TAVI (Transcatheter Aortic Valve Implantation)

Definition

Transcatheter aortic valve implantation (TAVI) is a catheter-based technique for delivering and deploying a bioprosthetic aortic valve without open-heart surgery. Most commonly performed via transfemoral access. TAVI has progressively expanded from inoperable/high-risk patients to intermediate- and low-risk populations.

Key Concepts

Evidence Base — TAVI vs SAVR

• High-risk patients: TAVI non-inferior or superior to SAVR (PARTNER 1, CoreValve High Risk)
• Intermediate-risk: TAVI non-inferior (PARTNER 2, SURTAVI)
• Low-risk patients: TAVI noninferior and directionally superior to SAVR at 1 year in DEDICATE-DZHK6 (HR 0.53); no significant difference in composite death/stroke/rehospitalization at 5 years in PARTNER 3 (HR 0.79; P=0.07) — early TAVI advantage attenuates over time
• TAVI advantages: lower AF, AKI, major bleeding, hospital stay; SAVR advantages: lower pacemaker rate, lower paravalvular leak sources/vhd-esc-2025 very high sources/tavr-dedicate-nejm-2024 very high

PARTNER 3 — 5-Year Outcomes in Low-Risk Patients (SAPIEN 3)

• n=1,000; mean STS-PROM 1.9%; 71 sites; transfemoral TAVR only; mean age 73 years sources/tavr-partner3-5yr-nejm-2023 very high
• First primary endpoint (death/stroke/rehospitalization): TAVR 22.8% vs SAVR 27.2%; difference −4.3 pp (P=0.07; HR 0.79 [0.61–1.02]) — trend favors TAVR but not statistically significant
• Second primary endpoint (hierarchical win ratio): 1.17 (P=0.25) — also non-significant
• Temporal attenuation: TAVR was superior at 1–2 years; by 5 years this advantage was lost; landmark years 1–5 analysis shows HR 1.17 for TAVR vs SAVR events (NS), with late deaths more common in TAVR
• Mortality: TAVR 10.0% vs SAVR 8.2% (odds ratio 1.24 [0.79–1.97]; NS) — mortality numerically higher in TAVR by year 5
• Stroke: TAVR 5.8% vs SAVR 6.4% (HR 0.87; NS)
• Rehospitalization: TAVR 13.7% vs SAVR 17.4% (HR 0.75; NS) — persistent benefit
• New-onset AF: TAVR 13.7% vs SAVR 42.4% — dramatically and persistently lower with TAVR
• Pacemaker implantation: TAVR 13.5% vs SAVR 10.4%
• Clinically significant valve thrombosis (VARC-3): TAVR 2.5% vs SAVR 0.2% — significantly more with TAVR; none fatal; 3 strokes among thrombosis patients
• Bioprosthetic-valve failure (VARC-3): TAVR 3.3% vs SAVR 3.8% (similar); structural SVD 1.4% vs 2.0%
• Paravalvular AR (mild+): TAVR 20.8% vs SAVR 3.2% — but not associated with higher TAVR mortality at 5 years
• Alive with normally functioning valve: TAVR 86.3% vs SAVR 87.4%
• Health status at 5 years: KCCQ-OS TAVR 86.2 vs SAVR 85.9; alive with KCCQ ≥75: 71.0% vs 71.9% — equivalent sources/tavr-partner3-5yr-nejm-2023 very high

DEDICATE-DZHK6 — Industry-Independent Pragmatic TAVI vs SAVR RCT (NEJM 2024)

