After the first TTVI was successfully implanted in an animal in 2005 [12], other devices were being developed and had undergone recent clinical studies. One of the first successful dedicated ATVI in a human native tricuspid valve annulus was reported by Jose L Navia in 2017 [13]. Although the prosthesis was first designed as a surgical implant, later modified as a catheter-based prosthesis, yet using a right mini-thoracotomy approach. One out of two cases of NaviGate valved-stent (NaviGate Cardiac Structures, Inc, Lake Forest, CA, USA) of the first-in-man (FIM) series was a Valve-in-Ring (ViR) procedure [13].

A few years later, at the Cardiovascular Research Technologies symposium 2020, Vinayak N. Bapat presented the Intrepid valve (Medtronic, Minneapolis, MN, USA) for severe TR FIM case experience [14]. The Intrepid system used was a surrogate of the Intrepid transcatheter mitral valve implantation system [15]. Similarly, EVOQUE (Edwards Lifesciences, Irvine, CA, USA) valve replacement system for TR is identical to their system for transcatheter mitral valve implantation reported its FIM series in 2021 [16]. Meanwhile, the Food and Drug Administration (FDA), United States designated the Cardiovalve system (Cardiovalve Ltd, Or Yehuda, Israel) as a breakthrough device, and it revealed the results of its preclinical testing [17]. The Lux valve (Ningbo Jenscare Biotechnology Co., Ltd, Ningbo, China) also reported its initial clinical success that year [18]. Trisol Valve (Trisol Medical Ltd. Inc. Yokneam, Israel) issued an FIM report later in 2021 to enable high-risk patients to avoid surgery [19]. The first successful use of the Topaz tricuspid heart valve (TRiCares SAS, Paris, France) was also announced in 2021 [20]. Azeem Latib presented the preclinical data for the VDyne Valve (VDYNE, Inc. Maple Grove, MN, USA) at the Transcatheter Cardiovascular Therapeutics Connect (TCTconnect) meeting in 2020 [21], but the VDyne has not yet reported a human use (Fig. 1, Ref. [20, 21, 22, 23, 24, 25]).

Fig. 1.

Transcatheter tricuspid replacement devices. Annular Tricuspid Valve Implantation (ATVI) devices are (A) Intrepid*, Medtronic Inc. (B) Navigate*, NaviGate Cardiac Systems Ltd. (C) Lux-Valve*, Jenscare Biotech Inc. (D) Topaz#, TriCares Inc. (E) EVOQUE*, Edwards Lifesciences Inc. (F) Trisol*, Trisol Medical; Caval Valve Implantation (CAVI) prosthesis are (G) Tric-SVCǂ, P+F Ltd. (H) Tric-IVCǂ, P+F Ltd. (I) Tricento*, New Valve Tech Ltd.; Non-dedicated devices are (J) Sapien XT*, Edwards Lifesciences Inc. (K) Melody§, MelodyVR, Inc. (L) MyVal€, Meril Life Inc.; Devices in development are (M) Cardiovalve*, Boston Medical Ltd. (N) VDyne¥, VDyne, Inc. *Adopted from Goldberg YH et al., (2021) [24]. #Adopted from Straubinger HJ (2021) [20]. ǂAdopted from Sharma NK et al., (2021) [25]. §Adopted from Riede FT, & Dähnert I (2012) [22]. €Adopted from Lu F et al. (2020) [23]. ¥Adopted from Latib A (2020) [21]. TTVI, transcatheter tricuspid valve implantation.

In a preclinical swine model, Alexander Lauten proposed percutaneous caval transcatheter valve implantation in the superior vena cava and inferior vena cava using a porcine pulmonary valve in 2010 [26]. The FIM application of CAVI was reported by the same group, however, they employed a specially constructed self-expanding valve [27]. It has now been included in the use of Sapien XT/3, an off-label transcatheter aortic valve (Edwards Lifesciences). TricValve (P&F Products Features Vertriebs, Weßling, Germany) was created using Sapien XT/3, and they published their first clinical series in 2018 [28]. A custom-made solution was proposed based on a similar concept, which has reported the successful use of the Tricento (NVT, Hechingen, Germany) valve in humans in the same year [29].

The initial experience with TTVI was restricted to the balloon-expandible Sapien valve Valve-in-Valve (ViV) or ViR procedures (Edwards Lifesciences), done in 2011 and 2014, respectively [30, 31]. The use of a non-dedicated Sapien valve (Edwards Lifesciences) was the first reported ATVI in the native annulus in 2014 [31]. In line with the trend, a pediatric patient had the transcatheter Melody pulmonary valve (MelodyVR, Medtronic, Fridley, MN, USA) implanted as a ViV for TTVI in 2012 [22].

