Subclinical Leaflet Thrombosis and Subclinical Aortic Valve Complex Thrombosis in TAVR

Antithrombotic therapy following TAVR remains a dynamic and rapidly evolving area of investigation. Recent clinical trials have explored various antithrombotic strategies, aiming to define optimal regimens that balance thromboembolic prevention with bleeding risk while supporting individualized patient care 73 ( Table 2 ). The current focus centers on tailoring early intensive antithrombotic therapy postimplantation while minimizing long-term bleeding risks. The dichotomy between favorable imaging findings and increased clinical adverse events with OAC compared to antiplatelet therapy in patients without a formal indication for OAC has been highlighted by key trials such as the ATLANTIS (Antithrombotic Strategy After Trans-Aortic Valve Implantation for Aortic Stenosis) 4D-CT and GALILEO-4D studies. 25 , 33 The GALILEO trial examined a rivaroxaban-based regimen for patients without long-term anticoagulation indication. The CT substudy demonstrated a reduction in HALT at 90 days post-TAVR, but the main trial revealed a concerning increase in mortality, thromboembolic events, and bleeding complications compared to an antiplatelet-based strategy. Although the ATLANTIS trial's 4D CT substudy demonstrated that apixaban reduced valve thrombosis compared with antiplatelet therapy, there was a significant increase in noncardiovascular mortality in the apixaban arm in stratum 2 (patients without an indication of OAC) in the main trial reported. These findings should be interpreted with caution and in light of the main ATLANTIS trial results, which showed that apixaban was not superior to the standard of care. The ADAPT-TAVR trial evaluated an edoxaban-based regimen (60 mg or 30 mg daily) vs dual antiplatelet therapy for the prevention of leaflet thrombosis following TAVR. At 6 months, there was an absolute reduction of 8.5% in leaflet thrombosis with edoxaban; but this difference did not reach statistical significance. Moreover, the reduction in leaflet thrombosis was not associated with a corresponding decrease in new cerebral lesions on magnetic resonance imaging or improvements in neurological or neurocognitive outcomes. The findings of this trial were limited by being underpowered, constraining the ability to draw definitive conclusions regarding clinical efficacy. Table 2 Comparative Overview of Designs and Key Outcomes in the GALILEO, LRT 2.0, ADAPT-TAVR, and ATLANTIS Trial Trial (Year) Implications for Practice Intervention Study Design Summary Principal Findings Interpretations GALILEO (2020) In the GALILEO trial, a rivaroxaban-based strategy post-TAVR led to increased ischemic and bleeding events compared with a clopidogrel-based strategy highlighting the potential harm of routine anticoagulation in the absence of a clear indication Rivaroxaban 10 mg daily/aspirin 75-100 mg daily vs clopidogrel 75 mg daily/aspirin 75-100 mg daily. At 90 d, rivaroxaban alone was continued in the experimental group, while aspirin alone was continued in the control group Population: 1,644 patients post-TAVR (mean age 80 y; 48% female) Follow-up Duration: 18 mo Self-expanding valve: 46%; balloon-expandable valve: 46%; mechanically-expandable: 5%; other: 3%; valve-in-valve: 6% Trial terminated early for safety; 42% had reached the primary outcome at closure The primary efficacy outcome: death, myocardial infarction, stroke, systemic thromboembolism, symptomatic valve thrombosis, or deep vein thrombosis/pulmonary embolism for Rivaroxaban vs Clopidogrel: occurred 12.7% vs 9.5% ( P = 0.04). Secondary outcomes for Rivaroxaban vs Clopidogrel • VARC-2 major, disabling, or life-threatening bleeding: 5.6% vs 3.8% ( P = 0.08). • All-cause mortality: 7.7% vs 4.6% ( P = 0.009) • Non-cardiovascular mortality: 3.5% vs 1.3% ( P < 0.05) • VARC-2 major bleeding: 3.6% vs 1.8% ( P < 0.05) • Leaflet thrombosis substudy (GALILEO-4D): ≥1 prosthetic leaflet with RLM: 2.1% vs 10.9% ( P = 0.014) ≥1 thickened leaflet: 12.4% vs 32.4% ( P < 0.05) • TAVR patients without an indication for oral anticoagulation - Rivaroxaban-based strategy associated with increased risk of death, thromboembolic events, and bleeding • Significantly higher composite of adverse clinical outcomes in the rivaroxaban group • Lower rates of subclinical valve thrombosis with rivaroxaban • Routine use of DOACs post-TAVR without a clear OAC indication may be harmful • Disconnection between imaging findings (HALT) and clinical outcomes LRT 2.0 (2021) LRT 2.0, though terminated early and underpowered, demonstrated a lower incidence of SLT with short-term warfarin plus aspirin compared to aspirin alone, but did not establish a clear clinical benefit Low-dose aspirin vs warfarin (target INR 2.5) plus low-dose aspirin for 30 d Population: 94 patients post-TAVR (mean age 73 y; 30% female) Follow-up Duration: 30 d Self-expanding valve: 43%; balloon-expandable valve: 57% Trial stopped early (Underpowered; early termination: n = 94; planned n = 200) The primary outcome, composite of HALT, reduced leaflet motion, hemodynamic dysfunction, and stroke/TIA at 30 days for aspirin vs aspirin + warfarin: 26.5% vs 7.0% ( P = 0.014). Secondary outcomes for aspirin vs aspirin + warfarin: HALT: 16.3% vs 4.7% ( P = 0.07) No excess bleeding at 30 d with anticoagulation. • TAVR patients without an indication for oral anticoagulation • Trial limitations: Underpowered and terminated early; relied on surrogate imaging endpoint (HALT) with unclear clinical significance • No systematic INR monitoring with potential subtherapeutic dosing in the warfarin arm • Short-term warfarin may reduce subclinical leaflet thrombosis in low-risk TAVR patients, but impact on long-term valve durability and clinical outcomes remains unproven ADAPT-TAVR (2022) ADAPT-TAVR, though underpowered, showed less valve thrombosis with edoxaban vs DAPT, but no reduction in cerebral events Edoxaban 60 mg daily (30 mg daily for renal insufficiency, body weight ≤60 kg, or use of certain P-glycoprotein inhibitors) vs dual antiplatelet therapy (aspirin 100 mg plus clopidogrel 75 mg daily) Population: 229 patients post-TAVR (mean age 80 y; 66% female) Follow-up Duration: 6 months Self-expanding valve: 10%; balloon-expandable valve: 90%; valve-in-valve: 1.7% The primary outcome, incidence of valve thrombosis at 6 mo for edoxaban vs DAPT: 9.8% vs 18.4% ( P = 0.08). Secondary outcomes for edoxaban vs DAPT: New cerebral lesions on brain magnetic resonance imaging: 25.0% vs 20.2% ( P = 0.4) Worsening of neurological function: no difference • TAVR patients without an indication for oral anticoagulation • Lower rates of subclinical leaflet thrombosis with edoxaban vs DAPT • No difference in new cerebral lesions or neurocognitive function • No correlation between HALT and cerebral outcomes • Limitations: Underpowered; single time-point imaging; open-label design • ADAPT-TAVR trial supports a class effect of DOACs in reducing HALT, but questions the clinical benefit of routine anticoagulation post-TAVR in this population ATLANTIS (2022) Among patients undergoing TAVR, full-dose apixaban was not superior to standard of care (VKA if OAC indicated; APT if not), despite a lower incidence of subclinical leaflet thrombosis compared with APT Apixaban 5 mg or 2.5 mg twice daily vs aspirin (75-100 mg once daily) or dual antiplatelet therapy (aspirin 75-100 mg plus clopidogrel 75 mg daily) in the setting of a coronary indication Population: 1,500 patients post-TAVR (mean age 82 y; 53% female) Stratification: Based on OAC indication • Stratum 1 (OAC indicated; n = 451, 21%): Apixaban 5 mg BID vs VKA • Stratum 2 (No OAC indication; n = 1,049, 79%): Apixaban 5 mg BID vs APT (single 15%, dual 57%) Follow-up duration: 1 y Self-expanding valve: 53%; balloon-expandable valve: 47%; valve-in-valve: 5% The primary outcome, time to death, stroke, myocardial infarction, systemic emboli, intracardiac or valve thrombosis, deep vein thrombosis/pulmonary embolism, or major bleeding, for apixaban vs standard of care: 18.4 vs 20.1% (HR: 0.92; 95% CI: 0.73-1.16; P = 0.43). Secondary outcomes for apixaban vs standard of care: Death/MI/stroke: 10.5% vs 8.3% (HR: 1.32; 95% CI: 0.95-1.85) All-cause mortality: 7.2% vs 5.5% ( P > 0.05) Major bleeding: 6.7% vs 6.4% Bioprosthetic thrombosis: 1.1% vs 4.7% ( P < 0.05) • Leaflet thrombosis substudy (n = 762): ≥1 prosthetic leaflet with RLM grade 3/4 or HALT grade 3/4 at 90 d, for apixaban vs standard of care: 8.9% vs 