Transcatheter aortic valve replacement for structural degeneration of previously implanted transcatheter valves (TAVR-in-TAVR): a systematic review

Abstract

Open in new tabDownload slide

OBJECTIVES

Transcatheter aortic valve replacement (TAVR) represents a valid treatment for patients with aortic valve stenosis and high or intermediate surgical risk. However, biological transcatheter valves can also experience a structural degeneration after years, and a redo-TAVR procedure (TAVR-in-TAVR) can be a valid option. We revised the current available literature for indications, procedural and technical details and outcome on TAVR-in-TAVR procedures for degenerated TAVR valves.

METHODS

A systematic search was conducted in the public medical database for scientific articles on TAVR-in-TAVR procedures for degenerated transcatheter valves. Data on demographics, indications, first and second transcatheter valve type and size, mortality, complications and follow-up were extracted and analysed.

RESULTS

A total of 13 studies (1 multicentre, 3 case series, 9 case reports) were included in this review, with a total amount of 160 patients treated with TAVR-in-TAVR procedures for transcatheter valve failure. The mean age was 74.8 ± 7.8 with 84 males (52.8%). The mean elapsed time from the first TAVR procedure was 58.1 ± 23.4 months. Main indication for TAVR-in-TAVR was pure stenosis (38.4%, with mean gradient of 44.5 ± 18.5 mmHg), regurgitation (31.4%), mixed stenosis and regurgitation (29.5%) and leaflet thrombosis (8.8%). Procedural success rate was 86.8%, with second TAVR valve malposition occurred in 4 cases (2.5%). The hospital mortality rate was 1.25% (2/160). Post-procedural echocardiographic control showed moderate regurgitation in 5.6% of patients (9/160) and residual transvalvular mean gradient ≥20 mmHg in 5% of cases. Postoperative complications included major vascular complications (8.7%), new pacemaker implantation (8.7%), acute kidney failure (3.7%), stroke (0.6%) and coronary obstruction (0.6%). The mean follow-up time was 6 ± 5.6 months with 1 non-cardiovascular death reported.

CONCLUSIONS

TAVR-in-TAVR represents a valid alternative to standard surgery for the treatment of degenerated transcatheter valves in high-risk patients. Despite these promising results, further studies are required to assess durability and haemodynamic performances of the second TAVR valve.

INTRODUCTION

In the last 10 years, the transcatheter aortic valve replacement (TAVR) has become the preferred treatment option for high-risk patients with symptomatic severe aortic valve stenosis and is a valid alternative to surgical aortic valve replacement (SAVR) in patients at intermediate risk [1, 2]. Based on favourable results of 2 recent randomized controlled trials including low-risk patients undergoing TAVR, the use of this minimally invasive approach is rapidly expanding towards patients below 80 years of age. Thus, even in the absence of concerns about transcatheter heart valve (THV) durability, a substantial proportion of contemporary TAVR patients are expected to live sufficiently long to experience the degeneration and failure of the transcatheter bioprosthesis [3, 4]. Therapeutic options for the treatment of structural THV degeneration are limited to open-heart surgery and redo-TAVR. Concerning standard surgery, patients considered at high surgical risk during the first TAVR procedure remain at high surgical risk also for the second TAVR procedure and therefore other alternative approaches, such as the redo-TAVR, appear to be more suitable. The TAVR-in-TAVR procedures for degenerated transcatheter valves or for valve thrombosis have so far been scarce due to the advanced age and intrinsic frailty of patients treated in the early days of TAVR. However, some reports already exist and in this systematic review, we analyse the available evidence on indications, procedural characteristics, technical details and short-term outcomes of TAVR-in-TAVR for the treatment of structural THV degeneration.

METHODS

This systematic review was performed and reported in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [5]. MEDLINE, PubMed, EMBASE, Cochrane Library, Google Scholar, Science Direct and Web of Science were investigated with time points set to the end of September 2020 using 5 medical subject headings and text words supplemented by scanning the bibliographies of selected articles. ‘Transcatheter treatment’ and ‘degenerated transcatheter valve’ were combined using the Boolean search. Similar search strategies were used regarding the terms ‘TAVI in TAVI’, ‘TAVR in TAVR’, ‘transcatheter valve in transcatheter valve’, ‘treatment after degenerated transcatheter valve’ and ‘valve in valve’. The language was limited to articles written in English. Two co-authors (M.G. and A.G.) reviewed and selected relevant articles for inclusion. Differences were resolved in consensus discussions.

