Carotid artery stenting: Identification of risk factors for poor outcomes
Article Outline
Objectives
Age greater than 80 has been identified as a risk factor for complications, including stroke and death, in patients undergoing carotid artery angioplasty and stenting (CAS). This study evaluates other potential predictors of perioperative complications in patients undergoing CAS.
Methods
All cerebrovascular endovascular procedures performed by the vascular surgery division at our university hospital between July 2003 and December 2005 were retrospectively examined. During the course of 212 admissions, 198 patients underwent 215 procedures. Patient age, comorbidities, and admission status were analyzed as independent (predictor) variables. Complication rate, discharge disposition, and length of hospital stay were considered dependent (outcome) variables. Logistic regression and Fisher exact test or Student t test were performed, as appropriate.
Results
Complications included major and minor stroke, myocardial infarction, femoral artery pseudoaneurysm, and death. The rates of perioperative major and minor stroke were 0.5% and 2.8%, respectively. Chronic renal insufficiency was a predictor of perioperative complications, including stroke: patients with serum creatinine greater than 1.3 mg/dL had a 37% complication rate and a 11.1% stroke rate, while those with normal renal function had a 13% complication rate (P = .003) and a 0.6% stroke rate (P =.001). Similar association was seen between creatinine clearance and both stroke and complications. Obesity was a risk factor for complications, but not stroke: obese patients had a complication rate of 28%, while others had a 16% complication rate (P = .024). Emergency admission predicted both extended hospital stay (P < .001) and requirement for further inpatient care in a rehabilitation or nursing facility (P = .007). There was no significant difference in complication rate or stroke rate between octogenarians and others.
Conclusion
This experience demonstrates that chronic renal insufficiency, obesity, and emergent clinical setting are risk factors for patients undergoing CAS.
Carotid artery stenting (CAS) is increasingly viewed and employed as an alternative to carotid endarterectomy (CEA) for both asymptomatic and symptomatic carotid artery stenosis. Prior evidence has suggested an increased periprocedural risk of both stroke and death in octogenarians.1 Other predictors of perioperative complications have been examined and reported, including obesity and chronic renal insufficiency.2, 3
The aim of this study was to review a single-institutional experience with CAS, and to identify clinical predictors of poor outcome, including stroke, death, myocardial infarction, and other perioperative complications. Length of hospital stay and discharge disposition were considered additional endpoints of the study and were examined for correlation with clinical predictors.
Methods
Study design
Between July 2003 and December 2005, 198 patients underwent 215 consecutive carotid endovascular diagnostic and therapeutic procedures at the Hospital of the University of Pennsylvania. Each procedure was performed by one of four attending vascular surgeons. By review of the inpatient medical record, pre- and postprocedure office notes, and operative dictation, the patients and procedures were examined retrospectively for complications and for potential patient-related and procedure-related risk factors. These sources were interrogated for any complication, including stroke, occurring within 1 month postprocedure. Post hoc “intention-to-treat” analyses were also performed in an attempt to eliminate any procedures, which were evidently and exclusively planned as diagnostic procedures.
Operative procedure
All procedures were carried out percutaneously via puncture of the femoral artery. Weight-adjusted heparin (70 U/kg) was administered prior to wire or catheter manipulation in the aortic arch. Activated clotting times were measured as needed during more extended procedures. Selective angiography of the stenotic carotid artery was performed in at least two projections. If clinically indicated, the contralateral carotid artery was also selectively imaged. Procedures were performed via a long 6F sheath. All interventions were performed over a 0.014 inch system. Uncovered self-expanding stents were used exclusively (Accunet, Guidant/Abbott, Abbott Park, Ill or Wallstent, Boston Scientific, Natick, Mass). Stent diameter and length were selected according to the vessel being treated. Balloon predilation was performed infrequently, as was angioplasty without stenting. Distal embolic protection devices were used whenever possible. Patients were preferentially started on aspirin and clopidogrel preoperatively, and continued on that regimen for at least 1 month postoperatively.