13.0% ( P for interaction = 0.038). Patients with presence of thrombus at 90 d: 19.2% vs 25% ( P = 0.011) • Apixaban not superior to standard care (VKA if OAC indicated; APT if not) post-TAVR • Reduced leaflet thrombosis with apixaban vs APT, but no clinical benefit • Higher noncardiovascular mortality with apixaban vs APT in patients without OAC indication • Thrombosis rates similar between apixaban and VKA • Findings consistent with GALILEO (rivaroxaban) • Apixaban may be a viable VKA alternative when OAC is indicated APT = antiplatelet therapy; BID = Bis in Die (latin for “twice a day”); DOAC = direct oral anticoagulant; VARC = Valve Academic Research Consortium-3; VKA = vitamin K antagonist; other abbreviations as in Table 1 . Comparative Overview of Designs and Key Outcomes in the GALILEO, LRT 2.0, ADAPT-TAVR, and ATLANTIS Trial VARC-2 major, disabling, or life-threatening bleeding: 5.6% vs 3.8% ( P = 0.08). All-cause mortality: 7.7% vs 4.6% ( P = 0.009) Non-cardiovascular mortality: 3.5% vs 1.3% ( P < 0.05) VARC-2 major bleeding: 3.6% vs 1.8% ( P < 0.05) Leaflet thrombosis substudy (GALILEO-4D): ≥1 prosthetic leaflet with RLM: 2.1% vs 10.9% ( P = 0.014) ≥1 thickened leaflet: 12.4% vs 32.4% ( P < 0.05) ≥1 prosthetic leaflet with RLM: 2.1% vs 10.9% ( P = 0.014) ≥1 thickened leaflet: 12.4% vs 32.4% ( P < 0.05) TAVR patients without an indication for oral anticoagulation - Rivaroxaban-based strategy associated with increased risk of death, thromboembolic events, and bleeding Significantly higher composite of adverse clinical outcomes in the rivaroxaban group Lower rates of subclinical valve thrombosis with rivaroxaban Routine use of DOACs post-TAVR without a clear OAC indication may be harmful Disconnection between imaging findings (HALT) and clinical outcomes TAVR patients without an indication for oral anticoagulation Trial limitations: Underpowered and terminated early; relied on surrogate imaging endpoint (HALT) with unclear clinical significance No systematic INR monitoring with potential subtherapeutic dosing in the warfarin arm Short-term warfarin may reduce subclinical leaflet thrombosis in low-risk TAVR patients, but impact on long-term valve durability and clinical outcomes remains unproven TAVR patients without an indication for oral anticoagulation Lower rates of subclinical leaflet thrombosis with edoxaban vs DAPT No difference in new cerebral lesions or neurocognitive function No correlation between HALT and cerebral outcomes Limitations: Underpowered; single time-point imaging; open-label design ADAPT-TAVR trial supports a class effect of DOACs in reducing HALT, but questions the clinical benefit of routine anticoagulation post-TAVR in this population Stratum 1 (OAC indicated; n = 451, 21%): Apixaban 5 mg BID vs VKA Stratum 2 (No OAC indication; n = 1,049, 79%): Apixaban 5 mg BID vs APT (single 15%, dual 57%) Leaflet thrombosis substudy (n = 762): ≥1 prosthetic leaflet with RLM grade 3/4 or HALT grade 3/4 at 90 d, for apixaban vs standard of care: 8.9% vs 13.0% ( P for interaction = 0.038). Patients with presence of thrombus at 90 d: 19.2% vs 25% ( P = 0.011) ≥1 prosthetic leaflet with RLM grade 3/4 or HALT grade 3/4 at 90 d, for apixaban vs standard of care: 8.9% vs 13.0% ( P for interaction = 0.038). Patients with presence of thrombus at 90 d: 19.2% vs 25% ( P = 0.011) Apixaban not superior to standard care (VKA if OAC indicated; APT if not) post-TAVR Reduced leaflet thrombosis with apixaban vs APT, but no clinical benefit Higher noncardiovascular mortality with apixaban vs APT in patients without OAC indication Thrombosis rates similar between apixaban and VKA Findings consistent with GALILEO (rivaroxaban) Apixaban may be a viable VKA alternative when OAC is indicated APT = antiplatelet therapy; BID = Bis in Die (latin for “twice a day”); DOAC = direct oral anticoagulant; VARC = Valve Academic Research Consortium-3; VKA = vitamin K antagonist; other abbreviations as in Table 1 . The GALILEO-4D, ATLANTIS 4D CT, and ADAPT-TAVR trials underscore the ongoing need for adequately powered trials to elucidate the role of anticoagulation strategies in reducing thrombotic complications and improving neurological outcomes following TAVR. The ACASA-TAVI 74 trial aims to compare antifactor Xa direct oral anticoagulants (apixaban, edoxaban, and rivaroxaban) with antiplatelet therapy by assessing HALT incidence at 12 months. The NOTION-4 trial 75 aims to address a critical knowledge gap in antithrombotic management following TAVR by enrolling patients without an indication for OAC. The trial compares lifelong single antiplatelet therapy against an early 3-month course of direct oral anticoagulant therapy followed by single antiplatelet therapy. Serial CT evaluations at 3 months, 1 year, and 5 years postrandomization are expected to provide deeper insights into the natural history of HALT and its clinical implications. In addition, this trial will assess whether an early intensive antithrombotic approach can yield sustained clinical benefits, balancing thrombotic protection with bleeding risk over time. Indeed, recent analyses indicate that HALT arises not only from acute thrombus formation but also from organizing or organized thrombus (ie, pannus), with progressive evolution over time. As thrombi mature, their composition shifts from fibrin and platelets (ie, the targets of antithrombotic therapy) to smooth muscle cells and extracellular matrix, likely rendering them resistant to treatment which underscores the potential benefit of early detection and timely intervention. 6 The NOTION-4 trial will be instrumental in refining the evidence base for tailored antithrombotic strategies in TAVR recipients. In contrast, the rationale for long-term anti thrombotic therapy, and single antiplatelet therapy in particular, after TAVR is increasingly being questioned due to the limited supporting evidence and the fact that current data do not account for recent advances in THV design and procedural refinements. For example, in the OCEAN-TAVI registry, 293 patients (8.2% of the cohort) were discharged without any antiplatelet therapy most likely due to a high bleeding risk. 76 Notably, these patients experienced no differences in all-cause or cardiovascular mortality, stroke, or myocardial infarction, and valve performance remained comparable to those receiving single antiplatelet therapy, but experienced a lower incidence of bleeding events. In addition, the 8.5% incidence of leaflet thrombosis in the nonantithrombotic group is somewhat reassuring. This finding is in line with previous studies demonstrating that HALT following TAVR is not linked to the pharmacodynamic response to clopidogrel, 17 and that the extent of platelet reactivity does not predict SLT. 20 , 77 Furthermore, these findings support the need for further clinical trials to evaluate a nonsingle antiplatelet therapy strategy in selected patients. Currently, the key challenge in the prevention and management of SLT lies in determining when the benefits of OAC outweigh its associated risks. Addressing this issue necessitates a deeper understanding of the long-term natural history and clinical significance of leaflet thrombosis in relation to valve durability. Waksman et al. found that HALT observed at 30 days postprocedure did not correlate with structural valve deterioration over a 4 years of follow-up. 78 In contrast, Hein et al. reported opposing results at 3 years, with HALT being associated with symptomatic hemodynamic valve deterioration. 35 The association between the absence of anticoagulation and structural valve deterioration, 79 , 80 along with histological observations of THV fibrosis and calcification occurring exclusively in cases of thrombosis, 72 , 6 further support the concept of a pathophysiological continuum between thrombosis and structural valve deterioration. Indeed, another recent report suggested that OAC at discharge marginally predicted structural valve dysfunction. 60 However, extended follow-up data are essential to draw definitive conclusions. Furthermore, future research strategies should focus on identifying specific patient subgroups who might benefit most from OAC or could safely avoid it. Closure time-ADP (CT-ADP), a surrogate marker of high-molecular-weight multimer defects, may serve as a valuable biomarker for accurately identifying patients at high bleeding risk. Indeed, prolonged CT-ADP (>180 s) has been shown to strongly predict bleeding events in TAVR patients. 