Listen to this manuscript’s

audio summary by

Editor-in-Chief

Dr. Valentin Fuster on

JACC.org.

J O U R N A L O F T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y V O L . 7 4 , N O . 7 , 2 0 1 9

ª 2 0 1 9 T H E A U T H O R S . P U B L I S H E D B Y E L S E V I E R O N B E H A L F O F T H E A M E R I C A N

C O L L E G E O F C A R D I O L O G Y F OU N D A T I O N . T H I S I S A N O P E N A C C E S S A R T I C L E U N D E R

T H E C C B Y - N C - N D L I C E N S E ( h t t p : / / c r e a t i v e c o mm o n s . o r g / l i c e n s e s / b y - n c - n d / 4 . 0 / ) .

ORIGINAL INVESTIGATIONS

Permanent Percutaneous Carotid ArteryFilter to Prevent Stroke inAtrial Fibrillation Patients

The CAPTURE Trial

Vivek Y. Reddy, MD,a,b Petr Neuzil, MD, PHD,a Tom de Potter, MD,c Jan van der Heyden, MD,d Selma C. Tromp, MD,e

Benno Rensing, MD,d Eva Jiresova, MD,a Libor Dujka, MD,a Veronika Lekesova, MDa

ABSTRACT

ISS

Fro

Me

Ca

An

Me

00

to

BACKGROUND Patients with high stroke risk and atrial fibrillation who are unsuitable to oral anticoagulants (OACs)

require other stroke prevention strategies. A novel permanent coil filter directly placed into both common carotid arteries

(CCAs) was designed to capture emboli >1.4 mm in diameter.

OBJECTIVES The multicenter, nonrandomized, first-in-human clinical CAPTURE (Carotid Artery Implant for Trapping

Upstream Emboli for Preventing Stroke in Atrial Fibrillation Patients) trial sought to determine the feasibility and safety

of bilateral CCA filter placement.

METHODS Eligible patients had atrial fibrillation, CHA2DS2-VASc (Congestive heart failure, Hypertension, Age 75 years,

Diabetes, Stroke/transient ischemic attack, Vascular disease, Age 65 to 74 years, Sex category) $2, OAC unsuitability, CCA

size 4.8 to 9.8 mm, and no carotid stenosis>30%. Under ultrasound guidance, after direct transcutaneous carotid puncture

with a 24-gauge needle, a motorized unit expels the filter to unfurl in the artery. Patients received aspirin/clopidogrel for

3 months, and aspirin thereafter. Primary endpoints were: 1) procedural success—bilateral, properly positioned CCA filters;

and 2) 30-day incidence of major adverse events—death, stroke, major bleeding, filter migration, CCA thrombus, or stenosis.

Carotid ultrasounds were conducted post-procedure, pre-discharge, at 1 week, and at 1, 3, 6, and 12 months.

RESULTS At 3 centers, 25 patients were enrolled: age 71 � 9 years, CHA2DS2-VASc ¼ 4.4 � 1.0, prior embolism in 48%.

Procedure success was 92% (23 of 25 patients); 1 patient had unilateral deployment. There were no device/procedure-

related major adverse events; minor puncture site hematomas/edema occurred in 5 of 25 (20%). After 6-month mean

follow-up, asymptomatic thrombi were detected in 4 patients (1 bilateral, 4 unilateral), adjudicated as captured (n ¼ 3),

unclassified (n ¼ 2), or in situ (n ¼ 0). In all patients, the thrombi dissolved with subcutaneous heparin. In 1 patient,

2 device/procedure-unrelated minor strokes occurred.

CONCLUSIONS Permanent carotid filter placement for stroke prophylaxis is technically feasible and safe.

(Carotid Artery Implant for Trapping Upstream Emboli for Preventing Stroke in Atrial Fibrillation Patients [CAPTURE];

NCT03571789) (J Am Coll Cardiol 2019;74:829–39) © 2019 The Authors. Published by Elsevier on behalf of the

American College of Cardiology Foundation. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

N 0735-1097 https://doi.org/10.1016/j.jacc.2019.04.035

m the aDepartment of Cardiology, Homolka Hospital, Prague, Czech Republic; bDepartment of Cardiology, Icahn School of

dicine at Mount Sinai, New York, New York; cDepartment of Cardiology, OLV Ziekenhuis, Aalst, Belgium; dDepartment of

rdiology, Sint-Antonius Ziekenhuis, Nieuwegein, the Netherlands; and the eDepartment of Clinical Neurophysiology, Sint-

tonius Ziekenhuis, Nieuwegein, the Netherlands. This trial was supported by the manufacturer of the carotid coil filter, Javelin

dical Ltd. Drs. Neuzil, Jiresova, and Dujka are supported by a scientific grant from the Czech Ministry of Health (DRO NNH

023884 IG 180504). Dr. Reddy has served as a consultant for and has stock in Javelin Medical; has served as a consultant

Abbott, Acutus Medical, Affera, Apama Medical, Aquaheart, Autonomix, Axon, Backbeat, BioSig, Biotronik, Cardiofocus,

ABBR EV I A T I ON S

AND ACRONYMS

AF = atrial fibrillation

ASA = aspirin

CCA = common carotid artery

MAE = major adverse event

NOAC = nonwarfarin oral

anticoagulation

OAC = oral anticoagulation

SQ = subcutaneous

Cardionom

Farapulse,

medical, V

BioSig, Circ

LuxCath, M

Neuzil, de P

reported th

Manuscript

Reddy et al. J A C C V O L . 7 4 , N O . 7 , 2 0 1 9

Permanent Carotid Coil Filter for Stroke Prevention in AF A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9

