Early Leaflet Thickening, Durability and Bioprosthetic Valve Failure in TAVR

INTRODUCTION

Historically, surgical aortic valve replacement (SAVR) was the standard of care for the treatment of symptomaticpatientswithsevere aortic stenosis (AS). However, over the last decade, transcatheter aortic valve replacement (TAVR) has become the established alternative to surgery for the treatment of patients with symptomatic, severe AS. Large, multicenter trials have shown the noninferiority, and even superiority, of TAVR to SAVR in high-risk surgical patients^1,2 and intermediate-risk surgical patients.^3,4

In 2019, the Food and Drug Administration (FDA) approved TAVR for patients with a low surgical risk after multiple large multicenter trials established the role of TAVR for the treatment of low-risk patients with severe AS.^5–7 As younger patients, with a typical life expectancy of more than 10 years, receive TAVR they will most likely outlive their bioprosthetic valve. Indeed, all bioprosthetic valves, both surgical and transcatheter, have a finite lifespan before their leaflets inevitably degenerate, leading to stenosis or regurgitation. The unresolved question of TAVR valves is the long-term valve durability. The paucity of data regarding the durability of transcatheter heart valves (THVs) is often underscored as a weakness of TAVR, but the data for surgical bioprosthesis durability are also poor in quality. This review will discuss early leaflet thickening, valve durability, and bioprosthetic valve failure (BVF) in TAVR patients.

SURGICAL VALVE DURABILITY

Historically, surgery has been the standard of care for the treatment of severe AS. Given the fact that they have been around for longer, there is a larger number of studies reporting data on surgical aortic valve durability. However, there is a lot of variability in the study methodologies and definitions as well as inadequate follow-up. The heterogeneity of these studies makes the interpretation of the data challenging.

Largely, structural valve degeneration (SVD) in SAVR studies has been defined as valve reintervention with redo surgery, with SVD rates of 2% to 10% at 10 years and 10% to 20% at 15 years. However, this only accounts for valve failure that proceeds to redo surgery and does not account for patients who die or suffer from SVD treated medically. This leads to a substantial underestimate of the incidence of valve deterioration, hemodynamic compromise, and overall valve failure. This underestimation has been validated by studies in which SVD definitions include hemodynamics assessed by echocardiography, and which reported significantly higher SVD rates of 10% at 5 years and 30% at 10 years post-SAVR.⁸

A recent systematic review of preexisting literature on the actuarial freedom from SVD in SAVR patients and outcomes of SVD, yielding 167 studies and 12 FDA reports.⁹ Authors found that 11 different definitions of SVD were used and only 11 studies used a core laboratory to collect data. Furthermore, mean follow-up ranged from less than 1 year all the way to 14 years with only 0% to 37% of patients remaining at risk at maximum follow-up. Thus, there is substantial variability in reporting SVD for surgical valves, with different definitions and inadequate long-term systematically collected core laboratory data. Rigorously collected long-term data with standardized definitions for surgical valves are needed to provide a benchmark for the durability of rapidly evolving transcatheter valves.

LEAFLET THICKENING

Leaflet thrombosis has been observed with both surgical and transcatheter bioprostheses. Symptomatic obstructive valve thrombosis results in an increase in the transvalvular gradient and reduction in the effective orifice area of the valve. However, overall, this event is rare in TAVR patients, with a rate of about 0.5%. Furthermore, there is a signal for a slightly higher rate of leaflet thrombosis in annular valves as compared with supra-annular valves. However, this needs to be further validated in larger, prospective studies.¹⁰

In contrast, subclinical thrombosis—meaning asymptomatic—causes thickening and reduced leaflet motion of the bioprosthetic valve. This finding could be missed on echocardiography as the transvalvular gradients often remain within the normal range. Commonly, this finding is only detected on computed tomography (CT) scan, which is not routinely performed outside of clinical trials. Early leaflet thickening (30 days) is found to be more frequent in THVs, as compared with surgical valves, with an estimated incidence rate of 5% to 40% of patients.¹¹ It is thought to be more frequent in THVs given the incomplete expansion of the THV in an oval aortic annulus and also the metallic nature of the struts.¹² That being said, there is no clear signal yet that leaflet thickening is associated with excess cerebrovascular accidents or premature structural valve deterioration. Recently, several clinical trials have attempted to address these questions further.