81 A tailored algorithm incorporating CT-ADP could help optimize antithrombotic management following TAVR, especially in the early phase, as proposed in Figure 5 . Figure 5 Post-TAVR Antithrombotic Guidelines and CT-ADP Algorithm A) Consensus-based algorithm stratifying antithrombotic therapy according to indication for long-term OAC and recent coronary stenting. Patients are categorized by 1) presence vs absence of an indication for OAC and 2) history of recent PCI. This panel summarizes current expert consensus on selection and duration of antithrombotic regimens after TAVR. B) Proposed risk-tailored algorithm incorporating CT-ADP measured 1 day postprocedure. CT-ADP >180 seconds has demonstrated discriminative power for identifying patients at heightened risk for late bleeding. In this schema, CT-ADP serves as a bleeding-risk marker to guide de-escalation or continuation of antithrombotic therapy after TAVR. The aim is to integrate established guideline-based recommendations with emerging, individualized risk stratification (via CT-ADP) to optimize patient outcomes. CT-ADP = closure time with adenosine diphosphate; DAPT = dual antiplatelet therapy; OAC = oral anticoagulation; PCI = percutaneous coronary intervention; SAPT = single antiplatelet therapy; TAVR = transcatheter aortic valve replacement. Post-TAVR Antithrombotic Guidelines and CT-ADP Algorithm A) Consensus-based algorithm stratifying antithrombotic therapy according to indication for long-term OAC and recent coronary stenting. Patients are categorized by 1) presence vs absence of an indication for OAC and 2) history of recent PCI. This panel summarizes current expert consensus on selection and duration of antithrombotic regimens after TAVR. B) Proposed risk-tailored algorithm incorporating CT-ADP measured 1 day postprocedure. CT-ADP >180 seconds has demonstrated discriminative power for identifying patients at heightened risk for late bleeding. In this schema, CT-ADP serves as a bleeding-risk marker to guide de-escalation or continuation of antithrombotic therapy after TAVR. The aim is to integrate established guideline-based recommendations with emerging, individualized risk stratification (via CT-ADP) to optimize patient outcomes. CT-ADP = closure time with adenosine diphosphate; DAPT = dual antiplatelet therapy; OAC = oral anticoagulation; PCI = percutaneous coronary intervention; SAPT = single antiplatelet therapy; TAVR = transcatheter aortic valve replacement. To date, no readily available circulating biomarker has shown sufficient predictive accuracy or causal association for SLT, leaving risk stratification unvalidated and underscoring the need for focused biomarker discovery, mechanistic studies, and prospective validation. Based on our literature review, we propose a nonexhaustive series of potential studies to inform and guide future research in this evolving field ( Table 3 ). Table 3 Proposed Investigations to Advance Understanding of Subclinical Leaflet Thrombosis after Transcatheter Aortic Valve Replacement Prospective Biomarker–Imaging TAVR Cohort Rationale: Identify circulating predictors of SLT to enable early risk stratification. Methods: Baseline and serial blood sampling for candidate markers: Lp(a), inflammatory cytokines (eg, IL-6, hsCRP), coagulation markers (eg, D-dimer, thrombin generation assays). Imaging: Cardiac CT angiography at prespecified intervals (eg, 30 d, 3 mo, 6 mo, and 12 mo) for detection of HALT/RLM. Transthoracic echocardiography at matching time points for transvalvular gradient and leaflet motion assessment. Endpoints Primary • Incident SLT on follow-up CT (presence of HALT ± RLM). Secondary • Change in transvalvular gradient or leaflet motion over time. • Relationship between biomarker kinetics and SLT. • Interaction between baseline antithrombotic regimen and biomarker levels. Mechanistic Imaging Study Using 18 F-NaF PET Rationale: Determine whether prosthetic leaflet microcalcific activity precedes or coexists with SLT; identify PET uptake thresholds predictive of thrombosis risk. Methods: Quantification of leaflet uptake by 18 F-NaF PET and Cardiac CT angiography (HALT/RLM) and transthoracic echocardiography (gradient) at prespecified intervals (eg, 30 d, 3 mo, 6 mo, and 12 mo) Endpoints Primary • Correlation between baseline or early 18 F-NaF uptake and subsequent SLT on CT. Secondary • Change in gradient or leaflet motion relative to PET uptake. • Temporal sequence: whether elevated uptake precedes SLT. • Regional leaflet differences in uptake vs localized HALT. Randomized Antithrombotic Strategy Trial Stratified by Bleeding-Risk Marker Rationale: Test whether anticoagulation regimen based on CT-ADP threshold of 180 s reduces bleeding events without excessive SLT compared to standard care Methods: Cardiac CT angiography at prespecified intervals (eg, 30 d, 3 mo, 6 mo, and 12 mo) for detection of HALT/RLM. Echocardiography for gradient monitoring. Monitoring of MACE and bleeding events (VARC-3 definitions). Endpoints Primary • Incidence of SLT on follow-up cardiac CT imaging. Secondary • Net clinical benefit composite (SLT prevention vs major bleeding). • Change in transvalvular gradient or leaflet motion. Evaluation of Sodium–Glucose Cotransporter-2 Inhibitors on Bioprosthetic Leaflet Biology Rationale: Emerging evidence suggests that SGLT2i may slow aortic stenosis progression 82 and prevent bioprosthetic valve dysfunction. 83 Test SGLT2i effect on post-TAVR leaflet calcific activity and SLT. Methods: Initiate SGLT2i at day 1 after TAVR vs standard care. Imaging: Cardiac CT angiography at prespecified intervals (e.g., 30 d, 2 mo, 6 mo, 12 mo) for detection of HALT/RLM 18 F-NaF PET/CT at baseline (early post-TAVR) and follow-up to assess change in microcalcification. Echocardiography for gradient monitoring Endpoints Primary • Difference in change of 18 F-NaF uptake between SGLT2i and control groups. Secondary • SLT incidence on cardiac CT. • Change in transvalvular gradient or leaflet motion. • MACE, heart failure rehospitalization Additional Axes of Research Omics and Biomarker Discovery Uncover novel circulating factors and pathways implicated in SLT Longitudinal Durability Follow-Up Study Determine whether incident SLT predicts structural valve deterioration or adverse clinical outcomes over the long term (>5-10 y), and whether early intervention modifies risk. Analyses of Antithrombotic Effects on Valve Biology Clarify mechanistic differences among anticoagulant classes (direct oral anticoagulants vs vitamin K antagonists vs antiplatelet agents) on leaflet biology (eg, calcification, inflammation, motion). 18 F-NaF = fluorine-18 sodium fluoride; CT-ADP = closure time with adenosine diphosphate; hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6; Lp(a) = lipoprotein(a); PET = positron emission tomography; SGLT2i = sodium–glucose cotransporter-2 inhibitor. Proposed Investigations to Advance Understanding of Subclinical Leaflet Thrombosis after Transcatheter Aortic Valve Replacement Incident SLT on follow-up CT (presence of HALT ± RLM). Change in transvalvular gradient or leaflet motion over time. Relationship between biomarker kinetics and SLT. Interaction between baseline antithrombotic regimen and biomarker levels. Correlation between baseline or early 18 F-NaF uptake and subsequent SLT on CT. Change in gradient or leaflet motion relative to PET uptake. Temporal sequence: whether elevated uptake precedes SLT. Regional leaflet differences in uptake vs localized HALT. Incidence of SLT on follow-up cardiac CT imaging. Net clinical benefit composite (SLT prevention vs major bleeding). Change in transvalvular gradient or leaflet motion. Difference in change of 18 F-NaF uptake between SGLT2i and control groups. SLT incidence on cardiac CT. Change in transvalvular gradient or leaflet motion. MACE, heart failure rehospitalization 18 F-NaF = fluorine-18 sodium fluoride; CT-ADP = closure time with adenosine diphosphate; hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6; Lp(a) = lipoprotein(a); PET = positron emission tomography; SGLT2i = sodium–glucose cotransporter-2 inhibitor. SLT is complex due to its dynamic natural history characterized by spontaneous regression or progression over time, and the evolving balance between thrombotic and hemorrhagic risk in patients, which is influenced by age, comorbidities, and life events. Understanding these fluctuating risk factors is crucial for optimizing prevention and, tailoring them to individual patient profiles.