• Design: Investigator-initiated, randomised, open-label noninferiority trial; 38 German sites; May 2017–September 2022; 1,414 patients sources/tavr-dedicate-nejm-2024 very high
• Key distinctive feature: All CE-marked prostheses permitted — operator/heart team freely selected device after randomisation. No industry sponsorship (funded by German Center for Cardiovascular Research + German Heart Foundation). Most pragmatic and generalisable TAVI–SAVR RCT.
• Population: Age ≥65 years (mean 74±4); 57% male; median STS-PROM 1.8% (predominantly low surgical risk); bicuspid AV excluded; clinically significant untreated CAD excluded
• Primary endpoint (death from any cause or fatal/nonfatal stroke at 1 year): TAVI 5.4% vs SAVR 10.0%; HR 0.53 (95% CI 0.35–0.79; P<0.001 for noninferiority) — upper 95% CI 0.79 is well below both the null (1.0) and the noninferiority margin (1.14)
• All-cause death: 2.6% vs 6.2% (HR 0.43; 95% CI 0.24–0.73); CV death: 2.0% vs 4.4% (HR 0.47)
• Stroke: 2.9% vs 4.7% (HR 0.61; NS); disabling stroke: 1.3% vs 3.1% (HR 0.42)
• New-onset AF: 12.4% vs 30.8% (HR 0.36) — dramatically lower with TAVI, consistent with all prior trials
• Permanent pacemaker: 11.8% vs 6.7% (HR 1.81) — significantly higher with TAVI
• Major/life-threatening bleeding: 4.3% vs 17.2% (HR 0.24); vascular access complications: 7.9% vs 0.7% (HR 10.64)
• Prosthetic-valve dysfunction: 1.6% vs 0.6% (HR 2.44; NS trend higher with TAVI)
• Valve hemodynamics at 1 year: comparable mean gradients (10 vs 11 mmHg) and effective orifice area (1.6 cm² both); AR ≥moderate 2.8% vs 1.0% — higher with TAVI, requires longer follow-up
• Procedural: TAVI median 48 min; 97.3% transfemoral; 74.7% home discharge. SAVR median 165 min; 50.9% full sternotomy; 40.4% home discharge
• Context: SAVR death rate (6.2%) notably higher than PARTNER 3 (~2% at 1 year) — attributed to COVID-19 pandemic enrollment (majority of patients enrolled during pandemic) and higher proportion of women (associated with higher post-SAVR mortality). TAVI death rate (2.6%) also slightly higher than PARTNER 3, consistent with older DEDICATE population.
• 5-year primary outcome reassessment planned — critical for durability and long-term mortality assessment

Expanded Indications (2025)

• Symptomatic severe AS: All risk levels with appropriate anatomy (Heart Team decision)
• Asymptomatic severe AS with early intervention (new Class IIa): Alternative to watchful waiting in patients with LVEF ≥50%, normal exercise test, low procedural risk — supported by EARLY TAVR, RECOVERY, AVATAR, EVoLVeD trials
• BAV stenosis at high surgical risk (new Class IIb B): If anatomy suitable per Heart Team — historically excluded from major RCTs
• Severe AR in inoperable patients (new Class IIb B): If anatomy suitable — off-label use of non-dedicated devices carries ~10% risk of second valve implantation or surgical conversion; dedicated TAVI devices associated with high pacemaker rate (~24%)
• Moderate AS + HFrEF: insufficient evidence for routine TAVI at this time sources/vhd-esc-2025 very high

EARLY TAVR Trial — TAVR vs Surveillance for Asymptomatic Severe AS (NEJM 2025)

• Design: Prospective, multicenter, open-label, 1:1 RCT; 75 US/Canadian centres; 901 patients; March 2017–December 2021; median follow-up 3.8 years sources/tavi-earlytavr-nejm-2025 very high
• Device: Transfemoral SAPIEN 3 or SAPIEN 3 Ultra (Edwards Lifesciences); median time to procedure 14 days
• Population: Age ≥65 years (mean 75.8); STS-PROM 1.8%; 83.6% low surgical risk; LVEF ≥50% (mean 67.4%); confirmed asymptomatic by negative treadmill stress test in 90.6%; peak AV velocity 4.3 m/s; KCCQ 92.7
• Primary endpoint (death + stroke + unplanned CV hospitalisation): 26.8% TAVR vs 45.3% surveillance; HR 0.50 (95% CI 0.40–0.63; P<0.001) — superiority demonstrated
• All-cause death: 8.4% vs 9.2% (HR 0.93; NS) — no mortality benefit at median 3.8 years
• Stroke: 4.2% vs 6.7% (HR 0.62; 95% CI 0.35–1.10; NS) — directionally lower, not significant
• Unplanned CV hospitalisation: 20.9% vs 41.7% (HR 0.43; P<0.001) — dominant composite driver; includes 105 early conversions (<6 months) counted as events
• HF hospitalisation (exploratory): HR 0.32 (0.18–0.58) — strongly favours early TAVR
• Favourable outcome at 2yr (alive + stable KCCQ ≥75): 86.6% vs 68.0% (P<0.001)
• Integrated LV/LA health at 2yr: 48.1% vs 35.9% (P=0.001) — superior cardiac remodelling with early TAVR
• Surveillance crossover: 87.0% of surveillance arm underwent AVR (median 11.1 months); 39.2% had advanced symptoms before conversion; 11 deaths before conversion; declining KCCQ (−14.8 pts), rising NT-proBNP (298→462 pg/mL), and falling 6MWT (−46.4 m) confirm cardiac damage accumulates during watchful waiting
• No peri-procedural mortality (30-day): 0% CV death in both groups; 0.9% vs 1.8% 30-day stroke (NS) sources/tavi-earlytavr-nejm-2025 very high