Definitions
All strokes were diagnosed clinically and radiologically by a neurology team. A major stroke was defined as a new neurologic deficit persisting at least 1 month and increasing the NIH Stroke Scale by at least four points. Patients transferred into our hospital emergently (according to the admissions department), or admitted from the emergency department, or operated on emergently were considered to have “emergent procedures.” A complicated discharge was defined as any discharge to a nursing home or other inpatient nursing facility, or any discharge to home requiring arrangement for ongoing visiting nursing care. Carotid stenosis was considered “symptomatic” in patients with symptoms referable to an ipsilateral carotid distribution within the prior 1 year. Contralateral carotid status was classified as patent (<50% stenosis), stenotic (>50% stenosis), or occluded. Tortuosity was determined according to the operative dictation of the attending surgeon. Patients with body mass index (BMI) greater than 30 were considered obese. Patients with preoperative serum creatinine greater than 1.3 were considered to have chronic renal insufficiency (CRI). However, all analyses were done in parallel, using creatinine clearance (Cockroft-Gault equation) in place of serum creatinine.
Statistical analysis
Logistic regressions were used to analyze the dependence of all categorical and binary outcomes (complications, stroke, stroke-death, and discharge disposition) on proposed predictor variables. Multiple regression was utilized to analyze the dependence of length of hospital stay on all proposed predictor variables. Fisher exact test was used to interrogate the relationship between categorical and binary predictor and outcome variables, if statistical dependence was suggested by logistic regression. Similarly, Student t test was used to interrogate the relationship between continuous outcome and binary predictor variables, if statistical dependence was suggested by multiple regression. Student t test was also utilized to quantify the association between continuous independent variables and binary outcomes. All probabilities were interpreted as two-tailed, with significance taken at P < .05 level. All results are presented as mean ± standard deviation.
Results
The mean age of the patients was 70.7 (± 10.0) years; there were 105 (53%) males and 35 (18%) octogenarians. Patient demographics and comorbidities are summarized in Table I. Characteristics of carotid lesions and arterial anatomy are summarized in Table II.
Table I. Patient demographics and comorbidities
| Age (y) | 70.7 ±10.0 |
| Gender (male) | 105 (53%) |
| Obese (BMI > 30) | 58 (29%) |
| BMI (kg/m2) | 27.6±6.0 |
| Creatinine clearance (mL/min/1.73 m2) | 53 ± 22 |
| CRI (serum Cr > 1.3) | 53 (27%) |
| Prior stroke | 59 (27%) |
| COPD | 30 (14%) |
| History of CHF | 33 (15%) |
Table II. Lesions and anatomy
| Recurrent stenosis | 63(29%) |
| Symptomatic | 87(41%) |
| Contralateral occluded | 36(17%) |
| Contralateral stenotic | 46(21%) |
| Tortuous | 19(9%) |
Hospital admissions or procedures were classified as “emergent” in 38 instances (18%). Of the 215 procedures, 170 were therapeutic and 45 limited to diagnostic measures. Of the purely diagnostic procedures, 10 patients had an ipsilateral carotid occlusion, 29 had no hemodynamically significant stenosis, and six patients underwent subsequent carotid endarterectomy. Of those patients undergoing CEA, four were converted preferentially; in the other two patients, endovascular attempts at crossing the carotid lesion were unsuccessful. Distal embolic protection devices were used in 92% of therapeutic procedures; reasons for nonutilization were: tortuous internal carotid artery or narrow culprit lesion not allowing passage of device, tortuous distal internal carotid artery not deemed suitable for distal placement of device, or narrow and diseased distal internal carotid not deemed suitable for deployment of device.
