon distribution. However, it was not possible to render one of the neuropsychological tests normal (Choice Reaction Time Test), and caution must be used in interpreting the results of this test. The Kolmogorov-Smirnov test is considered to be a stringent assessment of normality. It was also not possible to render the microemboli count normal, and these data were analyzed non-parametrically.
RESULTS
Of the 200 patients recruited into the study, 5 (2.5%) were excluded before randomization. Of the 195 participants, 177 (91%) completed both pre- and postoperative neuropsychological assessments. Clinical outcomes and attendance at follow-up are given in Table 1
. Comparisons were made of patient characteristics, baseline neuropsychological performance, and baseline mood state between those attending the follow-up neuropsychological assessment and those not attending. The only significant difference was in the EuroSCORE (p = 0.007), with those not completing having a significantly a higher EuroSCORE (mean, 6.06) than those completing (mean, 4.49). There were no significant preoperative differences between the groups in demographic characteristics, premorbid intelligence quotient, neuropsychological performance, or mood state (Tables 2 and 3). There was no correlation between anxiety (r = 0.135, p = 0.122) or depression (r = 0.032, p = 0.717) and neuropsychological performance. Intraoperative data comparisons showed a significantly longer crossclamp time (p < 0.001) in patients who had cardioplegia, with no differences in operation time, CBP time, number of grafts, or internal mammary grafts. There was also no difference between groups in the number developing postoperative atrial fibrillation (Table 4). Epiaortic ultrasonography and transesophageal echocardiography resulted in the site of cannulation being changed in 18 patients to avoid atheroma. Recordings were obtained from the middle cerebral artery in 166 patients. The numbers of microemboli detected are given in Table 4 (interrater reliability, r = 0.91). In one case, 3,601 microemboli were detected (fibrillation group). Therefore, all analyses involving microemboli were conducted both with and without this outlier; no differences were found between these analyses. There was no significant difference in total microemboli count between groups, or in microemboli occurring at any stage of the operation, including periods of no manipulation (Table 5). Changes in each test of neuropsychological performance and total neuropsychological score changes are given in Table 6; no significant differences were found between groups. Although there was a significant decline in both anxiety and depression postoperatively, there was no difference between the 2 groups. There was no correlation between mood state (anxiety, r = –0.007, p = 0.930; depression, r = 0.022, p = 0.791) and neuropsychological change scores. Spearman’s rho correlations were conducted between total microemboli and neuropsychological change scores, revealing no correlation either as a whole group (rho = 0.159, p = 0.05) or separately by group (fibrillation: rho = 0.063, p = 0.581; cardioplegia: rho = 0.244, p = 0.036).
|
|
|
|
|
|
The generation of microemboli during manipulation of the aorta has often been cited as a cause for concern when considering intermittent crossclamp fibrillation as a technique for myocardial protection. There is evidence that in some interventions in cardiac surgery with CPB, microemboli (assessed by transcranial Doppler) can be reduced.20 The finding of no overall difference in microemboli between the 2 myocardial protection techniques in an adequately powered and randomized study with well-balanced groups is reassuring to those who practice either technique. In particular, it denies concerns regarding intermittent fibrillation, and the strong views that some hold about this technique. Our findings confirm those of the previous small study by Musumeci and colleagues.6 There are some important differences between these 2 studies: the participants in our study were older and none were excluded if they had aortic disease; also, body temperature and flow rate during CPB did not differ between the 2 groups. Both groups were comparable in terms of preoperative characteristics, but ischemic time was shorter in the fibrillation group. Shorter crossclamp times have frequently been found in patients undergoing intermittent fibrillation, compared to those receiving cardioplegia, but this has not been noted in all studies.3,4,7 It has often been claimed that CPB time is shorter with the fibrillation technique, making this better for the brain than cardioplegic arrest.4,6 No CPB time difference was found in this study, possibly because all surgeons were skilled in both techniques, resulting in short bypass times (mean < 74 min) in both groups.
A number of studies have demonstrated that the occurrence of microemboli can be linked to various aspects of surgery, particularly the onset of bypass, aortic cannulation, and removal of clamps.9 Given the difference in the 2 procedures, it is likely that the number of emboli generated at each maneuver may vary. In this study, although more microemboli were detected in the fibrillation group on removal of the crossclamp, this difference was not significant. However, during the other 8 manipulations, the number of microemboli was higher (albeit not significantly) in the cardioplegia group. As with previous studies, although showers of microemboli were detected at the time of the 9 surgical events, especially initiation of CPB and proximal anastomosis, the majority occurred spontaneously.9,15 In this study, the number of spontaneous microemboli constituted approximately 80% of those in each group, resulting in no difference in total microemboli between groups. These findings differ from those of Musumeci and colleagues6 who found that patients undergoing intermittent fibrillation produced more microemboli during surgical manipulation, whereas with cardioplegic arrest, the majority of emboli occurred spontaneously. This difference may be due to methodological factors regarding microemboli detection and the criteria for considering microemboli to be related to a surgical event. We rated microemboli as associated with a surgical event if they occurred within 1 min of a maneuver, a definition used previously;8,9,13 the time-frame used by Musumeci and colleagues6 was not stated. It is of note that no differences were found between groups in the occurrence of microemboli at each surgical event, despite increased manipulation of the aorta in the fibrillation group.
Both groups showed overall improvement in neuropsychological performance postoperatively. This improvement was also reported by Musumeci and colleagues6 who employed a different set of neuropsychological tests (Luria Nebraska Neuropsychological Battery). We chose neuropsychological tests that are sensitive to change and thus amenable to repeated administration. One measure of cerebral damage is the inability to learn a task or strategy, which is the underlying basis of practice effects. The analysis used in this study takes into account any learning or failure to learn postoperatively, and a general improvement would be expected. The tests were performed with a widely used battery, and produced no significant differences between groups in overall performance or any individual test. As the overall neuropsychological change score was the primary endpoint of the study, the findings suggest that the impact of these 2 procedures on brain function was equivalent, at least under the conditions and testing procedures performed here. It has been argued that mood prior to cardiac surgery, especially anxiety, may influence neuropsychological performance.6 However, the groups were well matched on mood, and more importantly, no association was seen between neuropsychological performance and anxiety and depression. Therefore, the lack of a difference in overall neuropsychological performance was not related to mood.
Although microembolization is considered to be an important mechanism in the etiology of neuropsychological changes after CABG, it is well accepted that this is a multifactorial problem and the changes are not attributable to microemboli alone.12 This study suggests that there is no difference in the generation of micro-emboli between 2 common methods of myocardial protection, and that these techniques do not result in different degrees of neuropsychological change. Given the equivalence of effect of cardioplegic arrest and fibrillation on microemboli and neuropsychology, consideration of which form of myocardial protection to employ should perhaps focus more on which method affords most protection to the heart.
ACKNOWLEDGMENTS
This study was made possible through a generous grant from the Reta Lila Weston Trust for Medical Research.
REFERENCES
Asian Cardiovasc Thorac Ann 2009;
17:259-265
© 2009 by SAGE Publications
DOI: 10.1177/0218492309104749
This article has been cited by other articles:
|
|
 |
|
  M. Ugurlucan, M. Basaran, T. N. Ogus, and O. Isik Fibrillation without Cross-Clamping in the Pesence of Sclerotic Ascending Aorta Asian Cardiovasc Thorac Ann, January 1, 2010; 18(1): 94 - 95. [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 |
