Cerebral Embolic Protection during TAVI
Definition
Cerebral embolic protection (CEP) refers to intravascular filter devices deployed in the supra-aortic vessels during transcatheter aortic valve implantation (TAVI) to capture debris released during the procedure and prevent it from reaching the cerebral circulation. The Sentinel device (Boston Scientific) is the only CEP device approved for clinical use in the United States and Europe. It is delivered via right radial artery access, deploying a distal filter in the left common carotid artery and a proximal filter in the right innominate artery.
Key Concepts
Rationale for CEP
- TAVI-related stroke occurs via three mechanisms: embolism (debris from valve/aorta), hemorrhage, and cerebral hypoperfusion
- Embolic debris is consistently captured in CEP filters in observational studies; DWI-MRI detects new cerebral lesions in >70% of patients after TAVI regardless of clinical stroke
- CEP was designed to intercept the embolic component — the most mechanistically targetable pathway
- sources/cep-tavi-bhfprotecttavi-nejm-2025
very high
Clinical Trial Evidence
BHF PROTECT-TAVI (NEJM 2025) — n=7,635 — DEFINITIVE NULL
- Largest RCT of CEP; 33 UK centres; October 2020–October 2024; BHF-funded
- Primary endpoint (stroke within 72h): 2.1% CEP vs 2.2% control; diff −0.02 pp; 95% CI −0.68 to 0.63; P=0.94
- Disabling stroke (mRS ≥2 + ≥1-point increase at 6–8 weeks): 1.2% vs 1.4% (diff −0.2 pp; NS)
- Severe stroke (NIHSS ≥10): 0.5% vs 0.5% (NS)
- Death within 72h: 0.8% vs 0.7% (NS); death within 8 weeks: 2.1% vs 1.9% (NS)
- Serious adverse events: 0.6% vs 0.3% (numerically doubled with CEP)
- Access-site complications at 6–8 weeks slightly higher with CEP (0.8% vs 0.4%; p<0.05) — primarily minor
- Stopped for futility: lower 99% CI excluded ≥40% relative risk reduction
- CACE analysis (adjusting for non-adherence): concordant — no treatment effect
- Device fully deployed (both filters) in 81.2%; ≥1 filter in 87.5%
- sources/cep-tavi-bhfprotecttavi-nejm-2025
very high
PROTECTED TAVR (NEJM 2022, Kapadia et al.) — n=3,000 — NULL
- 51 sites (North America, Europe, Australia); stopped early per prespecified enrollment rules
- Stroke within 72h: NS (incidence 2.6% overall — higher than BHF PROTECT-TAVI)
- Disabling stroke: numerically lower with CEP (not statistically significant for primary outcome)
- Consistent with BHF PROTECT-TAVI on the primary endpoint
- More restrictive eligibility than PROTECT-TAVI; definition of disabling stroke (mRS) differed from BHF trial stroke definition
- sources/cep-tavi-bhfprotecttavi-nejm-2025
very high(cited comparator)
Why CEP May Not Work Clinically Despite Capturing Debris
- Only embolic stroke component is intercepted; hemorrhagic and hypoperfusion strokes are not prevented by filters
- CEP covers only two of four supra-aortic vessels — vertebrobasilar territory is unprotected
- Subclinical MRI lesions captured in observational data may not translate to clinically detectable stroke
- Patient selection: less restrictive eligibility in BHF PROTECT-TAVI may have included complex anatomy where filter deployment is suboptimal
- Stroke timing: majority of strokes occur within 24h, but late embolic events and AF-related stroke are not reduced by procedural CEP
Device Deployment Rate
- BHF PROTECT-TAVI: both filters 81.2%; ≥1 filter 87.5% — no learning curve effect detected across quartiles of recruitment
- PROTECTED TAVR: higher deployment rates with more restrictive anatomy criteria
- Imperfect deployment in ~13–19% of patients could attenuate any true CEP benefit
Contradictions / Open Questions
- DWI-MRI subclinical lesions vs clinical stroke discordance: CEP consistently reduces DWI-MRI-detected cerebral lesions in observational studies and small randomized trials. Yet two large RCTs show no reduction in clinical stroke. The disconnect may reflect: (1) subclinical lesions rarely cause detectable deficits; (2) non-embolic stroke mechanisms dominate; (3) filter gaps (vertebrobasilar territory) miss a major embolic pathway. sources/cep-tavi-bhfprotecttavi-nejm-2025
very high - Disabling stroke signal in PROTECTED TAVR vs BHF PROTECT-TAVI: PROTECTED TAVR showed a numerically lower disabling stroke rate with CEP (pre-specified secondary); BHF PROTECT-TAVI did not replicate this (1.2% vs 1.4%; NS). Definition of disabling stroke differed (mRS threshold, ascertainment timing), making cross-trial comparison imprecise. sources/cep-tavi-bhfprotecttavi-nejm-2025
very high - Subgroup analyses: Both BHF PROTECT-TAVI prespecified subgroup analyses and PROTECTED TAVR failed to identify a population enriched for CEP benefit. Whether high-risk anatomical subgroups (large aortic arch debris burden, calcified annulus, atherosclerotic aorta) might benefit remains exploratory and unvalidated.
- Serious adverse events doubled with CEP (0.6% vs 0.3%): The mechanism is unclear. Access-site complications at 6–8 weeks (0.8% vs 0.4%; p<0.05) likely reflect the additional radial artery puncture. Whether the doubling of serious adverse events reflects this or other device-related harm warrants monitoring. sources/cep-tavi-bhfprotecttavi-nejm-2025
very high
Connections
- Related to concepts/TAVI — primary procedural context
- Related to concepts/Aortic-Stenosis — underlying disease
- Related to concepts/Stroke-Prevention — clinical goal