Determining valve durability in the era of TAVI

Paul Brennan and Mark Spence

Transcatheter aortic valve implantation (TAVI) has progressed from being a procedure that was solely reserved for patients who were too high risk for surgery to one that is now being considered for lower risk populations. In this commentary, Paul Brennan and Mark Spence review current definitions of bioprosthetic valve durability and how these can be applied to the durability of transcatheter heart valve.

The first “in-human” TAVI was performed in Rouen in 2002. Since then, TAVI has evolved to become an alternative intervention to surgical aortic valve replacement for high-risk patients with severe symptomatic aortic stenosis.1,2 Recent studies have demonstrated the non-inferiority of TAVI to surgery in intermediate-risk patients.3,4 The favourable outcomes of the pivotal trials have led to changes in the class of recommendation for TAVI and, alongside improved valve system technology, we have seen the use of TAVI progressively rise over the last 15 years.5

To aid surgeons and patients choosing between surgery and TAVI, a useable definition of aortic transcatheter heart valve durability and structural valve deterioration is required. This will also provide a firm basis for a coherent discourse for clinical investigators.

Long-term outcomes after bioprosthetic surgical valves have been widely published. Historically, surgical studies have used freedom from reoperation as a primary endpoint, and marker of structural valve deterioration, without defining the specific haemodynamic or imaging criteria required to diagnose deterioriation.6

In a recent 2015 registry, bioprosthetic aortic valve replacement durability was reviewed in 12,569 patients after implantation of the Edwards Lifesciences Carpentier-Edwards Perimount bioprosthesis between 1982 and 2011. The primary endpoint was time to explant for valve deterioration. In this registry, 354 bioprostheses were explanted during the follow-up period with 44% attributable to deterioration. Younger age was a statistically significant risk factor for explant due to deterioration. Actuarial estimates of risk of explant for deterioration at 10, 15 and 20 years—for patients under 60 years age—were 5.6%, 20% and 45%, respectively. In patients aged 60–80 years, risk of explant for deterioration was 1.5%, 5.1% and 8.1%, respectively, for the aforementioned time intervals.7 Freedom from deterioration, observed in a Hancock II registry in 2010, at five, 10, 15, and 20 years was 99.7%, 97.6%, 86.6% and 63.4% respectively.8 Similar outcomes were seen in a St Jude Medical Biocor registry in 2015, with freedom from deterioration at five, 10, 15 and 20 years being 97.9%, 92.1%, 84.8% and 67% respectively.9

There is, however, a paucity of data with respect to the long-term outcomes of bioprosthetic surgical aortic valve replacement for the patient group for whom TAVI was originally offered.

Several groups have now published five-year outcomes with regards to transcatheter bioprosthetic deterioration. The Medtronic CoreValve and Cribier Edwards/ Edwards Sapien/ Sapien XT bioprostheses were respectively associated with a 1.4% and 1.67% incidence of deterioration at the five-year follow-up point.10,11 Structural valve deterioration was defined in both groups using the surgical definition, of freedom from re-intervention, and also precise echocardiographic criteria according to the VARC (Valve Academic Research Consortium)-1 criterion.

A systematic review of 8,914 patients, across 13 observational studies, focusing on prognosis after TAVI, found that the pooled incidence rate of deterioration was 28.08 per 10,000 patient years, with 12% of patients experiencing deterioration proceeding to reintervention.12

The current VARC-2 definition of deterioration defines it as valve-related dysfunction or the need for a repeat procedure ≥30 days after TAVI and/or not due to endocarditis. Valve related dysfunction, as assessed by echocardiography, is diagnosed in the presence of a mean aortic gradient ≥20mmHg, a rise in gradient of 10mmHg, a dimensionless valve index <0.35 and/or an effective orifice area of ≤0.9-1.1cm2 and/or the presence of moderate-to-severe regurgitation.13 This definition of deterioration does not include valve thrombosis—an important cause of leaflet restriction and increasing gradient.

Not all patients with an increasing transprosthetic gradient, on echocardiographic follow-up, will be symptomatic or have significant prosthetic valve stenosis; and in this scenario, deterioration is further characterised as “isolated haemodynamic dysfunction”. Conversely, patients may demonstrate early anatomical changes, without haemodynamic dysfunction, such as leaflet thickening or reduced leaflet mobility, and this is characterised as “morphological surgical valve deterioration”.14

The term “bioprosthetic valve failure”—applied historically to surgical bioprostheses—differs from deterioration in that while it includes deterioration and the associated clinical consequences, it also extends to valve thrombosis, paravalvular leak and endocarditis.14 Deterioration, currently, remains the focal outcome, though, with regards to long-term durability of the aortic transcatheter valve and surgical aortic valve.15 Theoretically, certain procedural-related complications, specific to TAVI, may result in early deterioration—e.g. crimping-related structural damage, paravalvular leak post implantation, potentially resulting in turbulent transcatheter valve flow and a predisposition to accelerated deterioration.

Long-term follow-up after bioprosthetic aortic valve replacement primarily focuses on mortality and structural valve deterioration incidence. As previously stated, younger age at the time of implantation has historically been associated with increased mortality. A recent study compared outcomes after mechanical or bioprosthetic aortic and mitral valve replacement, in all eligible patients, between 1996 and 2013.16 Interestingly, in a study population of 9,942 patients, they found that the mortality benefit associated with mechanical aortic valve replacement persisted until 53 years of age, after which there was no was statistically significant mortality benefit with mechanical aortic valve replacement. As seen in previous studies, mechanical aortic valve replacement was associated with a statistically significant lower risk of reoperation, at the expense of an increased risk of bleeding and stroke, when compared to bioprosthetic aortic valve replacement.

Decision making, with respect to valve choice, in younger patients at low-intermediate surgical risk requires a detailed knowledge of the existing literature and available valve technologies. Bagdur et al propose the application of a valve durability (years) to life expectancy (years) ratio when making an informed decision between TAVI and surgery.17 They suggest that we choose a bioprosthesis with a ratio of close to, or greater than, one; i.e. a bioprosthetic valve likely to outlast the longevity of the patient.

Ten-year real world outcomes, will provide further insight regarding the long-term performance of the aortic THV in high risk patients. This long term durability data, alongside ongoing clinical trials, will guide clinicians in the application of TAVI for patients at intermediate and low surgical risk.

In summary, predicted aortic bioprosthetic valve durability and structural valve deterioration, alongside a patient-orientated approach, are fundamental guides in the decision making process for patients with severe symptomatic aortic stenosis.

Paul Brennan and Mark Spence are both at the Department of Cardiology, Royal Victoria Hospital Belfast, Northern Ireland.


  1. Leon et al. N Engl J Med 2010; 363: 1597–1607.
  2. ESC Task Force. European Heart Journal 2017; 00: 1–53.
  3. Leon et al. N Engl J Med 2016; 374: 1609–20.
  4. Reardon et al. N Engl J Med 2017; 376: 1321–31.
  5. Nishimura et al. Circulation 2017; Epub.
  6. Rodriguez-Gabriella et al. J Am Coll Cardiol 2017; 70 (8): 1013–28.
  7. Johnston et al. Ann Thorac Surg 2015; 99: 1239–47.
  8. David et al. Ann Thorac Surg 2010; 90: 775–81.


Please enter your comment!
Please enter your name here