The Correlation Between Carotid Stenosis and Perioperative Stroke During Heart Surgery: Is There Real Evidence?

CONTENT

1 Introduction

2 Methods
2.1 Search method
2.2 Evaluation criteria for the articles
2.2.1 Inclusion criteria
2.2.2 Outcome measures

3 Results
3.1 Study results
3.2 Evaluation of the studies
3.3 Comparison of the studies

4 Discussions and Conclusion

5 References

ABSTRACT

The goal of this systematic review was to find evidence for a correlation between carotid artery stenosis and perioperative stroke in patients undergoing cardiac surgery. Articles were searched in NCBI Pubmed, and appraised based on several criteria, to explore the evidence for a correlation between carotid stenosis and perioperative stroke. A total of 96 articles were found with the search strategy. Also references as well as related articles were screened. A total of six articles matched the criteria and results were accurately analysed. Concluding, the incidence of carotid stenosis is signifiantly higher in stroke patients, comparted to patients without stroke. However, carotid stenosis is the aetiological factor of stroke in just a little number of patients, who suffer from stroke and is more likely an additional co-morbidity, without an independent correlation.

1 INTRODUCTION

One of the major complications during cardio pulmonary bypass (CPB) surgery is stroke, with an incidence of approximately 1,7%.1 The prevalence of significant carotid stenosis in patients undergoing CPB surgery is high, ranging from 2-26%.2 Carotid stenosis is thought to be an independent risk factor of peri- and postoperative stroke during cardiac surgery,3-5 not without controversy about the exact role that carotid stenosis plays in the aetiology of postoperative morbidity and mortality. In this systematic review a literature research was done to analyse the correlation between severity of stenosis and incidence of peri- and postoperative stroke.

2 METHODS

A literature search was performed in Pubmed (until April 2009), combining the search terms “carotid artery”, “stenosis”, “cardiac surgery”, “pre-operative screening” and “post-operative stroke”. For all search terms, alternative search-terms were used as well as the “Mesh terms”, when possible (see table 1). Besides this, also the references list of articles and related articles (in Pubmed) were used in the search to augment the pool of literature. A criterion of limitation was that the article is published in the English, German or Dutch language.

illustration not visible in this excerpt

Table 1: Search strategy.

2.2 Evaluation criteria for the articles

Several criteria were used to screen and evaluate articles to answer the research question.

2.2.1 Inclusion criteria

Study type

Articles were included, when in the majority of patients, undergoing heart surgery, preoperative carotid artery screening was performed without interventions of the carotid arteries, like carotid endarterectomy (CEA).

Study size

The group size needed to contain at least 500 individuals because of a low perioperative stroke incidence.

Outcome

Postoperative stroke rate needed to be reported in the study.

2.2.2 Outcome measures

In the evaluation of the main points of this review, zero to two points (“+”) were given to every criterion, depending on the degree of fulfilment. The main points that were evaluated were: method, statistical analysis, sample size, in- and exclusion criteria, patient characteristics, aetiology of disease and internal as well as external validity.

Method

Selected articles were appraised on the description of the criteria: spectrum of patients, screening method, definition of carotid stenosis, definition of stroke, aetiology of stroke and time of follow-up.

Statistical analysis

Here it was evaluated if a uni- and multivariate analysis was performed on factors, related to stroke and carotid stenosis and if the 95% confidence limits were given.

Sample size

The sample size (screened carotid arteries) is not the same as the number of patients, who are enrolled in the study because not every patient was screened in every study. Here points were given for sample size, totality of patients screened and if there was a random selection on screening.

In- and exclusion criteria

For mentioning in- and exclusion criteria one point was given. In addition a point was given, when a reason was mentioned and when the criteria seem to be conductive to answer the research question.

Patient characteristics

It was further taken into account if patient characteristics were mentioned. A point was given for age and gender, co morbidities and intra-operative variables. When all three were present, an extra point was given for completeness.