Clinical outcomes, including complications, of all procedures are summarized in Table III. In addition, Table III indicates the complication rates for the cohort reanalyzed according to an “intention-to-treat” algorithm; none of those patients undergoing purely diagnostic arteriograms suffered any complications. Overall, there were 53 complications in 41 procedures. There was a 0.9% (two patients) death rate, a 1.4% (three patients) MI rate, and a 3.3% (seven patients) stroke rate. There was only one major stroke (0.5%). All strokes were ischemic or embolic in nature; there were no hemorrhagic strokes. Of the six strokes occurring in the setting of therapeutic procedures, four were ipsilateral and two contralateral; one stroke occurred in the setting of a diagnostic cerebral arteriogram. Narrowing the cohort to only those patients undergoing therapeutic intervention, the stroke rate was 3.5% (6/170). A distal embolic protection device was used in all patients undergoing CAS who suffered strokes.
Table III. Clinical outcomes – including complications
| Outcome | Number | Percent | Number (ITT) | Percent (ITT) |
|---|---|---|---|---|
| Complications | 41 procedures (53 total) | 19.1% | 41 procedures (53 total) | 19.8% |
| Strokes | 7 | 3.3% | 7 | 3.4% |
| Major | 1 | 0.5% | 1 | 0.5% |
| Minor | 6 | 2.8% | 6 | 2.9% |
| Death | 2 | 0.9% | 2 | 1.0% |
| MI | 3 | 1.4% | 3 | 1.4% |
| Local groin complications | 17 | 7.9% | 17 | 8.2% |
| Hematoma | 9 | 4.2% | 9 | 4.3% |
| Pseudoaneurysm | 6 | 2.8% | 6 | 2.9% |
| Bleeding | 2 | 0.9% | 2 | 1.0% |
| Length of hospital stay (d) | 3.1 ± 3.9 (mode = 1) | 3.1 ± 3.9 (mode = 1) | ||
| Discharge disposition | 28 complicated | 13.0% | 25 | 12.1% |
The clinical milieu of each of the seven strokes is summarized in Table IV. Of the independent variables included in the logistic regression, only chronic renal insufficiency was predictive of stroke (odds ratio [OR] 30.9; 95% confidence interval [CI] 1.8-528.9; P = .017). Of the 54 patients with chronic renal insufficiency, six suffered strokes (11.1%); of the 161 patients with normal renal function, only one suffered a stroke (0.6%, P = .001). Creatinine clearance was also predictive of stroke by logistic regression (P = .015). Those patients suffering stroke had creatinine clearances of 30.3 ± 3.6; others had creatinine clearances of 53.9 ± 21.5 (P = .004). The influence of selected independent variables on the advent of stroke is summarized in Fig 1.
Table IV. All strokes observed in this seriesa
| Event | Indication and procedure | Timing | Nature of stroke | Contralateral carotid |
|---|---|---|---|---|
| Major stroke | ||||
| 1 | CAS for symptomatic 85% stenosis | Intraop | Ipsilateral temporal-parietal | Patent |
| Minor strokes | ||||
| 1 | CAS for asymptomatic high grade stenosis | Intraop | Ipsilateral parietal | Patent |
| 2 | CAS for symptomatic > 90% stenosis | POD 1 (at home) | Ipsilateral cerebellar | Patent |
| 3 | CAS for asymptomatic recurrent stenosis | PACU | Contralateral parietal | Occluded |
| 4 | CAS for symptomatic recurrent stenosis | POD 6 | Ipsilateral parietal | Occluded |
| 5 | CAS for symptomatic high-grade stenosis | POD 1 | Contralateral temporal | Stenotic |
| 6 | Angiogram for stroke, MRA-US discordance | POD 1 | — | Patent |
aA distal embolic protection device was used in all patients undergoing CAS who suffered strokes. |
Fig 1.
Odds ratios for stroke given various clinical conditions, according to logistic regression (LR). N = 7 affected, 184 without stroke.