The calcified, native aortic valve, once deemed an inert bystander postreplacement, has increasingly been recognized as a significant contributor to the prothrombotic milieu following TAVR 59 ( Figure 4 ). Since the native valve is structurally and functionally obsolete due to the newly implanted bioprostheses, it was originally thought that it would remain biologically inactive and pose no further risk. Emerging evidence has challenged this assumption, suggesting that the retained native aortic valve tissue may continue to exhibit biological activity even after TAVR. Indeed, the calcified native valve may contribute to thrombotic events and impact bioprosthetic valve efficacy and durability, making it a potential therapeutic target in post-TAVR management. Recent analyses underscore that aortic valve calcium complex volume is an independent predictor of structural valve failure, potentially due to the effects of ongoing biological activity, valve underexpansion, asymmetry, or alterations in valve rheology. 60 Figure 4 TAVR and the Native Valvulo-Thrombosis Axis This figure illustrates the interaction between retained native valve material and subclinical thrombosis of the aortic valve complex post-TAVR. Abbreviations as in Figure 2 . TAVR and the Native Valvulo-Thrombosis Axis This figure illustrates the interaction between retained native valve material and subclinical thrombosis of the aortic valve complex post-TAVR. Abbreviations as in Figure 2 . Initial indication of the potential adverse role of the retained native valve emerged from observations of a higher incidence of subclinical thrombosis following TAVR compared to surgical aortic valve replacement, regardless of the antithrombotic regimen at 1-month follow-up. 10 , 22 This finding suggested that the native valve itself may exert a deleterious effect, unlike in SAVR counterparts as the native valve is removed. Crucial imaging studies using flourine-18-sodium fluoride ( 18 F-NaF) positron emission tomography (PET), which captures both biological activity and anatomical details, have demonstrated persistent 18 F-NaF uptake in the retained native aortic valve in TAVR patients. 61 , 62 This ongoing metabolic activity was significant and independent of valve motion and mechanical stress. In addition 18 F-NaF uptake correlated positively with implantation duration, thus potentially predicting subsequent valve dysfunction. Histological analysis of explanted TAVR valves further revealed continued activation of procalcific markers in the native valve tissue, with increased staining for osteopontin and Runx-2, highlighting the sustained biological activity post-TAVR. 61 However, further investigation is needed to determine the correlations between serum biomarkers (lipoprotein(a), metalloproteinase-3, matrix metalloproteinase-9, and osteopontin and PET/CT findings. 62 , 63 18 F-NaF PET provides a novel approach for assessing aortic valve microcalcification, which occurs as part of the healing response to valvular inflammation. This approach is useful for detecting ongoing native valve disease activity and the early stages of bioprosthetic valve degeneration. Indeed, 18 F-NaF PET has recently been acknowledged for its capacity to detect subtle degenerative changes in bioprosthetic valves before they become apparent on conventional imaging such as echocardiography. 61 , 64 , 65 , 66 18 F-NaF PET enables the real-time assessment of disease activity in the native aortic valve and facilitates the targeted evaluation of specific therapeutic interventions. 18 F-NaF PET requires a relatively small numbers of patients to demonstrate a particular treatment effect, making it a valuable tool for clinical research. Although 18 F-NaF PET imaging provides the key mechanistic insights, the small size and rapid motion of TAVR leaflets predispose to partial volume and motion artifacts that compromise quantification. Moreover, the lack of standardized acquisition and analysis protocols limits reproducibility across centers. Finally, the high cost and limited availability preclude its routine application for broad risk stratification. Further to these imaging studies, emerging evidence suggests that the calcified regions of aortic stenosis valves serve as active reservoirs of thrombotic material that are enriched with procoagulant cell-derived extracellular vesicles (EVs). These AS-associated EVs act as potent biological mediators, transforming valvular endothelial cells into a prothrombotic, proadhesive, and proinflammatory surface that facilitates inflammatory cell recruitment and promotes thrombogenicity. 67 This localized and site-specific effect aligns with findings that patients with severe AS exhibit a distinctive phenotype and plasma profile, characterized by elevated proinflammatory cytokines (interleukin-1β, interleukin-6, tumor necrosis factor-alpha) and increased factor Xa activity. This inflammatory plasma signature exerts harmful effects on aortic valvular endothelial cells, inducing reactive oxygen species formation, endothelial dysfunction, and a prothrombotic, proinflammatory, and proadhesive phenotype. 68 Whether TAVR contributes to an improvement in these pathological features, as well as the timing of such effects, remains uncertain. Nevertheless, the numerical increase in stroke rate observed in the conventional arm of the early TAVR trial suggests a reduction in thrombotic propensity after TAVR. 69 Consistently, reduced oxidative stress, 70 decreased procoagulant EVs, 71 and improvements in endothelial function have been described after TAVR. These mechanisms play a key role in regulating vascular homeostasis and preventing clot formation. The native valvulo-thrombosis axis represents a promising avenue for further investigation. Future research should focus on elucidating the association between the extent of active calcified regions within the retained native valve—assessed through imaging modalities such as 18 F-NaF PET—and serum biomarkers indicative of ongoing disease activity. Examining the links between persistent disease activity and the native valve, microthrombosis, and early degenerative changes in bioprosthetic valves will be essential for identifying effective therapeutic strategies. A refined understanding of the native valve’s role in post-TAVR complications could have significant implications for patient management and long-term outcomes.

There is limited research on the histopathologic processes underlying leaflet thrombosis, whether subclinical or clinical, and their correlations with CT findings. A recent study by Sato et al. 6 has attempted to bridge this gap by investigating the histological characteristics of explanted self-expanding THVs and comparing micro-CT findings from individuals with suspected HALT with the histological features of valve thrombosis. Key findings include the high rate of leaflet thickening of any degree (46.5% of leaflets) and HALT on micro-CT (43.4% of explanted leaflets). Histological analysis confirmed that HALT results not only from acute thrombus but also from organizing or organized thrombus, with thrombus organization progressing over time. Notably, pannus formation, calcification, and structural changes increased with implant duration, whereas leaflet thrombosis and inflammation were not associated with implant duration. Finally, thrombus thickness and the extent of leaflet involvement did not vary with implant duration, although histological analysis indicated evolving thrombus characteristics over time. These findings align with earlier insights from a smaller autopsy cohort of THVs, 72 where Sellers et al. observed a time-dependent progression of fibrosis and calcification, present in all valves explanted after 60 days and 4 years, respectively. The authors proposed a stepwise progression from thrombus formation to sequential fibrosis and calcification, resulting in leaflet thickening and valve deterioration, with abnormal endothelial cell formation contributing to this process. Both pathological studies of explanted THVs converge on the finding that valves implanted for over 1 year predominantly exhibit organizing or organized thrombi, raising questions about the long-term efficacy of OAC in cases of HALT.

