Incidence and Outcome of Filter Occlusion during Carotid Artery Stent Procedure
Article Outline
Recent reviews of device-specific complications using neuroprotection have addressed technical difficulties during delivery as well as adverse outcomes, intraoperative and 30-day. Little has been written, however, regarding the relevance of filter occlusion during the carotid stent procedure. A retrospective review was conducted of patients undergoing carotid artery stent procedures using a variety of neuroprotection devices from 2003 to 2007. Prospective databases from two institutions were examined for incidence and management of filter occlusions during procedures as well as adverse neurological events (intraoperative and 30-day) associated with filter occlusion. There were 283 carotid artery stent procedures performed on 256 patients (163 male, 93 female): 177 (62.5%) arteries were asymptomatic and 106 were symptomatic. Neurological adverse events occurred in six patients (2.1%); three of these resolved completely at 72 hr. Neuroprotection was used in 95% of all patients, and filters were used in 221 stent procedures: Boston Scientific Filter Wire (n = 81), Guidant Accunet (n = 100), Angioguard (n = 17), and Abbot Emboshield (n = 23). Filter occlusion occurred in 11 patients (4.9%) in whom this form of neuroprotection was employed: Angioguard (n = 5), Accunet (n = 2), Emboshield (n = 2), and EPI Filter wire (n = 2). Two of the 11 patients with filter occlusions suffered a neurological event. There was no correlation between filter occlusion and gender, symptoms, stent, or filter type (p > 0.05). Filter occlusion was managed with export catheter-directed aspiration in seven patients and with prompt filter retrieval in five patients. Filter occlusion is an infrequent event that does not appear to be filter-specific and can be managed successfully by catheter-directed aspiration or filter retrieval. The majority of patients with filter occlusion do not suffer from atheroemboli as a result of this occlusion.
Introduction
Carotid artery angioplasty and stenting (CAS) has gained substantial momentum in part due to technological advancement. Tapered stent designs, lower profile delivery system, and the development of embolic protection devices (EPDs) have all contributed to the evolution of CAS by allowing for easier and safer delivery and deployment of the stent as well as improving neuroprotection techniques. The growing concern over the potential of atheroemboli during CAS has led to the development and widespread use of EPDs.1 In fact, prospective trials suggest the use of EPDs is associated with a low procedural complication rate.2 There are currently three types of EPDs in use: wire-based filters, proximal balloon occlusion with flow reversal, and distal balloon occlusion.
Despite the absence of level-one data, there is recent literature, including a large meta-analysis, that suggests that EPDs are associated with decreased perioperative adverse events.3 Furthermore, there appears to be no difference in carotid stenting outcome between any of the three specific EPDs available.2, 4 Apart from all the specific EPD data, there is paucity in the literature regarding the relevance of filter occlusion during CAS. A review of 453 carotid interventions by Casserly et al.5 found a 9.3% (n = 42) incidence of “slow-flow” phenomenon. As the majority of slow-flow events occurred post-stent angioplasty, the authors proposed that microembolization contributes to slow flow and related stroke.
Our study examines a prospective database of patients who underwent CAS and its incidence and outcomes of embolic filter-related occlusion.
Methods
Data Collection
Data for the present analysis were obtained from an institutional CAS database compiled by two participating centers. Data were prospectively collected from all 283 carotid stents, including reoperations, performed on 256 patients at the two institutions between 2003 and 2007 at New York University School of Medicine and Mt. sinal University School of Medicine.
Patient data included patient demographics, preoperative neurological symptoms, and postoperative neurological outcome. Operative data consisted of operative side, percent stenosis, presence of bovine anatomy, type of stent and cerebral protection device, presence of filter occlusion, and management of filter occlusion.
Operative Procedure
All patients were screened with preoperative duplex ultrasound and either computed tomographic angiography or magnetic resonance angiography to identify lesions suitable for treatment. CAS was performed by gaining access to the common carotid artery with either a 6F or 7F sheath and delivering an EPD according to instructions for use. The decision for pre- and poststent angioplasty was user-dependent. Predilatation was generally performed only if a stenosis was considered too severe to deliver the stent easily. In general, postdilatation was performed for residual stenosis >30%. The choice of EPD was also user-dependent. Filter protection was utilized only if the distal internal carotid artery permitted an adequate zone for deployment of the filter (i.e., straight portion >3-4 cm).
All patients were treated perioperatively with Clopidogrel (Plavix) antiplatelet therapy and for a minimum of 30 days postoperatively. All patients underwent preoperative, perioperative, and 30-day neurological evaluations by either neurologists (if enrolled in a clinical trial) or treating physicians. Carotid stent and cerebral protection device used were determined by the operating physician and subject to clinical trial requirements.
