How long will my TAVI valve last, doctor?

KEYWORDS:

• Aortic Stenosis
• Transcatheter
• aortic valve replacement
• durability

1. Introduction

1.1. Aortic stenosis

Aortic stenosis is the most common clinically relevant valve pathology, affecting 1% of people over 65 years old, rising to 4% in those over 85 years [1,2]. Once symptoms develop, the prognosis of the condition is poor, with median survival less than 2 years after the onset of heart failure [1,2]. In fact, recent population data indicate that the prognosis for those with severe AS, even in the absence of symptoms, is significantly worse than their peers [3]. Historically, the only treatment option has been surgical aortic valve replacement, with either a mechanical or tissue prosthesis [4]. Mechanical valves have excellent durability but require lifelong anticoagulation to prevent valve thrombosis. By contrast, bioprosthetic valves, which do not require warfarin, are substantially less durable [4].

1.2. The development of TAVI

The development of transcatheter aortic valve implantation (TAVI) has dramatically changed the treatment landscape of severe aortic stenosis since the first-in-human implant in 2002 [5]. TAVI is the percutaneous replacement of the aortic valve, predominantly performed via the femoral artery under local anesthetic and involves a short hospital stay. It was therefore initially seen to offer a minimally invasive option to suitable patients who are deemed inoperable or at high surgical risk. Based on the results of pivotal randomized trials and changes in the major guideline recommendations, the use of TAVI has increased exponentially worldwide [6,7]. The European Society of Cardiology (ESC) currently recommends that patients 75 years or older who are suitable for both procedures receive TAVI rather than surgical aortic valve replacement (SAVR), while the American Heart Association guidelines are more inclusive, with a Class I indication for both transfemoral TAVI or SAVR for patients aged 65 or older [6,7]. This discrepancy is mainly due to the lack of long-term data available on TAVI valve durability.

1.3. The challenge of defining durability of TAVI

There has been a historical lack of a standardized definition of valve durability which has caused difficulty when comparing the treatment modalities. For example, in the early trials of surgical valves, many used the clinical end points of ‘freedom from reoperation’ or death rather than actual valve degeneration and it is quite possible that many patients with degenerated tissue valves were not reoperated on due to frailty. Furthermore, many did not assess echocardiographic follow-up and therefore most probably underestimated the true incidence of SVD. This highlights the fundamental challenge of determining whether to define the durability of a device in terms of its functional/anatomical status or whether it requires replacement. From a patient’s point of view, the latter is much more likely to be important, but from an academic perspective, they are both essential parameters. In 2017, the European Association of Percutaneous Cardiovascular Interventions consensus has now produced a standardized definition of structural valve degeneration (SVD) and bioprosthetic valve failure (BVF) to enable more scientific reporting of durability for future studies [8]. It is based on a combination of altered hemodynamic parameters, most commonly due to chronic inflammation and pannus formation, and morphological criteria such as leaflet restriction/fracture or any strut frame abnormality. BVF is defined as a patient-orientated clinical endpoint and encompasses any of: 1) autopsy-derived bioprosthetic valve dysfunction, most probably related to the cause of death, or ‘valve-related death,’ which implies death caused by bioprosthetic valve dysfunction without autopsy; 2) valve reintervention (e.g. valve-in-valve TAVI, paravalvular leak closure, or SAVR); or 3) severe hemodynamic SVD. More recently, the VARC-3 committee provided an updated definition of SVD that combines intrinsic valve morphology with valve hemodynamics to be able to differentiate true SVD from high residual gradients as a result of severe patient prosthesis mismatch [9].

1.4. TAVI versus SAVR

Several landmark trials comparing SAVR and TAVI are now available across the whole surgical risk spectrum, extending from inoperable/high surgical risk patients via intermediate risk patients to low risk surgical patients [10–13]. Table 1 contains a summary of the four landmark trials with outcome to at least 5 years. Most RCTs have shown similar outcomes, or even improved hemodynamics with TAVI, especially in patients with small aortic annuli (23 mm) [14]. This is most probably due to some TAVI devices having suprannular designs which enable relatively large calculated valve areas post implant [14]. In addition, we also have access to observational population data that tracks real world outcomes after TAVI, which add value from a completely different perspective.

