Summary
Background
Extracranial internal carotid artery stenosis is a risk factor for perioperative stroke in patients undergoing coronary artery bypass surgery (CAB). Although selective and non-selective methods of preoperative carotid screening have been advocated, it remains unclear if this screening is clinically relevant.
Aim
To test whether selective carotid screening is as effective as non-selective screening in detecting significant carotid disease.
Methods
The case records of patients consecutively undergoing CAB were reviewed. Patients were stratified retrospectively into high- or low-risk groups according to risk factors for significant carotid stenosis and perioperative stroke: peripheral vascular disease (PVD), carotid bruit, diabetes mellitus, age >70 years and/or history of cerebrovascular disease. Prevalence of carotid stenosis detected by ultrasonography, surgical management and perioperative stroke rates were determined in each group.
Results
Overall, 205 consecutive patients underwent preoperative carotid screening. The prevalence of significant carotid stenosis was 5.8%. Univariate analysis confirmed that PVD ( P = 0.005), carotid bruit ( P = 0.003) and diabetes mellitus ( P = 0.05) were significant risk factors for stenosis. Carotid stenosis was a risk factor for stroke ( P = 0.03). Prevalence of carotid stenosis was higher in the high-risk group (9.1%) than the low-risk group (1.2%) ( P < 0.05). All concomitant or staged carotid endarterectomies/CAB (5/205) and all patients who had perioperative strokes (5/205) were in the high-risk group ( P = 0.01).
Conclusion
In our cohort, selective screening of patients aged >70 years, with carotid bruit, a history of cerebrovascular disease, diabetes mellitus or PVD would have reduced the screening load by 40%, with trivial impact on surgical management or neurological outcomes.
Résumé
Objectifs
Les sténoses de l’artère carotide interne sont un facteur de risque d’accident vasculaire cérébral (AVC) chez les patients bénéficiant d’une chirurgie coronaire. Faut-il pour autant dépister ces sténoses chez tous les patients ou une sélection est-elle judicieuse ?
Méthodes
Les dossiers des patients ayant bénéficié de pontages coronaires ont été revus. Ces patients ont été rétrospectivement classés en « haut » et « bas » risque d’AVC en fonction des facteurs de risque identifiés dans la littérature. La prévalence des sténoses carotidiennes au doppler, ses conséquences sur le management des patients et le taux d’AVC periopératoire étaient déterminés dans chaque groupe.
Résultats
Deux cent cinq patients consécutifs ont bénéficié de l’évaluation carotidienne préopératoire par échoDoppler. La prévalence des sténoses carotidiennes significatives au doppler était de 5,8 %. L’analyse univariée a confirmé que l’AOMI ( p = 0,005), le souffle carotidien ( p = 0,003) et le diabète ( p = 0,005) étaient des facteurs de risque significatifs de sténose. La sténose carotidienne était un facteur de risque d’AVC ( p = 0,003). La prévalence des sténoses était supérieure dans le groupe à « haut » risque (9,1 % contre 1,2 % ; p = 0,05). Tous les patients ayant bénéficié d’un geste sur les carotides avant ou pendant la chirurgie coronarienne (5/205) et tous les patients ayant souffert d’un AVC (5/205) étaient dans le groupe à « haut » risque indépendamment du fait qu’ils aient une sténose ou pas ( p = 0,01).
Conclusion
Dans notre série, le dépistage sélectif chez les patients de plus de 70 ans, ayant un souffle carotidien, un antécédent cérébrovasculaire, un diabète ou une AOMI aurait réduit le nombre d’échodoppler de 40 % quasiment sans impact sur la prise en charge et les AVC.
Background
Perioperative stroke is one of the major complications of coronary artery bypass surgery (CAB), with a reported incidence of 2.1–5.2% and a related mortality of 0–38% . Significant extracranial internal carotid artery stenosis (i.e. ≥70% luminal narrowing) is a well-established risk factor for perioperative stroke in patients undergoing CAB . To prevent this serious complication, carotid endarterectomy (CEA) has been recommended in patients undergoing CAB in a staged or concomitant approach; CEA/CAB studies have been conducted since the 1970s . Although the benefits of CEA/CAB remain uncertain, some of these studies reported reductions in stroke rates, prompting the notion that preoperative screening for carotid stenosis in all CAB patients is necessary to reduce perioperative and long-term stroke rates . Such systematic, non-selective carotid screening does, however, add considerable time and expense to preoperative workups.
Alternatively, some investigators have identified risk factors for carotid disease that could be used for more selective screening. These risk factors include older age , carotid bruit , previous neurological event , previous carotid surgery , peripheral vascular disease (PVD) , hypertension, diabetes, dyslipidaemia and smoking . Unfortunately, there are neither consensus criteria to provide guidelines for centres looking to optimize their carotid screening practices nor prospective management outcome studies.
