Neurologic Outcomes with Embolic Protection Devices in Patients Undergoing Transcatheter Aortic Valve Replacement: A Systematic Review and Meta-Analysis of Randomized Controlled

TrialsGennaro Giustino, MD*, Roxana Mehran, MD, FAHA*, Roland Veltkamp, MD†, Angela Del Giudice,

MD†, Michela Faggioni, MD*, Jaya Chandrasekhar, MD*, Usman Baber, MD,* and

George D. Dangas, MD, PhD, FAHA*.

*Interventional Cardiovascular Research and Clinical Trials, The Zena and Michael A. Wiener

Cardiovascular Institute, Icahn School of Medicine at Mount Sinai (New York City, New York);

†Department of Medicine, Division of Brain Sciences, Imperial College (London, UK).

Running TitleEmbolic Protection during TAVR

Word Count: 5,673

Corresponding AuthorGeorge D. Dangas MD, PhD, FACC.Mount Sinai Hospital,

One Gustave L. Levy Place, Box 1030

New York, New York 10029

Tel: 212-241-7014; Fax: 212-241-0273

E-mail: George.Dangas@mountsinai.org

1

AbstractBackground: Cerebral embolism is a common complication after transcatheter aortic valve

replacement (TAVR). Randomized controlled trials (RCTs) investigating the efficacy of embolic

protection (EP) devices during TAVR were relatively underpowered. We investigated the effect of

intraprocedural EP on neurologic outcomes after TAVR.

Methods: We performed a systematic review and study-level meta-analysis of RCTs that tested the

efficacy and safety of EP during TAVR. RCTs using any type of EP and TAVR vascular access were

included. Primary imaging efficacy endpoints were total lesion volume (TLV; in mm3) and number of

new ischemic lesions. Primary clinical efficacy endpoints were any deterioration in National Institute of

Health Stroke Scale (NIHSS) and Montreal Cognitive Assessment (MoCA) score at hospital discharge.

Primary analyses were performed with the intention-to-treat approach.

Results: A total of 4 RCTs (total n=252) have been included. Use of EP was associated with lower TLV

(standardized mean difference [SMD]: -0.83; 95% confidence interval [CI]: -1.57 to -0.10; p=0.03) and

lower number of new ischemic lesions (SMD: -1.06; 95% CI: -1.77 to -0.35; p = 0.003). EP was

associated with a trend to lower risk of deterioration in NIHSS at discharge (RR: 0.55; 95% CI: 0.27 to

1.09; p=0.09) and higher MoCA score (SMD: +0.40; 95% CI: +0.04 to +0.76; p = 0.03). Risk of overt

stroke and all-cause mortality were non-significantly lower in the EP group.

Conclusions: Use of EP seems to be associated with a reduction of imaging markers of cerebral

infarction and early clinical neurologic effectiveness in patients undergoing TAVR.

Key Words: TAVR; Embolic Protection; Stroke.

2

IntroductionTranscatheter aortic valve replacement (TAVR) has emerged as a standard of care to treat

degenerative aortic stenosis in patients deemed at high- or prohibitive-risk for surgical aortic valve

replacement (SAVR).1, 2 Patients undergoing TAVR are often elderly, frail and affected by multiple

comorbidities, implying a significant risk for thromboembolic cerebrovascular events.1-3 Overt stroke is

the most feared complication of TAVR being associated with a strong effect on morbidity and mortality.3,

4 Early studies suggested that TAVR is associated with an increased risk of stroke compared with

medical treatment or surgical aortic valve replacement.1, 2 Additionally, several studies demonstrated a

very high incidence of new cerebral ischemic lesions on post-procedural diffusion-weighted magnetic

resonance imaging (DW-MRI), and of high-intensity transient signals evaluated with transcranial

Doppler, respectively.5 In current TAVR practice the rate of overt stroke during or early after TAVR is

relatively low (≤ 2%);6 however the frequency and burden of micro-embolization and cerebral ischemic

injury may still have a substantial impact on mid- and long-term cognitive function.5, 7 Therefore, in order

for TAVR to expand to lower risk patients, measures to mitigate neurologic risk are warranted.

Additional, several studies in surgical cohorts demonstrated greater cognitive function impairment post-

SAVR compared with patients undergoing coronary artery by-pass graft surgery8, 9.

The application of EP has been explored in several small observational and few randomized

studies in TAVR but its efficacy and safety remains inconclusive. Therefore, by pooling study-level data

from randomized controlled trials (RCTs), in the present metanalysis we sought to investigate imaging

and clinical neurologic outcomes associated with intraprocedural EP in patients with severe aortic

stenosis undergoing TAVR.