The pathophysiological mechanisms underlying subclinical aortic valve complex thrombosis following TAVR remain incompletely understood but appear to involve a multifactorial interplay of anatomical, hemodynamic, and device-specific factors. The native valve may act as a localized prothrombotic site, initiating thrombus formation. In addition, sinus and neosinus hemodynamics, characterized by low shear stress, reduced flow velocity, and suboptimal washout, create an environment conductive to platelet activation and thrombus development. Device-related structural factors, including underexpansion, asymmetric deployment, and the use of larger prosthetic valves, may further amplify flow disturbances and promote flow stagnation. Taken together, these interrelated mechanisms establish a prothrombotic milieu within the aortic valve complex that contribute to the formation of both valvular and perivalvular thrombi. Understanding these processes is essential for guiding future preventive and therapeutic strategies in patients undergoing TAVR.

SLT is characterized by hypoattenuated leaflet thickening (HALT), which may or may not be accompanied by reduced leaflet motion on contrast-enhanced cardiac CT. HALT is defined as a visually apparent thickening of the bioprosthetic leaflet, exhibiting a progressive gradient of thickness from its basal insertion to the free edge. 4 The extent of HALT is assessed using a semiquantitative grading scale, which describes the percentage of leaflet involvement beginning at its insertion point 5 ( Figure 1 ). Histologically, leaflet thickening corresponds to the presence of a thrombus and reflects thrombus progression. 6 Importantly, HALT may occur with or without concomitant reduced leaflet motion. The true incidence of SLT is unclear, primarily due to the lack of systematic imaging surveillance protocols, as well as the potential underreporting of SLT in the literature. The reported incidence rates of SLT vary widely, 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 as shown in Table 1 and Figure 2 , with symptomatic thrombosis representing the most severe manifestation within the spectrum of bioprosthetic aortic valve thrombosis. In a series of 106 explanted self-expanding transcatheter aortic bioprostheses obtained via autopsy or surgical explant, HALT was detected by micro-CT in 43.4% of the valves. 6 According to a meta-analysis of 25 studies, involving a total of 11,098 patients, 42 the median incidence of SLT was 6% at a median follow-up of 30 days. Furthermore, the use of intra-annular valves was associated with a two-fold greater risk of developing SLT compared to supra-annular valves. No significant difference in SLT risk was observed between single and dual-antiplatelet therapy; however, oral anticoagulation (OAC) was associated with a 58% relative risk reduction in SLT. Finally, leaflet thrombosis was linked to a 2.6-fold higher risk of stroke or transient ischemic attack. Figure 1 Grading of Hypoattenuated Leaflet Thickening Grading of HALT based on percentage of affected leaflet. HALT appears as a meniscal-shaped lesion with low attenuation on the aortic surface of the bioprosthetic valve leaflets, best visualized during diastole. Grading is based on the extent of leaflet involvement. Grade 0: No visible thickening. Grade 1: Mild HALT affecting 50% of the leaflet but without restriction in motion. Grade 4: HALT with associated reduced leaflet motion. HALT = hypoattenuated leaflet thickening. Table 1 Incidence of Hypoattenuated Leaflet Thickening After TAVR as Reported in Major Studies With Computed Tomography Across the Literature First Author, Year Design N Type of TAVR Bioprosthetic Valve Time Interval from TAVR to CT Imaging Assessment HALT Subclinical Thrombosis of the Aortic Valve Complex Outcomes Makkar et al, 7 2015 Pooled analysis combining data from one RCT and 2 observational registries: RCT: PORTICO IDE Study (n = 55) Observational Registries: RESOLVE registry and SAVORY registry (n = 105) 200 RCT: Portico valve, Sapien XT valve. Observational registries: N/A RCT 32 d (IQR 28-37). Observational registries 86 d (range 7-1851) Global: 23.1%. RCT: 40% (22/55). Observational registries 14% (15/105) No significant differences in clinical outcomes were observed between treatment groups in the PORTICO IDE trial. In the pooled RESOLVE and SAVORY cohorts, patients with RLM experienced a higher incidence of stroke or TIA compared with those exhibiting normal leaflet motion (18% [3 of 17] vs 1% [1 of 115]; P = 0.007), although the overall number of events was notably low Pache et al, 8 2016 Nonrandomized observational single-center study 156 Sapien 3 5 d 10.3% (16/156) No association with MACE. Higher mean pressure gradient at the time of CT (11.6 + 3.4 vs 14.9 + 5.3 mm Hg, P = 0.026) Hansson et al, 9 2016 Nonrandomized observational single-center study 405 Sapien XT, Sapien 3 1-3 mo 5.7% (23/405) N/A Chakravarty T et al, 10 2017 Pooled analysis from 2 Observational Registries: RESOLVE Registry and SAVORY Registry 752 Edwards-Sapien (22), Sapien XT (122), Sapien 3 (309), CoreValve (70), Evolut (75), Lotus (83), Portico (50), Centera (7), Symetis (8), Direct flow (6) 58 d (IQR 32-236) 13% (101/752) Increased TIA rates (4.18 vs 0.60 per 100 person-years; P = 0.0005) and stroke/TIA rates (7.85 vs 2.36 per 100 person-years; P = 0.001 Vollema et al, 11 2017 Nonrandomized observational single-center study 128 Sapien (76), SAPIEN XT (52) 35 d (IQR 19-210) 12.5% (16/128) No assocation between HALT and stroke/TIA Sondergaard et al, 12 2017 SAVORY Registry: Nonrandomized observational single-center study 84 CoreValve, EvolutR, Lotus, Portico, Sapien 3 First CT 140 ± 152, and second CT 298 ± 141 d 38.1% (32/84).Progression in 15.5% (n = 13) and regression in 10.7% (n = 9) No association with clinical events Fuchs A, et al, 13 2017 Nonrandomized observational single-center study 75 CoreValve, Evolut R, Portico 30 d N/A Ruile et al, 14 2018 Nonrandomized observational single-center study 754 SAPIEN XT, SAPIEN 3, CoreValve, Evolut-R, Lotus, Portico, Symetis ACURATE, and JenaValve 5 d (IQR 4-6) 15.9% (120/754) No association with all-cause mortality and composite of stroke and TIA at a median follow-up period of 406 d Basra et al, 15 2018 Nonrandomized observational single-center study 55 Sapien 3, Sapien XT, Sapien, Corevalve, Evolut, Direct Flow 14.23 mo (range 5 d-130.9 mo) in SLT patients 30.9% (17/55) No association with clinical events Marwan et al, 16 2018 Nonrandomized observational single-center study 78 Sapien XT, Sapien 3, Portico, Symetis Acurate 4 mo (IQR 1 mo) 23% (18/78) N/A Nührenberg et al, 17 2019 Nonrandomized observational single-center study 200 SAPIEN 3,SAPIEN XT, CoreValve, Evolut-R, Lotus, Portico, Symetis 5 d (IQR 4-6) 18% (36/200) N/A Yanagisawa et al, 18 2019 OCEAN-TAVI substudy: multicenter prospective registry 485 Sapien XT, SAPIEN 3, CoreValve 3 d, 6 mo, 1, 2, and 3 y 9.3% (45/485) at 3 d, 7.1% (22/310) at 6 mo, 11.3% (23/203) at 1 y, 12.7% (18/142) at 2 y, and 16.9% (11/65) at 3 y Early SLT detected on 3-d CT imaging showed no impact on cumulative event rates for the composite endpoint of death, stroke, or heart failure-related rehospitalization over a mean follow-up of 1.8 y Koo et al, 19 2019 Nonrandomized observational single-center study 138 Sapien (3), Sapien XT (56), Sapien 3 (11), CoreValve (62), Evolute-R (6) 17.5 d (IQR 3-390.8 d) 18% (25/138) Hypoattenuating subvalvular thickening: 23% (n = 32); Sinus of Valsalva: 4% (n = 5) N/A Jimenez et al, 20 2019 Nonrandomized observational single-center study 85 SAPIEN XT, SAPIEN S3, CoreValve, Evolut-R 114 d (65-205) 35% (30/85) No association with clinical events Karády et al, 21 2020 RETORIC study: Nonrandomized observational single-center study 111 CoreValve 19 mo (IQR 11-29) 21.6% (24/111) No association with clinical events Makkar et al, 22 2020 PARTNER 3 CT substudy: multicenter, prospective, RCT. PARTNER 3 trial: TAVR vs SAVR in low-risk patients with aortic stenosis Treatment Arms: Intervention: TAVR Group (n = 221) Control: SAVR Group (n = 214) 221 SAPIEN 3 30 d and 1 y 13% (22/165) at 30-d CT; 28% (42/153) at 1-y CT SLT more frequent TAVR at 30 d (13% vs 5%; P = 0.03), but not at 1 y (28% vs 20%; P = 0.19) Association with increased pooled thromboembolic event rates of stroke, TIA, and retinal artery occlusion in TAVR Both 30 d HALT and 1 y HALT had significantly increased aortic valve gradients at 1 y (17.8 2.2 mm Hg vs 12.7. 