830

P reventing stroke is arguably the mostimportant clinical management goalin the treatment of atrial fibrillation

(AF). In AF patients with CHA2DS2-VASc(Congestive heart failure, Hypertension,Age 75 years, Diabetes, Stroke/transientischemic attack, Vascular disease, Age 65 to74 years, Sex category) $2 not treated withoral anticoagulants (OAC), the averageannual stroke risk weighted by the observedrisk distribution is w5%/year (2% and 11%/

year in CHA2DS2-VASc 2 and 9, respectively) (1,2).Warfarin and nonwarfarin oral anticoagulants(NOACs) reduce this risk by approximately 65% (3).However, in high-risk patients (CHA2DS2-VASc $4 orhistory of stroke), the annual stroke risk despite tak-ing OACs remains excessive at 2% to 4% (4,5).Furthermore, AF-related stroke more often involvesthe anterior circulation, and it is typically more se-vere—17% versus 8% mortality in non-AF patients (6).

SEE PAGE 840

The pathogenesis of ischemic stroke in AF is mostoften an embolus emerging from the heart or largearteries and lodging into a cerebral artery. The size ofthe infarct and extent of brain damage are directlyrelated to the size of the embolus. In AF patients,approximately 80% of strokes are total or partialanterior circulation strokes (major strokes) caused byocclusions of the main branches of the carotid ar-teries, M1-2 and A1-2 (6). The diameter of thesebranches is typically >1.5 mm (7–9). Since an emboluswith a given size occludes an artery of similar orsmaller size, preventing emboli >1.4 mm fromreaching the cerebral anterior circulation may reducethe risk of major stroke.

Accordingly, a novel, permanent, bilateral pair ofcommon carotid artery (CCA) filters was developed tocapture such emboli >1.4 mm, thereby preventingmajor cerebral stroke. They are placed into the CCAsusing 24-gauge needles that directly puncture thevessel. This technology is intended for high-risk AFpatients who may or may not be receiving OAC. Inpreclinical in vitro and ovine studies: 1) the filter wasreproducibly implantable; 2) it was nonthrombogenic,

ic, CardioNXT/AFTx, Circa Scientific, Corvia Medical, East End M

Impulse Dynamics, Keystone Heart, LuxCath, Medlumics, Med

alcare, and VytronUS; and has equity in Acutus Medical, Affera,

a Scientific, Corvia Medical, East End Medical, EPD, Epix Thera

anual Surgical Sciences, Medlumics, Middlepeak, Newpace, Nu

otter, van der Heyden, Tromp, and Rensing have received grant

at they have no relationships relevant to the contents of this pap

received April 8, 2019; revised manuscript received April 29, 20

nonocclusive, and mechanically stable; 3) simulatedemboli exceeding 1.4 mm were captured; 4) autolo-gous thromboemboli did not disintegrate upon cap-ture; and 5) captured thromboemboli either did notprogress or completely regressed over a period of21 days (Yodfat et al., unpublished data). Herein, wereport the results from an ongoing first-in-humanclinical trial of the feasibility and safety of the CCAfilter, evaluated in AF patients unsuitable for OAC.

METHODS

TRIAL DESIGN. The CAPTURE (Carotid ArteryImplant for Trapping Upstream Emboli for PreventingStroke in Atrial Fibrillation Patients) (NCT03571789)clinical trial is a first-in-human, multicenter, non-randomized feasibility study of the CCA filter. An in-dependent data safety monitoring board oversaw thesafety of the trial, and an independent clinical eventscommittee adjudicated endpoint events (OnlineAppendix). The trial was funded by Javelin MedicalLtd., the developer of the CCA filter. The study wasconducted by Javelin with independent monitoringby a contract clinical research organization (GENAE,Antwerp, Belgium). The trials were approved by eachcenter’s local ethics committee and the correspond-ing national regulatory agencies.

STUDY POPULATION. Patients with persistent orpermanent AF (later amended to any AF) unsuitablefor OAC and CHA2DS2-VASc $4 (later amended toCHA2DS2-VASc $2) were included in the trial. Patientswith CCA atherosclerosis or carotid stenosis >30%were excluded. The detailed inclusion and exclusioncriteria provided in the protocol appear in the OnlineAppendix. Written informed consent was obtainedfrom all patients.

PERMANENT CAROTID COIL FILTER. The permanentcarotid coil filter (Vine, Javelin Medical Ltd.,Yokneam, Israel) is made from a super-elastic nitinolwire with a circular cross-section (diameter 240 mm).In the undeployed state, the filter assumes a sub-stantially linear shape within the lumen of a 24-gaugeinsertion needle. Upon deployment, the filter unfoldsinto a helix that resides within the CCA lumen,

edical, EBR, EPD, Epix Therapeutics, EpiEP, Eximo,

tronic, Middlepeak, Nuvera, Philips, Stimda, Ther-

Apama Medical, Aquaheart, Autonomix, Backbeat,

peutics, EpiEP, Eximo, Farapulse, Keystone Heart,

vera, Surecor, Valcare, Vizara, and VytronUS. Drs.

support from Javelin Medical. All other authors have

er to disclose.

19, accepted April 30, 2019.

FIGURE 1 The Carotid Coil Filter

Pulling Wire

Internal Anchor

Supporting Coils

External Anchor

FilteringPortion

FilteringPortion

Leading Coils Direction of Blood Flow

Leading coilsSupporting coils

A B

(A) The coil filter is shown with the various aspects of the filter including: 1) the supporting and leading coils; 2) the filtering portion, which is

the primary location where clots are captured; 3) the anchors that fix the device in position as the filter wire traverses the carotid arterial wall;

and 4) the pulling wire that can be used to remove the entire device if necessary, or is otherwise cut if placed in a good position. (B) The coil

filter is visualized by transcutaneous ultrasound in an ovine carotid artery (Yodfat et al., unpublished data).