Outcomes
Filter occlusions were identified as either slow-flow or complete occlusion and analyzed by filter type. Management of occlusions was by either catheter-directed aspiration or prompt filter retrieval. Retrieved filters were examined for presence of clot and/or debris. Adverse events were defined as death, transient ischemic attack (TIA, neurological deficit lasting <24 hr), and stroke (neurological deficit lasting >24 hr). Events were chronologically identified as intraoperative or 30-day (defined as occurring during the procedure and up to 30 days after).
Statistics
Statistical analysis was conducted using SPSS (SPSS, Inc., Chicago, IL). p < 0.05 was deemed statistically significant. Primary analysis compared the incidence of filter occlusion with regard to gender, presence of preoperative symptoms, type of stent, or type of filter. Secondary analysis compared the adverse event rates between patients with and without an intraoperative filter occlusion. Comparison of filter occlusion rates with the primary analysis was accomplished using logistic regression. Comparison of event rates between different groups was based on a two-sided χ2 test as in the secondary analysis. Risk ratios were not calculated due to the small sample size.
Results
There were 283 CAS procedures performed on 256 patients (163 male, 93 female) from 2003 to 2007. Average age was 72.4 (±9) years, and 177 (62.5%) of the procedures were performed for asymptomatic stenosis. Adverse neurological events occurred in six patients overall (2.1%). Three were permanent and three resolved at 72 hr. Neuroprotection was used in 95% of all patients: filters (n = 221), distal balloon PercuSurg GuardWire (n = 49), and reversal of flow (n = 4). The most frequently used filters were Accunet (n = 100, 45%), EPI FilterWire (n = 81, 37%), Emboshield (n = 23, 10%), and Angioguard (n = 17, 8%). Stents used included Acculink (n = 132, 46%), Wall stent (n = 67, 24%), Xact (n = 28, 10%), Precise (n = 24, 8%), and Nexstent (n = 20, 8%) (Table I).
Table I. Currently available EPDs—filter type8
| Device manufacturer | Filter size (mm) | Crossing profile | Pore size (μm) | Status |
|---|---|---|---|---|
| Rx Accunet (Abbott Vascular, Abbott Park, IL) | 4.5-7.5 | 3.5-3.7F | 150 | FDA-approved |
| EmboShield (Abbott Vascular) | 3-6 | 2.9-3.3F | 140 | FDA-approved |
| FilterWire EZ, EX (Boston Scientific, Watertown, MA) | 3.5-5.5 mm vessel | 3.2F | 110 | Outside US |
| Rubicon Filter (Boston Scientific) | Outside US | |||
| SpidRx (ev3, Plymouth, MN) | 3-7 | 3.2F | 36 | Outside US |
| Fiber Net (Lumen Biomedical, Plymouth, MN) | 1.75-7 mm vessel | 1.7-3.2F | 40 | Outside US |
| Interceptor PLUS (Medtronic, Sunnyvale, CA) | 4.25-6.25 mm vessel | 3.6F | 100 | Outside US |
Filter Occlusions
Filter occlusions occurred during 11 carotid stent procedures when a filter EPD was used (4.9%). This represents the study cohort. The majority of occlusions occurred when using the Angioguard filter (n = 5, 46%). There was no statistical correlation between filter type, stent type, gender, presence of symptoms, and occlusion (p > 0.05) (Fig. 1). Two of 11 filter occlusions (one Angioguard, one Accunet) suffered an adverse neurological event (18%, p = 0.2). Eight of 11 (72%) filter occlusions occurred following poststent angioplasty. The remaining three occurred following stent deployment. The latter group of patients did not undergo poststent angioplasty but underwent prompt filter retrieval.
Fig. 1
Filter occlusions occurred during 11 CAS procedures when a filter EPD was used (4.9%). There was no statistical correlation between filter type, stent type, gender, presence of symptoms, and occlusion (p > 0.05).
One patient had a permanent neurological deficit when the filter occluded subsequent to poststent angioplasty for a 30% residual stenosis. Flow was sluggish and then ceased completely. The filter was promptly retrieved without attempting aspiration. A second patient who occluded the filter with complete cessation of flow underwent catheter-directed aspiration. This patient had a transient neurological deficit, which resolved in 24 hr. Filter occlusion was not associated with increased risk of a neurological event (p = 0.2).
Management of Filter Occlusion
Filter occlusions were managed with either export catheter-directed aspiration (n = 6) or prompt filter retrieval without aspiration (n = 5). Aspirate was then irrigated onto a PercuSurg filter for inspection. Decision as to management of filter occlusion was physician-dependent. An example of a filter occlusion is depicted in Figure 2. Only two occluded filters had evidence of visible debris.
Fig. 2
CAS procedure showing filter occlusion. a Asymptomatic severe internal carotid artery stenosis. b EPD deployed with sluggish flow. c Complete filter occlusion with stent deployed. d Restoration of brisk flow through filter following catheter-directed aspiration.