In the present study, we sought to report the results of our single-centre routine experience in non-selective preoperative carotid screening of CAB patients over a 5-year period. Our hypothesis was that selective carotid screening is as effective as non-selective screening in detecting significant carotid stenosis and does not result in higher perioperative stroke rates. We also studied whether selective screening would result in significant changes in surgical management.
Materials and methods
Patient selection
We retrospectively reviewed the files of all consecutive patients undergoing isolated de novo CAB from January 2003 to December 2008, who fulfilled the necessary criteria. Inclusion criteria were: patients undergoing CAB with no other concomitant cardiac procedure (such as valve replacement/repair, aneurysmectomy, atrial septal defect closure); carotid screening by ultrasonography performed exclusively in our centre; and assessment of carotid bruit by at least one of the senior physicians in our department. Exclusion criteria were: aortic stenosis even if not significant (bruit of aortic stenosis can hide a carotid bruit); need for emergency surgery; and carotid evaluation performed in another centre.
Data collection
Prespecified preoperative, operative and postoperative clinical data were extracted independently by two investigators (D.L.S., J.-C.C.) from all patients’ charts using a standardized form. Information discrepancies were resolved by consensus or by retrieving further information from additional medical records. Preoperative variables included demographic data, smoking status, diabetes mellitus diagnosed as a documented history of diabetes or use of any antidiabetic medication, hypertension, history of previous stroke, carotid bruit, cerebrovascular disease (CVD) and PVD. Patients were considered as having PVD if they had intermittent claudication, a history of peripheral revascularization or duplex ultrasound showing significant arterial stenosis.
Evaluation of internal carotid stenosis was performed with duplex ultrasound. The degree of stenosis was expressed as the percentage of luminal narrowing estimated by ipsilateral internal common carotid artery flow velocity ratios (duplex ultrasound). Carotid artery stenosis was considered “significant” when there was ≥70% luminal narrowing of the affected internal carotid artery, which was determined by duplex ultrasonography in accordance with widely accepted clinical guidelines. In our institution, CEA is considered if carotid stenosis is >70% in asymptomatic patients; surgery is decided on a case-by-case basis .
Postoperative data were extracted from a neurological assessment/outcome database initiated at our centre to prospectively monitor the neurological progress and clinical outcomes of all patients after cardiac surgery. These data were collected on a daily basis and included death and stroke ratios. A cerebral vascular accident or “stroke” was defined as an acute neurological event resulting from cerebral circulatory impairment and lasting >24 hours. The outcome of postoperative stroke was defined as the clinical diagnosis of stroke and confirmed by brain imaging (head computed tomography, magnetic resonance imaging or both). A transient ischaemic attack was defined as a temporary neurological deficit attributable to circulatory impairment and lasting <24 hours. Mortality was defined as any death occurring during the same hospital stay.
Operative technique
All patients underwent median sternotomy. The anaesthetic technique was standardized and consisted of low-to-intermediate doses of narcotics, inhalational agents and paralytics. Cardiopulmonary bypass was performed with myocardial protection achieved by anterograde and/or retrograde cardioplegia and topical hypothermia. Cardiotomy suction was routinely returned to the cardiopulmonary bypass circuit. Off-pump CAB was performed according to surgeon preference.
When performed concomitantly with CAB, CEA was completed before sternotomy. Using uniform operative techniques, vascular surgeons from the division of vascular surgery conducted each endarterectomy. Partial heparinization, common to internal carotid intraluminal shunting and, when indicated, patch carotid arteriotomy closure, were used in each case.
Risk stratification
In our cohort, according to previously established risk factors for stroke , patients with at least one of the following features were retrospectively stratified into the high-risk group: patients with either CVD or PVD, diabetes mellitus, carotid bruit and/or aged >70 years. Patients without any of these risk factors were included in the low-risk group. We determined the prevalence of significant carotid stenosis, the number of CEAs performed and the number of perioperative strokes in the high-risk and low-risk groups. We retrospectively applied the screening algorithm (high-risk and low-risk groups) to our cohort of CAB patients who underwent routine carotid screening and then determined the prevalence of carotid stenosis in each group. Finally, the predictive value of the selective screening strategy based upon these risk factors was estimated.
Statistical analysis
Preoperative, operative and postoperative outcome data were reviewed. Continuous and dichotomous variables were compared using Student’s t test and the Chi 2 test, respectively. Fisher’s exact test was used for comparisons in which at least one cell value was <5. All probabilities were two-tailed with P < 0.05 regarded as significant. Statistical analysis was performed using the SPSS statistical software package (SPSS Inc., Chicago, IL, USA).
Results
Univariate risk factor analysis
We included 205 patients in the study group. Among these patients, 12 (5.8%) had significant carotid stenosis, unilateral in all cases. Univariate analysis confirmed, in accordance with the literature, that PVD ( P = 0.005), diabetes mellitus ( P = 0.05) and carotid bruit ( P = 0.003) are risk factors for significant carotid stenosis ( Table 1 ).
