Standardized definitions of structural deterioration and valve failure in assessing long-term durability of transcatheter and surgical aortic bioprosthetic valves: a consensus statement from 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)

Despite continuing efforts during the last decades, there is no ‘ideal prosthetic valve substitute’. Every valve prosthesis invokes new pathophysiological processes, including the risks of thromboembolism, prosthetic endocarditis, and structural valve deterioration (SVD) or non-structural valve deterioration with consequent need for reintervention (Figure 1). Bioprostheses are now increasingly used in preference to mechanical valves in the aortic position but valve dysfunction may occur over time. The literature concerning surgical prostheses has taught us that bioprosthetic valve dysfunction is a complex phenomenon whose understanding requires more than the reporting of reintervention. Further research must encompass biological, pathological and haemodynamic mechanisms, use of contemporary non-invasive imaging, evaluation of the true incidence while avoiding methodological pitfalls, and identification of clinical, technical, and prosthesis-specific predictors. Causes of bioprosthetic valve dysfunction. Since introduction in 2002 and broader clinical use in 2007, penetration of transcatheter aortic valve implantation (TAVI) has grown exponentially as a result of accruing evidence demonstrating safety and efficacy, and reduced invasiveness compared with surgery. TAVI is now the recommended therapy in elderly patients with aortic stenosis who are inoperable or at increased surgical risk1 and recent evidence has demonstrated at least its equivalence to surgery in intermediate and high-risk cohorts.2–4 However, our knowledge concerning the clinical outcomes of TAVI beyond 5 years is still limited. Although SVD is likely to be the main mechanism of bioprosthetic valve dysfunction in the longer term, definitions of SVD vary and follow-up studies are scarce. While it is possible to draw lessons from longer term experience with surgical bioprostheses, there are fundamental differences between TAVI and surgical aortic valve replacement (SAVR) (i.e. remaining valve calcification, mechanical stress, crimping of the valve tissue, valve leaflet geometry, balloon expansion or dilation, differences in haemodynamic profile, and patient-prosthesis mismatch), which may impact on the natural history of SVD (see Supplementary material online, Appendix). Critically, extended knowledge of the durability of TAVI is essential as we enter the time (>5 years after implantation) when SVD starts to occur in surgical bioprostheses. This knowledge assumes even greater importance as we consider expanding the indications for TAVI to lower risk and younger patients. As such, standardizing the definitions of valve- and patient-oriented durability outcomes is of paramount importance to enable objective evaluation of existing and novel TAVI prostheses, and their comparative efficacy vs. SAVR. In this context, the European Association of Percutaneous Cardiovascular Intervention (EAPCI) determined that improved characterization of long-term TAVI outcomes was timely. Two face-to-face meetings (September 2016, London; January 2017, Frankfurt) involving members of the EAPCI, the European Society of Cardiology (ESC), and the European Association for Cardio-Thoracic Surgery (EACTS) representing interventional cardiology, clinical cardiology, imaging and surgery, provided much of the discussion to inform the present document. Herein, we present the available evidence on TAVI SVD, addressed in terms of existing definitions, predictors, and detection. In parallel, we present a standardized definition of SVD and a new patient-oriented clinical end point named bioprosthetic valve failure (BVF) for use in future studies, which aims to capture the clinically relevant manifestations and consequences of SVD or other forms of bioprosthetic valve dysfunction. This effort precedes a registry initiated within the ESC European Observational Registries Programme (EORP) which will evaluate the incidence, presentation, mode, and timing of bioprosthetic valve dysfunction in a contemporary real-world setting. The ultimate goals of this multidisciplinary collaboration are to improve the characterization of SVD and BVF in line with similar ongoing efforts by the Valve Academic Research Consortium (VARC) 3 and optimize the future utilization of TAVI. Survival without valve reintervention or explant for SVD is an outcome still used by some published series to assess the durability of surgical bioprostheses.5 However, surgical guidelines for event reporting after cardiac valve interventions have not supported this approach since 2008, and stipulate that SVD should also be defined by clinically detectable measures other than the need for reoperation for a failing bioprosthesis (i.e. using echocardiographic criteria).6 In 2009, Zoghbi et al . 