Microembolic Signals in Patients Undergoing Coronary Artery Bypass Grafting: Effect of Aortic Atherosclerosis

Abstract

The aim of this prospective study was to determine whether aortic atherosclerotic plaques are associated with increased frequency of microembolic signals and stroke in patients who undergo coronary artery bypass grafting. A total of 69 such patients were monitored by transcranial Doppler ultrasonography for 30 minutes before and after surgery. To our knowledge, this study is the 1st in which in vivo pathologic analysis of aortic plaques was systematically performed—and microembolic signals monitored—before and after open-heart surgery. Plaques were assessed by transesophageal echocardiography and by biopsy of materials taken during surgery. The frequency of microembolic signals was evaluated with regard to the occurrence of postoperative stroke. In the preoperative phase, only 10 of 48 patients with aortic plaques had microembolic signals, and the mean count of microembolic signals was 3.2 ± 1.2 per hour. At the conclusion of 24 postoperative hours, 29 patients (42%) displayed such signals (mean count, 9.8 ± 3.1/h). Seven of the 48 patients (15%) with aortic atherosclerosis had cerebral ischemic events, but none of those with normal aorta (21 patients) experienced stroke during the postoperative phase. During postoperative monitoring, patients with stroke had higher microembolic-signal counts than did those with normal aorta (17.4 ± 3.3/h vs 5.9 ± 3.1/h; P <0.05). Our findings suggest that microembolic signals can be a marker of severe aortic atherosclerosis and that monitoring these signals should enable the application of appropriate surgical methods to coronary artery bypass patients who are at higher risk of stroke.

The aortic arch is known to have a predilection for atherosclerotic plaque. Further, it is well known that these aortic plaques can be a source of cerebral emboli either spontaneously or during invasive procedures such as arteriography or cardiac surgery. ^1–7

In recent years, pathologic and transesophageal echocardiographic (TEE) studies have shown that atherosclerotic disease of the aortic arch is an independent risk factor for ischemic stroke and carries a high risk of recurrent vascular events. 1,2,8–10 It might account for brain infarcts of unknown cause, which have no carotid or cardiac sources of emboli.

Many studies that have used transcranial Doppler (TCD) ultrasonography to examine the occurrence of microembolic signals (MES) distal to the atheroscle-rotic lesions suggest that clinically silent microembolism correlates with recent symptomatic cerebral ischemia. ^5,11,12 In this study, we postulate that MES may demonstrate indirectly the embolization of aortic plaque and thereby predict postoperative cerebral ischemic events, in patients who have undergone cardiovascular surgery.

We prospectively studied patients who did or did not have aortic plaques (as detected by perioperative TCD monitoring and by postoperative pathologic examination), in order to assess the emboligenic potential of these aortic lesions.

Patients and Methods

Patients

A total of 69 recruited patients who underwent coronary artery bypass grafting (CABG) in our institution were entered into this open, prospective study, after they gave informed consent. All patients were monitored for 30 minutes by TCD 24 hours before and again 24 hours after the surgical procedure, to compare the perioperative stroke group with the non-stroke group. We delayed TCD for 24 hours after CABG to avoid recording microembolic signals due to intraoperative manipulation such as aortic cannulation and decannulation, insertion and removal of the cardioplegia cannula, and cross-clamping of the aorta.

All patients underwent detailed preoperative neurologic examinations and serial, detailed clinical assessments during the first 30 postoperative days. In all patients, we performed noninvasive vascular investigations (Doppler and duplex ultrasonography, as well as TEE) to identify and evaluate atherosclerotic lesions in the aorta and in the extra- and intracranial carotid system. Atherosclerotic plaques were defined by TEE or by biopsy material taken during surgery as grade I (1 or more plaques with an irregular surface and having a thickness of <1.5 mm); grade II (fibrous plates with a laminated or layered deposition along the involved intimal surface and having a thickness of 1.5 to 3.9 mm); or grade III (advanced calcified lesions with ulceration and mobile lesions suggesting débris or free-floating thrombus, and having a thickness of ≥4 mm). Ulcerated plaque in the aortic arch was defined as 1 or more ulcerations with disruption of the intimal surface, located in the ascending or the horizontal aorta between the leaflets of the aortic valve and the ostium of the left subclavian artery. The plaques in the ascending aorta and the arch were considered as a potential emboligenic source for the brain.

