Endovascular Treatment of Acute Carotid Stent Occlusion: Aspiration Thrombectomy and Angioplasty

Introduction Acute carotid stent occlusion (CSO) is a rare complication of endovascular carotid stent placement that requires emergent intervention. We describe angioplasty or combined angioplasty and aspiration thrombectomy as a new endovascular technique for CSO treatment. The technique is compared to others previously described in the literature. Methods We performed a retrospective cohort study of all patients who underwent endovascular treatment (ET) of acute symptomatic CSO from January 2008 to March 2018 at our neurovascular referral center. Patient demographics, endovascular treatment details, and outcome data were determined from the electronic medical record. Primary outcome was successful stent recanalization and cerebral reperfusion (modified thrombolysis in cerebral infarction (mTICI) score IIB-III). Secondary outcomes were National Institutes of Health Stroke Scale (NIHSS) shift from presentation to discharge, mortality, and modified Rankin Scale (mRS) score at 3 months. Additionally, a literature review (years 2008-2019) was performed to characterize other techniques for ET of CSO. Results Four patients who underwent ET of acute CSO were identified. ET treatment by angioplasty (n = 1) or combined aspiration thrombectomy and angioplasty (n = 3) resulted in carotid stent recanalization in all patients. Tandem intracranial occlusions were present in three patients (75%), and successful cerebral reperfusion was achieved in all patients. Patient symptoms improved (mean NIHSS shift -5.3 ± 7.2 at discharge). One patient died of a symptomatic reperfusion hemorrhage and another died of cardiac complications by 3-month follow-up. The mRS scores of the surviving patients were 1 and 3. Previously described studies (n = 14) using different and varied techniques had moderate recanalization rates and outcomes. Conclusion Combined aspiration thrombectomy and angioplasty for the neurointerventional treatment of acute CSO leads to high rates of stent recanalization and cerebral reperfusion. The recanalization rate here is improved compared to previously reported techniques. Further multicenter studies are required to risk-stratify patients for specific ET interventions.


Introduction
Acute carotid stent occlusion (CSO) is a rare complication of endovascular carotid stent placement that requires emergent intervention. We describe angioplasty or combined angioplasty and aspiration thrombectomy as a new endovascular technique for CSO treatment. The technique is compared to others previously described in the literature.

Methods
We performed a retrospective cohort study of all patients who underwent endovascular treatment (ET) of acute symptomatic CSO from January 2008 to March 2018 at our neurovascular referral center. Patient demographics, endovascular treatment details, and outcome data were determined from the electronic medical record. Primary outcome was successful stent recanalization and cerebral reperfusion (modified thrombolysis in cerebral infarction (mTICI) score IIB-III). Secondary outcomes were National Institutes of Health Stroke Scale (NIHSS) shift from presentation to discharge, mortality, and modified Rankin Scale (mRS) score at 3 months. Additionally, a literature review (years 2008-2019) was performed to characterize other techniques for ET of CSO.

Results
Four patients who underwent ET of acute CSO were identified. ET treatment by angioplasty (n = 1) or combined aspiration thrombectomy and angioplasty (n = 3) resulted in carotid stent recanalization in all patients. Tandem intracranial occlusions were present in three patients (75%), and successful cerebral reperfusion was achieved in all patients. Patient symptoms improved (mean NIHSS shift -5.3 ± 7.2 at discharge). One patient died of a symptomatic reperfusion hemorrhage and another died of cardiac complications by 3-month follow-up. The mRS scores of the surviving patients were 1 and 3. Previously described studies (n = 14) using different and varied techniques had moderate recanalization rates and outcomes.

Conclusion
Combined aspiration thrombectomy and angioplasty for the neurointerventional treatment of acute CSO leads to high rates of stent recanalization and cerebral reperfusion. The recanalization rate here is improved compared to previously reported techniques. Further multicenter studies are required to riskstratify patients for specific ET interventions.

Introduction
Acute carotid stent occlusion (CSO) is a rare cause of acute ischemic stroke that is associated with significant morbidity and mortality [1]. CSO occurs in 0.05% to 0.8% of patients with the internal carotid artery (ICA) or common carotid artery stents and is caused by antiplatelet medication noncompliance or discontinuation, antiplatelet medication resistance, overlapping stent placement, or intrinsic prothrombotic disorders [1][2][3][4]. In addition, procedural events and complications, such as dissection, atheroma perturbation, or ICA kinking after stent placement, may predispose a stent to occlusion [5].
We present the first report of combined angioplasty and aspiration thrombectomy for the treatment of acute CSO. All patients underwent angioplasty to promote thrombus disruption and restore antegrade flow through the occluded carotid stent. Residual in-stent thrombus was removed using aspiration thrombectomy. This technique is described in detail, and its effectiveness is compared to the literature for ET of CSO.