7 published a series of recommendations for the evaluation of prosthetic valves using echocardiography and Doppler ultrasound. Possible stenosis was defined as peak prosthetic aortic jet velocity 3–4 m/s, mean gradient 20–35 mmHg, and effective orifice area 0.8–1.2 cm2. Significant stenosis was defined as peak prosthetic aortic jet velocity >4 m/s, mean gradient >35 mmHg, and effective orifice area <0.8 cm2. The 2012 ESC guidelines, written in collaboration with the EACTS, recommend annual echocardiography beyond the first 5 years following bioprosthetic valve implantation (and earlier in young patients) to detect early evidence of ‘SVD, leaflet stiffening, calcification, reduced effective orifice area, and/or regurgitation’.1 Based on these guidelines, the transprosthetic gradients should be interpreted in comparison with the baseline values. This requires an early postoperative assessment to set up a reference point for future investigations and to detect important conditions such as patient-prosthesis mismatch and left ventricular dysfunction. Reoperation is recommended in symptomatic patients with a significant increase in transprosthetic gradient or severe regurgitation (Class I, Level of Evidence C) and should be considered in asymptomatic patients with significant bioprosthetic valve dysfunction, provided they remain at low-surgical risk (Class IIa, Level of Evidence C). The VARC-2 recommendations also suggest echocardiography as the principal imaging modality for assessment of bioprosthetic valve function immediately before initial hospital discharge (to establish baseline parameters) and at 6 months, 1 year, and annually thereafter.8 VARC-2 defines SVD as (i) valve-related dysfunction (mean aortic gradient ≥20 mmHg, effective orifice area ≤0.9–1.1 cm2, and/or dimensionless valve index <0.35, and/or moderate or severe prosthetic valve regurgitation) or (ii) need for a repeat procedure (TAVI or SAVR). Lancellotti et al . 9 suggested incorporating an increase in mean gradient during stress echocardiography or at follow-up (possible obstruction 10–19 mmHg; significant obstruction ≥20 mmHg). In a recent surgical series, Bourguignon et al . 10 defined SVD using strict echocardiographic criteria independent of symptomatic status, including severe aortic stenosis (mean transvalvular gradient >40 mmHg) and severe aortic regurgitation (effective regurgitant orifice area >0.30 cm2, vena contracta >0.6 cm). Of note, this definition relies on the systematic implementation, recording and reporting of echocardiographic data at pre-defined follow-up intervals, which make data interpretation problematic if these conditions are not observed.10 The clinical course of patients with bioprosthetic valves should be monitored periodically, with the interval between routine follow-up visits determined according to cardiac status, comorbidities, and other clinical factors. Various imaging techniques are available for detection of bioprosthetic valve dysfunction. These include 2D/3D echocardiography, multi-detector computed tomography (MDCT) and magnetic resonance imaging (MRI).11–13 Echocardiography is a ‘functional’ imaging modality and superior for the demonstration of valve haemodynamics (i.e. increased transvalvular gradient, valve regurgitation), whereas MDCT provides more ‘anatomical’ and structural information. MRI has the potential to combine anatomical and functional information but is not always readily available and experience in the assessment of bioprosthetic valve dysfunction is limited. These considerations have implications for the application of different imaging modalities in the assessment of bioprosthetic valve durability. Periodic echocardiographic surveillance is currently the reference standard for detection of SVD in cases unidentified at reoperation or autopsy. Stenosis or regurgitation of the bioprosthetic valve should be reported using validated quantitative or semi-quantitative methods.9 The term deterioration implies changes intrinsic to the valve (including wear, fracture, calcification, leaflet tear, and/or disruption of any component). Transoesophageal imaging can improve visualization of morphological aspects of the valve prosthesis and the additional role of 3D echocardiography in this setting is yet to be defined. Multi-detector computed tomography may be more sensitive than echocardiography in detecting valve thrombosis, particularly at early stages of the process (i.e. subclinical leaflet thrombosis without haemodynamic consequences).14 , 15 Multi-detector computed tomography criteria for TAVI thrombosis include hypo-attenuated leaflet thickening (with or without reduced leaflet motion of one or more leaflets, identifiable in two or more multiplanar curved reconstructions).11 Specific MDCT measurements include stent