0.3 mm Hg; P = 0.04). Blanke et al, 23 2020 LTI substudy: an substudy of the Evolut Low Risk trial: multicenter, prospective, RCT Evolut Low Risk trial: TAVR vs SAVR in low-risk patients with aortic stenosis Treatment Arms: Intervention: TAVR Group (n = 725) Control: SAVR Group (n = 678) 197 CoreValve, Evolut R, Evolut PRO 30 d and 1 y 17.3% (31/179) at 30-d CT; 30.9% (47/152) at 1-y CT No association between the presence or severity of HALT or RLM and the occurrence of death, stroke, transient ischemic attack, or myocardial infarction up to 1 y postprocedure. Xiong et al, 24 2020 Nonrandomized observational single-center study 179 Medtronic, Venus-A, VitaFlow, Lotus 6 d (Median) 10.6% (19/179) N/A De Backer et al, 25 2020 GALILEO-4D Study: A multinational, multicenter, prospective RCT substudy of the GALILEO trial. Treatment Arms: Intervention: Rivaroxaban 10 mg + Aspirin 75-100 mg daily Control: Aspirin 75-100 mg + Clopidogrel 75 mg daily 199 Balloon-expandable (106), Self-expandable (125) 90 ± 15 d Global 22.6% (n = 55/199). In the intention-to-treat analysis, HALT of at least one leaflet 12.4% (12/97) in the rivaroxaban group and 32.4% (33/102) in the antiplatelet group (difference, −20.0 percentage points; 95% CI: −30.9 to −8.5) Risk of death or thromboembolic events and the risk of life-threatening, disabling, or major bleeding were higher with rivaroxaban (HR 1.35 and 1.50, respectively). Stroke: 1.8% vs 1.7% ( P > 0.90) Zhu et al, 26 2021 Nonrandomized observational single-center study 351 N/A 30 d and 1 y 11.1% (39/351) at 30-d CT. 35.8% (113/315) at 1-y CT. No association with clinical events Rashid et al, 27 2021 Nonrandomized observational single-center study 172 Lotus, CoreValve, Sapien 3 6 wk (IQR: 30-37) 14% (24/172) No association with clinical events Waksman et al, 28 2021 Low-Risk TAVR Substudy Low Risk TAVR: Comparison of TAVR vs SAVR in patients with low surgical risk. Treatment Arms: Intervention: TAVR Group (n = 200) Control: SAVR Group (n = 719) (Inverse probability weighting–adjusted cohort) 193 Sapien 3, CoreValve, Evolut R, Evolut PRO 30 d 14% (27/193) HALT observed at 30 d had no impact on valve hemodynamics at 2 y and was not associated with cerebrovascular events or structural valve deterioration at the 2-y follow-up Imaeda S et al, 29 2022 Nonrandomized observational single-center study 124 SAPIEN-XT 1 y 21.8% (27) No correlation with all-cause or cardiac mortality, stroke, heart failure readmissions, valve performance, or structural valve deterioration over 5 y (TTE-assessed) Garcia et al, 30 2022 Nonrandomized observational single-center study 638 Balloon-expandable (398), self-expanding valves (240) 31 d (30-37) 12% (79/638) HALT was independently associated with long-term mortality at 2.2 y of follow-up. Bak et al, 31 2022 Nonrandomized observational single-center study 94 Ballon expandable (50), Self-expendable (23), Lotus (1) 367 d (353-393) 21.3% (20/94) No association with clinical events Park et al, 32 2022 ADAPT-TAVR Trial: A multinational, multicenter, prospective, RCT Treatment Arms: Intervention: Edoxaban 60 mg or 30 mg daily Control: DAPT (Aspirin 100 mg + Clopidogrel 75 mg daily) 229 SAPIEN 3; Evolut R; CoreValve; Evolut PRO; ACURATE Neo 6 mo Global 14.2% (30/2011). 9.8% in the edoxaban group compared with 8.5% in the dual antiplatelet therapy group ( P = 0.076). Leaflets with leaflet thickening >50% involvement 2% vs 2.5% ( P = 0.72) No significant association between leaflet thrombosis presence or extent and new cerebral lesions, neurological changes, or neurocognitive function. Death: 2.7% vs 1.7% ( P = 0.68). Stroke: 1.8% vs 1.7% ( P > 0.90). New cerebral lesions on imaging: 25.0% vs 20.2% ( P = 0.4). Montalescot et al, 33 2022 ATLANTIS Trial CT Substudy: An international, randomized RCT comparing standard care with Apixaban Treatment Arms Intervention: Apixaban 5 mg or 2.5 mg twice daily Control: Aspirin 75-100 mg daily or DAPT (Aspirin 75-100 mg + Clopidogrel 75 mg daily for coronary indications) 762 Self-expendable (416), balloon-expandable (343), Valve in Valve (38) 3-6 mo after randomization Global 11% (84/762). 8.9% in the apixaban group and 13.0% in the standard-of-care group MACE: No association with the composite of death, myocardial infarction, stroke, or systemic embolism at 1 y. Bleeding: increase in bleeding Fukui et al, 34 2022 Nonrandomized observational single-center study 565 SAPIEN 3; Evolut-R 30 d 19.1% (108/565) HALT was independently associated with increased 1-year mortality risk, including all-cause mortality (HR: 2.98; 95% CI: 1.57-5.63; P = 0.001), cardiac death (HR: 4.58; 95% CI: 1.81-11.6; P = 0.001), and a composite of all-cause mortality and heart failure hospitalization (HR: 1.94; 95% CI: 1.14-3.30; P = 0.02), after adjusting for age, sex, and comorbidities Hein et al, 35 2022 Nonrandomized observational single-center study 804 SAPIEN XT, SAPIEN 3, CoreValve Evolut R/Pro, Lotus, Portico valve, Symetis ACURATE, JenaValve 5 d (IQR 4-6) 16.0% (115/804) No association with clinical events at a median follow-up of 3.25 y/HALT associated with a higher 3-year event rate of symptomatic hemodynamic valve deterioration (9.4% vs 1.5%, P < 0.001). Choi et al, 36 2023 Substudy of the ADAPT-TAVR Trial ADAPT-TAVR Trial: A multinational, multicenter, prospective, RCT Treatment Arms: Intervention: Edoxaban 60 mg or 30 mg daily Control: DAPT (Aspirin 100 mg + Clopidogrel 75 mg daily)" 211 SAPIEN 3 (188), Evolut R (10), CoreValve (1), Evolut PRO (8), ACURATE Neo (4) 6 mo 14.2% (30/211) Thrombus involvement in the aortic valve complex: 43.1% (n = 91), including leaflet thrombus in 14.2% (n = 30), perivalvular thrombus in 37% (n = 78), and both leaflet and perivalvular thrombus in 8.1% (n = 17). No association with new cerebral thromboembolism, neurologic or neurocognitive dysfunction, or adverse clinical outcomes Brunner et al, 37 2023 Nonrandomized observational single-center study 50 ACURATE neo/neo2 6-mo 16% (8/50) 18% (9/50) thrombosis of the sinus of Valsalva N/A Zhou et al, 38 2024 Nonrandomized observational single-center study 61 Self-expandable (52), Mechanical (9) Long-term annual cardiac CTA follow-up (≥5 y) 54.1% (33/61) 37.7% with hypoattenuated filling defects in the aortic sinus. 