J A C C V O L . 7 4 , N O . 7 , 2 0 1 9 Reddy et al.A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9 Permanent Carotid Coil Filter for Stroke Prevention in AF

831

connected to a linear stem that traverses the CCA walland 2 anchors—internal and external (Figures 1 and 2).The helix includes 3 segments: supporting coils,filtering portion, and leading coils. The filteringportion has the outline of a cone with the apex facingupstream. The distance between consecutive coils isapproximately 1.0 mm. The filter is inserted using theCCA coil Inserter (Figure 3) under ultrasound guid-ance: the 24-gauge insertion needle punctures theskin in the neck atop the CCA, and the filter is auto-matically deployed upon pressing the CCA Inserteroperating button. The filter may be retrieved up to 4 hfollowing implantation using a pulling wire that isconnected to the stem and extends outside the pa-tient’s skin. The filter is available in 11 sizes (0.5-mm

FIGURE 2 Illustration of the Carotid Coil Filter

Clot

Supporting Coils FilteringPortion

Leading Coils

Direction of Blood Flow

A

These schematics show the filter in both side (A) and front (B) views, a

increments) to accommodate CCA diameters between5 and 10 mm.PROCEDURAL WORKFLOW. Immediately prior to theprocedure, patients received low-molecular-weightheparin IV (25 to 50 U/kg incrementally until reach-ing activated clotting time >200 s) and clopidogrel(loading dose 600 mg). Following CCA diametermeasurement, a filter oversized with respect to theCCA by 0.2 to 0.7 mm was selected and was deployedunder local anesthesia with ultrasound guidance(Figure 3, Online Video 1). If proper deployment wasnot achieved, the filter was pulled out using thepulling wire and immediately replaced. Pressure wasthen applied for several minutes to prevent bleedingfrom the skin puncture site. The procedure was then

Internal AnchorExternal Anchor

Pulling WireB

s well as the location of where a thrombus may be captured.

FIGURE 3 The Carotid Coil Filter Insertion Device

Operating button

Needle

Ultrasoundprobe

Inserter

CCA

Insertionneedle

A

B C

(A) The inserter is shown pre-loaded with a common carotid artery (CCA) filter. Once the 24-gauge needle is inserted into the carotid artery,

the operating button is depressed to activate the motorized deployment of the coil into the vessel. (B) During implantation, an ultrasound

probe is used to guide direct CCA puncture. (C) On the transverse ultrasonic view of a human carotid artery, the inserter needle in shown in the

center of its lumen—the point at which the operating button is depressed to expel the CCA filter (Online Video 1).

Reddy et al. J A C C V O L . 7 4 , N O . 7 , 2 0 1 9

Permanent Carotid Coil Filter for Stroke Prevention in AF A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9

832

repeated for the contralateral CCA. Following bilat-eral deployment, proper position of implants wasverified by x-ray, the pulling wires were cut, and theskin was lifted over the stubs of these wires to “hide”the ends under the skin. Pressure bandages were thenapplied, and the patient was released to recovery. Thefirst 5 enrolled patients had staged implantationprocedures 4 weeks apart, with the filter implanted inthe right CCA during the first procedure and in the leftCCA during the second procedure.

POST-PROCEDURE AND FOLLOW-UP. Following theprocedure, patients were prescribed clopidogrel75 mg daily for 3 months and aspirin (ASA), 81 to100 mg daily for the duration of the study. Patientswere hospitalized overnight following the procedure.They had bilateral ultrasound imaging at 0.5 to 4 hfollowing the procedure. Ultrasound imaging andclinical and neurological follow-up were scheduled at1 day, 1 week, and 1, 3, 6, and 12 months followingthe procedure.

ENDPOINTS. The detailed list of endpoints of theCAPTURE study is presented in the Online Appendix.The primary safety endpoint, evaluated at 30 days,was device- or procedure-related major adverse

events (MAE), defined as the composite of death,stroke, major bleeding, CCA stenosis >70%, filtermigration, CCA thrombus, or any CCA complicationrequiring endovascular treatment or surgery torepair. The primary feasibility endpoint, also evalu-ated at 30 days, was procedure success, defined as theabsence of device- or procedure-related MAEs, alongwith proper filter positioning in each CCA. Properfilter position is defined as: 1) having the supportingcoil in contact with the CCA wall; 2) the absence ofdevice migration, fracture, or coils outside of the CCAlumen; and 3) no entangled or overlapping coilsinside of the CCA lumen. Secondary endpointsincluded the absence of device-related MAEs andproperly positioned filter in each CCA, both evaluatedat 3, 6, and 12 months, and successful filter deploy-ment attempts. Thrombi on the filter adjudicated bythe CEC as “definitely” or “probably” device-relatedwere counted toward the endpoint, while thoseadjudicated as “unrelated” or “possibly” device-related were not.