MethodsStudy Design. We performed a systematic review and study-level meta-analysis of RCTs that tested

the efficacy of EP devices during TAVR according to the Preferred Reporting Items for Systematic

Reviews and Meta-Analyses (PRISMA) guidelines.10 RCTs investigating the efficacy of EP with any

device and for any TAVR vascular access were included. All non-randomized studies reporting

outcomes with EP during TAVR were excluded. We opted to include only RCTs in order to reduce the

selection and confounding bias of observational pilot studies. The two study groups were defined by the

randomized assignment to either intraprocedural EP or not. The pre-specified imaging neurologic

endpoints were total lesion volume (TLV) in mm3 assessed with DW-MRI and number of new ischemic

lesions. The pre-specified primary clinical neurologic endpoints were any clinical deterioration from

baseline in National Institute of Health Stroke Scale (NIHSS) and the Montreal Cognitive Assessment

(MoCA) score at discharge. Secondary endpoints were: percentage of patients with new cerebral

3

ischemic lesions, mean number of new cerebral ischemic lesions, clinically overt stroke at follow-up, all-

cause mortality at follow-up, fluoroscopic time and acute kidney injury. Imaging endpoints were

evaluated as well according to the type of valve implanted (balloon-expandable [BE] versus self-

expandable [SE]), where available. All endpoints were estimated at the maximum time of follow-up

reported according to the intention-to-treat (ITT) principle. The study was performed according to the

PRISMA statement.10

Research strategy. MEDLINE, Scopus, the Cochrane Library, and TCTMD.org were searched for

abstracts, manuscripts, and conference reports published until December 31, 2015. There were no

language restrictions. The following key words were used for the search: “TAVR embolic protection”,

“TAVR embolic protection randomized controlled trial”, “TAVR stroke”, “TAVR Claret”, “TAVR Triguard”,

and “TAVR Embol-X”. Two investigators (GG and MF) independently reviewed the studies and reported

the results in a structured dataset. Disparities between investigators regarding the inclusion of each trial

were resolved by consensus by a third independent investigator (GD). Pre-specified data elements

were extracted from each trial and included in a structured dataset; these elements included type of EP

device, baseline characteristics, TAVR access site, type of transcatheter heart valve device implanted,

risk of bias, and outcome measures, including imaging endpoints (TLV, mean number of new ischemic

lesions, number of patients with new ischemic lesions), clinical endpoints (any worsening in NIHSS,

MoCA score, clinically overt stroke, all-cause mortality, acute kidney injury) and procedural variables

(fluoroscopic time). Additionally, device success (defined as the correct deployment and retrieval of the

device) was captured. Endpoints were collected according to both the ITT and the per-treatment (PT)

principles. Considering the potential effectiveness of EP devices on intra-procedural (acute) events, the

primary analysis included all events reported as close as possible to the date of the index TAVR

procedure in each RCT. Risk of bias in each trial for both the primary imaging (TLV) and clinical (risk for

NIHSS deterioration at discharge) endpoints was evaluated with the Cochrane’s tool as described by

Higgins et al11; the following elements potential source of bias have been evaluated: random sequence

generation (selection bias), allocation concealment (selection bias), blinding of participants and

personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome

data (attrition bias) and selective reporting (reporting bias). For each element, a qualitative attribution of

bias was given (low risk, intermediate risk or high risk for bias) by two independent investigators (GG

and MF). Disparities between investigators regarding the risk of bias were resolved by a third

independent investigator (GD).

4

Study Devices.

The following EP devices have been used in the included RCTs:

The Claret Montage system is composed of two polyurethane mesh filters with 140-mm pores

mounted on a nitinol frame, and delivered through transradial access. This is a selective EP device,

with 1 filter deployed in the brachiocephalic trunk and one filter in the left common carotid artery. This

device allows for capture and retrieval of embolic material.

The Triguard embolic deflection device is a nitinol-based, temporary, non-selective filter delivered

through transfemoral access and positioned in the aortic arch in order to cover the ostia of the

innominate artery, the left common carotid and the left subclavian while maintaining blood flow to the

cerebral vessels through 130 mm pores while deflecting larger emboli to the descending aorta.

The EMBOL-X device is a filtration system composed of a self-expanding nitinol-based frame

covered by a semi permeable polyester mesh with 120-mm pore diameter. It is delivered through the

radial or brachial artery and positioned within the ascending aorta where it captures embolic material

while allowing blood flow. This device allows for capture and retrieval of embolic material.

Statistical Analysis. We estimated risk ratios (RRs) and standardized mean differences (SMD) with

95% confidence intervals (CIs) for all available categorical and continuous variables, respectively.

Given the possible heterogeneity in outcomes ascertainment across trials, we opted to use SMD as this

is a more conservative summary statistic that expresses the size of the intervention effect in each study

relative to the variability observed in that study. During data extraction, continuous variables reported

as median with low- and high-end of the range were converted to mean and standard deviations

according to the method of Hozo et al.12 If ranges were not directly reported, these were extracted by

visual estimation of the plots. Primary analytic method was the more conservative random effect model

according to DerSimonian and Laird. The primary and secondary analytic approaches (for the imaging

and neurologic endpoints) were according the ITT and PT principles, respectively. Publication bias for

the primary imaging and clinical endpoints was estimated via visual inspection of the Funnel Plot.

