Research · Mar 24, 2026
Thrombosis after TAVI: frequency, clinical impact and valve design
The two sources
- Jacobsen MR, Overduin DC, Sørensen R, Andreotti F, Schulte C, Huber K, et al. Transcatheter aortic valve thrombosis. EuroIntervention. 2026;22(6):e331-e346. doi:10.4244/EIJ-D-25-00512. State-of-the-art review.
- Yamamoto M, Yashima F, Shirai S, Tada N, Naganuma T, Yamawaki M, et al. Performance and outcomes of the SAPIEN 3 Ultra RESILIA transcatheter heart valve in the OCEAN-TAVI registry. EuroIntervention. 2024;20(9):579-590. doi:10.4244/EIJ-D-23-00996.
Frequency and clinical impact
On 4D CT, hypoattenuated leaflet thickening (HALT) appears in 12% to 18% of patients at 3 months and 25% to 30% at 1 year. Clinically significant valve thrombosis ranges from under 1% to 2.8%. The central problem is the gap between detection and clinical impact.
Most studies find no link between subclinical HALT and mortality or stroke. One signal stands out: bioprosthetic valve failure at 3 years was higher with early HALT, HR 6.10 (95% CI 2.59 to 14.29). To me, durability may be the real concern.
Mechanisms
The mechanisms follow a modified Virchow triad:
- flow stasis: implant height, neosinus volume;
- hypercoagulability: von Willebrand factor upregulation after TAVI;
- endothelial injury: overexpansion, commissural misalignment.
All three are potentially modifiable.
Valve design
In a systematic review, subclinical thrombosis was lower with supra-annular prostheses (7.0%) than with intra-annular ones (12.1%). Underexpansion is a consistent HALT predictor and may partly explain the negative ACURATE IDE result, with about 20% underexpansion.
SAPIEN 3 Ultra RESILIA in OCEAN-TAVI
Yamamoto et al. compared 618 consecutive patients with SAPIEN 3 Ultra RESILIA (S3UR) to 8750 historical SAPIEN 3 controls. In the propensity-matched cohort (618 per group):
| Outcome | S3UR | SAPIEN 3 |
|---|---|---|
| Paravalvular leak ≥ mild | 13.0% | 23.3% |
| Mean gradient ≥ 20 mmHg | 1.6% | 5.1% |
| Any patient-prosthesis mismatch | 5.7% | 14.1% |
The largest gains were with 20 and 23 mm valves, historically the weak spot of balloon-expandable platforms.
Less mismatch and lower gradients mean less turbulence and better neosinus washout. In the Midha et al. simulator, deeper implantation increased the stagnant flow area 7-fold. In theory, better hemodynamics could lower thrombosis risk. The Yamamoto study did not measure thrombosis.
Inflammation and treatment
In Co-STAR, colchicine around TAVI reduced leaflet thickening from 54.2% to 27.1% against placebo. This suggests a role for inflammation in early thrombus formation and a treatment path beyond anticoagulation.
Anticoagulation resolves HALT, but GALILEO and ATLANTIS showed harm from routine anticoagulation after TAVI in patients without another indication. Today we treat clinically significant valve thrombosis and watch subclinical HALT.
Ongoing trials
| Trial | n | Comparison |
|---|---|---|
| POPular ATLANTIS | 2500 | CT-guided personalized therapy |
| NOTION-4 | 324 | DOAC vs aspirin |
| REACT-TAVI2 | 1206 | Ticagrelor vs aspirin |
Results are expected between 2027 and 2029.
Conclusion
Thrombosis after TAVI is common, and its clinical impact is uncertain, especially for durability. Valve design with better hemodynamics may lower the risk. Randomized trials linking prosthesis features to hard endpoints are still missing.