The LRT (Low-Risk TAVR) trial was an investigator-initiated, prospective, multicenter study, and was the first FDA-approved Investigational Device Exemption trial to evaluate the feasibility of TAVR in low-risk patients. In the LRT trial, leaflet thickening was a secondary endpoint. Subjects in the LRT trial underwent follow-up time-resolved contrast-enhanced cardiac CT at 30 days. Hypoattenuated leaflet thickening (HALT) was observed in 14.0% of subjects (n = 27 or 193) with an evaluable CT or transesophageal echocardiography at 30 days. This rate is similar to other studies in low-risk TAVR patients, for example, the PARTNER 3 trial (13%)⁶ and the Evolut Low-Risk Trial (17.3%).⁷ Additional, larger trials including the SAVORY (Subclinical Aortic Valve Bioprosthesis Thrombosis Assessed With Four-Dimensional Computed Tomography) and RESOLVE (Assessment of Transcatheter and Surgical Aortic Bioprosthetic Valve Thrombosis and Its Treatment With Anticoagulation) observational registries (13%)¹³ showed similar rates in all surgical risk cohorts.

Further analysis in the LRT trial demonstrated that reduced leaflet motion was observed in 11.2% of subjects. Interestingly, in this analysis, HALT was observed in subjects who received balloon-expandable valves only, and not self-expanding valves. Furthermore, at 30 days, mean valve area and dimensionless index were lower in subjects with HALT, with a trend toward higher mean gradients. However, at 1 year, these differences appeared to resolve. Finally, clinically at 1 year, there was a numerically higher rate of stroke in subjects with HALT (3.8% vs 1.9%; P = .53), although the absolute number of events was small in both groups (1 of 27 with HALT vs 4 of 166 with no HALT).¹⁴

Next, a follow-up LRT Trial was performed, which was an investigator-initiated, prospective, multicenter study, and was the first and only U.S. FDA–approved investigational device exemption trial to evaluate the feasibility of TAVR with either balloon-expandable or self-expanding valves in low-risk patients with bicuspid AS. Baseline and follow-up echocardiography and CT to detect leaflet thickening were analyzed in an independent core laboratory.

At 30 days, 60 of 61 subjects underwent contrast-enhanced CT scans, which were analyzed for subclinical leaflet thrombosis. HALT was present in 6 patients (1 with a self-expanding THV and 5 with balloon-expandable THVs). HALT was observed in 10% of bicuspid TAVR patients at 30 days, which is similar to the previously published tricuspid LRT cohort (14%),¹⁴ the PARTNER 3 trial (13%),⁶ and the larger SAVORY (Subclinical Aortic Valve Bioprosthesis Thrombosis Assessed With Four-Dimensional Computed Tomography) and RESOLVE (Assessment of Transcatheter and Surgical Aortic Bioprosthetic Valve Thrombosis and Its Treatment With Anticoagulation) observational registries (13%).¹³ Furthermore, the stroke rate in the bicuspid arm of the LRT trial was extremely low (only 1 nondisabling stroke at 30 days), so it was not possible to correlate THV leaflet thrombosis with clinical events.

At the current time, there is no evidence that leaflet thickening results in worsening valve durability. In the tricuspid arm of the LRT trial, patients with HALT had inferior THV hemodynamic status compared with those without HALT at 30 days. Patients with HALT had smaller aortic valve areas and Doppler velocity index (0.4±0.1 vs 0.5±0.1; P = .003) than those without HALT at 30 days. However, hemodynamic differences disappeared at 1 year and clinical outcomes were similar at the 1-year mark.¹⁵ Longer follow-up of larger cohorts with serial echocardiography will be needed to determine whether THV leaflet thrombosis/thickening affects long-term THV hemodynamic status, durability, and rate of late thromboembolic events.

