Transcatheter Treatment of Aortic Valve Disease Clinical and Technical Aspects

Abstract

Degenerative aortic valve disease is the most common heart valve disease in western countries. After the onset of symptoms, the prognosis of aortic stenosis is poor, despite optimal medical therapy. In recent years transcatheter aortic valve implantation has been affirmed as a viable treatment for patients with high to low surgical risk. Patient screening and procedural planning are crucial for minimizing complications and achieving procedural success. In the last decade, we have seen a progressive technological development in the percutaneous approach, allowing for expanding indications even in low-risk populations. Here we report a brief review summarizing patient screening and procedural planning in patients with aortic valve disease undergoing a transcatheter approach.

1. Introduction

Alongside the progressive aging of the general population, aortic valve disease is currently one of the most common heart valve diseases worldwide, and its management is going to have a central role in public health, with an expected doubling of the cases in the next 50 years [1, 2]. This topic is particularly significant in developed countries, as in the United States, 4.2 to 5.6 million (approximately 2.5% of the population) are estimated to have a clinically relevant form of heart valve disease in which aortic valve diseases account for 35% of cases [3]. Notwithstanding worldwide primary etiology of aortic stenosis is rheumatic fever. In western countries, regular access to antibiotic therapy and the aging of the population made calcific aortic valve disease, its most common cause, giving account to the rise of aortic stenosis as the most significant heart valve disease in the elderly population [4].

Transcatheter aortic valve implantation (TAVI) has progressively emerged as a valid alternative and choice since 16 April 2002, the date of the world’s first TAVI, performed by Alain Cribier in Lyon (France). In the European Society of Cardiology (ESC) 2012 Guidelines TAVI was limited to surgical high/prohibitive risk patients [5], mainly based on the results of the Placement of Aortic Transcatheter Valves I (PARTNER I; NCT005308944) trial. This trial investigated patients with severe aortic stenosis who were not suitable candidates for surgery; TAVI, as compared with standard therapy, significantly reduced the rates of death from any cause (71.8% vs. 93.6%, Hazard Ratio 0.50; 95% CI 0.39–0·65; p < 0.0001), cardiovascular death (57.3% vs. 85.9%; p < 0.0001), repeat hospitalizations (47.6% vs. 87.3%; p < 0.0001), and cardiac symptoms in terms of New York Heart Association (NYHA) class improvement (NYHA III-IV 14.3% vs. 40%; p = 0.531) [6]. In 2016 PARTNER II trial (NCT01314313), comparing SAVR and TAVI in a randomized trial considering intermediate-risk, patients concluded that TAVI had a similar rate of the primary endpoint (death and disabling stroke) at 2 years follow-up in the overall cohort (Hazard Ratio 0.87; 95% CI 0.71–1.07; p = 0.18) and lower in the transfemoral-access cohort (Hazard Ratio 0.78; 95% CI 0.61–0.99; p = 0.04). On the other hand, surgery demonstrated fewer major vascular complications and less paravalvular aortic regurgitation compared to transcatheter approach [7]. In 2019 results from PARTNER III trial (NCT02675114), comparing SAVR and TAVI in surgical low-risk, patients confirmed that the rate of the composite of death, stroke, or rehospitalization at 1 year was significantly lower with TAVI than with conventional surgery (8.5% vs. 15.1%) [8]. In parallel to PARTNER trials, based on a balloon-expandable prosthesis (Edwards Sapien valve), surgical replacement and transcatheter aortic valve implantation (SURTAVI - intermediate-risk; NCT01586910) and Evolut (low-risk; NCT02701283) trials were conducted, comparing standard surgical therapy to transcatheter implantation of a self-expandable aortic prosthesis (Medtronic Evolut valve). In both studies, TAVI was not inferior to surgery in reducing the primary endpoint of death from any cause or disabling stroke at 24 months [9, 10]. In 2019, based on the evidence generated by these clinical trials, the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA) approved TAVI for the treatment of symptomatic severe aortic stenosis in surgical low-risk patients (Figure 1) [11, 12].

Figure 1.

Overview of the most important clinical trials, stratified according to surgical risk scores.