STATISTICAL ANALYSES. CAPTURE is a feasibilitystudy with no formal hypothesis testing, and there-fore, no power analyses were performed. Subjects arefollowed on an intent-to-treat basis. Study results are

TABLE 1 Patient Baseline Characteristics (n ¼ 25)

Age, yrs 71.3 � 9.4

Prior MI 2 (8)

History of spontaneous or clinicallyrelevant bleeding

4 (16)

AF type

Paroxysmal 2 (8)

Persistent 19 (76)

Permanent 4 (16)

CHA2DS2-VASc score 4.4 � 1.0

CHA2DS2-VASc score components

Female 9 (36)

Congestive heart failure 7 (28)

Hypertension 25 (100)

Age 65–74 yrs 9 (36)

Age $75 yrs 11 (44)

Diabetes 12 (48)

Stroke/TIA/thromboembolism 12 (48)

Vascular disease 4 (12)

Recurrent stroke 3 (12)

Carotid atherosclerosis 2 (8)

Prior use of warfarin >30 days 8 (32)

Prior use of DOAC >30 days 7 (28)

Abnormal renal function 0 (0)

Medication at the time of enrollment

Amiodarone 5 (20)

Beta-blocker 13 (52)

Aspirin 1 (4)

Clopidogrel 0 (0)

Digoxin 0 (0)

Calcium-channel blocker 6 (24)

Statin 11 (44)

NSAID 0 (0)

Gastric antacid 5 (25)

Values are mean � SD or n (%).

AF ¼ atrial fibrillation; CHA2DS2-VASc ¼ Congestive heart failure, Hypertension,Age 75 years, Diabetes, Stroke/transient ischemic attack, Vascular disease, Age 65to 74 years, Sex category; DOAC ¼ direct oral anticoagulant; MI ¼ myocardialinfarction; NSAID ¼ nonsteroidal anti-inflammatory drug; TIA ¼ transient ischemicattack.

TABLE 2 Procedural Characteristics (n ¼ 25)

Major adverse events: major bleeding, CCA stenosis>70%, filter migration, CCA thrombus, or any CCAcomplication requiring endovascular treatment orsurgery to repair

0 (0)

Puncture site hematoma/edema 5 (20)*

2 properly positioned implants w/o device/procedure-related MAEs

23 (92)

Filter deployment

Bilateral 23 (92)

Unilateral 1 (4)

None 1 (4)

Values are n (%). *Resolved without treatment.

CCA ¼ common carotid artery; MAE ¼ major adverse event.

J A C C V O L . 7 4 , N O . 7 , 2 0 1 9 Reddy et al.A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9 Permanent Carotid Coil Filter for Stroke Prevention in AF

833

presented using descriptive statistics. For continuousvariables, the results include number, mean, and SD,where pertinent. Presented data for categorical vari-ables include the number and percent of subjects ineach category.

RESULTS

PATIENTS. A total of 36 patients signed a consentform and underwent ultrasound screening: 8 patients(22%) were excluded due to prohibitive carotid arte-rial atherosclerosis (n ¼ 4), stenosis in the internal,external, or common carotid artery (n ¼ 2), excessiveCCA lumen diameter variability (n ¼ 1), and CCAs outof the requisite 4.8- to 9.8-mm range. Three addi-tional patients withdrew consent. The remaining 25

patients constituted the study cohort, and weretreated between March 2, 2018, and November 2,2018. The mean age was 71.3 years, 16 (74%) weremale, and the mean CHA2DS2-VASc score was 4.4(Table 1). Nearly one-half of the patient cohort (48%)had a history of stroke, TIA, or thromboembolism,with 12% having a history of multiple strokes. PriorOAC use, either warfarin or a DOAC, occurred in 60%of the patients.

PROCEDURAL CHARACTERISTICS. Of the 25-patientcohort, 23 (92%) received properly positioned filtersbilaterally (Table 2, Figure 4). One patient received aproperly positioned filter unilaterally due to a CCAatheromatous plaque in the contralateral CCA thatwas not recognized during the screening ultrasound.The final patient underwent an unsuccessful pro-cedure attempt because of poor ultrasonic visibility.

In addition to the 47 successful deployments, therewere 9 additional attempts that did not deploycorrectly, were all (100%) successfully retracted usingthe pulling wire, and were immediately reimplantedsuccessfully; thus, the overall rate of successfuldeployment was 47 of 56 (84%).

There were 0 (0%) MAEs during the procedure(Table 2)—including no cases (0%) of death, stroke,major bleeding, CCA stenosis, filter migration, CCAthrombus, or any CCA complication requiring endo-vascular treatment or surgery to repair. Five patients(20%) had hematoma with or without concomitantedema immediately following the procedure. Thesepatients were just managed conservatively, as allresolved over time without treatment.

PRIMARY ENDPOINTS. All 25 patients were fullyevaluable for the primary endpoints at $30 days post-procedure. The primary safety endpoint of absence ofdevice or procedure-related MAEs (death, stroke,major bleeding, CCA stenosis >70%, CCA thrombus,or any CCA complication requiring endovasculartreatment or surgery to repair), evaluated at 30 days,

FIGURE 4 The Carotid Coil Filter In Situ

A properly positioned common carotid artery filter is shown in a patient’s carotid artery visualized by transcutaneous ultrasound imaging (A), and bilaterally in both

carotid arteries as visualized on x-ray imaging (B and C).

TABLE 3 Primary En

Primary Safety Endpoi(major bleeding, CCCCA complication r

Primary Feasibility Enddevice/procedure-r

Values are n (%).