Data were recorded on the age and sex of the patients and on the following risk factors: hypertension, diabetes mellitus, hypercholesterolemia, cigarette smoking, peripheral arterial disease, and cardiac arrhythmia. The control group consisted of 60 age- and sex-matched volunteers without neurologic abnormality. The carotid arteries and aorta were normal in all control subjects, as determined by duplex and TEE scanning.

Surgical Procedure

All patients who underwent standard CABG under general anesthesia were monitored continuously with the Life Scan Brain Activity Monitor (Diatek Patient Management System, Inc.; San Diego, Calif). Naso-pharyngeal and rectal temperatures were measured during the operation. Patients received a dose of bovine heparin (approximately 300 IU/kg) before cannulation, and this was reversed by protamine sulfate. During surgery, the aortic arch was palpated to locate atherosclerotic plaques. The plaques were then measured and biopsy material was taken to determine their composition. In patients who underwent CABG, the proximal anastomoses were performed after release of the aortic cross-clamp, with the use of a side-clamp. Postoperatively, patient management included arterial pressure monitoring and mechanical ventilation overnight, with extubation on the 1st postoperative day. Postoperative ischemic stroke was defined as neurologic deficit indicated by such focal neurologic signs as limb paralysis, aphasia, and visuospatial neglect (right hemisphere syndrome), after an initial normal neurologic evaluation in the recovery room. Computed tomography was performed in cases of postoperative stroke to determine the nature of the lesion. Anticoagulation with low-molecular-weight heparin was started in all patients after surgery. Neurologic complications due to cerebral embolization during the first 30 postoperative days were treated jointly by a surgeon and a neurologist.

Monitoring of Embolic Signals

Bilateral TCD monitoring was performed over the middle cerebral arteries for 30 minutes per patient or control subject, using 2-MHz probes of a pulsed Doppler machine (Multi-Dop X-4; DWL Elek-tronische Systeme GmbHz; Sipplinger, Germany), with multirange embolus detection software (TCD-8 for MDX, version 8.00K; DWL Elektronische Systeme GmbH) at a depth of 50 to 55 mm, using 160 mW/cm² of power, a sample volume of 6 mm, a sweeping speed of 6 sec, and a 128-point FFT (fast Fournier transformer). Two examiners, who were present throughout the study, monitored the middle cerebral arteries online and evaluated the results.

The criteria for the detection of microemboli were those of the International Consensus Group on Microembolus Detection: ¹³ 1) characteristic acoustic properties; 2) short duration (<0.3 msec); 3) random appearance in the cardiac cycle; 4) unidirectional signal; and 5) an intensity increase at least 9 dB above background noise. Differentiation of embolic signals from artifacts was also performed in accordance with International Consensus Group criteria. Signals were assumed to be artifacts if they registered simultaneously left and right, or above and below the baseline. Data were stored on the hard disk for subsequent analysis. The examiners were blinded to the status of all subjects and worked independently of one another in evaluating MES counts in the middle cerebral arteries. They were asked to note both the exact position of each MES and the total MES counts. Each MES count was repeated by a 2nd observer to assure interobserver agreement at the same point in time. Only signals recognized as MES by both observers were accepted. The numbers of MES were presented as the total median numbers of MES within a range, during preoperative and postoperative phases.

Pathologic Analysis

During the operation, samples for biopsy were drawn from the ascending and aortic arch, and these were analyzed postoperatively in the pathology laboratory. For lipid accumulation, Oil-Red-O dye (solvent red 27) was used. The other samples were dyed with hematoxylin-eosin stain and examined under standard light microscopy, after being fixed with 10% buffered formalin. We classified the severity of atherosclerotic lesions into the 3 subgroups described above.