Discussion
Early critics of CAS cautioned that traversing a carotid plaque with wires, catheters, balloons, and stents would result in inevitable distal embolization and thus prohibitive neurological adverse events. Indeed, using bench-top flow models consisting of ex vivo carotid plaque treated with angioplasty and stent, Ohki et al.6 demonstrated that all specimens produced significant emboli. Echolucent (heterogeneous) plaque morphology as well as high-grade stenoses appeared to confer greatest risk for emboli. Transcranial Doppler studies have confirmed the phenomenon of diffuse cerebral embolization during CAS. Jordan et al.7 examined a series of 105 patients (115 bifurcations) with transcranial Doppler (TCD) during carotid endarterectomy (n = 75) as well as during unprotected CAS procedures (n = 40). They found eight times the number of high-intensity transient signals consistent with emboli occurring during CAS compared to endarterectomy. This and other studies supported the advent of EPDs consisting of either internal carotid artery filter or balloon occlusion (with or without reversal of flow).
A meta-analysis by Kastrup et al.8 included a total of 2,537 CAS procedures performed without EPD and 896 procedures using cerebral protection devices (CPDs). The combined stroke/death rate at 30 days was significantly lower in patients where CPD was used (1.8% vs. 5.5%, p < 0.001). While no randomized trials exist comparing CAS with and without neuroprotection, their use has become the standard of care when performing CAS procedures in the United States. Indeed, a selection of EPDs including distal balloon occlusion, filters, and flow reversal has quickly become available in a short time; each carries its unique advantages and disadvantages.9
Filter wires are perhaps the most commonly used of the three types of EPD. Currently, there are eight commercially available filter devices (Table I). Unlike distal balloon occlusion EPDs (e.g., PercuSurge guard wire, TriActive Proguard, and MoMa), the filter wires allow for continual antegrade flow to the brain throughout the procedure. Angiography can be performed with filter protection in place. Moreover, patients with an incomplete circle of Willis who would otherwise be intolerant to interruption of flow can be treated with a filter EPD. Disadvantages of the filter EPD include a large crossing profile, risk of crossing the lesion unprotected first, spasm and/or dissection, incomplete filter wall opposition, and filter occlusion due to thrombosis or debris burden.
There are a few data comparing outcomes when using one EPD or stent type with another. Rubartelli et al.10 compared 31 patients undergoing TCD during CAS with either distal balloon occlusion (GuardWire) or filter EPD (FilterWire EX device) and found the total amount of microembolic signals during the procedure was higher when the filter device was employed (183.0 ± 42.1 vs. 31.7 ± 12.0, p < 0.0001). A larger retrospective study by Iyer et al.2 reviewed 3,160 CAS procedures using nine EPDs and found no significant difference in procedural adverse events between any of the EPDs when compared to the most commonly used filter wire. Moreover, there was no significant difference in risk between eccentric and concentric filters. Others, however, have found that concentric filters may be associated with a greater risk of neurological adverse events at 30 days.11 In a study of 701 consecutive CAS procedures, Hart et al.11 found that the symptomatic subgroup (n = 304) with an open cell stent design and use of concentric filter EPD yielded odds ratios of 4.1 (95% confidence interval [CI] 1.4-12, p = 0.0136) and 3.3 (95% CI 1.016-10, p = 0.0525), respectively, for 30-day stroke/death/TIA.
Although the relative influence of EPD type on outcome for CAS remains unclear, filter occlusion itself appears to confer an increased risk for complication.
Filter occlusions may be related to numerous factors unique to filter design and other patient-specific or anatomy-specific variables. A study by Powell et al.4 examined technical difficulties encountered during CAS procedures and found that filter occlusion occurred in two of 42 patients when a filter EPD was used (4.7%). In our study, the incidence of filter occlusion was similarly low, occurring in only 11 CAS procedures (4.5%). While the majority occurred when using the Angioguard filter (n = 5), there was not a statistically significant correlation between filter type and occlusion. Moreover, filter occlusion was not associated with an increased risk of neurological event in our series (p = 0.2). Others have reported filter occlusion to be associated with a significant increase in 30-day death and stroke. In their series of 42 patients with slow flow during carotid stent procedures, Casserly et al.5 found a 9.5% incidence of stroke in patients with slow flow vs. 1.7% in patients with normal flow (p = 03). Predictors of slow flow included a recent history (6 months) of stroke or TIA, increased stent diameter, and increased patient age.
We were unable to identify predictors for filter occlusion in our study. Moreover, the mechanism for filter occlusion is unclear. Eight of 11 filter occlusions (72%) occurred following poststent angioplasty, consistent with the theory that disrupted plaque may “toothpaste” into the lumen through the stent interstices following angioplasty. While the majority of the occluded filters were not full of large debris/thrombus burden, current techniques for visualizing debris are imperfect and microembolic material is likely to have contributed to occlusion.