Further information was provided by the pivotal industry-sponsored low-risk TAVR trials with the balloon-expandable and a self-expanding THVs. First, a substudy of the PARTNER 3 (The Safety and Effectiveness of the SAPEIN 3 Transcatheter Heart Valve in Low-Risk Patients With Aortic Stenosis)¹⁶ enrolled 435 patients with low-surgical-risk AS who received either TAVR (n = 221) or surgery (n = 214) and who underwent serial 4-dimensional CT at 30 days and 1 year. They found that the incidence of HALT increased from 10% at 30 days to 24% at 1 year. Spontaneous resolution of 30-day HALT occurred in 54% of patients at 1 year, whereas new HALT appeared in 21% of patients at 1 year. HALT was more frequent in transcatheter versus surgical valves at 30 days (13% vs 5%; P = .03), but not at 1 year (28% vs 20%; P = .19). The presence of HALT did not significantly affect aortic valve mean gradients at 30 days or 1 year. Furthermore, patients with HALT at both 30 days and 1 year, compared with those with no HALT at 30 days and 1 year, had significantly increased aortic valve gradients at 1 year.

Second, a similar substudy was conducted evaluating patients who received a self-expanding THV in the Evolut Low-Risk trial. In this substudy patients from the Evolut Low-Risk trial who were not on oral anticoagulation underwent computed tomographic imaging at 30 days and 1 year after TAVR (n = 179) or SAVR (n = 139). At 30 days, the frequency of HALT was 17.3% for TAVR and 16.5% for surgery; and at 1 year, the frequency of HALT was 30.9% for TAVR and 28.4% for surgery. Aortic valve hemodynamic status was not influenced by the presence or severity of HALT. They concluded that the presence of computed tomographic imaging abnormalities on aortic bioprostheses was frequent but dynamic in the first year after self-expanding transcatheter and SAVR, but these findings did not correlate with aortic valve hemodynamics after aortic valve replacement in patients at low risk for surgery. Clearly, based on these studies, the impact of HALT on thromboembolic complications and SVD needs further assessment.

Naturally, it would be of interest to identify anatomic features that would place a THV at an increased risk for HALT and subsequent leaflet degeneration. A recent subanalysis of the LRT trial tried to determine anatomic characteristics associated with HALT, which may contribute to early THV degeneration.¹⁷ The authors found that in patients who developed HALT, THV implantation depth was shallower than in patients who did not develop HALT. In addition, there were more patients in the HALT group with commissural malalignment, but this did not reach statistical significance. Furthermore, in a univariable regression model, no predetermined variables were shown to independently predict the development of HALT. Future, larger, prospective trials are needed to potentially identify anatomic risks associated with HALT.¹⁷

LEAFLET DEGENERATION

Just as with native aortic valve, bioprosthetic leaflet degeneration occurs when leaflet calcification develops and results in either valve stenosis, regurgitation, or both.¹² There are three components specific to a THV, as compared to a SAVR, which may impact early leaflet degeneration. First, when loading/crimping the THV into the delivery catheter, there is potential for trauma to the bioprosthetic leaflets. Second, after the THV is deployed, noncircular expansion could affect normal valvular functioning. Noncircularity may occur due to excessive native aortic leaflet calcification, ellipticity of the left ventricular outflow tract, and prosthetic valve oversizing. This is more commonly seen with self-expanding THVs, although the supra-annular position of the leaflets in the Evolut valve may mitigate the impact of noncircularity at the annulus. Finally, the presence of turbulence in the aortic root due to the combined presence of bulky calcium nodules in the sinuses of Valsalva and presence of the THV, which could potentially affect blood flow patterns, resulting in chronic mechanical stresses on the bioprosthetic leaflets, and in turn leading to early degeneration. A recent systematic review and meta-analysis identified younger age, patient-prosthesis mismatch, body surface area, and smoking as risk factors for SVD.¹⁸

VALVE DURABILITY

Failure modes of bioprosthetic aortic valves may be different for TAVR versus SAVR. When an SAVR valve fails, the etiology is usually due to a leaflet tear (leading to valve incompetence), calcification (leading to valve restenosis), or pannus formation. When a TAVR valve fails, the etiology could be from leaflet tear or calcification, but it could also be due to valve dislocation (albeit very rare) or worsening paravalvular leak.¹⁹ Historically, the lack of standard definitions of SVD has made it difficult to analyze studies on the durability of both surgical and transcatheter bioprostheses.