Abbreviations as in Tabl

Reddy et al. J A C C V O L . 7 4 , N O . 7 , 2 0 1 9

Permanent Carotid Coil Filter for Stroke Prevention in AF A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9

834

was achieved in all 25 (100%) patients (Table 3). Theprimary feasibility endpoint of procedural success(2 properly positioned filters and absence of device/procedure-related MAEs at 30 days) was achieved in23 (92%) patients. One patient missed the primaryfeasibility endpoint due to receiving only 1 implant,and the other patient missed the endpoint due toreceiving no implants.SECONDARY ENDPOINTS. Of the 24 patientsreceiving at least 1 implant, the median follow-up was6 months (range 3 to 12 months; interquartile range: 3to 6 months). All patients (100%) who completed eachof these follow-up visits met the secondary safetyendpoint of absence of device-related MAEs (Table 4).The secondary performance endpoint of having2 properly positioned implants was achieved in23 of 24 (96%), 11 of 12 (92%), and 3 of 4 (75%) of thepatients who completed the 3-, 6-, and 12-monthfollow-up visits, respectively. The patient with uni-lateral filter placement is the only one who did notmeet these secondary performance endpoints. Thepatient without both filters was excluded from thestudy after completing the 1-month follow-up. Finally,of the 47 implants that were properly positioned at theend of the procedure, 47 (100%) remained properlypositioned at all subsequent follow-ups.

dpoints (n ¼ 25)

nt: Absence of device/procedure-related MAEs at 30 daysA stenosis > 70%, filter migration, CCA thrombus, or anyequiring endovascular treatment or surgery to repair)

25 (100)

point: Procedural success (2 properly positioned filters w/oelated MAEs at 30 days)

23 (92)

e 2.

OTHER OUTCOME ASSESSMENTS. In addition to theprimary and secondary safety endpoints, additionaloutcome assessments were made (Table 5). Three pa-tients died of causes unrelated to the procedure or thedevice (pneumonia, brain cancer, and heart failuredecompensation). One patient had 2 consecutive mi-nor strokes adjudicated by the clinical events com-mittee as device-unrelated. In both instances,cardiovascular magnetic resonance revealed involve-ment in both posterior and anterior circulation terri-tories. This is most consistent with a shower of embolioriginating proximal to the devices. The device didnot stop the anterior circulation thrombi, likelybecause they were <1.4 mm in size. Consistent withthe small size of these emboli, and minor natureof the strokes, the patient’s symptoms haveresolved completely.

One patient experienced a device-unrelated majorbleed (intracranial, subdural, and intracerebral hem-orrhage) and traumatic brain injury from a fall at8 months; this patient largely recovered, with onlyslight residual aphasia. This patient also had an inci-dental finding of thrombus on the filter, withoutsymptoms of stroke. Due to the coincident nature ofthis fall and thrombus, the fall was classified by theCEC as “possibly device-related,” although no causalrelationship between the 2 was demonstrated. Nopatient (0%) had carotid stenosis or occlusion.

There were 6 occurrences of thrombi on the CCAfilter (1 bilateral, 4 unilateral) that were identifiedby ultrasound imaging in 4 patients during thestudy (Figure 5). All thrombi were reported by thesites as most probably captured, and not in situthrombi; the CEC adjudicated 3 thrombi ascaptured, and undetermined in 2. The thrombi

TABLE 4 Secondary Endpoints

3 Months(n ¼ 24)

6 Months(n ¼ 12)

12 Months(n ¼ 4)

Absence of device-related MAEs 24 (100) 12 (100) 4 (100)

2 properly positioned filters 23 (96) 11 (92) 3 (75)

Values are n (%).

MAE ¼ major adverse event.

TABLE 5 Other Outcome Assessments

1 Day to1 Week

1 Week to1 Month

1–3Months

3–6Months

6–12Months

Total Number ofPatients With

Events

Number of patients in study(beginning of period)

25 25 24 24 23 NR

Number of patients in study(end of period)

25 25 24 23 21 NR

Number of patients whocompleted follow-up(end of period)

25 25 24 15 4 NR

Death 0 0 0 1 2 3

Stroke 0 1 0 0 0 1

Major bleeding 0 0 0 0 1* 0

CCA stenosis or occlusion(filter section)

0 0 0 0 0 0

Number of patients with firstthrombus on filter

0 3 0 0 1 4

Number of thrombi on filter 0 4 1 0 1 NR

*Intracranial, subdural, and intracerebral hemorrhage resulting from traumatic brain injury due to fall.

CCA ¼ common carotid artery; NR ¼ not relevant.

J A C C V O L . 7 4 , N O . 7 , 2 0 1 9 Reddy et al.A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9 Permanent Carotid Coil Filter for Stroke Prevention in AF

835

varied in thickness between 0.15 and 2.0 mm, andin length between about 2.0 and 10 mm. One pa-tient had bilateral thrombi at 1 month and anadditional unilateral thrombus at 2.5 months. Twopatients had unilateral thrombi at 1 month, and 1 pa-tient had a unilateral thrombus at 9 months. Allthrombi were without concomitant symptoms ofstroke. Five of the thrombi resolved with subcutane-ous (SQ) low-molecular-weight heparin (8,000 to10,000 U/12 h) within 0.5 to 5 months. The lastthrombus is more recent, and it is almost completelyresolved with SQ heparin. No thrombus caused aninterruption to blood flow by color Doppler ultrasound.

DISCUSSION

The present study suggests that deployment of apermanent coil filter in the human CCA is feasible andsafe (Central Illustration). The filter captured 6thromboemboli in 4 patients, none of whom devel-oped stroke symptoms. Furthermore, 5 of the 6thrombi have resolved with SQ heparin, and thelast is almost completely resolved. The per-patientprocedural success rate was 92% (23 of 25 patients),and the rate of successful filter deployment was 84%(47 of 56 patients). The 30-day device-related MAErate was 0%. Minor AEs—puncture site hematoma/edema—occurred in 20% (5 of 25 patients). During anaverage follow-up of 6 months, there was no evidenceof in situ thrombus formation or CCA stenosis.

Implantation failures (9 of 56) were related toimplant/CCA sizing mismatch and poor visibility ofthe CCA and needle in cases of very deep (>4 cmbelow skin) CCAs. In all cases of implantation failure,the implant was retracted immediately—an importantcapability that had been part of the design of thedevice. In 7 cases, another properly-sized filter wasthen successfully placed; however, in 1 patient, noCCA filter was placed after 2 consecutive failed im-plantations and successful retractions.

