TrialsgrambetaRCTs are beautiful
TCT 2026Presented Tue 3 Nov · 12:38 GMT-8Late-Breaking Clinical Science: Session III, in Collaboration with the Journal of the American Medical AssociationAll Late-Breaking trials →

PROTEMBO

2026

Cerebral Protection in Transcatheter Aortic Valve Replacement: The PROTEMBO IDE Trial

In patients with symptomatic severe aortic stenosis undergoing TAVR, the ProtEmbo cerebral protection device did not reduce total new lesion volume on brain MRI compared with the Sentinel device or no cerebral embolic protection.

Result

Primary endpoint: MACCE at 30 days; total new lesion volume on diffusion-weighted MRI at 36 hours

MACCE at 30 days with ProtEmbo 7.6%, with an upper one-sided 99.6% credible bound of 15.0%. Median total new lesion volume 499 mm3 with ProtEmbo vs 340 mm3 with Sentinel vs 329 mm3 with no cerebral embolic protection, not significantly different across the three arms. Any stroke 3.5% vs 2.8% vs 5.8%, with disabling stroke 1.4% in all three groups. Life-threatening or disabling bleeding 2.8% in each protection arm vs 4.3% with no protection. ProtEmbo deployment was successful in 97.8%.

Original publication
Subtopic
Cerebral embolic protection
Design
Multicentre (20 sites in the US, Germany and Poland), double-blind randomised IDE trial (Protembis); TAVR with the ProtEmbo cerebral embolic protection device versus the Sentinel device, with an unprotected control arm
Enrolled
305 patients
Source
Kodali S, et al. Presented at TCT 2026, 3 November. Reported by TCTMD (CRF). Not yet published.

Discussion & critique

The trial shows that ProtEmbo met its 30-day safety goal but did not reduce total new lesion volume on MRI compared with Sentinel or no protection. It was stopped early, so it is smaller than planned and the efficacy comparison rests on an imaging surrogate rather than on clinical stroke. Stroke rates were numerically lower in the protection arms, but events were few and the report gives no formal test for them. The report does not describe blinding or full subgroup data, and the trial was sponsored by Protembis.