However, since 2017, the European Association of Percutaneous Cardiovascular Interventions (EAPCI), endorsed by the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS), proposed a consensus definition, applicable to both transcatheter and surgically implanted bioprosthetic valves. When describing the mechanism of valve failure, the etiology can be separated into 2 broad categories—(1) structural valve deterioration and (2) non-structural valve deterioration.²⁰

SVD is the most common type of bioprosthetic valve dysfunction.²¹ It is characterized by permanent intrinsic changes to the valve. Examples of structural valve deterioration include leaflet calcification, leaflet tear, valve seam disruption, and stent fracture. These abnormalities result from both patient-related factors and prothesis-related factors. The patient-related factors include dyslipidemia, diabetes, hypertension, metabolic syndrome, phosphocalcic dysregulation, and/or a component of an immune reaction. The prosthesis-related factors include absence of antimineralization treatment, flaws in the bioprosthesis design, severe prosthesis-patient mismatch, and/or small prosthesis size. These two factors lead to a spectrum of increased leaflet mechanical stress all the way to abnormal valvular flow leading to SVD.

Examples of nonstructural valve deterioration include leaflet thrombosis, endocarditis, and paravalvular leakage. These abnormalities tend to happen from TAVR specific factors including leaflet injury (due to crimping, loading, dilatation), abnormal trans-and/or-paravalvular flow patterns, and/or noncircular, irregular, incomplete stent deployment^8,22 (Fig. 1).

Fig. 1.

Mechanisms of bioprosthetic valve dysfunction.

Once either SVD or nonstructural valve dysfunction is present, the next step in the cascade is hemodynamic valve deterioration. The ESC/EACTS created standardized definitions for this as well. Morphologic SVD needs at least multidetector CT to be diagnosed, whereas the diagnosis of hemodynamics is based on echocardiography or invasive hemodynamics. They define moderate hemodynamic valve deterioration is defined as a mean transprosthetic gradient ≥20 to <40 mm Hg or ≥10 to 20 mm Hg change from baseline. Alternatively, moderate intraprosthetic aortic regurgitation, new or worsening (>11/4+) from baseline.²⁰ After this, the patient then develops severe hemodynamic valve deterioration and BVF. Severe hemodynamic valve deterioration is defined as a mean transprosthetic gradient ≥40 mm Hg or ≥20 mm Hg change from baseline or severe intraprosthetic aortic regurgitation. Alternatively, it can be defined by autopsy with findings likely related to the case of death, valve-related death, or repeat intervention.²⁰ BVF is used when there are clinical symptoms and signs of bioprosthetic valve deterioration. Furthermore, ESC/EACTS define BVF as a patient-oriented composite outcome of death (confirmed by autopsy or by clinical diagnosis of bioprosthetic valve dysfunction before death), repeated intervention (including valve-in-valve TAVR, paravalvular leak closure or surgery), or severe hemodynamic structural valve deterioration (Fig. 2). In addition, it can be defined as early (≤30 days) or late (>30 days) after valve replacement.

Fig. 2.

European Society of Cardiology/European Association for Cardio-Thoracic Surgery Definition of Hemodynamic Valve Deterioration.