TAVI vs SAVR — Selection Factors

Heart Team decision integrating:

• Age and life expectancy: SAVR preferred in younger patients (<65 years aortic; <70 years mitral) with longer life expectancy; TAVI acceptable in older patients with limited life expectancy
• Anatomical factors: Narrow aortic root, low coronary ostia, commissural misalignment, severe LVOT calcification, BAV morphology → favour SAVR or require careful TAVI planning
• Procedural risk: EuroSCORE II or STS-PROM; high surgical risk favours TAVI
• Patient preference and values: Informed patient preference is Class I consideration sources/vhd-esc-2025 very high

Cerebral Embolic Protection (CEP) during TAVI — Routine Use NOT Recommended

• CEP devices (Sentinel, Boston Scientific) deploy filters in the left common carotid artery and right innominate artery via right radial access to capture embolic debris during TAVI
• BHF PROTECT-TAVI (NEJM 2025): n=7,635; 33 UK centres; largest RCT of CEP; stopped for futility — 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 1.2% vs 1.4% (NS); death 0.8% vs 0.7% (NS); serious adverse events 0.6% vs 0.3% (doubled with CEP); lower 99% CI excluded ≥40% RRR sources/cep-tavi-bhfprotecttavi-nejm-2025 very high
• PROTECTED TAVR (NEJM 2022): n=3,000; 51 international sites; primary stroke endpoint NS; disabling stroke numerically lower (not significant); concordant with BHF PROTECT-TAVI
• Both trials confirm: routine CEP does not reduce clinical stroke after TAVI — no role for routine CEP based on current evidence
• See concepts/Cerebral-Embolic-Protection-TAVI for full evidence synthesis, device mechanism, and open questions

CAD Management Before TAVI

• CCTA sufficient to rule out significant CAD if procedural planning CT is of sufficient quality (new Class IIa)
• PCI Class IIa B: ≥90% stenosis in vessels ≥2.5 mm before TAVI — informed by NOTION-3 sources/PCI-TAVI-NOTION3-NEJM-2024 very high
• PCI Class IIb B: ≥70% stenosis in proximal segments of main vessels before any transcatheter valve intervention
• Consider coronary access post-TAVI: valve frame height, commissural alignment, narrow aortic root may limit future coronary access sources/vhd-esc-2025 very high

NOTION-3 Trial — PCI + TAVI vs Conservative + TAVI (NEJM 2024)

• International open-label RCT; n=455; 12 Nordic-Baltic sites; September 2017–October 2022; median 2-year follow-up sources/PCI-TAVI-NOTION3-NEJM-2024 very high
• Inclusion: Severe symptomatic AS + ≥1 coronary stenosis with FFR ≤0.80 or ≥90% diameter stenosis in vessel ≥2.5 mm; median age 82; STS-PROM 3%; SYNTAX score 9 (low complexity)
• Primary MACE endpoint (death/MI/urgent revascularisation at 2 years): PCI 26% vs conservative 36%; HR 0.71 (95% CI 0.51–0.99; P=0.04) — significant superiority of PCI
• MI: HR 0.54 (P<0.05); Urgent revascularisation: HR 0.20 (P<0.01) — both significantly reduced; all-cause death HR 0.85 (NS)
• Safety: Bleeding higher in PCI group (28% vs 20%; HR 1.51; P<0.05); acute kidney failure lower (5% vs 11%; HR 0.45; P<0.05)
• PCI was staged before TAVI in 74% of patients; complete revascularisation in 89%
• Supersedes the earlier ACTIVATION trial (n=235; 1-year; anatomic CAD; inconclusive) by using FFR guidance, larger sample, and longer follow-up
• See concepts/PCI-Before-TAVI for full evidence synthesis, ACTIVATION comparison, and FFR reliability issues in severe AS

Procedural Planning

• CCT mandatory: annulus dimensions, aortic root anatomy, coronary ostia height, LVOT calcification, optimal fluoroscopic projections, iliofemoral access assessment
• TOE or CMR as alternatives when CCT is difficult (renal failure, arrhythmia)
• Commissural alignment optimisation important for future coronary access