Table 1. Currently available mid term outcomes of TAVI Trials.

Trial	Reference	Surgical Risk	Mean STS	FU Duration	TAVI Device	Year	Patients	5 year outcome
CoreValve U.S. Pivotal High Risk Trial	13	high risk	7.40%	5 year	1st generation CoreValve (Medtronic)	2018	797	No significant difference (death 55.3% TAVI vs 55.4% SAVR; Stroke 12.3% TAVI vs 13.2% SAVR)
PARTNER 2 Trial	10	Intermediate	5.80%	5 year	2nd Generation SAPIEN XT (Edwards)	2020	2032	No significant difference (death or disabling stroke 47.9% versus 43.4%, p = 0.21)
SURTAVI Trial	14	Intermediate	4.50%	5 year	SE CoreValve (84%) Evolut R (16%) Medtronic	2022	1660	No significant difference (death or disabiling stroke 31.3% versus 30.8%, p = 0.85)
NOTION	15	low risk	3%	8 year	1st generation CoreValve (Medtronic)	2021	280	No significant difference (death 51.8% TAVI vs 52.6% SAVR; Stroke 8.3% TAVI versus 9.1% SAVR)

*STS = Society of Thoracic Surgeons; TAVI = transcatheter aortic valve implantation; SE = self expandable.

To date, randomized data for TAVI exist up to 8 years maximum versus 15–20 years for conventional surgical valves [4,15]. This is due to TAVI technology being relatively new and also that the life expectancy of patients in the early major landmark trials of high/extreme risk cohorts being low, with only approximately 30% of patients surviving to 5 years [12]. In fact, 5-year durability data were analyzed in the 424 surviving patients from the original PARTNER trial, and low rates of SVD were reported, with mean gradients of approximately 10 mmHg in both arms and reintervention rates of less than 1% [16].

Recently, pooled data on durability, from the CoreValve US High Risk Pivotal and SURTAVI trials have reported lower rates of valve degeneration with TAVI at 5 years (7.8% (TAVI) vs. 14.2% (SAVR), p < 0.001) [13,14,17]. Notably, bioprosthetic valve degeneration (including SAVR/TAVI) was associated with a 1.5-fold increased risk of all-cause (p = 0.004) and cardiovascular mortality (p < 0.001) at 5 years [14].

Regarding observational data to 5 years, the FRANCE-2 Registry has reported the largest dataset of 4201 patients treated with self-expanding (34%) or balloon-expandable (66%) valves and revealed a rate of severe and moderate/severe SVD of 2.5% and 13.3%, respectively, at 5 years. Higher rates were noted with BE valves (8.9% SE versus 13.8% BE) [18].

Regarding longer term outcomes, randomized data from 1 trial exist to 8 years in low-risk patients. In the NOTION trial, 280 patients were randomized and the estimated risk of the composite outcome of all-cause mortality, stroke, or myocardial infarction was 54.5% after TAVI and 54.8% after SAVR (P = 0.94) [15]. All-cause mortality (51.8% vs. 52.6%; P = 0.90) was high in both cohorts, reflected probably by the mean age of patients in the trial being approximately 80 (mean age 79.1) [15]. Rates of SVD were again lower after TAVI than after SAVR (13.9% vs. 28.3%; P = 0.0017), whereas the risk of BVF was similar (8.7% vs. 10.5%; P = 0.61) [15]. When looking specifically at aortic valve reintervention rates, similar rates were observed (3.6% TAVI vs 2.3% SAVR, p = 0.51).