Heterogeneity among trials for each outcome was estimated with Chi2 test and quantified with I2

statistics (with I2 > 75% indicating substantial heterogeneity).13 If a trial did not report one of the pre-

specified primary and efficacy endpoint, this was excluded from that specific analysis. Analysis for the

imaging neurologic endpoints was stratified by type of valve (BE and SE TAVR devices) with

subsequent formal interaction test. Analyses were conducted with Cochrane’s Review Manager

(RevMan) version 5.3.

5

ResultsSystematic Review.

The research flow diagram according to the PRISMA guidelines is illustrated in Supplementary

Figure 1. Out of more than 200 screened article, 4 RCTs (N = 252) that met the inclusion criteria were

found (Table 1).

The CLaret Embolic Protection ANd TAVI - Trial (CLEAN-TAVI; NCT01833052)14 was a double-

blind randomized controlled trial that assigned 100 patients to either EP (n = 50) with the Claret

MontageTM dual-filter or to no EP (n = 50). Patients were all treated with femoral access and a SE

device. All patients underwent DW-MRI at baseline, day 2, day 7 and day 30 (not available for data

extraction) after TAVR. All patients underwent serial assessment with NIHSS at 2 days, 7 days and 30

days. DW-MRI endpoints at 7 days were included in this analysis. The drop-out rate for DW-MRI

assessment at 7 days was 13% (87 / 100). NIHSS results at 2 days were included. The drop-out rate

for NIHSS assessment was not reported. Device success was achieved in 96% of patients (48 / 50).

The A Prospective, Randomized Evaluation of the TriGuard™ HDH Embolic Deflection Device

During TAVI (DEFLECT-III; NCT02070731) study15 was a single-blind randomized controlled trial in

which 85 patients were randomized to either EP (n = 46) with the deflector TriGuard HDH or no EP (n =

39). Ninety-six percent of patients underwent TAVR through a transfemoral approach and 4% through a

transapical approach. A BE valve was implanted in 63% of patients and a SE in 31%. DW-MRI was

performed in all patients at day 4 ± 2 and day 30 ± 7 after TAVR. All patients underwent serial

neurologic assessment with NIHSS, MoCA and computerized Cogstate Research Test. DW-MRI

endpoints at 4 days were included. Drop-out rate for DW-MRI assessment at 4 days was 30% (33/46 in

the EP group and 26/39 in the no EP group). NIHSS and MoCA assessment at discharge were

included. Drop-out rate for NIHSS and MoCA assessment at discharge was 6% (5/85). Device success

was achieved in 88.9% of patients (40 / 45).

The Intraprocedural Intraaortic Embolic Protection with the EmbolX Device in Patients

Undergoing Transaortic Transcatheter Aortic Valve Implantation (TAo-EmbolX; NCT01735513) trial16

randomized 30 patients to either EP (n = 14) with Embol-X or no EP (n = 16). All patients underwent

TAVR through transaortic approach and all patients received a BE valve. DW-MRI was performed at

baseline (pre-TAVR) and within a week after TAVR. No specific serial neurologic assessment was

performed. Drop-out rate for DW-MRI assessment was 0% (0/30). Device success was of 100%

(14/14).

The MRI Investigation in TAVI with Claret (MISTRAL-C; NTR4236) study17 was a multicenter,

double-blind randomized trial that randomly assigned 65 patients to TAVR with (n = 32) or without (n =

33) EP with the Sentinel Cerebral Protection System. All patients underwent DW-MRI at baseline (pre-

6

TAVR) and 5 days after TAVR. All patients underwent serial neurocognitive assessment with NIHSS,

MoCA, mini-mental status examination and Center for Epidemiological Studies-Depression scale by

trained neurologist blinded to allocation at baseline and 5 days after TAVR. Both DW-MRI and clinical

neurologic evaluation at 5 days were included. The drop-out rates for DW-MRI and neurologic clinical

assessment were 43% (28 / 65) and 26% (17 / 63), respectively. Device success was achieved in 94%

of patients (30 / 32).

Neurologic Imaging Endpoint.TLV was reported in all 4 RCTs, number of new ischemic lesions in 3 RCTs and number of

patients with new ischemic lesions in 3 RCTs. At ITT analysis, use of EP during TAVR was associated

with lower TLV (Figure 1A; SMD: -0.83; 95% CI: -1.57 to -0.10; I2=84%; p = 0.03), lower mean number

of new ischemic lesions (Figure 1B; SMD: -1.06; 95% CI: -1.77 to -0.35; I2=74%; p = 0.003) and a trend

to lower number of patients with new ischemic lesions (Figure 1C; 72.4% vs. 82.5%; RR: 0.87; 95% CI:

0.73 to 1.04; I2=0%; p = 0.12). There was no evidence of publication bias for TLV and mean number of

new ischemic lesions endpoints at visual inspection of the Funnel plot (Supplementary Figure 2).