Statistical Analysis

The data were evaluated by using SPSS for Windows, Release 8.0 (SPSS, Inc.; Chicago, Ill). The statistical analysis of MES-positive and MES-negative patients with and without stroke, in regard to the severity of aortic lesions and other risk factors, was done by χ² and logistic regression analysis. All statistical significance was assumed at the P lt;0.05 level.

Results

In the study group, there were 48 men (mean age, 58.4 ± 10 years; range, 37 to 77 years) and 21 women (mean age, 58.2 ± 14.4 years; range, 29 to 71 years). Hypertension was present in 56% of patients, a history of smoking in 61%, hypercholesterolemia in 28%, diabetes in 18%, and nonsignificant carotid stenosis (<50%) in 23% (Table I). The prevalence of aortic atherosclerosis was 70% (48 of the 69 patients), and ulcerated plaques (grade III) were present in 15%. The severity of atherosclerosis of the aortic arch was grade I in 27 patients (39%), grade II in 11 patients (16%), and grade III in 10 patients (15%); and there was no abnormality in 21 patients (30%) (Table II). Preoperatively, the incidence of MES in patients with aortic atherosclerosis was 15% (total, 10 patients: grade III in 6 patients and grade II in 4 patients), and the mean of the MES count was 3.2 ± 1.2 per hour (range, 2 to 8/h). Patients positive for MES were significantly more likely to have histories of smoking, to have aortic atherosclerotic disease, and to experience postoperative stroke than were patients negative for MES. During this preoperative phase, none of the patients had neurologic signs or symptoms (Table I). After the first 24 postoperative hours, 29 patients (42%) displayed MES, and the mean number of MES was 9.8 ± 3.1 per hour (range, 2 to 28/h). All patients who displayed MES preoperatively also displayed MES postoperatively. Preoperative and postoperative MES were significantly higher in patients who had aortic atherosclerosis than in those who did not (χ²=19.8; P lt;0.001). Among patients with postop-erative MES, 9 had grade III aortic atherosclerosis (31%), 9 had grade II (31%), 6 had grade I (21%), and 5 had normal aorta (17%). The incidence of postoperative stroke was 10%. Seven patients (10%) had minor stroke (sensory-motor stroke in 5; pial infarction, in the territory of the middle cerebral artery, in 1; and vertebrobasilar infarction in 1), and clinical signs were resolved within 2 weeks in all patients. Patients with MES had more frequent stroke than did those without (χ²=43.84; P lt;0.001). All patients with stroke had preoperative and postoperative MES: the mean preoperative MES count in the 7 patients with stroke was 1.6 ± 0.8 per hour (range, 1 to 3/h), which increased to 17.4 ± 3.3 per hour (range, 2 to 28/hr) in the 24th hour of the postoperative phase (Table III). Patients with stroke had grade II (2 patients) or grade III (5 patients) atherosclerotic disease of the ascending aorta and aortic arch. These 7 patients with stroke had higher MES counts than did the 21 patients with normal aorta (17.4 ± 3.3/h vs 5.9 ± 3.1/h; P lt;0.05). In 1 patient, cerebral computed tomography demonstrated a new ischemic lesion in the territory of the middle cerebral artery.

TABLE I. Data Pertaining to Patients with and without Pre- and Postoperative Microembolic Signals

TABLE II. Severity of Aortic Plaque in Patients with and without Postoperative Stroke

TABLE III. Counts of Microembolic Signals per Hour during Pre- and Postoperative Phases, in 7 Patients with Stroke

Discussion

The present study confirms that plaques in the ascending aorta and aortic arch impart a high risk of cerebral ischemic events in patients who undergo CABG surgery. Furthermore, the risk of microembolization and cerebral ischemic events increases with the severity of aortic atherosclerosis. The combination of thickness and ulceration was predictive of a higher mean MES count, especially during the postoperative phase, when compared with the absence of aortic plaques. ^14,15

