Abstract and Introduction
Abstract
Objective: The benefit of carotid endarterectomy (CEA) over medical therapy in patients with asymptomatic carotid artery stenosis is predicated upon a life expectancy of at least 5 years after the procedure. The goal of this study was to create a scoring system for prediction of 5-year survival after CEA that can be used to triage patients with ACAS.
Methods: All patients who underwent CEA for severe asymptomatic carotid stenosis from 1989 to 2005 were identified. Long-term survival was determined by a review of hospital records and the social security death index. Because all patients had at least 5-year follow-up, a logistic regression of predictors of survival at 5 years was performed and the odds ratios associated with particular significant comorbidities were used to create a scoring system to predict survival. The scoring system was then validated within the cohort using the Hosmer-Lemeshow Test and a derivation/validation receiver operating characteristic (ROC) curve.
Results: There were 2004 CEA performed in 1791 patients. The average follow-up was 130 ± 49 months. The clinical profile of the cohort data included 84% hypertension, 56% coronary artery disease (CAD), 24% diabetes, and 71% on statins. The 30-day stroke rate was 1.1% and the death rate was 0.7%. The actual 5-year survival was 73%. Logistic regression yielded the following predictors of mortality: age (by decade) (odds ratio [OR] = 1.8, P < 0.0001), CAD (OR = 1.5, P = 0.0007), chronic obstructive pulmonary disease (OR = 2.5; P < 0.0001), diabetes (OR = 1.7, P < 0.0001), neck radiation (OR = 2.6, P = 0.005), no statin (OR = 2.1, P < 0.0001), and creatinine more than 1.5 (OR = 2.6, P < 0.0001). These variables were then assigned a hierarchal point scoring system in accordance with the OR value. The 5-year survival based on the scoring system was as follows: 0 to 5 points = 92.5%, 6 to 8 points = 83.6%, 9 to 11 points = 63.7%, 12 to 14 points = 46.5%, and more than 15 points = 33.8%. The Hosmer-Lemeshow test validated the scoring system (P = 0.26) and there was no difference in the ROC curves (C statistic = 0.74 vs 0.73).
Conclusions: This validated scoring system can be a useful tool for determining which patients are likely to benefit most from CEA based on the probability of long-term survival. Given that the 5-year survival of patients in the medical arm of the asymptomatic CEA trials was 60% to 70%, it is reasonable to conclude that patients who score 0 to 8 points are excellent candidates for CEA whereas most patients with >=12 points should be managed with medical therapy alone.
Introduction
Multiple prospective, randomized trials support the superiority of carotid endarterectomy (CEA) over medical therapy for the prevention of stroke in patients with asymptomatic severe (>70%) carotid artery disease. This benefit is predicated upon the maintenance of low perioperative stroke and death rates (<3) after CEA and a projected patient survival of at least 5 years after the procedure. The emergence of alternatives to CEA, namely carotid artery stenting, and the variable interpretation of subsequent comparative trials have refueled the debate over the preferred management of patients with severe asymptomatic carotid disease. Indeed, a recent document from the Society for Vascular Surgery intended to guide the clinical research goals for the next decade identified optimal management of asymptomatic carotid stenosis as their top priority.
As the proponents of carotid artery stenting continue to make the case that it is at least equivalent, if not superior, to CEA in selected patients, a parallel argument that asymptomatic carotid stenosis is best treated with modern medical therapy alone has been advocated by some. Although this position has been driven by recent data from pooled prospective trials of statin therapy that showed a significant reduction in stroke in patients taking statins, and epidemiologic evidence that the lifetime risk of stroke in the general population is decreasing compared to 20 years ago, there exists no level 1 evidentiary support of this claim. This is in particular true for patients with severe albeit asymptomatic carotid stenosis. For example, contemporary data from the Reduction of Atherothrombosis for Continued Health registry showed that the presence of a carotid stenosis more than 70% significantly increases the risk of stroke independent of statin use.
There was a 2% annual carotid-related risk of stroke in the medically treated arm of the Asymptomatic Carotid Atherosclerosis Study (ACAS) that remained consistent in the more contemporary Asymptomatic Carotid Surgery Trail despite an 80% statin usage in the later years of the study. The Asymptomatic Carotid Stenosis and Risk of Stroke study group attempted to stratify the risk of cerebrovascular events in asymptomatic patients with more than 50% carotid stenosis and found that there was a linear correlation between degree of stenosis and neurologic risk. Other investigators have attempted to stratify patients as to stroke risk based upon plaque morphology as determined by high-resolution magnetic resonance imaging or duplex ultrasonography. But, these techniques are difficult to duplicate across institutions and are currently not widely used. In the absence of reliable morphologic data for risk stratification, it is intuitive that life expectancy should be a major driver in clinical decision making referable to patients with severe asymptomatic carotid stenosis. The goal of this study was to create an objective scoring system for the prediction of 5-year survival after CEA that can be used in conjunction with other clinical data to triage patients with asymptomatic severe carotid stenosis.