The observed low procedural risk is consistent withevidence that inadvertent carotid artery punctureduring jugular cannulation is innocent, despite itsrelatively common occurrence (w3% of cannulations)(10). It is also consistent with data showing that CCApunctures with smaller than 18-gauge needles areinnocent (11). The risk of inadvertent plaque ruptureduring carotid puncture was mitigated by excludingatheromatous CCA segments by pre-procedure ultra-sound screening. Puncture site skin hematoma wasrelated to soft tissue needle trauma—no peri-CCAhematoma was observed in these cases at ultra-sound imaging. In 3 of these cases, there wasconcomitant neck edema (confirmed by computed

tomographic scanning), which most probably resultedfrom local reaction of soft tissue to blood oozing. Allcases of hematoma/edema resolved over timewithout treatment. These minor adverse events couldprobably be minimized by more meticulous applica-tion of local pressure at the puncture site for severalminutes following implantation.

Preclinical studies revealed no neointimal growthon the implant, including the supporting and leadingcoils that are in contact with the carotid arterial wall(Yodfat et al., unpublished data). This observation islikely related to the relatively low radial force exertedby the implant on the arterial wall (5% to 10% over-sized relative to CCA systolic diameter) comparedwith arterial stents (12). For example, a standard 8-mm stent spanning the carotid bifurcation from a 6-mm common carotid artery to a 4-mm internal ca-rotid artery results in w30% and 100% oversizing,respectively. The CCA filter is fixed in place by a stemtraversing the arterial wall and equipped with extra-luminal and intraluminal anchors. This serves toreduce or even eliminate migration irrespective of therelatively low friction between the implant and thearterial wall.

FIGURE 5 Captured Versus In Situ Thrombus

From the present study, a captured thromboembolus captured on the CCA filter detected at ultrasound follow-up 1-month post-implantation (A). This is quite similar in

appearance to an autologous thrombus embolized from upstream to and captured by a permanent carotid coil filter in sheep (B). Note that these thromboemboli have

a wire-like appearance and are mobile (Online Videos 2 and 3). In contrast, thrombi that formed on the filtering portion (C) or anchor (D) in other ovine experiments (13),

are more regular in shape and less mobile (Online Videos 4 and 5).

Reddy et al. J A C C V O L . 7 4 , N O . 7 , 2 0 1 9

Permanent Carotid Coil Filter for Stroke Prevention in AF A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9

836

Thrombi observed on the filter were reported ascaptured embolic material, not in situ formation. Thisobservation is consistent with the low rate (<1%) ofthrombus formation on “free floating” stent and graftstruts crossing coronary, renal, and carotid ostia(13–16). In this study, AF patients were treated withdual antiplatelet therapy for 3 months (ASA þ clopi-dogrel) followed by ASA alone; accordingly, the riskof device-related thrombus formation is likely low inboth AF patients who are unsuitable for OAC and AFpatients with planned (i.e., surgery) or unplannedinterruptions of OAC.

The appearance of the thrombi in all cases waselongated and mobile, resembling the string-likeappearance of captured thrombi in ovine studies in 5of 6 cases (Yodfat et al., unpublished data). Thrombusthat formed spontaneously during preclinical ovinestudies appears very different—these are larger andmore regularly shaped echogenic masses (Figure 5).In addition, in situ thrombi observed in theseovine studies invariably appeared within 1 h of im-plantation, whereas all thrombi observed on filters in

the present study appeared between 1 week and9 months.

All 4 patients with thrombi observed on their filtershad a history of stroke, and 2 of them had a history ofrecurrent stroke. This is consistent with the reason-able hypothesis that they might be more likely toelaborate frequent emboli compared with AF patientswithout a history of stroke. The rate at which patientswere identified with thrombus on the CCA filters inthe present study is w30% per patient-year (4 pa-tients per 12.5 patient-years), which is approximately5 to 6 times the expected stroke rate in the studypopulation (1,2,17,18). This is consistent with the 2important observations. First, embolic showers arefrequent in AF patients, as evident from randomtranscranial Doppler monitoring of the middle cere-bral artery: 26% of AF patients during 2 consecutive1-h monitoring periods (19). Most of these emboli areharmless, but some cause symptomatic stroke ormultiple asymptomatic brain infarcts: silent strokesare observed in 45% of AF patients with no history ofstroke (20). Second, stroke studies of intraprocedural

CENTRAL ILLUSTRATION A Percutaneous Carotid Coil Filter for Stroke Prevention in Atrial Fibrillation

Reddy, V.Y. et al. J Am Coll Cardiol. 2019;74(7):829–39.

The permanent carotid coil filter is implanted bilaterally in the common carotid arteries, by direct skin puncture in the neck atop the common carotid arteries using a

24-gauge needle inserter. The filter is configured to capture emboli that exceed 1.4 mm in size originating below the neck, including the heart or aorta, thereby

protecting the brain against ischemic stroke of (cardio)embolic source. In the first-in-human CAPTURE clinical trial, implantation of the coil filter was feasible with high

success rates, safe, and effective in capturing thromboemboli. Furthermore, these captured thromboemboli resolved over time, at least in part due to the action of

endogenous thrombolytic enzymes. CHA2DS2-VASc ¼ Congestive heart failure, Hypertension, Age 75 years, Diabetes, Stroke/transient ischemic attack, Vascular

disease, Age 65 to 74 years, Sex category.