Inclusion criteria

‘TAVR in degenerated transcatheter aortic valve’ was used as the initial retrieval. Only cases receiving TAVR for the treatment of a degenerated or thrombosed transcatheter aortic valve prosthesis were considered and included in the study. Only papers reporting overall morbidity and mortality data, definitive treatment modalities, THV details and configuration of catheter-based techniques were included in the review.

Exclusion criteria

Bailout TAVR-in-TAVR treatments at index procedure due to unsuccessful or suboptimal primary TAVR valve implantation were excluded from this study. Cases in which the TAVR-in-TAVR was performed in a patient with prior valve-in-valve procedure for a degenerated surgical bioprosthesis or aortic homograft were excluded as well. Correspondence, expert opinion, conference presentations, editorials and reviews were excluded. Duplicate articles were excluded, and data were used to achieve the completeness of data extraction.

Data extraction

All data, including baseline characteristics, technical details, as well as safety and efficacy outcomes, were extracted from the selected article texts, tables and figures and included in the analysis. The primary outcome was 30-day mortality. Other outcomes extracted included rates of stroke, myocardial infarction, successful device implantation, transcatheter valve malposition, coronary obstruction, major vascular complications (including access site vascular complications), acute kidney injury, permanent pacemaker implantation, residual aortic valve regurgitation more than moderate, mean TAVR valve mean gradient and postoperative gradient above 20 mmHg.

Statistical analysis

Data collected were organized on an Apple Numbers spreadsheet (Apple, Cupertino, CA, USA) (version 6.6.2). Descriptive statistics were used to describe demographic data, outcomes and TAVR-in-TAVR details. Continuous variables were described as mean ± standard deviation while dichotomous variables were expressed as absolute number with a percentage of the total.

RESULTS

The literature search yielded 1732 publications in MEDLINE, PubMed, EMBASE, Cochrane Library, Google Scholar, Science Direct, and Web of Science. Papers were screened by title/abstract and full text. Seven studies were excluded for duplication and 6 for TAVR-in-TAVR in a previous degenerated surgical bioprosthesis. Finally, 13 publications in English language focusing on TAVR-in-TAVR treatment for degenerated or thrombosed THV were identified and included in the present analysis, covering a period of time ranging from 2012 to 2020 [6–18] (Fig. 1).

Demographic data

A total amount of 160 TAVR-in-TAVR cases from 13 scientific articles (1 multicentre registry, 3 case series, 9 case reports) were included in this review [6–18]. The mean age was 74.8 ± 7.8 years and 52.8% (84/159 reported) of patients were male. Preoperative characteristics are described in Table 1. Dyslipidaemia was present in 65.3% of patients, while 32.2% suffered from diabetes, 79.7% had hypertension and 24.8% had chronic obstructive pulmonary disease. Previous history of surgical myocardial revascularization was present in 22.5% of patients. Congestive heart failure (New York Heart Association class III–IV) at the time of hospitalization was present in 78% of patients (Table 1).

Indications, procedures and devices

The second TAVR procedure was performed 58.1 ± 23.4 months after the first TAVR (data available for 21 cases) and beyond 1 year in 138 patients [6]. The main indication (38.4%) for the TAVR-in-TAVR was a severe stenosis of primary THV, with the mean transaortic peak and the mean gradients of 84 ± 43.8 mmHg and 44.5 ± 18.5 mmHg, respectively (Table 2). Other indications included severe THV regurgitation (31.4%), mixed stenosis and regurgitation (29.5%) and THV thrombosis resistant to the anticoagulation treatment (8.8%). The mean ejection fraction was 41 ± 10.2% and 2 patients had impaired left ventricular function (1 patient was equipped with a ventricular assist device).

Detailed data about valve type and size of the primary TAVR procedure were reported in 48 patients (Table 3 and Supplementary Material, Table S1). At first TAVR, 25 Corevalve (Medtronic, Minneapolis, MN, USA), 20 Sapien (Edwards LifeSciences, Irvine, CA, USA), 2 Jena Valve (Jena Valve Technology, Irvine, CA, USA) and 1 Direct Flow (Direct Flow Medical Inc., Santa Rosa, CA, USA) were implanted. At TAVR-in-TAVR, 27 Sapien, 19 Corevalve, 1 Portico (Abbott Laboratories, Abbott Park, IL, USA) and 1 Allegra (New Valve Technologies, Madrid, Spain) were used.

The combination of deployment mechanisms for the first and second THV is reported in Fig. 2, Table 3 and Supplementary Material, Table S2. The possible combinations include balloon-expandable (BE) in balloon-expandable valve (BE into BE; 20.8%; 10/48), balloon-expandable in self-expanding (SE) valve (BE into SE; 35.4%; 17/48), self-expanding in balloon-expandable valve (SE in BE; 20.8%, 10/48) and self-expanding in self-expanding valve (SE in SE; 23%, 11/48). Compared to the first THV, a larger sized THV was implanted in 25% of cases (12/48), while in 58.3% of patients (28/48), a same sized THV was used, and in 16.7% (8/48) a smaller sized THV was implanted (Supplementary Material, Tables S1 and S3–S6).