frame expansion and eccentricity index, number of leaflets with hypo-attenuated leaflet thickening, as well as degree of leaflet thickening, motion reduction and calcification.11 Importantly, MDCT cannot determine aortic valve gradients and is therefore of diminished utility for the diagnosis of SVD. Several large series have reported the long-term outcomes of SAVR bioprostheses with mixed results (Table 1). Importantly, the age of patients undergoing SAVR in these studies was on average lower than that of patients included in TAVI series, which makes cross-study comparisons inappropriate on the ground of long-term durability. As noted above, some of the surgical series evaluate durability in terms of survival or survival without reintervention; others expand the definition of SVD with criteria of haemodynamic progression. In a large series evaluating 2405 Carpentier-Edwards bioprostheses, survival without reintervention was 98 ± 0.2%, 96 ± 1%, and 67 ± 4% at 5, 10, and 20 years, respectively.18 Bourguignon et al . 10 evaluated 2758 Carpentier-Edwards bioprostheses using clinical and echocardiographic criteria, and reported SVD in 157 patients (123 of whom required reintervention) over a cumulative follow-up of 18 404 valve-years. All cases of SVD were late events and actuarial freedom from SVD at 15 and 20 years was 78.6 ± 2.2% and 48.5 ± 4.6%, respectively. In the Johnstone et al . 5 series assessing SVD in 12 569 patients (81 706 patient-years), actuarial of explant for SVD at 10 and 20 years were and bioprostheses have also demonstrated long-term durability in patients years or while an of SVD was with the prosthesis in of of patients with a bioprosthesis reported freedom from reoperation in and at 10 and 15 years, and freedom from reoperation for SVD in and outcomes vary with different surgical bioprostheses as demonstrated in recent surveillance of valves in and durability after surgical aortic valve replacement ± and 7 ± at 20 and years, from ± 4% at 20 years from ± 4% at 20 years and at 10 and 15 years, from and at 10 and 15 years, from reoperation for and at 10 and 15 years, ± ± ± and ± at 5, 10, and 20 years, from 98 ± 0.2%, 96 ± 1%, and 67 ± 4% at 5, 10, and 20 years, at 20 years from at 5 years ± at 20 years ± 4% at 20 years from ± and ± at 15 and 20 years, from explant to ± and ± at 15 and 20 years, of explant to and at 10 and 20 years, ± and 7 ± at 20 and years, from ± 4% at 20 years from ± 4% at 20 years and at 10 and 15 years, from and at 10 and 15 years, from reoperation for and at 10 and 15 years, ± ± ± and ± at 5, 10, and 20 years, from 98 ± 0.2%, 96 ± 1%, and 67 ± 4% at 5, 10, and 20 years, at 20 years from at 5 years ± at 20 years ± 4% at 20 years from ± and ± at 15 and 20 years, from explant to ± and ± at 15 and 20 years, of explant to and at 10 and 20 years, SVD, structural valve as any in function from any valve or as of aortic transprosthetic gradient with a effective orifice area or aortic regurgitation as severe aortic stenosis (mean transvalvular gradient >40 mmHg) or severe aortic regurgitation (effective regurgitant orifice area >0.30 cm2, vena contracta >0.6 cm). durability after surgical aortic valve replacement ± and 7 ± at 20 and years, from ± 4% at 20 years from ± 4% at 20 years and at 10 and 15 years, from and at 10 and 15 years, from reoperation for and at 10 and 15 years, ± ± ± and ± at 5, 10, and 20 years, from 98 ± 0.2%, 96 ± 1%, and 67 ± 4% at 5, 10, and 20 years, at 20 years from at 5 years ± at 20 years ± 4% at 20 years from ± and ± at 15 and 20 years, from explant to ± and ± at 15 and 20 years, of explant to and at 10 and 20 years, ± and 7 ± at 20 and years, from ± 4% at 20 years from ± 4% at 20 years and at 10 and 15 years, from and at 10 and 15 years, from reoperation for and at 10 and 15 years, ± ± ± and ± at 5, 10, and 20 years, from 98 ± 0.2%, 96 ± 1%, and 67 ± 4% at 5, 10, and 20 years, at 20 years from at 5 years ± at 20 years ± 4% at 20 years from ± and ± at 15 and 20 years, from explant to ± and ± at 15 and 20 years, of explant to and at 10 and 20 years, SVD, structural valve as any in function from any valve or as of aortic transprosthetic gradient with a effective orifice area or aortic regurgitation as severe aortic stenosis (mean transvalvular gradient >40 mmHg) or severe aortic regurgitation (effective regurgitant orifice area >0.30 cm2, vena contracta >0.6 cm). aortic valve implantation has available since and used in elderly in whom data concerning long-term durability are (Table annual echocardiography in the TAVI using a prosthesis with SAVR in high-risk demonstrated transvalvular gradient and aortic valve area up to 5 years patients at risk at follow-up data of a using a TAVI prosthesis vs. SAVR are available up to 3 years, more valve haemodynamics for TAVI without differences in In this severe patient-prosthesis mismatch was more in patients with SAVR than with