Patient information
The study was approved by the Institutional Review Board (IRB) and complied with the Health Insurance Portability and Accountability Act. The need for informed consent was waived the IRB. We retrospectively reviewed our neurointerventional database to identify consecutive patients who underwent ET for acute CSO treatment between January 2008 and March 2018. Patient demographics, endovascular treatment details, and outcome data were determined from the electronic medical record.
Among patients who underwent pre-interventional perfusion imaging, automated post-processing was performed using RApid processing of PerfusIon and Diffusion (RAPID) software (iSchemaView, Menlo Park, CA). Core infarct and penumbral volumes (defined as the volume of tissue with time-to-maximum (Tmax) >6 seconds) were determined using RAPID. Patients had pre-interventional computed tomography angiography (CTA) or magnetic resonance angiography (MRA). In one patient, the CTA was non-diagnostic due to technical issues. In this patient, a virtual CT angiogram was reconstructed from the perfusion source images.
All patients who undergo carotid stent placement at our institution undergo surveillance CTA at 3, 6, and 12 months to evaluate for in-stent stenosis. However, patients who present with CSO before these follow-up appointments were not screened for an in-stent stenosis. Therefore, only one patient in this series underwent follow-up imaging due to delayed presentation (patient four). Antiplatelet resistance in patients with verified medication compliance, and no other identified cause of CSO was verified using thromboelastography, with secondary testing by a hemostasis platelet function assay-100 (PFA-100) system (Siemens, Tarrytown, NY).

Endovascular treatment of CSO
All patients underwent ET in a biplane neuroangiography suite (Axiom Artis, Siemens) under either monitored anesthesia care or general anesthesia. Common femoral artery access was obtained using standard techniques, and an 8-or 9-French sheath was placed in the descending thoracic aorta. Access into the common carotid artery of the affected hemisphere was obtained with a 5-French Berenstein angiographic catheter (Cordis, Milpitas, CA), which was placed through a 6-French shuttle sheath ( The affected common carotid artery was accessed ( Figure 1A), and an aspiration catheter was advanced and withdrawn through the thrombosed stent into the more distal cervical ICA under continuous aspiration to at least partially recanalize the CSO ( Figure 1B). Next, an embolic protection device (Accunet, Abbott, Abbott Park, IL) was placed distal to the stent over a guidewire ( Figure 1C) in three patients. The embolic protection device could not be navigated through the partially occluded stent in the fourth patient. A balloon catheter was then advanced within the partially recanalized stent and stepwise angioplasty was performed to macerate the residual thrombus against the stent wall ( Figure 1C). Different balloon catheters were used for each patient: a 4x12mm monorail balloon catheter (Boston Scientific, Marlborough, MA), a non-compliant Trek 5x15mm balloon microcatheter (Abbott, Abbott Park, IL), a Trek rapid exchange 5x12mm balloon catheter (Abbott, Abbott Park, IL), and a Viatrac 6x20mm balloon catheter (Abbott, Abbott Park, IL). Post-angioplasty DSA was performed demonstrating improved stent caliber with minimal residual thrombus ( Figure 1D).

FIGURE 1: Schematic of endovascular treatment of CSO
Schematic for endovascular recanalization of CSO. After obtaining access to the affected common carotid artery (A), an aspiration catheter is advanced through the thrombosed stent under continuous aspiration and removed (B). An embolic protection device is placed distal to the stent, and angioplasty is performed to macerate the residual thrombus against the stent wall (C). Post-angioplasty DSA demonstrates stent recanalization with minimal residual thrombus (D).
CSO, carotid stent occlusion; DSA, digital subtraction angiography Tandem intracranial occlusions were treated with a combination of IA-tPA, aspiration, and mechanical thrombectomy. Detailed ET descriptions for each patient are presented in the Appendix Text.

Technique effect and outcome metrics
The primary outcome was successful CSO recanalization and cerebral vascular reperfusion (modified thrombolysis in cerebral infarction, (mTICI) score IIB-III). Secondary outcome measures were National Institutes of Health Stroke Scale (NIHSS) shift from presentation to discharge, modified Rankin Scale (mRS) score at 3 months and mortality [15][16].