By multivariate analysis, the predictors of complications were obesity (OR 3.1; 95% CI 1.2-8.4), emergent admission or procedure status (3.9; 1.3-11.4), and CRI (3.4; 1.6-7.1). The influence of selected independent variables on complication rate is summarized in Fig 2. Both CRI (P = .002) and obesity (P = .049) predicted complications by univariate analysis as well (Fisher exact test). Creatinine clearance was also predictive of complications by logistic regression (P = .032).
Fig 2.
Odds ratios for complications given various clinical conditions, according to logistic regression (LR). N = 36 affected, 155 without complication.
Obese patients suffered a 28% complication rate (16/58, with 19 total complications), while those with BMI less than 30 had complications in only 16% (25/157) of cases (P = .024). Groin complications (bleeding, hematoma, or pseudoaneurysm) accounted for slightly less than one half of all complications in the obese.
Age did not confer a significantly increased risk of stroke or complications, by logistic regression using age as a continuous variable or as a binary variable (octogenarians versus others). Likewise, age was not associated with any adverse outcome by either Student t test (age as continuous variable) or by Fisher exact test (octogenarian status). The incidence of stroke and of perioperative complication, segregated by age, are summarized in Table V,which demonstrates that there was a tendency towards higher stroke, stroke-death, and complication rates in octogenarians in this study.
Table V. Influence of age on complications, including strokes
| Octogenarian | Younger | Sig. | |
|---|---|---|---|
| Number | 35 | 180 | — |
| Stroke | 2(6%) | 5(3%) | .319 |
| Stroke/Death | 3(9%) | 6(3%) | .165 |
| Complications | 9(26%) | 32(18%) | .171 |
By multivariate analysis, the predictors of complicated discharge were emergent admission or procedure status (OR 7.9; 95% CI 2.4-26.3), prior stroke (3.1; 1.0-9.3), and congestive heart failure (CHF) (5.5; 1.6-19.4). All three conditions predicted complicated discharge by univariate analysis as well, as summarized in Table VI.
Table VI. Predictors of complicated discharge (to subacute care facility or to home with requirement for ongoing professional care)
| Complicated discharge | Uncomplicated discharge (home) | Significance | |
|---|---|---|---|
| Admission and procedure status | |||
| Emergent | 12(31%) | 27 | P=.007 |
| Elective | 16(9%) | 160 | |
| History of CHF | |||
| Yes | 10(30%) | 23 | P=.004 |
| No | 28(13%) | 187 | |
| Prior stroke | |||
| Yes | 13(22%) | 46 | P=.022 |
| No | 15(10%) | 141 |
Emergent admission or procedure status was the only independent variable which predicted extended hospital stay by both univariate and multivariate analysis. Those 39 patients treated emergently had inpatient hospital stays of 7.8 ± 5.8 days (postoperative length of stay 4.6 ± 4.3 days), while others had stays of 2.1 ± 2.3 days (P < .001). Reasons for emergent admission or procedure included: unstable angina, CHF, cardiogenic shock, recent transient ischemic attack (TIA), crescendo TIA, recent stroke, and syncope.
Discussion
Carotid artery angioplasty and stenting, with the concomitant use of distal embolic protection device, is a purported alternative to carotid endarterectomy (CEA) for the treatment of carotid stenosis, and subsequent prevention of stroke.4, 5, 6, 7, 8 It is an attractive alternative to CEA in the setting of medical or surgical comorbidities. Specifically, CAS has been demonstrated to be a viable alternative to CEA in the setting of high-risk patients.9, 10
The current report summarizes the experience of a vascular surgery group at an academic medical center over 2 and half years. The patient population was heterogeneous: 29% of patients were obese, 27% of patients suffered from CRI, and 27% of patients had a history of prior stroke. Overall, the 2.8% minor stroke rate and 0.5% major stroke rate detected in the current study are similar to prior data and published results.11 Meanwhile, the overall complication rate of 19% reflects, at least in part, a relatively large number of complications related to femoral arterial access (7.9% rate).