and with durability after transcatheter aortic valve implantation at 5 years at 5 years to at 5 years at 5 years valve at 5 years at 5 years at 5 years to at 5 years at 5 years valve at 5 years SVD, structural valve durability after transcatheter aortic valve implantation at 5 years at 5 years to at 5 years at 5 years valve at 5 years at 5 years at 5 years to at 5 years at 5 years valve at 5 years SVD, structural valve The and Registries demonstrated valve gradients over 5 years and of SVD of and , However, it should be noted that patients were still at 5 years their age and significant at the time of valve implantation) and that definitions of SVD were not two series currently data on durability (>5 in patients before et et at Valve In the series SVD was defined as mean transvalvular gradient ≥20 an increase over time and/or aortic regurgitation that was not present following valve this 1 SVD TAVI for and 3 asymptomatic patients SVD (mean gradient and increase in comparison with patients a gradient >40 In the series freedom from SVD defined as need for reintervention was while freedom from SVD defined as severe regurgitation or need for reintervention was at patients were years after TAVI with no of SVD in two in In standardized definitions for the of future studies, the on the following should be between SVD principal and BVF clinical valve deterioration dysfunction whereas other pathological of bioprosthetic valve dysfunction (i.e. thrombosis, are and should be and However, the or process as a of BVF if it to or bioprosthetic valve dysfunction. valve dysfunction (i.e. or prosthesis patient-prosthesis late may occur early after TAVI as a result of valve dysfunction in valve-related and haemodynamic dysfunction severe new or aortic as a of Echocardiography is the principal imaging modality for the detection of SVD and the and to detect changes in valve gradients should be determined in at least two measurements to for detection and to the different of bioprosthesis TAVI and echocardiography should be before discharge or within after valve implantation (i.e. baseline at 1 after valve implantation and annually (with additional follow-up and/or of other imaging modalities as and/or determined by the (Figure at echocardiographic haemodynamic multi-detector computed SVD, structural valve valve deterioration intrinsic changes of the valve (i.e. leaflet tear, calcification, or to and/or dysfunction, which in may result in stenosis or regurgitation (Table valve deterioration can be using imaging studies or at the time of reoperation or and can in symptomatic and asymptomatic patients. valve deterioration can be as and/or valve deterioration SVD, structural valve valve deterioration SVD, structural valve The diagnosis is on haemodynamic changes in valve function by of echocardiography, even without evidence of morphological SVD haemodynamic SVD may be in patients with haemodynamic SVD by echocardiography or other imaging the two of haemodynamic SVD and detection of haemodynamic dysfunction of clinical SVD is defined as (i) mean gradient ≥20 and and/or and from baseline discharge or within of valve implantation) and/or (ii) moderate new or aortic haemodynamic SVD is defined as (i) mean gradient and/or ≥20 from baseline discharge or within of valve implantation) and/or (ii) severe new or aortic The diagnosis is on imaging of reintervention is In of the diagnosis of morphological SVD should be and or on the pathological SVD of the following leaflet (i.e. or regurgitation), leaflet (i.e. pathological thickening and/or stenosis or regurgitation), leaflet function (i.e. in stenosis and/or regurgitation), and (i.e. or The term BVF severe SVD (i.e. the with its clinical consequences avoiding of valve-related outcomes in asymptomatic patients with no clinical and is recommended by the as the main outcome of in studies assessing the long-term of TAVI and SAVR (Figure Importantly, BVF may occur in the setting of SVD but also as the of pathophysiological to SVD, such as thrombosis, or non-structural valve dysfunction. BVF any of the (i) bioprosthetic valve dysfunction at likely to the of or defined as any by bioprosthetic valve dysfunction in the of (ii) aortic valve reintervention (i.e. or and severe haemodynamic SVD. Based on the degree of BVF can be as (i.e. severe haemodynamic or (i.e. valve-related and early (i.e. up to or late (i.e. according to the timing of after valve valve failure surgical aortic valve SVD, structural valve transcatheter aortic valve valve failure surgical aortic valve SVD, structural valve transcatheter aortic valve assessment of bioprosthetic valve failure (BVF) in outcome studies of transcatheter aortic valve implantation (TAVI) or surgical aortic valve replacement SVD, structural valve the durability of prostheses important and a number of should BVF be considered a or outcome is the of BVF in an elderly is there