Literature review
A comprehensive review of PubMed and Embase was performed to identify all eligible acute CSO treatment studies published since January 2008, with the last update of literature review on July 2019. Search terms used included: "carotid stent" AND "occlusion", "carotid artery stent" AND "occlusion", "carotid stent" AND "thrombosis", "carotid artery stent" AND "thrombosis", or "carotid" AND "stent" AND "occlusion". Inclusion criteria were: English language, acute onset of CSO (i.e. presentation of acute stroke symptoms leading to discovery of the CSO) irrespective from time of original stent placement, and publication dates January 2008 -July 2019. In total, 65 non-duplicate studies were identified and 51 were excluded due to: absence of complete CSO, non-acute CSO, no surgical or ET intervention performed or described, incomplete or absent characterization of CSO onset, etiology, and/or outcome ( Figure 2). Two authors reviewed each study that met all inclusion criteria and did not meet exclusion criteria.

Endovascular treatment of CSO
In our institutional neurointerventional database, we identified four patients who underwent ET for symptomatic acute CSO. Patient demographic and treatment data are described in Table 1. The mean presentation NIHSS score was 14.5 ± 3.9. Three patients underwent perfusion imaging prior to treatment, and, in these patients, the mean infarct core was 0.7 ± 0.7 ml and penumbra (Tmax > 6) was 110.5 ± 59.7 ml. Three patients were treated with aspiration thrombectomy and angioplasty, and the fourth underwent angioplasty alone for CSO treatment. these patients had evidence of these occlusions on pre-ET imaging ( Figure 3, Appendix Figure 6). All intracranial occlusions were successfully revascularized with mTICI IIB-III reperfusion (100%; Table 1).     Following ET, one patient expired after a cerebral reperfusion hemorrhage. At discharge, the mean NIHSS shift in the remaining three patients was -5.3 ± 7.2. By three months after ET, one additional patient had expired from a myocardial infarction, and the two surviving patients had 3-month mRS scores of 1 and 3 ( Table 1).

Literature review of acute CSO treatment
Acute CSO is most commonly secondary to antiplatelet resistance, intolerance, or noncompliance ( Table  2). Previously described ET techniques for CSO treatment apply methods used for non-stented carotid occlusions, and include IA thrombolysis, IA aspiration, heparinization, and mechanical thrombectomy [5,[8][9][10][11][12]14,[17][18][19][20][21][22][23][24]. A total of 23 patients were described, with most achieving recanalization (approximately 85-90%). However, outcomes reported range from a general single classification of either improved or not improved, or a spectrum of significant new functional deficits ( Table 2). The nonuniform reporting of TICI score, mRS, and mortality data limits precise comparative analytics between the studies and their techniques.  Twenty-three patients were described; however, the success rates of ET as measured by mTICI score nor mRS are not uniformly reported.