There is some question as to the most informative and correct definition of the current cohort, for calculation of complication rates and the determination of risk factors. The authors have chosen to include all carotid endovascular interventions. This allows for the not insignificant incidence of stroke and other complications after purely diagnostic studies, and obviates having to distinguish among studies, which were clearly intended to be diagnostic exclusively, those that were undertaken as “possible” carotid interventions, and those that were planned as therapeutic procedures but were not completed based on unexpected angiographic findings or inability to cross a lesion. The complication and stroke rates have also been calculated, however, both according to an “intention-to-treat” formulation (207 procedures) and incorporating only the ultimately therapeutic procedures (170). The key results of the study are largely invariant across these three different formulations of the cohort.
This study endeavors to identify, from among a comprehensive collection of potential predictor variables, those independent variables which confer increased risk of perioperative stroke, complication, complicated discharge, or prolonged hospital stay in patients undergoing CAS.
Among all patient-related and procedure-related variables, only chronic renal insufficiency was predictive of stroke by multivariate analysis. In this series, CRI conferred an 18-times greater risk of stroke compared with patients with normal renal function. CRI was also found, along with obesity and emergent admission or procedure status, to confer a higher risk of complications in general. One previous reports support this evidence: Saw et al identified CRI as an independent risk factor for stroke in patients undergoing CAS.2 Meanwhile, CRI has been implicated as a risk factor for perioperative mortality in patients undergoing CEA.12 Therefore, these results may not appreciably influence the decision-making process in cases wherein a patient is eligible for either CEA or CAS; they do, however, reinforce the need for a cautionary approach to CAS in those patients who have impaired renal function. The rationale for increased rates of stroke and other complications in patients with CRI undergoing CAS is elusive. While it is expected that these patients have significant cardiovascular pathology related to their kidney disease, all patients in this cohort, with or without impaired renal function, clearly have cardiovascular risk factors (ie, atherosclerotic carotid artery disease). Meanwhile, other potential risk factors for poor outcome, which are likewise associated with CRI (coronary artery disease, CHF, history of MI) failed to predict complications in this cohort; CRI presumably, then, is not exerting its influence through these associated conditions.
The present results also implicate obesity as a predictor of perioperative complications. Approximately one half of the complications in the obese patient population were related to femoral arterial access. (Complications related to groin arterial access accounted for nearly one-third of all complications in this series.) The increased soft tissue density within the groin area in the obese patients likely prevented adequate compression of the femoral artery against the femoral head. Furthermore, these patients may have a more difficult time laying flat after sheath removal. Our institution does not use closure devices; the use of these devices might decrease the incidence of this complication.13, 14 This study is limited by its retrospective nature. Retrospective studies are necessarily susceptible to selection bias and missing data. The current cohort does not have an appropriate control group, comprised of open endarterectomies or otherwise. However, its primary conclusions, especially those regarding the increased risk of stroke and complication in renal failure, are statistically robust. Therefore, we suggest that these correlations be examined in larger, prospective databases.
CAS has been considered a less invasive and less stressful procedure compared with CEA. Thus, we sought to determine whether any clinical features were predictive of extended hospital stay or complicated discharge. Emergent admission or procedure status conferred an increased risk of complicated discharge, largely because of the propensity for complications in these patients; prior stroke and CHF also predicted complicated discharge, likely owing to the attendant disabilities in these patients. Meanwhile, emergent intervention predicted extended length of hospital stay: most of these patients had extended inpatient stays either because of the need for preoperative workup or for treatment of associated conditions (including other vascular surgical conditions, the need for cardiac surgery, or for stabilization of cardiac status).