73.8% with structural filling defects within the stent Sinus filling defect was an independant predictor of MACE (stroke, cardiac re-hospitalization, and bioprosthetic valve dysfunction) Moscarelli M et al, 39 2024 Nonrandomized observational single-center study 100 Evolut R 6 mo 18% (18/100) 24% (24/100) with thrombosis of the anatomic sinus.23% (23/100) subvalvular thrombosis with partial or complete circumferential involvement of the prosthesis inner skirt No association with hemodynamic structural valve dysfunction, neurological events, and re-hospitalization Adrichem et al, 40 2024 Rotterdam Edoxaban (REDOX) Study: A prospective, single-arm, open-label trial Treatment Arms: Intervention: 3-month edoxaban monotherapy in patients without indications for OAC Control: DAPT 50 Self-expandable (33), balloon-expandable (17) 91 d (90-94) 12% (6/50) No association with clinical events Nagasaka et al, 41 2024 Nonrandomized observational single-center study 229 SAPIEN 3, SAPIEN Ultra 30 d 24.5% (56/229) THV deformation, annular and LVOT calcifications were significantly associated with a higher rate of all-cause mortality at 3 y and HALT at 30 d after TAVR in patients with Bicupsid aortic valves. CT = computed tomography; DAPT = dual antiplatelet therapy; HALT = hypoattenuated leaflet thickening; LVOT = left ventricular outflow tract; MACE = major adverse cardiovascular events; OAC = oral anticoagulation; RCT = randomized controlled trial; RLM = reduced leaflet motion; SLT = subclinical leaflet thrombosis; TAVR = transcatheter aortic valve replacement; THV = transcatheter heart valve; TIA = transient ischemic attack; TTE = transthoracic echocardiography. Figure 2 Incidence of HALT Reported in Major Computed Tomography Studies This figure illustrates the incidence of HALT as reported in major CT studies across the literature. It highlights the percentage of the overall cohort affected, the mean time from TAVR to CT, and the corresponding cohort sizes, providing a comprehensive overview of both the prevalence and the scale of the studies. CT = computed tomography; TAVR: transcatheter aortic valve replacement; other abbreviations as in Figure 1 . Grading of Hypoattenuated Leaflet Thickening Grading of HALT based on percentage of affected leaflet. HALT appears as a meniscal-shaped lesion with low attenuation on the aortic surface of the bioprosthetic valve leaflets, best visualized during diastole. Grading is based on the extent of leaflet involvement. Grade 0: No visible thickening. Grade 1: Mild HALT affecting 50% of the leaflet but without restriction in motion. Grade 4: HALT with associated reduced leaflet motion. HALT = hypoattenuated leaflet thickening. Incidence of Hypoattenuated Leaflet Thickening After TAVR as Reported in Major Studies With Computed Tomography Across the Literature CT = computed tomography; DAPT = dual antiplatelet therapy; HALT = hypoattenuated leaflet thickening; LVOT = left ventricular outflow tract; MACE = major adverse cardiovascular events; OAC = oral anticoagulation; RCT = randomized controlled trial; RLM = reduced leaflet motion; SLT = subclinical leaflet thrombosis; TAVR = transcatheter aortic valve replacement; THV = transcatheter heart valve; TIA = transient ischemic attack; TTE = transthoracic echocardiography. Incidence of HALT Reported in Major Computed Tomography Studies This figure illustrates the incidence of HALT as reported in major CT studies across the literature. It highlights the percentage of the overall cohort affected, the mean time from TAVR to CT, and the corresponding cohort sizes, providing a comprehensive overview of both the prevalence and the scale of the studies. CT = computed tomography; TAVR: transcatheter aortic valve replacement; other abbreviations as in Figure 1 . To enhance the understanding of the incidence and natural history of SLT, two CT substudies were conducted within large randomized clinical trials. 5 , 7 In the PARTNER 3 cardiac CT substudy, a pooled analysis revealed higher thromboembolic event rates, including stroke, transient ischemic attack, and retinal artery occlusion. 5 In the PARTNER 3 CT cohort, serial CT scans in the absence of anticoagulation revealed spontaneous SLT resolution in approximately half of the patients, new onset in 20%, and an increased frequency of SLT over time in about 25%. These findings highlight the dynamic nature of SLT, which complicates efforts to identify events and establish temporal associations between clinical outcomes and CT findings. The CT substudy of the Evolut Low Risk randomized trials confirmed that HALT is a frequent but dynamic process that occurs within the first year after TAVR. Specifically, HALT was observed in 17.3% of patients at 30 days, and 30.9% at 1 year, but with no association with aortic valve hemodynamic status or clinical outcomes. 23 SLT has evolved into a broader concept known as subclinical aortic valve–complex thrombosis ( Figure 3 ), which encompasses both valvular and perivalvular thrombi. Perivalvular thrombi, including supravalvular, subvalvular, and sinus of Valsalva regions, have been reported, refining the understanding of thrombosis affecting TAVR devices and the surrounding native environment. 19 , 36 , 37 , 38 , 39 A substudy of the ADAPT-TAVR (Anticoagulation vs Dual Antiplatelet Therapy for Prevention of Leaflet Thrombosis and Cerebral Embolization After Transcatheter Aortic Valve Replacement) trial, aimed to investigate the effect of edoxaban compared with dual antiplatelet therapy on preventing leaflet thrombosis and the associated risks of cerebral thromboembolism and neurological or neurocognitive dysfunction after TAVR. Specifically, Choi et al. 36 reported that 91 of 211 patients (43.1%) with CT scans had thrombus at any location within the aortic valve complex: 14.2% had a leaflet thrombus, 37.0% had a perivalvular thrombus, and 8.1% had both a leaflet and perivalvular thrombus. Moscarelli et al. 39 reported a 44% incidence of thrombosis involving any component of the aortic valve complex at 6-month follow-up. Moreover, Zhou et al. analyzed a small cohort of 61 patients who underwent long-term annual cardiac computed tomography angiography follow-ups for over 5 years after TAVR. 38 Their findings revealed that HALT occurred in 54.1% of cases, hypoattenuated filling defects in the aortic sinus occurred in 37.7% of cases, and structural filling defects within the stent occurred in 73.8% of patients. However, no study to date has comprehensively evaluated how valvular, perivalvular, or combined pathologies distinctly influence valve durability or the risk of adverse cardiovascular events, highlighting a critical gap in current cardiovascular research. Evidence linking SLT to long-term valve durability remains inconclusive, and conflicting data on antithrombotic strategies preclude routine anticoagulation for durability enhancement without an individualized bleeding risk assessment. 43 Figure 3 Subclinical Thrombosis of the Aortic Valve Complex After TAVR This figure shows the anatomical distribution and cardiac CT angiography features of subclinical aortic valve complex thrombosis post-TAVR, including valvular and perivalvular regions. Valvular thrombosis. HALT: Meniscus-shaped lesion on leaflet surface; graded by extent leaflet involvement. RLM: Impaired systolic excursion due to extensive HALT. Supravalvular thrombosis: Thrombus above the leaflet tips, within the valve frame or cuff region, which may disturb local hemodynamics or contribute to embolic risk. Subvalvular thrombosis: Thrombus located below the leaflet base, often within the left ventricular outflow tract, perivalvular thrombosis. Sinus of Valsalva thrombi: Located within the anatomic sinuses adjacent to the valve frame, often associated with localized flow stasis. Neosinus thrombi: Arise within the newly created neosinus compartment post-TAVR, a region susceptible to thrombosis due to altered flow dynamics and reduced washout. RLM = reduced leaflet motion; other abbreviations as in Figures 1 and 2 . Subclinical Thrombosis of the Aortic Valve Complex After TAVR This figure shows the anatomical distribution and cardiac CT angiography features of subclinical aortic valve complex thrombosis post-TAVR, including valvular and perivalvular regions. Valvular thrombosis. HALT: Meniscus-shaped lesion on leaflet surface; graded by extent leaflet involvement. RLM: Impaired systolic excursion due to extensive HALT. Supravalvular thrombosis: Thrombus above the leaflet tips, within the valve frame or cuff region, which may disturb local hemodynamics or contribute to embolic risk. Subvalvular thrombosis: Thrombus located below the leaflet base, often within the left ventricular outflow tract, perivalvular thrombosis. Sinus of Valsalva thrombi: Located within the anatomic sinuses adjacent to the valve frame, often associated with localized flow stasis. Neosinus thrombi: Arise within the newly created neosinus compartment post-TAVR, a region susceptible to thrombosis due to altered flow dynamics and reduced washout. RLM = reduced leaflet motion; other abbreviations as in Figures 1 and 2 .