To our knowledge, the current study is the 1st in which in vivo pathologic analysis of aortic plaques was systematically performed—and microembolic signals were monitored by TCD—before and after open-heart surgery. We observed, postoperatively, 7 patients with ischemic stroke, who also had severe aortic atherosclerosis. The MES were present in all patients with stroke during the preoperative phase, and they increased in those patients during the postoperative period. It seems that the severity of atherosclerotic aortic disease increases the risk of stroke, because half of the patients with grade II and III aortic atherosclerosis experienced stroke during the postoperative period. Case-control studies, using autopsy material and TEE, showed that atherosclerosis of the ascending aorta and aortic arch were a potential source of cerebral emboli and a major cause of perioperative stroke in patients undergoing CABG. 2,4,6,15 Cohen and colleagues ⁹ have shown that atherosclerotic plaques greater than 4 mm in thickness in the ascending aorta and the proximal arch, as detected by TEE, increase the relative risk of cerebrovascular events. ⁹ However, a causal relationship has never been demonstrated, and it has been thought that they may merely be a marker of atherosclerotic disease. Certain arguments suggest that aortic arch atherosclerosis is a source of cerebral embolism as observed during aortography, coronary arteriography, and cardiac surgery that requires cannulation of the aorta for extracorporeal circulation. ^6,15 Despite our study's small sample size, it showed a trend towards a higher frequency of MES in patients with stroke and severe aortic atherosclerotic lesions.

The reported incidence of perioperative stroke as a complication of CABG is 0.9% to 5.2%. ⁷ Causes that have been proposed for perioperative stroke include embolization from an atherosclerotic ascending aorta; microembolization of air, fat, or platelet aggregates formed during bypass; and carotid artery thrombo-embolic events. In recent years, atheromas of the ascending aorta have been identified as a major cause of perioperative stroke in CABG. In a prospective study performed with intraoperative TEE, Trehan and associates ¹⁵ were able to reduce the overall perioperative stroke rate to 0.76% by modifying the surgical procedure in accordance with the location and extent of thoracic aortic disease. Similarly, Duda and co-workers ¹⁴ used intraoperative surface aortic ultrasonography to detect atherosclerosis of the ascending aorta in patients undergoing CABG and found that simple modifications of surgical technique in accord with ultrasonographic findings reduced the stroke rate. In a TEE study of thoracic aortic plaque morphology in patients who had sustained brain infarctions, Cohen and colleagues ⁹ found that the risk of events was higher in patients without calcifications than in patients with calcifications, regardless of other morphologic features. In our study, we observed that the frequency of MES increased during the postoperative phase, which suggests that atherosclerotic plaque is disrupted by intraoperative manipulation or by postoperative platelet dysfunction that promotes fibrinolysis and thrombosis. ^3,16,17 In our study, we did not find significant carotid artery disease or cardiac arrhythmia, which are potential sources of cerebral embolism. Our observation of the microemboli count appeared to indicate vascular sources.

Several studies ^6,16,17 have suggested that neurologic and neuropsychological deterioration in a large number of patients after CABG is due at least in part to cerebral microembolism. A 1993 study ¹² demon-strated that postoperative MES counts of more than 50 signals per hour in patients who had undergone carotid surgery were significantly associated with the development of focal cerebral ischemia. We too observed that higher MES counts (>15/h) in patients with severe aortic atherosclerosis (especially those with grade III) were strongly associated with stroke. However, the clinical consequences of cerebral microemboli may be dependent not only on the number but also on the composition and size of emboli entering the vasculature of the brain during CABG surgery.

In conclusion, our findings suggest that aortic atherosclerosis with higher MES frequency may be a marker of postoperative stroke. In our judgment, thorough preoperative and intraoperative evaluation of patients to locate aortic plaque and to determine its morphology can enable surgeons to reduce the risk of cerebral embolism by better adapting their surgical methods to the individual patient.