TLV and percentage of patients with new ischemic lesions according to valve types were

reported in 2 RCTs. The effect of EP during TAVR was consistent between BE and SE devices with

non-significant interaction test for both TLV (Figure 2A; pinteraction = 0.99) and risk of new ischemic lesions

(Figure 2B; pinteraction = 0.25), and absence of within-subgroup heterogeneity for both endpoints (I2=0%

for TLV in SE and BE subgroups; I2=0% for new ischemic lesions in SE and BE subgroups).

The direction of the effect estimates for the neurologic imaging endpoint were consistent at the

PT analysis, with an accentuation of the magnitude of the benefit for all three outcome metrics

(Supplementary Figure 3).

Neurologic Clinical Endpoint.NIHSS evaluation and MoCA score at discharge were reported in 3 and 2 RCTs, respectively.

Patients who were randomized to intraprocedural EP had a trend to lower risk of worsening in NIHSS at

discharge (8.3% vs. 16.8%; Figure 3A; RR: 0.55; 95% CI: 0.27 to 1.09; I2=0%; p = 0.09. Patients

randomized to EP had higher MoCA score at discharge (SMD: +0.40; 95% CI: +0.04 to +0.76; I2=0%; p

= 0.03). There was no evidence of publication bias for the risk of NIHSS deterioration and MoCA at

discharge at visual inspection of the Funnel plot (Supplementary Figure 2).

Neurologic clinical endpoint estimates were consistent at the PT analysis (Supplementary

Figure 4). Clinically overt stroke was reported in 3 RCTs. EP was associated with a non-significant

7

lower risk of stroke at 30-day of follow-up (Figure 3B; 2.2% vs. 4.5%; RR: 0.56; 95% CI: 0.11 to 2.82;

I2=0%; p = 0.49) at the ITT analysis.

Procedural Outcomes and Safety.Use of EP was associated with increased fluoroscopic time (Supplementary Figure 5A; SMD:

0.28; 95% CI: -0.02 to 0.58; I2=0%; p = 0.06). There were no differences in acute kidney injury (2.1%

vs. 5.6%; Supplementary Figure 5B; RR: 0.54; 95% CI: 0.05 to 6.11; I2=42%; p = 0.62). No evidence of

other major intraprocedural complications became evident in the systematic review of all RCTs.

Finally, EP was associated with non-significantly lower risk of all-cause mortality at 30 days of

follow-up (Figure 4; 1.4% vs. 5.1%; RR: 0.32; 95% CI: 0.08 – 1.34; I2=0%; p = 0.12).

DiscussionThe present metanalysis investigated the efficacy and safety of EP use during TAVR. The main

findings of the present study are: (i) EP is associated with significantly lower TLV and number of new

ischemic lesions after TAVR as assessed with DW-MRI; results were consistent at PT analysis, with an

accentuation of the magnitude of the benefit on all three neuroimaging endpoints (ii) EP resulted in

higher MoCA score and a trend to lower risk of any worsening in NIHSS at discharge; results were

consistent at PT analysis (iii) EP was associated with a non-significant reduction in stroke and all-cause

mortality; and (iv) use of EP was safe with no evidence of increased adverse events alongside a trend

to increased fluoroscopic time.

TAVR has become a standard of care for patients with severe aortic stenosis deemed at high or

prohibitive risk for surgery.1, 2 However, several concerns regarding its neurological safety rose early

after its introduction into clinical practice.18, 19 The incidence of clinically apparent neurological events

after TAVR is variable due to clinical endpoint definitions, ascertainment bias and underdiagnosis due

to lack of standardized, routine imaging and neurocognitive assessment.3, 20 Early after TAVR, cerebral

embolization is strongly related to technical and procedural factors such as retrograde crossing of the

degenerated stenotic aortic valve during diagnostic catheterization, catheter manipulation in an aortic

arch with severe atherosclerosis, preparatory balloon aortic valvuloplasty prior to bioprosthesis

implantation, eventual device malpositioning / dislodgment / embolization and need for valvular balloon

post-dilation in case of significant residual paravalvular leak. Histopathological studies revealed that

emboli composition is heterogeneous, with thrombotic material and tissue-derived debris identified in

74% and 63% of patients with embolization, respectively.20 As ischemic brain injury related to TAVR

spans a spectrum ranging between clinically overt strokes to seemingly silent ischemic lesions

identified with brain imaging studies and neurocognitive assessment, its evaluation is challenging.3, 5, 19