We have included published articles in English, which mentioned the results of experimental clinical studies in humans reporting first and early clinical trials by using TTVI prosthesis for tricuspid valve disease under appropriate clinical indication. “Transcatheter tricuspid Valve Implantation” was subject to ascertain via “topic”, “title” and “abstract” review during the enrolment process. A combination of the search terms and keywords as per the protocol-defined search strategy was implemented for the appropriate inclusion of a study. All the percutaneous and transcatheter tricuspid valve repair devices were excluded from this study. Hence, this review includes only the transcatheter tricuspid replacement devices. Transcatheter heart valves for other cardiac positions used concomitantly were also beyond the scope of this study as they may produce a confounding effect. Preclinical large animal experiment reports with TTVI and other concomitant cardiac procedures and non-clinical in vitro experiments studies were also excluded. Three reviewers screened and assessed the studies independently for inclusion by using the reference software EndNote${}^{\text{TM}}$ X9, (Clarivate, Philadelphia, PA, USA). The articles were first screened by their titles and abstracts. The full-text review was performed on articles for studies that have made it through the first stage, cases where a decision cannot be made, or if the reviewers were unable to confirm the relevance of the study for inclusion. A manual search by using the backward snowballing method was also done once further data verification was required. A detailed search strategy has been recorded in Supplementary Table 1.

GRADEpro Guideline Development Tool (McMaster University and Evidence Prime, Ontario, Canada quality of evidence assessment software) was used to evaluate the included studies as illustrated in chapter 11 of the Cochrane handbook of reviews [34]. The quality of the included manuscripts was further assessed for the risk of bias for inclusion by using ReviewManager 5.4 (Cochrane, England software, London, England) [35] in accordance with the guidelines in chapter 8 of the Cochrane handbook of reviews. In our study risk of bias in ramdomized controlled trial was also assessed according to guidelines from the Cochrane handbook. The risk of bias in nonrandomized observational studies was also assessed according to guidelines from the Cochrane handbook, risk of bias was evaluated using the Risk of Bias in Non-randomised Studies of Interventions (ROBINS-I), (Cochrane, London, England) tool [36].

The included studies were assessed by two authors independently, and details of the manuscripts were abstracted, including title, authors, year of publication, study type, number of patients, sex, age, TTVI device design description, method of anchoring, route, access of device deployment, periprocedural imaging, and early clinical outcome. The primary outcome measures were the procedural success rate and mortality. The secondary outcome measures were the indication of TTVI, all complications, device failure, all-cause mortality, and specific mortality, which is defined as mortality due to underlying cardiovascular causes. Data synthesis was done utilizing the ReviewManager 5.4 (Cochrane, England software, London, England) [35].

All the included studies were early clinical trials, of which most were the FIM clinical trial. Our risk of bias assessment showed five studies are categorized as high risk of performance bias [14, 19, 20, 22, 40] due to the nature of publication, namely a technical case presentation [14]. Press announcement of the technical success [20], and first case report [20, 22, 40]. All other included studies had an unclear risk of bias as inadequate information was available for blinding and randomization for a conclusion to be made. However, the importance and relevance of these articles were independently assessed by three authors, and the evidence provided by the included studies was found critical/important for inclusion (Supplementary Table 2). Characteristics of the included articles have been summarized in Table 1 (Ref. [14, 16, 19, 20, 22, 28, 29, 37, 38, 39, 40]).

All the transcatheter tricuspid prostheses were circular in design yet can be categorized based on the method of implantation (Fig. 1). With the exception of NaviGate (NaviGate Cardiac Structures Inc., Lake Forest, CA, USA), where a transatrial approach was necessary, and a transjugular approach for the same device was abandoned due to procedural complications [37], ATVI is the most prevalent, nearly invariably employing transfemoral access [14, 16, 19, 20, 37, 38]. Although the valve itself is circular, VDyne Valve (VDyne, Inc. Maple Grove, MN, USA) has a varying outer nitinol frame with a “Pop-off” to address afterload mismatch [21]. CAVI devices are also circular, and the caval mounting stent is spanning across the right atrium from superior to inferior vena cava. CAVI devices are classified depending on the location of the valve mounted, but all the available devices are delivered via femoral access. In addition, the Tricento (New Valve Technology, Hechingen, Germany) was a custom-made device [29]. Non-dedicated TTVI devices are made for and in use for other transcatheter therapies (i.e., Melody valve for pulmonary intervention [22]) and also come in a circular shape and are mostly used for ViV or ViR procedures (Table 2, Ref. [14, 16, 19, 20, 22, 28, 29, 37, 38, 39, 40]).