Historically, individual trials15, 16 have failed to find a significantly increased risk of stroke or death in octogenarians; however, these have been relatively underpowered to detect such differences.17 Likewise, the current series did not elucidate an increased risk of stroke or other complications in the octogenarian population. However, the increased risk in octogenarians was demonstrated in both Carotid Revascularization Enterectomy Versus Stenting Trial (CREST) and Carotid Acculink/Accunet Post Approval Trial to Uncover Rare Events (CAPTURE).1, 18 In addition, Kastrup et al showed that octogenarians undergoing CAS have higher stroke rates than those undergoing CEA, although this was only a retrospective analysis.19
In summary, while CAS offers an appealing alternative to CEA, continued analysis of CAS is critical. Recent data from the EVA-3S and SPACE trials has called into question the safety of CAS compared with CEA in patients with symptomatic carotid stenosis, with respect to the risk of stroke, death, and all complications.20, 21 It is therefore all the more important to continue to identify the risk factors associated with poor outcomes during CAS. In the present series, factors predictive of stroke, complications, extended hospital stay, and complicated discharge were identified. Although these risk factors may not preclude CAS, they should be considered in the clinical decision-making process when deciding between CAS and CEA.
Author contributions
References
- Carotid artery stenting is associated with increased complications in octogenarians: 30-day stroke and death rates in the CREST lead-in phase. J Vasc Surg. 2004;40:1106–1111
- . Effect of chronic kidney disease on outcomes after carotid artery stenting. Am J Cardiol. 2004;94:1093–1096
- . Impact of body mass index on outcome in patients undergoing carotid stenting. Am J Cardiol. 2005;96:1743–1745
- . Carotid angioplasty and stenting versus carotid endarterectomy: randomized trial in a community hospital. J Am Coll Cardiol. 2001;38:1589–1595
- . Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med. 1991;325:445–453
- . Carotid angioplasty and stenting versus carotid endarterectomy for treatment of asymptomatic carotid stenosis: a randomized trial in a community hospital. Neurosurgery. 2004;54:318–324
- Impact of cerebral protection devices on early outcome of carotid stenting. J Endovasc Ther. 2002;9:786–792
- . Early outcome of carotid angioplasty and stenting with and without cerebral protection devices: a systematic review of the literature. Stroke. 2003;34:813–819
- Protected carotid-artery stenting versus endarterectomy in high-risk patients. N Engl J Med. 2004;351:1493–1501
- . Endovascular versus surgical treatment in patients with carotid stenosis in the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS): a randomized trial. Lancet. 2001;357:1729–1737
- . SPACE: not the final frontier. Lancet. 2006;368:1215–1216
- . Carotid endarterectomy in patients with chronic renal insufficiency: a recent series of 184 cases. J Vasc Surg. 2005;41:24–29
- Rapid arterial hemostasis and decreased access site complications after cardiac catheterization and angioplasty: results of a randomized trial of a novel hemostatic device. J Am Coll Cardiol. 1685;25:1685–1692
- . A propensity analysis of the risk of vascular complications after cardiac catheterization procedures with the use of vascular closure devices. Am Heart J. 2007;153:606–611
- Is carotid artery stenting in octogenarians really dangerous?. J Endovasc Ther. 2006;13:302–309
- . Carotid artery stenting in older patients: is age a risk factor for poor outcome?. J Endovasc Ther. 2002;9:559–565
- . Carotid artery stenting in octogenarians: the jury is still out. J Endovasc Ther. 2006;13:310–311
- The CAPTURE registry: results of carotid stenting with embolic protection in the post approval setting. Catheter Cardiovasc Interv. 2007;69:341–348
- . Comparison of angioplasty and stenting with cerebral protection versus endarterectomy for treatment of internal carotid artery stenosis in elderly patients. J Vasc Surg. 2004;40:945–951
- Endarterectomy versus stenting in patients with symptomatic severe carotid stenosis. N Engl J Med. 2006;355:1660–1671
- 30 day results from the SPACE trial of stent-protected angioplasty versus carotid endarterectomy in symptomatic patients: a randomized noninferiority trial. Lancet. 2006;368:1239–1247
Competition of interest: none.
CME article
PII: S0741-5214(08)00205-X
doi:10.1016/j.jvs.2008.02.005
© 2008 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.