8

Previous studies suggest that new cerebral parenchymal ischemic lesions are common, usually

multiple,and distributed to both cerebral hemispheres.5 Clinically covert ischemic brain injury can result

in both acute and chronic cognitive and functional impairment which may have a substantial effect on

morbidity and mortality.21 Hence, intraprocedural strategies to mitigate cerebral embolization risk and

development of permanent neurologic deficit are essential in order to expand TAVR indication to lower

risk populations. Additionally, prior studies in surgical cohorts indicated higher risk of neurocognitive

deterioration in patients undergoing SAVR compared with age-matched patients undergoing coronary

artery by-pass graft surgery.9

EP devices have been introduced in TAVR practice. RCTs investigating the efficacy and safety

of EP during TAVR have been small and relatively underpowered for imaging and neurocognitive

endpoints.15 Additionally, TAVR studies with serial imaging follow-up are challenging due to technical

and logistic limitations in this elderly and sick patient population, with drop-out rate as high as 40% at

30 days.15, 17 Within this context, the present study-level metanalysis with enhanced statistical power

aimed to better define the role of EP during TAVR. EP was associated with significantly reduced TLV

and number of ischemic lesions compared with no EP, alongside higher MoCA score and a trend to

lower risk of NIHSS worsening at discharge. TLV is currently considered the most informative brain

imaging measure with excellent intra- and interrater concordance for DWI and fluid-attenuated inversion

recovery MRI and is one of the strongest predictors of supra-tentorial stroke outcomes.22, 23 The MoCA

and NIHSS are validated metric of neurocognitive function and neurologic dysfunction, respectively,

with excellent interrater reliability and a strong predictor of long-term outcomes after cerebrovascular

events.24-26 Additionally, we observed an accentuation of the benefits of EP on the neuroimaging

endpoint at the PT analysis. Possibly, this is consistent with the fact that when device success is

achieved intraprocedural EP is indeed effective in preventing cerebral embolization.

The effect of EP on neurologic imaging endpoints appeared to be uniform between SE and BE

valve types which differ significantly in term of design and implantation technique. While this analysis

could have been underpowered to detect statistically significant interactions, our results are consistent

with previous reports that failed to detect significant differences in stroke incidence and transcranial

Doppler-detected embolizations.18, 27

There was no evidence of safety concerns with intraprocedural EP, EP use was associated with

an overall device success of 97% and longer fluoroscopic time. The small, yet possibly clinically

relevant EP failure rate has to be interpreted carefully. Cardiac computed tomography (CT) scan has

become a standard of care in many TAVR centers as it allows the characterization of the ascending

aorta, aortic valve anatomy and calcifications, aortic root size and height of the coronary ostia from the

aortic annular plane.28, 29 Additionally, cardiac computed tomography (CT) angiography can be used to

9

assess peripheral vascular access sites, coronary and great epicardial vessels anatomy.29 Therefore

CTCA may play a pivotal role in screening patients suitable for EP and / or select the optimal type of EP

device according to the underlying anatomy.

We failed to detect a significant reduction in clinically overt stroke and all-cause mortality.

However both endpoints were numerically lower in the EP group. The present total sample size may be

insufficient to detect significant differences in overt stroke due to the rarity of this event. Larger RCTs

designed to detect differences in hard clinical endpoints and long-term neurocognitive assessment are

therefore warranted to provide conclusive evidence regarding the efficacy of EP during TAVR. For this

purpose, several larger RCTs investigating outcomes with EP are currently ongoing (NCT02214277;

NCT02536196). As a substantial amount of emboli during TAVR are of thrombotic origin

complementary antithrombotic strategies to EP are warranted. Recent studies demonstrated that both

alternative anticoagulation agents30 and anticoagulation regimens31 are safe in terms of neurologic

complications and possibly associated with a lower risk of bleeding compared with conventional

antithrombotic strategies. Conversely the role of antiplatelet agents in reducing the risk of intra- and

periprocedural cerebrovascular events has been poorly investigated. The role of antiplatelet agents to

prevent periprocedural thrombotic complications during TAVR may merit further investigation.

LimitationsThe present study has several limitations: (i) the present findings are subject to the inherent

limitations of the included RCTs due to study design, follow-up, imaging and neurocognitive

assessment drop-out and endpoint ascertainment; additionally, statistical heterogeneity was substantial

for some neuroimaging endpoints underscoring differences across trials; (ii) most of the valves

implanted in the included RCTs were SE or BE first-generation TAVR devices; as new-generation

TAVR devices seem to be associated with improved efficacy and safety compared with older

generations,32, 33 the magnitude of the benefit of EP may be attenuated with newer devices; (iii) due to

the relatively small sample size and low event rates the present study remains underpowered to detect

differences in hard clinical endpoints such as stroke and all-cause mortality; (iv) as longer term (≥ 1

year) follow-up was not available we cannot investigate the effect of intraprocedural EP on long-term

functional status.

ConclusionsNeuroprotection with EP devices during TAVR was associated with improved early imaging and

clinical neurologic outcomes. The neurologic benefits of EP appear to be consistent among valve types.