Antithrombotic Therapy After TAVI — ESC 2025

• No OAC indication:
  • ASA 75–100 mg/day for 12 months: Class I A (POPular TAVI cohort A)
  • Long-term ASA after 12 months: Class IIa C
  • DAPT: not recommended (Class III B) — no additional benefit, excess bleeding
  • Routine OAC: not recommended (Class III A) — GALILEO trial halted for harm (rivaroxaban + ASA)
• With OAC indication (AF, etc.): OAC alone Class I B; OAC + clopidogrel not recommended (excess bleeding, POPular TAVI cohort B)
• DOAC vs VKA after TAVI with OAC indication: no definitive superiority of either (ATLANTIS, ENVISAGE-TAVI AF) sources/vhd-esc-2025 very high

Antithrombotic Therapy After TAVI — ACC/AHA 2020

• No OAC indication: ASA 75–100 mg daily Class IIa B-R; DAPT (ASA + clopidogrel 75 mg) for 3–6 months Class IIb B-NR; VKA INR 2.5 for ≥3 months Class IIb B-NR
• Low-dose rivaroxaban 10 mg + ASA: Class III:Harm B-R — contraindicated without other OAC indication (GALILEO harm)
• Note: ESC 2025 upgraded ASA to Class I A and made DAPT Class III; ACC/AHA 2020 had less definitive evidence and rated ASA as Class IIa sources/VHD-AHA-2020 very high

Post-Procedural Follow-up

• TTE within 3 months, at 1 year, and annually thereafter
• Monitor for: structural valve deterioration (SVD), hypo-attenuated leaflet thickening (HALT), paravalvular leak, new conduction abnormalities
• Right heart catheterisation not routinely needed; CCT/PET if prosthetic valve thrombosis or endocarditis suspected

NCS Timing After TAVI

• NCS early after TAVI (COR 2a): Noncardiac surgery may proceed even within the first 30 days after a successful TAVI in patients with time-sensitive surgical conditions — in contrast to DES where NCS ≤30 days carries high thrombotic risk. (sources/periop-aha-2024, rating: very high)
• Rationale: Successful TAVI relieves the fixed LVOT obstruction; the predominant perioperative risk shifts back to the non-valvular comorbidities and general surgical risk rather than valve-related haemodynamics. (sources/periop-aha-2024, rating: very high)
• Continue aspirin perioperatively: COR 2a — aspirin should be maintained for antiplatelet coverage following TAVI during perioperative period; withholding risks early valve leaflet thrombosis. (sources/periop-aha-2024, rating: very high)
• NCS after MV TEER (transcatheter edge-to-edge repair, e.g., MitraClip): Proceeding with NCS after successful MV TEER is also reasonable (COR 2a); same principle as post-TAVI — resolution of the primary valvular pathology removes the key haemodynamic obstacle. (sources/periop-aha-2024, rating: very high)