Although no clear association has been shown between subclinical leaflet thrombosis and SVD, hypo attenuated leaflet thickening (HALT) has been observed in 13–17% of patients undergoing a CT after TAVI at 30 days and is associated with increased transvalvular gradients. In the most recent PARTNER 3, HALT was more common after TAVI at 30 days (13% versus 5%. P = 0.03) but not at 1 year (28% vs 20%, p = 0.19) [19]. By contrast, no differences were found at 30 days and 1 year in the EVOLUT low-risk substudy (1-year rates of 30.9% TAVI versus 27.0% SAVR) [20]. Both studies reported the dynamic pathophysiology of HALT, with spontaneous resolution of 30-day HALT in some cases but overall increases in prevalence at 1 year.

For longer term data up to 10 years, we have reviewed observational data from the pioneering centers involved in the first TAVI implants. Overall data are encouraging. For example, the UK TAVI registry from 2007 to 2011 on 221 patients (79.4 ± 7.3 years; 53% male) has reported median echocardiographic follow-up of 7.0 years (range 5 − 13 years) [21]. Severe SVD occurred in 13 patients (5.9%; median 7.8 years post-TAVI) and valve-related reintervention/death occurred in only five patients (2.3%) [21]. Ten-year data from a high-risk cohort of 234 consecutive patients treated with SAPIEN (77.4%), Cribier-Edwards (20.9%) or CoreValve (1.7%) devices have also been reassuring with SVD/BVF rates of just 6.5%. Out of the six patients with severe SVD, two had reintervention (one redo TAVR and one SAVR) [22]. Furthermore, Barbanti et al. reported a relatively low 8-year rate of moderate (5.9%) and severe (2.4%) SVD in 288 patients treated with CoreValve (82.3%) and SAPIEN XT (16.7%). BVF was reported in 11 patients (4.51%).²³ In another registry of 378 patients treated with BE valves, low rates of SVD (3.2%) and BVF (0.6%) were reported at 8-year follow-up [23]* [24]. Similar rates have been reported in another cohort of 990 patients from eight Italian Centers undergoing TAVI with CoreValve/Evolut-R devices at 8-year follow-up (moderate SVD 3.0%; severe SVD 1.6%; BVF 2.5%) [25].

Whilst the observational data are encouraging, we still need to wait for long term randomized data from the PARTNER III and the Evolut Low-Risk studies to be confident that TAVR will be comparable to conventional surgical aortic valve replacement, especially regarding the reintervention rates. It must be noted that not all surgical prostheses for SAVR have the same durability, for example, 10-year rates of reintervention have been reported at 3.6% with Perimount valves versus 12.2% with Mitroflow/Crown valves, and it is important as clinicians to take this into account when deciding on SAVR [26,27].

2. What happens when a TAVI valve fails? What should we recommend?

When a TAVI valve fails, treatment options include repeat TAVI or TAVI explant followed by a surgical valve replacement. TAVR explantation has, however, been reported to be associated with a very high 30-day mortality (19.4%), which may also include aortic root repair due to the aortic wall adhesions that develop after TAVI [28]. Repeat TAVI has been shown to be feasible with acceptable outcomes (30-day survival rate of 98.5% and 1-year rate of 88.3% in patients with late valve dysfunction undergoing redo TAVI) [29]. The pivotal role of CT scanning and computer modeling will enable clinicians to determine feasibility more accurately, as well as facilitate choice of device and at what depth to enable access to the coronary circulation if needed [30]. Looking toward the future, TAVI-in-TAVI procedures are projected to rapidly increase as inevitably more younger patients will undergo index TAVI, so much focus will be required on future valve designs to facilitate this procedure as safely and effectively as possible.

3. Conclusion

The currently available outcome data after TAVI are extremely promising from the point of view of requirements for reintervention and are competitive with surgical tissue AVR. More data are required both regarding longer-term outcomes and also concerning mechanisms of SVD and predicting outcomes in individual patients.

Declaration of interests

N Curzen has received unrestricted research grants from: Boston Scientific, Haemonetics, Heartflow, and Beckman Coulter. He has also received speaker fees/consultancy from Abbot Vascular, Heartflow, and Boston Scientific; and travel sponsorship from Edwards, Biosensors, Abbot, Lilly/D-S, St Jude Medical, Heartflow, and Medtronic.

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial relationships or otherwise to disclose.

Additional information

Funding

This paper was not funded.