Success rate, complications and follow-up

Successful TAVR-in-TAVR was achieved in 86.8% of cases (139/160), while 4 patients (2.5%) experienced THV malpositioning. In-hospital mortality was 1.25% (Table 4). Moderate aortic regurgitation was detected in 5.6% of cases (9/160) and a residual transvalvular mean gradient ≥20 mmHg was present in 5% of patients (18/160). Reported complications included stroke (0.6%), myocardial infarction (0.6%), coronary obstruction (0.6%), major vascular complications (8.7%), acute kidney injury (3.7%) and permanent pacemaker implantation (8.7%). At a mean follow-up time of 6 ± 5.6 months, the mean gradient of the second THV was 13.2 ± 5.9 mmHg. One patient died during the follow-up for non-cardiovascular cause (data limited to 20 patients).

DISCUSSION

In this systematic review, including 160 patients from 13 studies, we showed that TAVR-in-TAVR is a feasible and safe procedure associated with a high success rate and low rates of 30-day mortality, stroke, permanent pacemaker implantation, myocardial infarction and vascular complications [6–18]. With recent trends showing an increase in TAVR procedures worldwide [1, 2], the number of young, low-risk patients requiring reintervention for THV degeneration or thrombosis is also expected to increase in the years to come. In this setting, and following the experience of valve-in-valve procedures in degenerated bioprosthesis, the TAVR-in-TAVR appears to be a valid option in terms of morbidity and mortality as compared to surgical THV explant and subsequent SAVR, especially in high-risk patients [19–21].

Recently, THV structural degeneration was defined as an acquired deterioration of the leaflets or supporting structures resulting in thickening, calcification, tearing or disruption of the THV materials with the possible association of haemodynamic valve dysfunction manifested as stenosis or regurgitation [22]. Based on our data, most device failures were attributed to primary severe stenosis in 38.4% or severe regurgitation in 31.4%. Mixed stenosis and regurgitation (29.5%) and leaflet thrombosis (8.8%) are less common causes of device failure.

Our analysis shows an in-hospital mortality after TAVR-in-TAVR of 1.25% (2/160). This mortality rate favourably compares with reports of SAVR following early THV failure (median time 2.5 months) that shows an overall operative mortality of 17% [23]. Similarly, after TAVR-in-TAVR, there was a lower rate of stroke (0.6%) compared to valve-in-valve procedures after failed surgical bioprosthesis (1.7%) [23].

Despite these promising early results, there are still several concerns regarding TAVR-in-TAVR, such as the haemodynamic and rheological implications of having 2 stent-valves in aortic position, particularly in terms of risk of leaflet thrombosis and long-term THV durability.

The design and model of the TAVR valve have both been shown to be key determinants of flow and haemodynamic characteristics downstream of the valve. Hatoum et al. performed an ex situ analysis of 6 different TAVR-in-TAVR valve combinations (Evolut 23, Evolut 26 mm and 23 mm Sapien 3) in a pulse duplicator set to mimic the physiology of the left-sided heart over 100 cycles [24–27]. The largest effective orifice area (EOA) was obtained with a 23-mm Sapien 3 in a 26-mm Corevalve Evolut (2.07 ± 0.06 cm²) whereas the smallest was with a 23-mm Corevalve Evolut in a 23-mm Sapien 3 (1.50 ± 0.04 cm²). As a general principle, this study revealed that valves with a nitinol frame might provide the best haemodynamic profile after accepting a second transcatheter valve. These theoretical principles, when applied to the in vivo findings, may help explain some of the apparent preference for the Corevalve at the index procedure and the Sapien at TAVR-in-TAVR.