Patients in whom filter occlusion occurred did not appear to have tortuous distal internal carotid arteries nor were the filters undersized relative to target vessel diameter. Gender and symptomatology were not predictive of filter occlusion. In addition, while intrinsic hypercoagulable conditions were not assessed for each patient, activated clotting times were consistently in the therapeutic range throughout all procedures. Finally, completion angiogram failed to show any evidence of spasm and/or dissection following filter retrieval in the occluded cohort. Alternatively, procedure time may play a role in filter occlusion. Unfortunately, operative records at our institutions did not contain these times. Intuitively, the longer the filter is in place and wires are across the lesion, the greater the risk of thrombus formation as well as spasm.
Management of filter occlusion and other related EPD complications during CAS is not well represented in the literature. Powell et al.4 reviewed 141 CAS procedures performed using either balloon occlusion or filter EPD and found that EPD-related technical difficulties occurred in 15% of the balloon group and 31% of the filter group (p < 0.05). As expected, the balloon had a 10% incidence of neurological compromise during occlusion vs. 0% for the filter group (p = 0.002), whereas the filter group was unable to cross the lesion before predilatation more often (12%) than the balloon group (0%) (p = 0.001). Spasm requiring pharmacological treatment occurred more frequently in the filter group (12%). Two filter occlusions in their study were managed by catheter-directed aspiration and 100 μg nitroglycerin administered via the sheath, with no adverse sequelae. In our study, 11 filter occlusions were managed with either aspiration of the filter basket with an export catheter (PercuSurg export catheter, n = 6) or prompt filter retrieval (n = 5). Two patients suffered neurological adverse events (one transient, one temporary) following filter occlusion. In one of these patients catheter-directed aspiration was not performed prior to filter retrieval.
Although management strategy (prompt filter retrieval vs. catheter-directed aspiration) was not associated with increased risk of adverse event in this study, the small numbers in our series must be emphasized as they prohibit us from making strict recommendations regarding management of filter occlusion. Patients who demonstrate complete cessation of flow are at particular risk since the clot burden is impossible to determine, making filter retrieval alone unsafe. Indeed, one of the two patients who suffered a stroke demonstrated complete cessation of flow and had the filter retrieved without catheter-directed aspiration. While it would appear that prompt filter retrieval was safe and successful in the patients who demonstrated slow flow in the absence of complete filter occlusion, sluggish flow may be a harbinger to complete occlusion. Thus, it may be most prudent to aspirate all filters in the setting of sluggish and/or absent flow, providing this can be done expeditiously.
In summary, while CAS has become a viable alternative to CEA in a select group of patients, the risk of embolization, despite neuroprotection, continues to be a significant concern. Filter occlusion is an infrequent event, which does not appear to be filter-specific and can be managed successfully by catheter-directed aspiration or filter retrieval in select patients. The majority of patients with filter occlusion do not suffer from atheroemboli as a result of this occlusion. Larger studies comparing filter type and design as well evaluating management strategies for coping with filter occlusions are warranted.
References
- Protected carotid-artery stenting versus endarterectomy in high-risk patients. N Engl J Med. 2004;351:1493–1501
- The type of embolic protection does not influence the outcome in carotid artery stenting. J Vasc Surg. 2007;46:251–256
- . Assessment of short-term outcomes for protected carotid angioplasty with stents using recent evidence. J Vasc Surg. 2005;42:1094–1100
- Comparison of embolization protection device-specific technical difficulties during carotid artery stenting. J Vasc Surg. 2006;44:56–61
- Slow-flow phenomenon during carotid artery intervention with embolic protection devices: predictors and clinical outcome. J Am Coll Cardiol. 2005;46:1466–1472
- Ex vivo human carotid artery bifurcation stenting: correlation of lesion characteristics with embolic potential. J Vasc Surg. 1998;27:463–471
- Microemboli detected by transcranial Doppler monitoring in patients during carotid angioplasty versus carotid endarterectomy. Cardiovasc Surg. 1999;7:33–38
- Early outcome of carotid angioplasty and stenting with and without cerebral protection devices: a systematic review of the literature. Stroke. 2003;34:813–819
- . Embolic protection devices for carotid artery stenting: have they made a significant difference in outcomes?. Semin Vasc Surg. 2007;20:244–251
- Transcranial Doppler monitoring during stenting of the carotid bifurcation: evaluation of two different distal protection devices in preventing embolization. J Endovasc Ther. 2006;13:436–442
- Do device characteristics impact outcome in carotid artery stenting?. J Vasc Surg. 2006;44:725–731
PII: S0890-5096(08)00302-6
doi:10.1016/j.avsg.2008.08.004
© 2008 Annals of Vascular Surgery Inc. Published by Elsevier Inc All rights reserved.