The anatomical landmarks of the tricuspid valve and right heart are commonly assessed in the periprocedural period. TTVI devices are implanted to treat TR, needing the designs to accommodate a non-calcified construct that is both dynamic and D-shaped in one plane and saddle-shaped overall. On top of proper anchoring, TTVI devices need to conform to the native tricuspid annulus to apply the proper sealing required to prevent leakage through the interface of the valve stent and the native annulus, also as known as paravalvular leakage. In our study, we found a variety of different anchoring techniques have been proposed: using tethers to achieve counteracting axial forces (i.e., EVOQUE system (Edwards Lifesciences, Irvine, CA, USA) [16]); native leaflet grasping to fixate the prosthesis in place (i.e., CardioValve (Boston Medical, Shrewsbury, MA, USA) [17]); docking systems to allow radial forces sufficient enough for fixation (i.e., Trisol Valve (Trisol Medical Ltd. Inc. Yokneam, Israel) [19]). However, most CAVI devices are kept in situ by radial force to fix them into the mounting stent [28, 29] (Table 2).

Only 2 of the 11 TTVI systems analyzed in this review, the Edwards Sapien XT/3 (Edwards Lifesciences, Irvine, CA, USA) [39] and the Edwards EVOQUE system [16], were recommended for use in clinical trials. One valve system, Melody (MelodyVR, Medtronic, Fridley, MN, USA) [22], was used off-label, while the other nine were indicated for compassionate use in FIM case reports and series. All patients were high-risk cases with moderate to severe TR that was deemed to be of high surgical risk. The baseline characteristics of TTVI recipients are summarized in Supplementary Table 3. The mean age of patients was 76.1 years old, with one case study included a 12-year-old pediatric patient. 68.8% of the recipients were women, and the mean weight was 74.4 kg. Most patients (88.1%) were New York Heart Association (NYHA) class III or IV. The mean calculated European System for Cardiac Operative Risk Evaluation (EuroSCORE) II was 9.86%, ranging from 5.6% to 18.2%. A history of cardiac pathology, including atrial fibrillation (89.1%), and past valvular interventions (38.1%), was not uncommon. Other comorbidities such as chronic kidney disease (65.5%), diabetes mellitus (30.6%), and prior cerebrovascular events (17.1%) were also present in the studied populations. The average pulmonary artery systolic pressure was 39.2 mmHg (24.5–74 mmHg), and the mean left ventricular ejection fraction was 54.2% (51–65%). 9.4% of the patients had a primary TR diagnosis, 11.8% had a mixed pathology, and 76.4% of the patients had secondary TR. Transfemoral access for TTVI was used in 72.2% of patients, and the typical implantation time was 85.8 minutes (9.1–210 minutes). The majority of patients had their valves installed successfully, with a reported procedural success rate of 93.2%. Patients spent an average of 6.6 days in the hospital, which was the median length of stay (1–13.5 days). The procedural outcome has been summarized in Table 3 (Ref. [14, 16, 19, 20, 22, 28, 29, 37, 38, 39, 40]).

Results from the procedure and the post-procedure have been compiled in Supplementary Table 4. Early TTVI-related complications reported include bleeding (25.2%), major access site and vascular complications requiring intervention (5.8%), device migration or embolization (3.6%), and paravalvular leak (38%). 3 of 28 bleeding cases and 2 of 4 device migration cases occurred in the CAVI group. 8 (11.6%) patients required conversion to open surgery, of which six were from the CAVI group. Of these 6 cases, reported indications included valve migration (n = 2), valve-dislocations (n = 2), and cardiac tamponade from stent migration (n = 2). Cardiovascular mortality and all-cause mortality post-procedure was 2.3% and 11.2%, respectively. Notably, out of the 14 deaths post-procedure, nine occurred in the CAVI group.

Follow-up results are available for 7 out of 11 of the TTVI systems. Follow-up duration varied greatly among the studies, with a median of 2 months and ranging from 2 weeks to 12 months. 28 out of 62 (45.2%) all-cause mortality were reported from post-procedure till to follow-up. Parameters like mean left ventricular ejection fraction improved to 57.2% (55–70%), and mean pulmonary artery systolic pressure reduced to 33.9 mmHg (32.2–37.0 mmHg). There was an overall improvement in the NYHA class of patients. Of the four studies that reported NYHA class at follow-up, EVOQUE (78.8%) (Edwards Lifesciences, Irvine, CA, USA) [16], GATE (72.0%) (NaviGate Cardiac Structures Inc., Lake Forest, CA, USA) [37], and TricValve (52.7%) (P+F Products + Features, Vienna, Austria) [28] showed that a majority of their patients were NYHA class I or II at follow-up, as compared to class III/IV preoperatively. The remaining study, Edwards Sapien XT (Edwards Lifesciences, Irvine, CA, USA) [39], reported that 63% of their patients improved by 1 NYHA class.

Despite the advantages of a systematic investigation, our current review has a number of drawbacks. First off, due to their observational character, the FIM studies we included in our systematic review have built-in biases like selection bias. Additionally, some centers might have had financial constraints or concerns that influenced their choice of intervention. The lack of randomized controlled trials comparing ATVI and CAVI procedures in the literature was another obstacle. In order to provide more robust evidence for the best treatment plan for TTVI, further research, and analysis must be done with a larger patient cohort.