While the differences in overt stroke were not significant, use of intraoperative EP was associated with

10

a numeric stroke reduction which may become significant in larger RCTs powered for hard endpoints.

These results are hypothesis-generating and further prospective, adequately powered RCTs are

needed to establish the role of EP during TAVR.

Funding SourceNo external funding was available for this study.

DisclosuresDrs. Giustino, Faggioni, Chandrasekhar and Baber have no conflicts of interest to disclose. Drs.

Mehran and Dangas have received consultant and speaker honoraria (modest level) from Bristol-Myers

Squibb, Sanofi-Aventis, Eli Lilly, Daiichi Sankyo, Abbott Vascular, AstraZeneca, Boston Scientific, and

Johnson & Johnson. Dr. Veltkamp has received research support, consultant and speaker honoraria

from Bayer, Boehringer, BMS, Pfizer, Daiichi Sankyo, Medtronic, Morphosys, St. Jude medical,

Apoplex Medical technologies and Sanofi.

11

References1. Smith CR, Leon MB, Mack MJ, Miller DC, Moses JW, Svensson LG, Tuzcu EM, Webb JG,

Fontana GP, Makkar RR, Williams M, Dewey T, Kapadia S, Babaliaros V, Thourani VH, Corso

P, Pichard AD, Bavaria JE, Herrmann HC, Akin JJ, Anderson WN, Wang D, Pocock SJ.

Transcatheter versus surgical aortic-valve replacement in high-risk patients. N Engl J Med.

2011;364:2187-2198

2. Leon MB, Smith CR, Mack M, Miller DC, Moses JW, Svensson LG, Tuzcu EM, Webb JG,

Fontana GP, Makkar RR, Brown DL, Block PC, Guyton RA, Pichard AD, Bavaria JE, Herrmann

HC, Douglas PS, Petersen JL, Akin JJ, Anderson WN, Wang D, Pocock S. Transcatheter aortic-

valve implantation for aortic stenosis in patients who cannot undergo surgery. N Engl J Med.

2010;363:1597-1607

3. Giustino G, Dangas GD. Stroke prevention in valvular heart disease: From the procedure to

long-term management. EuroIntervention. 2015;11 Suppl W:W26-31

4. Nombela-Franco L, Webb JG, de Jaegere PP, Toggweiler S, Nuis RJ, Dager AE, Amat-Santos

IJ, Cheung A, Ye J, Binder RK, van der Boon RM, Van Mieghem N, Benitez LM, Perez S, Lopez

J, San Roman JA, Doyle D, Delarochelliere R, Urena M, Leipsic J, Dumont E, Rodes-Cabau J.

Timing, predictive factors, and prognostic value of cerebrovascular events in a large cohort of

patients undergoing transcatheter aortic valve implantation. Circulation. 2012;126:3041-3053

5. Mastoris I, Schoos MM, Dangas GD, Mehran R. Stroke after transcatheter aortic valve

replacement: Incidence, risk factors, prognosis, and preventive strategies. Clin Cardiol.

2014;37:756-764

6. Mack MJ, Brennan JM, Brindis R, Carroll J, Edwards F, Grover F, Shahian D, Tuzcu EM,

Peterson ED, Rumsfeld JS, Hewitt K, Shewan C, Michaels J, Christensen B, Christian A,

O'Brien S, Holmes D. Outcomes following transcatheter aortic valve replacement in the united

states. Jama. 2013;310:2069-2077

7. Kahlert P, Al-Rashid F, Dottger P, Mori K, Plicht B, Wendt D, Bergmann L, Kottenberg E,

Schlamann M, Mummel P, Holle D, Thielmann M, Jakob HG, Heusch G, Erbel R, Eggebrecht H.

Response to letters regarding article, "cerebral embolization during transcatheter aortic valve

implantation: A transcranial doppler study". Circulation. 2013;127:e591-592

8. Zimpfer D, Czerny M, Kilo J, Kasimir MT, Madl C, Kramer L, Wieselthaler GM, Wolner E, Grimm

M. Cognitive deficit after aortic valve replacement. Ann Thorac Surg. 2002;74:407-412;

discussion 412

9. Uekermann J, Suchan B, Daum I, Kseibi S, Perthel M, Laas J. Neuropsychological deficits after

mechanical aortic valve replacement. J Heart Valve Dis. 2005;14:338-343

12

10. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and

meta-analyses: The prisma statement. Bmj. 2009;339:b2535

11. Higgins JP, Altman DG, Gotzsche PC, Juni P, Moher D, Oxman AD, Savovic J, Schulz KF,

Weeks L, Sterne JA. The cochrane collaboration's tool for assessing risk of bias in randomised

trials. Bmj. 2011;343:d5928

12. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range,

and the size of a sample. BMC Med Res Methodol. 2005;5:13

13. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses.