Aetiology of disease

For the measurement of a correlation of patient characteristics with either stroke or carotid stenosis points were given.

Also a point was given for the evaluation of the origin of stroke.

Internal validity

The internal validity is depending on the comparability of the patients, included in the study. In this way a high homogeneity of age, gender and stadium of disease is necessary for a high internal validity.

External validity

Here it is evaluated, in how far results of the study can be projected to the generalized population. In this way it is important that the study sample is randomly chosen and reflecting the total population in terms of age, gender and stadium of disease.

3 RESULTS

3.1 Study results

In the primary Pubmed search, of a total of 93 results, articles were screened, based on the primary compiled inclusion criteria, as were related articles. In full text analysis, all references were included in the screening, when suggesting subject related literature. Of all articles, a total of six meet the inclusion criteria, to achieve adequate evidence for a conclusion about a correlation between the grade of stenosis and the incidence of peri- and postopertive stroke.

D´ Agostino et al.6 prospectively collected the data of 1835 patients, who had a first time CABG between 1990 and 1995. Of these patients, 21 had simultaneous defibrillator implantation, 34 CEA/CABG and 1 a multiple operation.

Scanning data for carotid stenosis was achieved for 1279 (69,7%). For all patients, 33 clinical characteristics were collected and tested for the correlation with either carotid stenosis or stroke. A total of 262 patients (20,5%) had significant stenosis, 172 (13,5%) uni- and 90 (7%) bilateral.

In the whole group, stroke appeared in 45 patients (2,5 %). 21 (50%) had stenosis >50%. The incidence increased with grade of stenosis from 2,2% (no stenosis) to 17,8% and 13% (bilateral moderate and severe stenosis respectively). Stenosis was observed in 50% of stroke patients, compared to 20% in patients without stroke (bilateral stenosis: 4,4% compared to 27%). The origin of stroke was further examined and related to aortic emboli, cardiac or cerebrovascular origin. Of stroke patients with measured carotid stenosis, not undergoing CEA/CABG (n=17), stroke could be related to cerebrovascular origin (carotid stenosis) only in six subjects. Of these six, four had history of stroke/TIA, three had unilateral stenosis and two bilateral. Just one patient was found to have no history of stroke/TIA and had bilateral stenosis. In the whole group, the incidence of stoke was the same in patients with a history of stroke/TIA without stenosis (4,6%) compared to patients with no history of strike/TIA with stenosis (4,6%). The highest incidence was seen in patients with both, stenosis and a history of stoke/TIA (18,2%). Similar results were achieved in a study of Hogue et al.,7 who followed 2972 patients, undergoing cardiac surgery between 1990 and 1996. After exclusion of patients with aortic root replacement, emergency operation or without available epiaortic ultrasound, a total of 2229 (75%) patients, who were older than 65 years or had a history of neurological events or carotid bruits were preoperatively scanned with Dublex. Stroke occurred in 48 patients (1,6%), of which 35% as early strokes and the rest as delayed. The stroke rate was higher in patients, who underwent valve surgery (3,4%), than in patients who underwent CABG alone (1,3%) or a combination of both (1,2%). In this study the cause of stroke was expected to be embolic in 29 patients (60,4%) and based on disturbed cerebrovascular hypoperfusion in the other 19. No specific subgroup analysis was performed in the article and just differentiated between moderate and significant stenosis. No correlation could be shown between grade of stenosis and incidence of stroke. In total 14 patient characteristics were measured for relation with the incidence of stroke. For all strokes, history of stroke was the strongest predictor (OR=14,0 (95% CI=6,7-29,4)). In a study of Bilfinger et al.,8 all patients with a stenosis >80% underwent combined CEA/CABG, so that just the effect of stenosis <80% on stroke incidence could be studied. In all patients, who underwent CABG between 1993 and 1997, 34 patients (1,7%) of the CABG-only group (n=1987) suffered from stroke.