Using the ESC/EACTS definitions, studies have demonstrated that the durability of THV is up to 7 and 8 years.^23–25 Previous analysis looked at clinical trials to try to assess long-term durability of TAVR. They concluded that the weighted incidence of SVD at 5 to 8 years was 1.3% (95% CI 0.7–1.9) and the weighted incidence of BVF at 6 to 8 years was 3.7% (95% CI 2.7–4.6).²⁶ An additional study looked at the 8-year durability of 990 patients treated with a self-expanding TAVR between 2007 and 2011 in Italy.²⁷ In this analysis, 78% of patients were alive at 8 years and the rate of moderate SVD was 3.0%, severe SVD was 1.6%, and BVF was seen in 2.5% of patients.²⁷ An additional trial, the UK TAVI Trial showed excellent long-term THV function. Between 5 and 10 years after implantation, 91% remained free of SVD with only 0.4% cases of severe SVD at 5.3 years after implantation and 8.7% cases of moderate SVD.²⁸ Finally, 2.4% patients demonstrated moderate to severe valve durability and 4.51% with BVF at 8 years in the local Italian REPLACE registry.²⁹ However, the main limitation of these earlier trials is that most of the studies to date have had a high rate of mortality at 5 years in the high and extreme risk patients treated with TAVR. This is due to the consequence of the advanced age and comorbidities of the high-risk population being treated, and not necessarily valve failure itself.

The NOTION trial provides data on bioprosthetic valve durability from a randomized clinical trial in patients at low surgical risk of mortality. At 6 years, all-comers (TAVR and surgical) showed that the rate of SVD was higher with SAVR than TAVR (24.0% vs 4.8%; P<.001), with similar rates of all-cause mortality (42.5% for TAVR vs 37.7% for SAVR, P = .58) and no differences in terms of non-SVD (57.8% vs 54.0%, P = .52) and endocarditis (5.9% vs 5.8%, P = .95). Furthermore, TAVR and SAVR patients experienced a similar degree of BVF through 6 years (7.5% vs 6.7%; P = .89).³⁰ A summary of these trials is outlined in Figs. 3–5.

Fig. 3.

Incidence of structural valve deterioration.

Fig. 5.

Computed tomography imaging of bioprosthetic valve thickening.

Lastly, in 2020, a PARTNER 2A substudy trial sought out to determine and compare the 5-year incidence of SVD, using the standardized definitions, in intermediate-risk patients with severe AS given annular TAVR (Second Generation SAPIEN-XT or Third Generation SAPIEN 3) or SAVR. The investigators found that compared with SAVR, the SAPIEN-XT TAVR cohort had a significantly higher 5-year exposure adjusted incidence rates (per 100 patient-years) of SVD (1.61 ± 0.24% vs 0.63 ± 0.16%), SVD-related BVF (0.58 ± 0.14% vs 0.12 ± 0.07%), and all-cause (structural or nonstructural) BVF (0.81 ± 0.16% vs 0.27 ± 0.10%) (P≤.01 for all). Alternatively, the 5-year rates of SVD (0.68 ± 0.18% vs 0.60 ± 0.17%; P = .71), SVD-related BVF (0.29 ± 0.12% vs 0.14 ± 0.08%; P = .25), and all-cause BVF (0.60 ± 0.15% vs 0.32 ± 0.11%; P = .32) in SAPIEN 3 TAVR were not significantly different to a propensity score-matched SAVR cohort. Finally, when comparing the 5-year rates of SVD and SVD-related BVF were significantly lower in SAPIEN 3 (third generation) versus SAPIEN XT (second generation) TAVR matched cohorts. Their finding highlights that SVD rates may be higher in TAVR patients, but advancements in valve design have decreased the overall risk.³¹

CORONARY ACCESS

When discussing valve durability, and eventual valve failure, the issue of coronary access after the deployment of the valve is important. The incidence of coronary artery disease, and associated acute coronary syndrome, is common with patients receiving TAVR for AS, given the fact that these patients share common comorbidities. Given the structure of the THV, coronary access with a guide catheter to perform a percutaneous coronary intervention can be more challenging. For example, in low-risk patients who received a balloon-expandable THVs, the most challenging anatomy for coronary access (THV frame above and commissural suture post in front of a coronary ostium) was observed in 9% to 13% of subjects.³² Furthermore, in low-risk patients who received a self-expanding THVs, CT simulation predicted sinus of Valsalva sequestration and resultant coronary obstruction during future TAVR-in-TAVR in up to 23% of patients. In addition, CT simulation predicted that the position of the pinned THV leaflets would hinder future coronary access in up to 78% of patients after TAVR-in-TAVR.³³