Bmj. 2003;327:557-560

14. Linke A. Clean-tavi: A prospective, randomized trial of cerebral embolic protection in high-risk

patients with aortic stenosis undergoing transcatheter aortic valve replacement. Transcatheter

Cardiovascular Therapeutics. 2014;Late breaking trial presentation

15. Lansky AJ, Schofer J, Tchetche D, Stella P, Pietras CG, Parise H, Abrams K, Forrest JK,

Cleman M, Reinohl J, Cuisset T, Blackman D, Bolotin G, Spitzer S, Kappert U, Gilard M, Modine

T, Hildick-Smith D, Haude M, Margolis P, Brickman AM, Voros S, Baumbach A. A prospective

randomized evaluation of the triguard hdh embolic deflection device during transcatheter aortic

valve implantation: Results from the deflect iii trial. Eur Heart J. 2015;36:2070-2078

16. Wendt D, Kleinbongard P, Knipp S, Al-Rashid F, Gedik N, El Chilali K, Schweter S, Schlamann

M, Kahlert P, Neuhauser M, Forsting M, Erbel R, Heusch G, Jakob H, Thielmann M. Intraaortic

protection from embolization in patients undergoing transaortic transcatheter aortic valve

implantation. Ann Thorac Surg. 2015;100:686-691

17. Van Mieghem NM. Mistral-c: Mri investigation in tavi with claret. Transcatheter Cardiovascular

Therapeutics. 2015;Late breaking trial presentation

18. Kahlert P, Al-Rashid F, Dottger P, Mori K, Plicht B, Wendt D, Bergmann L, Kottenberg E,

Schlamann M, Mummel P, Holle D, Thielmann M, Jakob HG, Konorza T, Heusch G, Erbel R,

Eggebrecht H. Cerebral embolization during transcatheter aortic valve implantation: A

transcranial doppler study. Circulation. 2012;126:1245-1255

19. Rodes-Cabau J, Puri R. Filtering the truth behind cerebral embolization during transcatheter

aortic valve replacement. JACC Cardiovasc Interv. 2015;8:725-727

20. Van Mieghem NM, El Faquir N, Rahhab Z, Rodriguez-Olivares R, Wilschut J, Ouhlous M,

Galema TW, Geleijnse ML, Kappetein AP, Schipper ME, de Jaegere PP. Incidence and

predictors of debris embolizing to the brain during transcatheter aortic valve implantation. JACC

Cardiovasc Interv. 2015;8:718-724

13

21. Sun X, Lindsay J, Monsein LH, Hill PC, Corso PJ. Silent brain injury after cardiac surgery: A

review: Cognitive dysfunction and magnetic resonance imaging diffusion-weighted imaging

findings. J Am Coll Cardiol. 2012;60:791-797

22. Vogt G, Laage R, Shuaib A, Schneider A. Initial lesion volume is an independent predictor of

clinical stroke outcome at day 90: An analysis of the virtual international stroke trials archive

(vista) database. Stroke. 2012;43:1266-1272

23. Luby M, Bykowski JL, Schellinger PD, Merino JG, Warach S. Intra- and interrater reliability of

ischemic lesion volume measurements on diffusion-weighted, mean transit time and fluid-

attenuated inversion recovery mri. Stroke. 2006;37:2951-2956

24. Cumming TB, Bernhardt J, Linden T. The montreal cognitive assessment: Short cognitive

evaluation in a large stroke trial. Stroke. 2011;42:2642-2644

25. Sivakumar L, Kate M, Jeerakathil T, Camicioli R, Buck B, Butcher K. Serial montreal cognitive

assessments demonstrate reversible cognitive impairment in patients with acute transient

ischemic attack and minor stroke. Stroke. 2014;45:1709-1715

26. Hinkle JL. Reliability and validity of the national institutes of health stroke scale for neuroscience

nurses. Stroke. 2014;45:e32-34

27. Webb JG, Barbanti M. Cerebral embolization during transcatheter aortic valve implantation.

Circulation. 2012;126:1567-1569

28. Panoulas VF, Montorfano M, Figini F, Spagnolo P, Contri R, Giustino G, Agricola E, Franco A,

Latib A, Colombo A. Unanticipated pseudocoarctation highlights the importance of visualizing

aortic arch anatomy before transfemoral transcatheter aortic valve implantation. Circ Cardiovasc

Interv. 2014;7:631-633

29. Chieffo A, Giustino G, Spagnolo P, Panoulas VF, Montorfano M, Latib A, Figini F, Agricola E,

Gerli C, Franco A, Giglio M, Cioni M, Alfieri O, Camici PG, Colombo A. Routine screening of

coronary artery disease with computed tomographic coronary angiography in place of invasive

coronary angiography in patients undergoing transcatheter aortic valve replacement. Circ

Cardiovasc Interv. 2015;8:e002025

30. Dangas GD, Lefevre T, Kupatt C, Tchetche D, Schafer U, Dumonteil N, Webb JG, Colombo A,

Windecker S, Ten Berg JM, Hildick-Smith D, Mehran R, Boekstegers P, Linke A, Tron C, Van

Belle E, Asgar AW, Fach A, Jeger R, Sardella G, Hink HU, Husser O, Grube E, Deliargyris EN,

Lechthaler I, Bernstein D, Wijngaard P, Anthopoulos P, Hengstenberg C. Bivalirudin versus

heparin anticoagulation in transcatheter aortic valve replacement: The randomized bravo-3 trial.