Contradictions / Open Questions

• NOTION-3: MACE benefit without mortality benefit and with bleeding excess: PCI in TAVI candidates reduces MACE (HR 0.71; P=0.04) driven by MI and urgent revascularisation, but all-cause death is neutral (HR 0.85; NS) and bleeding is higher (HR 1.51). The decision to perform PCI must be individualised accounting for bleeding risk. Urgent revascularisation (HR 0.20) — partly driven by open-label knowledge of untreated lesions — may overstate the true benefit. sources/PCI-TAVI-NOTION3-NEJM-2024 very high
• NOTION-3 vs ACTIVATION discordance: ACTIVATION (n=235; anatomic ≥70%; 1 year; inconclusive/futility) was neutral; NOTION-3 (n=455; FFR ≤0.80 or ≥90%; 2 years; positive). FFR guidance, larger sample, and longer follow-up are the most likely explanations. sources/PCI-TAVI-NOTION3-NEJM-2024 very high
• Long-term durability beyond 8–10 years remains the critical unknown — 10-year PARTNER 3 data are planned; valve-in-valve procedures for TAVI failure are feasible but associated with higher residual gradients
• Late mortality attenuation: PARTNER 3 showed TAVR superiority at 1–2 years reverting to non-significant difference at 5 years, with more late deaths in TAVR arm (10.0% vs 8.2%; HR 1.24 NS) — unclear whether this reflects valve biology, patient selection, differential comorbidities, or Covid-19 effect sources/tavr-partner3-5yr-nejm-2023 very high
• Valve thrombosis signal: TAVR 2.5% vs SAVR 0.2% by VARC-3 criteria in PARTNER 3 — reason for higher TAVI valve thrombosis remains speculative; possible differential anticoagulation patterns in early period; long-term implications for durability and stroke risk unknown sources/tavr-partner3-5yr-nejm-2023 very high
• DEDICATE high SAVR mortality vs prior trials: SAVR all-cause death at 1 year was 6.2% in DEDICATE vs ~2% at 1 year in PARTNER 3 (low-risk). Trial investigators attribute this to COVID-19 pandemic enrollment and higher female proportion. However, this inflates the apparent TAVI benefit and makes cross-trial comparison unreliable. sources/tavr-dedicate-nejm-2024 very high sources/tavr-partner3-5yr-nejm-2023 very high
• DEDICATE noninferiority vs claimed superiority: DEDICATE was designed as a noninferiority trial (margin HR ≤1.14). The primary HR 0.53 (upper CI 0.79) strongly exceeds noninferiority and is statistically consistent with superiority, but the trial was not powered or designed to formally claim it. Future interpretation should be cautious about the directional superiority language. sources/tavr-dedicate-nejm-2024 very high
• DEDICATE 70 SAVR-to-TAVI crossovers: 70 SAVR-allocated patients ultimately underwent TAVI (mostly patient preference). As-treated analysis concordant with ITT, but this crossover may attenuate the SAVR mortality estimate. sources/tavr-dedicate-nejm-2024 very high
• CEP null result despite debris capture: DWI-MRI detects new cerebral lesions in >70% of TAVI patients, and CEP consistently captures debris. Yet two large RCTs (PROTECTED TAVR n=3,000; BHF PROTECT-TAVI n=7,635) show no reduction in clinical stroke. The mechanistic disconnect likely reflects: (1) hemorrhagic and hypoperfusion strokes not prevented by filters; (2) vertebrobasilar territory unprotected; (3) subclinical MRI lesions rarely cause detectable deficits. Routine CEP is not supported. sources/cep-tavi-bhfprotecttavi-nejm-2025 very high
• TAVI for AR: limited evidence; dedicated devices promising but high pacemaker rate; no RCT vs SAVR
• TAVI for BAV: excluded from most major RCTs (including PARTNER 3); observational data suggest higher rates of paravalvular leak, second valve implantation, and pacemaker requirement
• Optimal antithrombotic strategy in the first 12 months after TAVI with concomitant OAC: VKA vs DOAC — still debated
• EARLY TAVR primary endpoint vs hard outcomes discordance: Composite primary (HR 0.50; P<0.001) substantially driven by unplanned CV hospitalisations (HR 0.43) including 105 early conversions counted as events. Death (HR 0.93; NS) and stroke (HR 0.62; NS) not individually significant. Trial effectively compares immediate vs delayed TAVR (median 11.1 months delay) rather than TAVR vs indefinite surveillance. sources/tavi-earlytavr-nejm-2025 very high
• EARLY TAVR vs RECOVERY mortality discordance: RECOVERY (10yr, SAVR, HR 0.42 all-cause mortality) vs EARLY TAVR (3.8yr, TAVR, HR 0.93 NS) — insufficient follow-up in EARLY TAVR; 87% crossover at 11.1 months dilutes surveillance-arm mortality; RECOVERY enrolled very severe AS (Vmax ≥4.5 m/s) vs mean 4.3 m/s in EARLY TAVR sources/tavi-earlytavr-nejm-2025 very high sources/as-recovery-nejm-2026 very high

Connections

• Related to concepts/Aortic-Stenosis
• Related to concepts/Aortic-Regurgitation
• Related to concepts/Valvular-Heart-Disease
• Related to concepts/Structural-Valve-Deterioration
• Related to entities/Atrial-Fibrillation
• Related to concepts/Perioperative-Cardiovascular-Assessment
• Related to concepts/PCI-Before-TAVI — concomitant CAD management in TAVI candidates; NOTION-3 evidence
• Related to concepts/Cerebral-Embolic-Protection-TAVI — routine CEP null result; two large RCTs
• Related to entities/Chronic-Coronary-Disease — stable CAD comorbidity in TAVI population
• Related to sources/periop-aha-2024

Sources

• sources/VHD-AHA-2020
• sources/periop-aha-2024
• sources/vhd-esc-2025
• sources/tavr-partner3-5yr-nejm-2023
• sources/tavr-dedicate-nejm-2024
• sources/PCI-TAVI-NOTION3-NEJM-2024
• sources/tavi-earlytavr-nejm-2025
• sources/cep-tavi-bhfprotecttavi-nejm-2025