Of note, the choice of the first THV will affect future TAVR-in-TAVR options, as there might be concerns with coronary access impairment after the redo-TAVR procedure. In fact, similar to TAVR for a degenerated SAVR valve, the leaflets of the first prosthesis are tilted up by the implantation of the second THV, thereby creating a cylindrical cage as high as the commissures of the first THV [28, 29]. If this barrier is higher than the coronary ostia and the space between the THV frame and the aortic wall is <2 mm (i.e. the minimum distance to fit a 6-Fr coronary catheter to engage the coronary), coronary access after TAVR-in-TAVR will be impossible and there is a risk of acute coronary obstruction with redo-TAVR. Studies performing coronary angiography and computed tomography after TAVR in consecutive patients have suggested that the risk of coronary access impairment after TAVR-in-TAVR could reach 30–50% [30, 31]. This risk could be even more pronounced in the setting of a bicuspid aortic valve, in which THV implantation is usually higher than in tricuspid anatomy [32]. Notably, patients with a low sinotubular junction and those with a supra-annular degenerated THV (given the greater neo-skirt height) are at a higher risk of coronary obstruction with TAVR-in-TAVR [30, 31, 33]. The implantation of a lower-frame, intra-annular device might be preferable at the time of the first TAVR implantation, as coronary access will be gained in most patients from above the stent-valve both after the index TAVR and after TAVR-in-TAVR [34, 35]. It could be speculated that a strategy of lower THV implantation would increase feasibility of coronary access after TAVR-in-TAVR. However, this would likely happen at the cost of higher rates of permanent pacemaker implantation [36]. Another theoretically helpful procedure to reduce the risk of coronary obstruction of TAVR-in-TAVR might be the Bioprosthetic Aortic Scallop Intentional Laceration to prevent Iatrogenic Coronary Artery obstruction (‘BASILICA’) [37]. However, this depends on the alignment of the THV with the coronary ostia, which is randomly oriented in most of the commercially available TAVR devices [33, 38]. Accordingly, optimal THV orientation should be attempted at the time of first TAVR procedure to avoid a commissural post to be placed in front of the coronary ostium, which would make BASILICA procedure less effective [39]. Moreover early-generation TAVR devices (SAPIEN XT and Lotus) exhibited wider splay angles and slit width after BASILICA in vitro than newer generation TAVR devices. The commissure of the new TAVR device might randomly align unfavourably and obstruct the splayed leaflet despite BASILICA laceration [40].

Overexpansion of a previously implanted TAVR has been promoted as a strategy to accommodate a larger second valve, yielding a larger EOA at the time of TAVR-in-TAVR and improving haemodynamic performance. Ex vivo investigations revealed that these valves, in particular the new Sapien 3 model, could be incrementally over expanded [41].

In this series, the incidence of coronary obstruction was as low as 0.6%. However, we should acknowledge that the reported data refer to a highly selected population of patients who have likely undergone thorough preprocedural screening. Therefore, we cannot exclude that subjects with a high probability of coronary access impairment with TAVR-in-TAVR (either because of a small sinotubular junction or implantation of a supra-annular device at index TAVR) might have been referred for surgical explant of the degenerated TAVR followed by SAVR.

Limitations

Limitations of this systematic review are related to those derived from the original literature, which included 1 multicentre registry, 3 case series and 9 case reports. Limitations of these sources include the relatively small sample size and unavailability of data for all procedural intervention and short-term outcomes.

CONCLUSION

The increasing adoption of TAVR will likely result in an increasing number of patients experiencing THV degenerations. These patients may be treated by TAVR-in-TAVR after properly selection based on Heart-Team decision. Continued study and experience with TAVR-in-TAVR are required to optimize haemodynamic performance and to assess long-term, durability.

IMPACT IN CLINICAL PRACTICE

TAVR-in-TAVR is a growing entity after the first transcatheter valve failure. At the time of the first TAVR implantation, a lower-frame, intra-annular device might be preferable implantation, as coronary access will be gained in most patients from above the valve stent both after the index TAVR and after TAVR-in-TAVR.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Conflict of interest: Enrico Ferrari received honoraria fees for consultancy and educational grants from Edwards LifeSciences, CERC and Somahlution. Ana Paula Tagliari received a Research Grant from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil (Capes)—Finance Code 001. All other authors declared no conflict of interest.

Data availability statement

All relevant data are within the manuscript and its Supporting Information files.

Author contributions

Michele Gallo: Conceptualization; Data curation; Formal analysis; Writing—original draft. Luca Nai Fovino: Data curation; Formal analysis; Writing—review & editing. David Blitzer: Data curation; Formal analysis; Writing—review & editing. Ilias P. Doulamis: Data curation; Formal analysis; Writing—review & editing. Alvise Guariento: Data curation; Formal analysis; Writing—review & editing. Loris Salvador: Supervision; Writing—review & editing. Ana Paula Tagliari: Data curation; Writing—review & editing. Enrico Ferrari: Conceptualization; Data curation; Validation; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Ignacio J Amat-Santos, Mizuki Miura and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

 
• BE

  Balloon-expandable

 
• SAVR

  Surgical aortic valve replacement

 
• SE

  Self-expanding

 
• TAVR

  Transcatheter aortic valve replacement

 
• THV

  Transcatheter heart valve