J A C C V O L . 7 4 , N O . 7 , 2 0 1 9 Reddy et al.A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9 Permanent Carotid Coil Filter for Stroke Prevention in AF

837

carotid filter placement suggest that the number ofparticles coursing through the carotids during theprocedure and exceeding 1 mm in size is an order ofmagnitude greater than the rate of clinical stroke. Incarotid artery stenting, emboli of size >1 and >2 mmwere observed in 50% and 40% of protection filters,respectively; however, the clinical stroke rate at30 days was only 2.9%—major at 0.5% and minor at2.4% (21,22). During transaortic valve implantation,emboli of size >1 and >2 mm were observed in 100%and 20% of temporary carotid filters, respectively,while the stroke rate at 30 days in the control (nofilter) arm was only 9.1% (23).

Of the 6 thrombi observed in the present study, 4originated between 1-week and 1-month post-procedure. All 4 patients with thrombi on the filterhad discontinued OACs within 30 days of the im-plantation procedure. Stroke risk in the 30 days afterrivaroxaban temporary (>3 days) or permanent(>30 days) discontinuation is 6% and 26%/year,respectively (24). It is therefore likely that patientswho stop OAC produce thromboemboli morefrequently right after OAC discontinuation than at“steady state.”

In the present study, all 4 patients with thrombi onfilter were treated with SQ heparin until completethrombus resolution. But the observed absence ofstroke in these patients was predicated by the rela-tively intensive ultrasound surveillance strategy,with subsequent SQ heparin; it is unknown if thenatural fibrinolytic process would have likewise beensufficient to ensure thrombus resolution. On theother hand, we should note that evidence to supportheparin treatment is also lacking: with IVC filters, at6 months, 60% of captured emboli regress and 30%do not progress, irrespective of heparin administra-tion (25). It is possible that regression of capturedthromboemboli within the arterial circulation wouldoccur regardless of heparin treatment.

There is overlap between the CAPTURE patientsand left atrial appendage closure candidates. Leftatrial appendage closure devices avoid some limita-tions of the CCA filter, but they do not prevent non-appendage origin emboli potentially addressable bythe filter. Moreover, the filter implantation procedureis less invasive, is faster (minutes), is performed by 1operator without general anesthesia, does not requiretransesophageal echocardiography, and will likely

PERSPECTIVES

COMPETENCY IN PATIENT CARE AND

PROCEDURAL SKILLS: Permanent coil filters

placed in the carotid arteries bilaterally can capture

thromboembolic material >1.4 mm in diameter

without causing stroke.

TRANSLATIONAL OUTLOOK: Larger randomized

trials are warranted to evaluate the safety and efficacy

of this stroke prevention strategy in patients with

atrial fibrillation.

Reddy et al. J A C C V O L . 7 4 , N O . 7 , 2 0 1 9

Permanent Carotid Coil Filter for Stroke Prevention in AF A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9

838

become an outpatient procedure in the future. Thisminimalistic nature of the procedure, combined withthe simple workflow, have prompted the design of arandomized trial studying the role of the CCA filter inaddition to optimal OAC in high-risk patients (see thefollowing section).

FUTURE WORK. Although this is only a 3-centerstudy of a relatively small number of patients, thefeasibility and safety observed in this study providesthe basis for larger safety and multicenter random-ized clinical trials to definitively establish the safetyand efficacy of this therapy. Accordingly, CAPTURE 2(NCT03892824) is an observational safety trialincluding patients at high risk for stroke (CHA2DS2-VASc $4 and a history of ischemic stroke) receivingOAC/DOAC and bilateral filter implantation. A ran-domized efficacy trial is also planned to compareOAC/DOAC therapy with OAC/DOAC therapy plusbilateral carotid filter placement with a primary effi-cacy endpoint of ischemic stroke in filter-protectedcarotid territories. One advantage of this strategy:because all patients with carotid filters will alsoreceive concomitant OAC, there is less need forperiodic ultrasound surveillance to identifycaptured thromboemboli.

STUDY LIMITATIONS. We studied relatively few pa-tients, with few operators. Indeed, this non-randomized first-in-human trial is best interpreted asa proof-of-concept study requiring further random-ized trials to adequately establish the therapy’s placein clinical practice. Although the evidence stronglysuggests that the 6 thrombi identified during follow-up were captured thromboemboli, and not in situthrombus development, this is difficult to definitivelyprove. One patient had 2 consecutive minorposterior circulation strokes involving multiple ter-ritories including posterior circulation territories.

This underscores a limitation of the technology—itprovides no protection against posterior circulationembolic strokes. Also, it does not protect againstsubmillimetric emboli slipping through the filteringportion. Although such particles are likely to be sub-clinical or only cause minor strokes, the overall risk-benefit ratio must be established in randomized tri-als. Finally, we did not collect the reasons for OACunsuitability in this study.

CONCLUSIONS

In this first-in-human study, permanent carotid coilimplantation was feasible and safe. The capture ofthromboemboli was also demonstrated.

ADDRESS FOR CORRESPONDENCE: Dr. Vivek Y.Reddy, Helmsley Electrophysiology Center, IcahnSchool of Medicine at Mount Sinai, One Gustave L.Levy Place, P.O. Box 1030, New York, NewYork 10029. E-mail: vivek.reddy@mountsinai.org.Twitter: @IcahnMountSinai.

RE F E RENCE S

1. Friberg L, Skeppholm M, Terént A. Benefit ofanticoagulation unlikely in patients with atrialfibrillation and a CHA2DS2-VASc score of 1. J AmColl Cardiol 2015;65:225–32.