Discussion
We describe a novel method of effective ET for symptomatic CSO using a combined aspiration thrombectomy and angioplasty technique. Our technique resulted in the successful recanalization of the CSO in 100% of patients described here and has become the standard intervention for CSO at our institution.
The etiology of CSO in our series was due to inadequate medical platelet inhibition secondary to poor medication compliance, intrinsic antiplatelet resistance, or unintentional anticoagulation reversal, which is similar to prior studies [1][2][3][4]. These etiologies most likely led to acute in-stent thrombus formation, which was amenable treatment using our technique. Furthermore, the variable timing of CSO in our series suggests that even carotid stents that are likely well endothelialized are at risk of CSO in the absence of adequate antiplatelet or anticoagulation protection.
Aspiration thrombectomy for CSO has been described as an effective treatment [12,23]. However, we found aspiration thrombectomy alone resulted in insufficient CSO recanalization in all three patients who underwent this technique before angioplasty. In our cohort, aspiration thrombectomy created a channel through the in-stent thrombus that allowed for the passage of a balloon catheter for subsequent angioplasty.
Combining other ET techniques with aspiration thrombectomy may have a good effect on full recanalization. Angioplasty for acute CSO has not been described, and we have applied this technique as an additional revascularization technique. Angioplasty is often reserved for placing and re-expanding stents that were not already fully secured and were associated with a thrombus [5,14]. Alternative techniques that combine mechanical and aspiration thrombectomy for CSO have been described as similarly effective [10][11]. However, pulling a stent retriever through a recently placed, nonendothelialized carotid stent may risk stent retriever detachment from the pusher wire.
An in-stent stenosis secondary to intimal hyperplasia may predispose to CSO. However, in our series, there was an easy passage of the balloon microcatheter and aspiration catheter through the stent and unremarkable serial surveillance CTA and CT perfusion scans, which suggests an absence of significant in-stent stenosis. In 75% of patients in our series, an embolic protection device was deployed to reduce the theoretical risk of secondary thromboembolism during angioplasty, though no debris was observed within the embolic protection devices at procedure end. Protection against emboli from the stent may be obtained with the use of a balloon-guide catheter, although these catheters are not routinely used at our institution.
Surgical management of CSO by stent explantation, carotid endarterectomy, or carotid artery bypass grafting is considered an alternative in patients with verified dual anti-platelet resistance or as a rescue procedure for failed ET [6,22]. As endovascular techniques continue to improve, we anticipate that the need for surgical treatment of CSO will be further reduced.
The frequency of tandem intracranial occlusions in the setting of CSO remains poorly described, and we found tandem occlusions in 3 of 4 of our patients. Of these patients, 2 had evidence of these occlusions prior to ET, and the third patient developed new tandem occlusions within the A2 segment of the left anterior cerebral artery and distal M2 segment of the left middle cerebral artery despite the use of a distal embolic protection device (Appendix Figure 4). This new embolic occlusion may have occurred prior to, during placement of the embolic protection device, or following angioplasty. In the absence of frequent interval intracranial imaging during ET, it is challenging to conclude if a component of the ET itself was causal for the distal emboli.
Tandem intracranial occlusions were successfully revascularized in all patients in our series using standard endovascular techniques [18,28]. Of note, IA-tPA is most commonly used for acute occlusions at the time of stent placement, for which it is an effective treatment [9]. IA-tPA was used in two patients in our series beyond six hours' time since last seen normal. The clinical benefit of IA-tPA for CSO treatment before or after six hours since last seen normal remains unclear, as in other types of stroke from large vessel occlusion in the form of carotid occlusion, it may reduce mortality but not change functional outcome [5,[17][18].
A single patient in our series experienced a fatal reperfusion hemorrhage, which is a risk inherent to all endovascular reperfusion therapies. Our series did note a high 50% mortality rate at 3 months, which partially reflects both the symptomatic nature of the CSO and the severe medical co-morbidities of our cohort. The odds of a poor outcome or death would likely be higher if the recanalization of the occluded stents was not achieved. Non-intervention for CSO has not been studied, but it is likely to result in poor outcomes; comparison to patients who do not undergo thrombectomy for large vessel occlusions supports this hypothesis [29]. A recent systematic literature review found in nearly 60% of patients treated for CSO, that there was either no improvement after therapy or outcomes were not reported [30]. The same is true for nonuniform reporting of outcomes in the literature reviewed here. Notably, some studies that reported more favorable outcomes characterized patients with an asymptomatic CSO.
No procedural complications or damage to the previously placed carotid stents occurred in our series. A meta-analysis or multicenter experience will likely be required to definitively describe the procedural risk of ET for CSO given the rarity of this event.

Limitations
Our study is limited by its retrospective design, single-center experience, and small sample size. Despite these limitations, this series is, to our knowledge, the largest reported series of effective endovascular treatment of CSO. While variability in the time of presentation, and likely varying degrees of stent endothelization exist within our cohort, it is representative of the spectrum of onset for CSO. We find that the heterogeneity in presentation reflects the reality of CSO occurrence and increases the generalizability of our findings to most patients with an acute CSO.

Conclusions
Combined aspiration thrombectomy and angioplasty is a viable technique for the neurointerventional treatment of CSO and results in high rates of stent recanalization and cerebral reperfusion, which are both improved from that of previous techniques. Patient symptoms as measured by mean NIHSS shift also improved; however, the severe non-neurological comorbidities within our cohort led to high mortality by 3-month follow-up. Further multicenter studies are required to risk-stratify patients for specific ET interventions.