J Am Coll Cardiol. 2015;66:2860-2868

14

31. Bernelli C, Chieffo A, Montorfano M, Maisano F, Giustino G, Buchanan GL, Chan J,

Costopoulos C, Latib A, Figini F, De Meo E, Giannini F, Covello RD, Gerli C, Franco A, Agricola

E, Spagnolo P, Cioni M, Alfieri O, Camici PG, Colombo A. Usefulness of baseline activated

clotting time-guided heparin administration in reducing bleeding events during transfemoral

transcatheter aortic valve implantation. JACC Cardiovasc Interv. 2014;7:140-151

32. Giustino G, Latib A, Panoulas VF, Montorfano M, Chieffo A, Taramasso M, Sato K, Agricola E,

Alfieri O, Colombo A. Early outcomes with direct flow medical versus first-generation

transcatheter aortic valve devices: A single-center propensity-matched analysis. J Interv

Cardiol. 2015;28:583-593

33. Gooley RP, Talman AH, Cameron JD, Lockwood SM, Meredith IT. Comparison of self-

expanding and mechanically expanded transcatheter aortic valve prostheses. JACC Cardiovasc

Interv. 2015;8:962-971

15

Figure LegendFigure 1. Neurologic Imaging Endpoints. Total lesion volume in mm3 (1A), mean number of new

ischemic lesions (1B) and patients with new ischemic lesions (1C). Results are reported as

standardized mean differences (with 95% confidence intervals [CI]) for continuous variables and risk

ratio (with 95% confidence intervals [CI]) for categorical variables.

Figure 2. Primary neurologic imaging endpoints per type of valve implanted. Total lesion volume

in mm3 (2A) and number of new lesions (2B).

Figure 3. Neurologic Clinical Endpoints. Any worsening in National Institutes of Health Stroke Scale

(NIHSS) at discharge (3A) and MoCA score at discharge (3B); clinically overt strokes at 30 days (3C).

Figure 4. All-cause Mortality. Risk of 30-day all-cause mortality.

16

Table 1. Characteristics of the included randomized controlled trials.

Trial Year

Device type

EP arm (n)

No EP arm (n)

Mean

age

Male

(%)

STS score

(%)

TAVR access

site

Type of THV

devicePrimaryendpoint

Type ofimaging

test

Timing of

imaging

Neurological

assessment

CLEAN-TAVI14

2014

Claret MontageTM dual-filter

50 50 80 43% 5.4%Femoral100% SE 100%

Number of ischemic brain lesions at 2 days vs baseline in protected territories

DW-MRI 3THigh-resolution T1-weighted anatomical image. Diffusion-weighted imaging (DWI) for ischemic lesions

Baseline andday 2-7-30 after TAVR

Serial assessment with NIHSS

DEFLECT-III15

2015

TriGuard HDH 46 39 82 45% 6.9%

Femoral96%transapical 4%

BE 63%SE 31%Other 6%

MACCE (all-cause mortality, all stroke,life-threatening bleeding, AKIstage 2 or 3, and major vascular complications)

DW-MRI

Baseline andDay 4±2 and 30 ±7 after TAVR

Serial assessment with NIHSS, mRS, MoCAcomputerized Cogstate Research Test

EMBOL-X16 2015

Claret Embol-X

14 16 82 40% 10.3% Transaortic 100%

BE 100%

Number and size of new ischemic cerebral lesion within 7 days after TAVR

DW-MRI 1.5T transversal fluid-attenuated inversion recover and transversal diffusion-weighted

Baseline andwithin a week after TAVR

Clinical observation

17

images of the whole brain

MISTRAL-C17

2015

Claret Sentinel 32 33 82 52% - NA BE 74%

SE 26%

New brain ischemic lesions 5 days after TAVR

DW-MRI 3T

Baseline andday 5

Serial assessment with CES-D, MMSE, MOCA

AKI=Acute Kidney Injury; BE=Balloon-Expandable; CES-D= Center for Epidemiological Studies-Depression; DW-MRI= Diffusion Weighted Magnetic Resonance Imaging; EP=Embolic Protection; MACCE= Major Adverse Cardiac and Cerebrovascular Event; MoCA= Montreal Cognitive Assessment; NA=Not Available; NIHSS= National Institute of Health Stroke Scale; nRS= modified Rankin Scale; SE= Self-Expandable; TAVR=Transcatheter Aortic Valve Replacement; THV=Transcatheter Heart Valve.

18