VALVE-IN-VALVE

When bioprosthetic valve degeneration does occur, repeat intervention is needed. The patients’ two options are SAVR or valve-in-valve (ViV) TAVR procedure. ViV procedures have demonstrated encouraging results in both patients with degenerated surgical aortic bioprostheses³⁴ and TAVR.³⁵ However, when TAV-in-TAV procedure is performed, the issue of coronary access becomes more problematic. In ViV TAVR procedures where a self-expanding THV is used, a recent study using CT simulation predicted that the position of the pinned THV leaflets would hinder future coronary access in up to 78% of patients.³³

YOUNGER PATIENTS

Among patients in whom a bioprosthesis is appropriate, there are certain scenarios when SAVR is preferred over TAVR. The recent 2020 ACC/AHA Guidelines for the Management of Patients with Valvular Heart Disease: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice guidelines provides guidance on these patients.³⁶ They still recommend SAVR as the preferred treatment among patients younger than 65 years or with a life expectancy more than 20 years. SAVR or TAVR is recommended after shared decision-making among symptomatic patients ages 65 to 80 years with no other contraindications. These recommendations appear to be based on studies comparing the long-term mortality benefit of patients who underwent aortic valve replacement or mitral valve replacement with a mechanical or biologic prosthesis (surgical or transcatheter). They found that there was a long-term mortality benefit that was associated with a mechanical prosthesis, as compared with a biologic prosthesis, and this persisted until 70 years of age among patients undergoing mitral valve replacement and until 55 years of age among those undergoing aortic valve replacement. In addition, the incidence of reoperation was significantly higher among recipients of a biologic prosthesis than among recipients of a mechanical prosthesis. Alternatively, patients who received mechanical valves had a higher cumulative incidence of bleeding and, in some age groups, stroke than did recipients of a biologic prosthesis.³⁷

It is important to note as well, anecdotally, that many patients do not want a mechanical valve, and commitment to anticoagulation, even if they are younger. It is important to have a shared decision with your patient to determine the management that is best for them. In addition, it is always a class I indication for all cases to be reviewed by a Heart Team to determine the aortic intervention. These age recommendations may change if we can better define the durability of both SAVR and TAVR in the future.

SUMMARY

Moving forward, establishing valve durability in larger cohorts and determining the relative freedom from structural valve deterioration of different valve designs is needed as TAVR is increasingly offered to and requested by younger patients with longer life expectancy. Ensuring adequate coronary access, especially following TAV-in-TAV, is imperative as more younger patients receive TAVR.

Fig. 4.

Incidence of bioprosthetic valve failure.

KEY POINTS.

• Heterogeneity in definitions, and follow-up, of studies evaluating the durability of SAVR.

• Leaflet thickening present in THV; however, further studies are needed to determine whether THV leaflet thickening affects hemodynamic status, durability, and rate of clinical thromboembolic events.

• New standardized definitions of structural valve deterioration allow for future studies to adequately evaluate TAVR durability.

• Coronary access, valve-in-valve procedures, and lifetime management strategy for younger patients still need to be determined.

CLINICS CARE POINTS.

• Large, multicenter trials have shown the noninferiority, and even superiority, of TAVR to SAVR in high-risk, intermediate and low surgical risk patients.

• Structural valve degeneration in SAVR studies has been defined as valve re-intervention with redo surgery, with rates of 2–10% at 10 years and 10–20% at 15 years. Heterogeneity and underestimation.

• Leaflet thickening, and leaflet degeneration, is seen in both SAVR and TAVR valves.

• Consensus definitions of Structural Valve Deterioration by societies provides standardization and framework for future clinical trials.

• Ensuring adequate coronary access, especially following TAV-in-TAV, is imperative as more younger patients receive TAVR.