Patient 1: Left Acute CSO Seven Days Post Stent Placement
A proximal left ICA stent was placed for severe symptomatic stenosis prior to coronary artery bypass surgery. Dual antiplatelet (DAPT) therapy with aspirin 81mg and clopidogrel 75mg daily was started one week prior to stent placement. Seven days after stent placement, the patient developed acute right-sided hemiparesis (NIHSS 4). The patient was evaluated at 0.5 hours after symptom onset, but the patient was not an IV-tPA candidate due to recent cardiac surgery. CT angiography and perfusion (CTA/P) demonstrated acute left CSO, no intracranial large vessel occlusion, a small core infarction, and significant tissue at risk of infarction ( Figure 3). The patient underwent ET.
Left common carotid artery access was obtained as described. Next, a 5-Max-Ace catheter (Penumbra, Alameda, CA) was advanced through the thrombosed stent under continuous aspiration and into the more distal cervical left ICA. Moderate resistance to 5-Max-Ace advancement was appreciated during this aspiration thrombectomy. The 5-Max-Ace was removed under continuous aspiration. DSA demonstrated partial stent recanalization with residual non-occlusive thrombus in the mid-section of the stent. Intracranial angiography following aspiration thrombectomy was not performed. Next, an Accunet distal embolic protection device (Cook Medical, Bloomington, IN) was placed in the left ICA distal to the stent. A 4x12mm monorail balloon catheter (Boston Scientific, Marlborough, MA) was advanced over the guidewire into the stent such that it bridged the area of residual thrombus, and angioplasty was performed to macerate the residual thrombus against the wall of the stent. Post angioplasty DSA demonstrated improved stent caliber with a minimal residual internal filling defect. No tandem intracranial occlusion was present. The Accunet was removed, and the procedure was concluded. ET for this patient is shown in Figure 3.
The retrospective chart review demonstrated inconsistent clopidogrel administration. Subsequent platelet mapping studies after therapeutic clopidogrel reloading demonstrated moderate aspirin resistance (arachidonic acid, 59.8% inhibition) and significant clopidogrel resistance (adenosine diphosphate, ADP, 2.8% inhibition). The DAPT regimen was changed to aspirin 81mg and prasugrel 10mg daily, and the patient had no further complications. Three months following ET, the patient's NIHSS was 0 and mRS score was 1.

Patient 2: Right Acute CSO Eight Hours Post Stent Placement with Tandem Middle Cerebral Artery Occlusion
A right ICA stent was placed for symptomatic stenosis in a patient with atrial fibrillation. The patient's coumadin was stopped two days prior to stent placement, with subsequent loading with clopidogrel 300mg and maintenance clopidogrel 75mg daily. Following stent placement, procedural heparinization was reversed with 100mg protamine to facilitate groin closure. Eight hours after stent placement, the patient had acute onset of left-sided hemiparesis and neglect (NIHSS 13). Virtual CT angiography (derived from CT perfusion images) revealed tandem occlusions of the right inferior M2 and A2 segments of the middle and anterior cerebral arteries. CT perfusion images were consistent with right CSO, no detectable core infarction, and 222 mL of tissue at risk (Tmax >6 sec, Figure 4). The time from deficit onset to the presentation was 0.25 hours; however, IV-tPA was deferred given multiple lower extremity vascular surgeries in the prior 24 hours. The patient underwent ET.
Right common carotid artery access was obtained with a 6-French shuttle sheath as described.
Aspiration thrombectomy of the CSO was performed with an ACE68 catheter (Penumbra, Alameda, CA). Post-aspiration DSA demonstrated partial stent recanalization without antegrade filling of the cerebral circulation. An attempt was made to place an Emboshield embolic protection device (Abbott, Abbott Park, IL) through the partially occluded stent, but the device could not be navigated through the stent due to resistance. Therefore, the stent was crossed with a Velocity microcatheter (Penumbra, Alameda, CA) over a Traxcess microwire (MicroVention, Tustin, CA) with a docking wire extension (MicroVention, Tustin, CA). The microwire was positioned in the right petrous ICA, and the microcatheter was exchanged for a non-compliant Trek 5x15mm balloon microcatheter (Abbott, Abbott Park, IL). Stepwise angioplasty of the stent was performed. Post-angioplasty DSA demonstrated improved stent and cervical ICA recanalization with residual partially occlusive thrombus in the cervical ICA and tandem right inferior M2 segment occlusion. Repeat stent and cervical ICA angioplasty plus aspiration thrombectomy was performed with subsequent complete recanalization of the cervical vessels.
In order to treat the inferior M2 segment occlusion without repeated crossing of the carotid stent, the ACE68 catheter was again advanced through the carotid stent, and the 6-French shuttle sheath was then advanced over the ACE68 such that it was positioned distal to the carotid stent. The inferior M2 segment occlusion was then treated by combined aspiration and mechanical thrombectomy using a 4x20mm Solitaire (Medtronic, Irvine, CA) and the ACE68. mTICI III reperfusion of the right MCA was achieved after two passes. ET for this patient is shown in Figure 4.
A thorough case review suggested that the most likely precipitant of the patient's CSO was intravenous protamine administration for heparin reversal in this patient with known atrial fibrillation. Platelet mapping studies demonstrated an appropriate response to aspirin and clopidogrel. The patient was discharged with an NIHSS of 3 on clopidogrel 75mg daily and coumadin for concomitant atrial fibrillation. Three months after ET, the patient was walking independently and had improved to an mRS score of 3. However, the patient subsequently developed heart failure, a urinary tract infection, acute kidney injury, and severe toxic metabolic encephalopathy and died 3.2 months after ET.