2. Vanassche T, Lauw MN, Eikelboom JW, et al.Risk of ischaemic stroke according to pattern ofatrial fibrillation: analysis of 6563 aspirin-treatedpatients in ACTIVE-A and AVERROES. Eur Heart J2014;36:281–8.

3. Hart RG, Pearce LA, Aguilar MI. Meta-analysis:antithrombotic therapy to prevent stroke in pa-tients who have non-valvular atrial fibrillation.Ann Intern Med 2007;146:857–67.

4. Connolly SJ, Ezekowitz MD, Yusuf S, et al.Dabigatran versus warfarin in patients with atrialfibrillation. N Engl J Med 2009;361:1139–51.

5. Yoshimura S, Koga M, Sato SK, et al. Two-yearoutcomes of anticoagulation for acute ischemicstroke with nonvalvular atrial fibrillation—SAMURAI-NVAF study. Circ J 2018;82:1935–42.

6. Saxena R, Lewis S, Berge E, Sandercock PA,Koudstaal PJ. Risk of early death and recurrentstroke and effect of heparin in 3169 patients withacute ischemic stroke and atrial fibrillation in theInternational Stroke Trial. Stroke 2001;32:2333–7.

7. Rai AT, Hogg JP, Cline B, Hobbs G. Cerebro-vascular geometry in the anterior circulation: ananalysis of diameter, length and the vessel taper.J Neurointerv Surg 2013;5:371–5.

8. Türe U, Yasargil MG, Al-Mefty O, Yasargil DC.Arteries of the insula. J Neurosurg 2000;92:676–87.

9. Müller HR, Brunhölzl C, Radü EW, Buser M. Sexand side differences of cerebral arterial caliber.Neuroradiology 1991;33:212–6.

10. Ruesch S, Walder B, Tramèr MR. Complicationsof central venous catheters: internal jugular versussubclavian access—a systematic review. Crit CareMed 2002;30:454–60.

11. Yilmazlar A, Bilgin H, Korfali G, Eren A,Ozkan U. Complications of 1303 central venouscannulations. J R Soc Med 1997;90:319–21.

12. Timmins LH, Miller MW, Clubb FJ Jr.,Moore JE Jr. Increased artery wall stress poststenting leads to greater intimal thickening. LabInvest 2011;91:955–67.

13. Yamawaki M, Muramatsu T, Araki M, et al.Natural history of side branches jailed by

J A C C V O L . 7 4 , N O . 7 , 2 0 1 9 Reddy et al.A U G U S T 2 0 , 2 0 1 9 : 8 2 9 – 3 9 Permanent Carotid Coil Filter for Stroke Prevention in AF

839

drug-eluting stents. J Interv Cardiol 2012;25:37–46.

14. England A, Butterfield JS, Ashleigh RJ. Inci-dence and effect of bare suprarenal stent strutscrossing renal ostia following EVAR. Eur J VascEndovasc Surg 2006;32:523–8.

15. Willfort-Ehringer A, Ahmadi R, Gruber D, et al.Effect of carotid artery stenting on the externalcarotid artery. J Vasc Surg 2003;38:1039–44.

16. de Borst GJ, Vos JA, Reichmann B, et al. Thefate of the external carotid artery after carotidartery stenting. A follow-up study with duplexultrasonography. Eur J Vasc Endovasc Surg 2007;33:657–63.

17. Diener HC, Eikelboom J, Connolly SJ, et al.Apixaban versus aspirin in patients with atrialfibrillation and previous stroke or transientischaemic attack: a predefined subgroup analysisfrom AVERROES, a randomized trial. Lancet Neu-rol 2012;11:225–31.

18. Gundlund A, Xian Y, Peterson ED, et al. Pre-stroke and post-stroke antithrombotic therapy inpatients with atrial fibrillation: results from a

nationwide cohort. JAMA Netw Open 2018;1:e180171.

19. Cullinane M, Wainwright R, Brown A,Monaghan M, Markus HS. Asymptomatic emboli-zation in subjects with atrial fibrillation not takinganticoagulants: a prospective study. Stroke 1998;29:1810–5.

20. Kalantarian S, Ay H, Gollub RL, et al. As-sociation between atrial fibrillation and silentcerebral infarctions: a systematic review andmeta-analysis. Ann Intern Med 2014;161:650–8.

21. Angelini A, Reimers B, Della Barbera M,et al. Cerebral protection during carotid arterystenting: collection and histopathologic anal-ysis of embolized debris. Stroke 2002;33:456–61.

22. Rosenfield K, Matsumura JS, Chaturvedi S,et al. Randomized trial of stent versus surgery forasymptomatic carotid stenosis. N Engl J Med2016;374:1011–20.

23. Kapadia SR, Kodali S, Makkar R, et al. Protectionagainst cerebral embolism during transcatheter

aortic valve replacement. J Am Coll Cardiol 2017;69:367–77.

24. Patel MR, Hellkamp AS, Lokhnygina Y, et al.Outcomes of discontinuing rivaroxaban comparedwith warfarin in patients with non-valvular atrialfibrillation: analysis from the ROCKET AF trial(Rivaroxaban Once-Daily, Oral, Direct Factor XaInhibition Compared with Vitamin K Antagonismfor Prevention of Stroke and Embolism Trial inAtrial Fibrillation). J Am Coll Cardiol 2013;61:651–8.

25. Ahmad I, Yeddula K, Wicky S, Kalva SP. Clin-ical sequelae of thrombus in an inferior vena cavafilter. Cardiovasc Intervent Radiol 2010;33:285–9.

KEY WORDS atrial fibrillation, carotidfilter, common carotid artery, embolicprotection, oral anticoagulation, strokeprevention

APPENDIX For supplemental tables andvideos, please see the online version of thispaper.