Of these patients, 8 had no, 22 <60% and four >60% carotid stenosis. The reason of stroke was evaluated and believed to be embolic in 29 and hypoxic (watershed or diffuse) in 5 (14,7%). In an overall analysis it was shown that carotid stenosis was not an independent risk factor for stroke, whereas was considered for extensively calcified aorta (OR=2,82 (CI=1,34-5,97)), hypertension (OR=2,67 (CI=1,22-5,83)), age (OR=1,09 (CI=1,09-1,13)) and bypass time (OR=1,01 (CI=1,00-1,02)). Here it is important to take under consideration that also CEA/CABG patients were included in the risk factor analysis, which makes the interpretation of these data difficult.

Tunio et al.9 performed CABG with or without valve replacement in 3344 patients between 1995 and 1998. In total, 182 (5,4%) had significant stenosis (60-99%) and 61 (1,8%) total occlusion, of whom 53 with non-significant contralateral stenosis and 8 with high-grade contralateral stenosis. In the other group no significant contralateral stenosis was recognized. 1,6% of patients with no significant stenosis, 3,8% with significant stenosis and 6,5% with an occlusion had perioperative stroke. In these groups the perioperative mortality was 3,6%, 6,6% and 8,6% respectively. In the whole group 72 patients underwent combined CEA/CABG, leaving 118 patients with significant stenosis, of whom 8 had contralateral occlusion. In this group a total of 5 patients had stroke (4,2%) and 10 died (8,5%). A further interpretation of these data was not possible because no exact subgroup analysis was possible. The incidence of stroke is higher in the stenosis than the no-stenosis group but besides a correlation between smoking and stroke, nothing is mentioned about the origin of stroke or other independent risk factors for stroke. For a more detailed view on patients, suffering stroke during CABG, Schoof et al.10 additionally performed transcranial Doppler sonography with carbon dioxide testing in patients with high-grade stenosis or occlusion to test the cerebrovascular reserve capacity.

Of all patients between 1996 and 2005 (n=7497), in 2797 preoperative carotid screening was performed (1995-1997 routinely, 1998-2005 selectively in “at-risk” patients). 528 patients (18,9%) had at least medium stenosis and 200 (7,2%) high-grade stenosis or occlusion. In total 67 patients (2,4%) experienced an anterior hemispheric stroke.

Of these patients 42 had no significant stenosis, 6 medium-grade, 1 with high-grade stenosis and 6 with occlusion. 3 strokes occurred contralateral to the side of stenosis/occlusion and 9 in the vertebrobasilar circulation. Of nearly all patients with at least high-grade stenosis, cerebrovascular reserve capacity was tested (see table 2).

illustration not visible in this excerpt

Table 2: Cerebrovscular reserve capacity correlated to grade of carotid stenosis.10

The risk of stroke was calculated for different subgroups of carotid stenosis. Patients with medium-grade stenosis as well as high-grade stenosis/occlusion without exhausted cerebrovascular reserve capacity are at even lower risk for suffering postoperative stroke (OR=0,73 and 0,35 respectively), compared to no/low-grade stenosis. Just patients with an exhausted cerebrovascular reserve capacity and high-grade stenosis/occlusion are at high risk for suffering postoperative stroke (n=3/11, OR=28,34 (CI=5,77-139,1)). Of importance in this article is that patients were selected for carotid screening, based on history of stroke, carotid bruits and peripheral vascular disease. This led to a patient population with a higher incidence of severe coronary artery disease and atherosclerotic disease in other organ systems. The correlation of grade of stenosis and incidence of stroke is given in table 3 and screening procedure and definition of stroke of all articles are presented in table 4. The incidence of carotid stenosis in patients, suffering from perioperative stroke, of other studies, where patient numbers were considered as to low for a sufficient conclusion are shown in table 5.

illustration not visible in this excerpt

Table 3: Correlation between of grade of carotid stenosis and incidence of perioperative stroke. *12 strokes were excluded because 3 were contra-lateral and 9 vertebrobasilar.

[...]