The relationship between subclinical thrombosis and hemodynamics remains poorly understood. Flow stasis in the native sinus and neosinus has increasingly been recognized as a critical factor, correlating with the severity and occurrence of early transcatheter heart valve (THV) thrombosis. 44 , 45 In particular, the neosinus is defined as the space between the leaflet of the TAVR valve and the valve cage, whereas the anatomic or native sinus refers to the space between the native valve and the aortic wall ( Central Illustration ). Understanding fluid dynamics in the lower portion of the native sinus and the neosinus remains a significant challenge due to limited optical accessibility. In addition, their characteristics and flow dynamics are influenced by several variables, including the type and size of the bioprostheses used, the precise deployment location, and native valve anatomy. Central Illustration Subclinical Thrombosis of the Aortic Valve Complex Encompasses All Thrombus Locations Within the Complex, Including Both Valvular and Perivalvular Areas The valvular component includes hypoattenuated leaflet thickening as the primary manifestation affecting the leaflets within the device 1, along with subvalvular 2 and supravalvular 3 thrombosis. Perivalvular thrombi involve the sinus of Valsalva regions, also known as the anatomic sinus 4. Potential modulators of aortic valve complex thrombosis include factors related to valve type, characteristics of the native valve, and the design features of the anatomic remnant sinus and neo-sinus. LVOT = left ventricular outflow tract; THV = transcatheter heart valve. Subclinical Thrombosis of the Aortic Valve Complex Encompasses All Thrombus Locations Within the Complex, Including Both Valvular and Perivalvular Areas The valvular component includes hypoattenuated leaflet thickening as the primary manifestation affecting the leaflets within the device 1, along with subvalvular 2 and supravalvular 3 thrombosis. Perivalvular thrombi involve the sinus of Valsalva regions, also known as the anatomic sinus 4. Potential modulators of aortic valve complex thrombosis include factors related to valve type, characteristics of the native valve, and the design features of the anatomic remnant sinus and neo-sinus. LVOT = left ventricular outflow tract; THV = transcatheter heart valve. The geometric configuration of THV and its interaction with coronary flow may play a role in subclinical aortic valve–complex thrombosis. Evidence suggests that a high risk of blood stagnation in the neosinus region, driven by inadequate blood flow washout during diastole is crucial in thrombosis. 46 , 47 , 48 Indeed, abnormal flow patterns—characterized by low shear stress, slow washout, and reduced velocities—create an ideal milieu for platelet activation and the initiation of thrombus formation. Work by Midha et al. 46 highlighted the critical role of the neosinus concept in TAVR outcomes, using CT scans of 72 patients from the RESOLVE registry (Assessment of Transcatheter and Surgical Aortic Bioprosthetic Valve Thrombosis and Its Treatment with Anticoagulation). Their findings revealed that reduced blood flow velocity and impaired clearance of particles in simulated neosinus scenarios contributed to thrombus formation. A supra-annular valve position reported has a lower risk of leaflet thrombosis compared to an intra-annular valve of an equivalent-size, due to faster neosinus washout via the coronary artery. Smaller TAVR neosinus volumes have been associated with an increased risk of HALT in vivo. 34 However, in vitro findings have challenged this association, suggesting that smaller neosinus volumes may reduce flow stagnation and particle residence, thereby lowering the risk of leaflet thrombosis. 49 This discrepancy underscores the importance of further investigation and highlights the potential value of proactively optimizing neosinus design in the development of future THV. A recent study demonstrated that the risk of blood stasis in the aortic root and sinus was significantly higher after TAVR and correlated with the preintervention calcium volume distributions of the leaflets. 50 In another study, using a left heart simulator under pulsatile physiological conditions, researchers assessed the hemodynamic performance of self-expanding and balloon-expandable THVs. The findings revealed that self-expanding supra-annular valves exhibited faster flow washout in both the sinus and neosinus regions compared to balloon-expandable THVs. 51 Finally, artificial intelligence and risk prediction models based 52 on morphological parameters, such as neosinus height and volume, hold significant promise in accurately predicting the occurrence of HALT in TAVR patients. Failure of uniform valve expansion, (eg, asymmetrical or underexpansion) is associated with flow disturbances across the THV 53 , 54 and asymmetrical stress, which both facilitate early thrombosis and leaflet degeneration. Asymmetrical balloon expansion of the THV, manifesting as stent frame deformation, significantly predicts impaired hemodynamic valve performance and may increase the occurrence of HALT. THV underexpansion leads to increased blood stasis on the surface of the THV leaflets. A previous study showed that regions where blood stasis persisted for over 0.5 s, which are highly susceptible to platelet activation, expanded linearly with the degree of TAV underexpansion. 55 Nonuniform expansion of TAVR prostheses has been associated with an increased risk of HALT. 13 , 34 , 55 Nagasaka et al. highlighted the clinical impact of THV deformation in patients with bicuspid aortic valves. 41 Their findings revealed a significant association between THV deformation—characterized by underexpansion and/or eccentricity—and all-cause mortality, particularly in patients with annular and left ventricular outflow tract calcification. This calcification pattern was further linked to an increased risk of HALT at 30 days and an elevated risk of major adverse cardiovascular events. The study highlights the critical role of calcium distribution and THV deformation in post-TAVR outcomes. The position and implantation depth of THVs have been associated with the occurrence of new conduction disturbances and paravalvular leakage, and THV depth specifically may be associated with the potential development of SLT. 56 , 57 A substudy from the ADAPT-TAVR (Anticoagulation Vs Dual Antiplatelet Therapy for Prevention) trial, supported by in vitro experiments, 57 revealed that native sinus thrombosis was more frequent in self-expanding valves and in cases where the THV implant depth was greater (OR: 1.2 [95% CI: 1.1-1.3]; P < 0.001). To date, no major trial or meta-analysis have demonstrated a definitive difference in SLT rates between balloon-expandable and self-expanding THVs, even after accounting for varying implantation depths. The noncoronary sinus of Valsalva is the most common site of thrombosis. Patients with supra-annular THV positioning, small native sinuses of Valsalva, low coronary ostium height, short sinotubular junction height, and narrow sinus inflow diameters ate at an increased risk of developing native sinus thrombosis. Commissural misalignment of the prosthetic valve, as determined by comparing native (pre-TAVR) and prosthetic (post-TAVR) aortic valve orientations on cardiac CT, was associated with the subsequent development of HALT. 58 Overall, these anatomical and procedural factors highlight the importance of precise THV placement in minimizing thrombotic complications and optimizing long-term valve durability.

Dr Morel has received grants in support of investigator and investigator-initiated studies from 10.13039/100004325 AstraZeneca , 10.13039/100004374 Medtronic and 10.13039/100001003 Boehringer Ingelheim , all outside of the submitted work; he has been awarded grants by “Fondation Cœur et Recherche” and “Endofrance,” 2 reputable charities in France committed to advancing research initiatives in cardiovascular disease in endometriosis. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.