a approach to these The following will these and on in survival for important is between valve and Valve outcomes to the intrinsic durability of the bioprosthesis (i.e. they the is the of this valve over time without In patients are more in their of a valve event during their remaining (i.e. is the of valve failing before Importantly, some valve outcomes (including haemodynamic are in which that they with time and not occur at a capture valve outcomes while it is important to consider the timing of or when assessing the or of haemodynamic SVD by of mean gradient using Doppler echocardiography, there is an important if the are while the changes data the other there is a risk of valve outcomes if the are (i.e. if echocardiography is at any time in symptomatic These are when are different with valve outcomes are more time in that they the of an event from the time of implantation to a (i.e. or a risk the risk of a valve to over time. In if the at a time when the valve is there is no to the valve have if the In other if BVF at some time during the end point is In if the with no bioprosthetic valve dysfunction, we cannot be the true durability of the prosthesis the for that valve to at a time This is more likely to occur in an and the is The term to the when the information an end point for a is a may be in a of TAVI durability (i) BVF not occur during the follow-up (ii) the before the end of the follow-up (i.e. or the is to of outcome studies is that can be or is Based on such an the survival experience of a who or is to follow-up may be by (i.e. and the outcome of as of a patients follow-up However, the of is in TAVI durability there is a between the risk of and BVF (i.e. in that (i) patients who before BVF are than who not and (ii) the of BVF is lower in patients. The relevant for a TAVI not to the intrinsic durability of the but to the of a clinical event to bioprosthetic valve dysfunction during the course of the remaining In this of actuarial may to since event an increasingly significant of the for survival from BVF as over the of (Figure may be for in the durability of a valve particularly if for is the of may for the of and a of true valve Importantly, for the reporting of should be (i) the number of patients at risk at time point the (ii) reporting and the survival when than of the initial is In to the actuarial the is the that should be used for clinical and This on a cumulative incidence provides lower than actuarial and have greater clinical utility in the of TAVI durability of the of using vs. actuarial for assessment of bioprosthetic valve failure (BVF) in studies of transcatheter aortic valve implantation (TAVI) durability. of 20 TAVI patients with follow-up 5 patients experience BVF and 15 patients with before any The and actuarial of BVF 15 patients and 5 patients experience with before any BVF (i.e. The actuarial provides a of BVF than the The of a new in the of valve using transcatheter techniques is The clinical of TAVI are increasingly well by and However, in the process of to of younger and lower risk it is important to the long-term durability of and future TAVI this we have and standardized definitions of SVD and BVF and recommendations for the timing and modalities of clinical and imaging follow-up the of these should also be extended to the evaluation of and future surgical bioprostheses, whose long-term efficacy and durability are currently addressed by a of information concerning bioprosthetic valve dysfunction and and their with bioprosthetic valve and techniques for valve implantation will data to new in and implantation and defined imaging will identification of bioprosthetic valve dysfunction to mechanical and While the process in and to that with surgical , recent evidence of valve leaflet thickening and thrombosis requires since (i) it these are of clinical and to and (ii) the for this is yet to be the aims to a large European registry of TAVI patients years by the European who TAVI at the early of this The registry will on two main aspects of data (i) of BVF at follow-up and (ii) of SVD in patients at different time important remaining in knowledge that need to be include the follow-up data beyond 10 years and our to the significant changes in and techniques over these the results of the registry will be in a understanding of results and the for TAVI in younger patients. they will a for the results of TAVI with of Supplementary material is available at European of from and from and and research from and from from and research from from and from and Lancellotti from and and from from and and research from and from and from and from and and research from and from and and research from and from and The and research from from and and research from and The to the of the following and role as of the from and research from and All the other have to

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.