2021 ESC/EACTS Guidelines (Figure 2) for the management of valvular heart disease have a more balanced approach, currently recommending TAVI for elderly (≥ 75 anni, STS-PROM/EuroSCORE II >8%) or patients unsuitable for surgery and SAVR for younger patients who are low risk for surgery (<75 years and STS-PROM/EuroSCORE II <4%) or in patients who are operable and unsuitable for transfemoral TAVI, leaving gray-zone context to the comprehensive evaluation of individual clinical, anatomical, and procedural factors by the Heart Team, which discussion is however recommended in every scenario [13].

Figure 2.

Management of patients with aortic stenosis [13].

2. Patients selection

Aortic stenosis is frequently associated with advanced age and numerous cardiovascular non-cardiovascular diseases. Because of that, the treatment choice is based on a careful and 360° patient evaluation.

2.1 General screening: symptoms and prognostic impact

A typical benign course characterizes aortic valve stenosis during most of its natural history. At the same time, a drastic prognostic worsening occurs after symptoms onset, with an event-free survival of only 30–50% at two years and with an average survival of just 2–3 years without aortic valve replacement [14, 15, 16, 17]. For that reason, looking for even vague symptoms and closer follow-up have a central role during the medical visit. Aortic stenosis typically manifests itself with effort angina, dyspnea, progressively evolving to congestive heart failure, pre-syncopal, and syncopal events. However, symptoms may be atypical, like fatigue or tiredness, especially in the elderly who, for concomitant reasons, are not able to perform relevant efforts. Usually, in western countries, the onset of the symptoms occurs between 7th and 9th decade of life as a consequence of progressive calcification of valvular cusps [18]. In elderly/complex patients, a critical effort should be to recognize the most likely cause of symptoms, especially in mild or moderate aortic stenosis, as symptoms normally occur in severe stenosis. Moreover, aortic stenosis shares the same risk factors and symptoms as other cardiac and noncardiac diseases. Dyspnea can be present in asthma, chronic obstructive pulmonary disease (COPD), anemia, renal failure, deconditioning, and coronary artery disease (CAD), which could also be manifest with angina and arrhythmias-related presyncope or syncope. In particular, CAD in aortic valve stenosis patients is highly-prevalent; it was found in 69.7% of patients addressed to TAVR in the PARTNER II trial and in 69.2% of patients assigned to SAVR in the SURTAVI trial [7, 9]. Coronary angiography is recommended in assessing each patient with severe aortic stenosis to identify patients that could benefit from contemporary coronary revascularization [13].

In general, aortic valve stenosis progression is constant, with an average annual reduction in the valvular aortic area of 0.03 ± 0.01 cm2/year and about 2.7 ± 0.1 mmHg in the mean transaortic pressure gradient [19]. To improve proper follow-up and identify the most suitable time to proceed to aortic valve replacement, In 2020, American Heart Association (AHA)/American College of Cardiology (ACC) guidelines classify patients into 4 stages according to the natural history phase of aortic valve stenosis: from those at risk of development aortic stenosis (Stage A), to progressive aortic stenosis with mild or moderate calcifications (Stage B), to asymptomatic severe aortic stenosis with normal or reduced left ventricular ejection fraction (LVEF) (Stage C), and to symptomatic aortic stenosis with normal or reduced LVEF (Stage D). This classification is useful in the management of patients because each stage is associated with a proper diagnostic-therapeutic iter; in particular, aortic valve replacement is recommended in all Stage D patients and in Stage C with reduced LVEF (< 50%) [20]. In fact, despite improving symptoms in the short term, medical therapy is not capable of changing the natural history of severe aortic stenosis; therefore, aortic valve replacement is the only effective therapy.

2.2 Risk stratification

According to 2021 ESC/EACTS guidelines for the management of valvular heart disease, aortic valve replacement is recommended for every symptomatic severe aortic stenosis (IB) and asymptomatic severe aortic stenosis with systolic left ventricular dysfunction (LVEF <50% IB; < 55% IIa B) without another cause, undergoing coronary artery bypass graft (CABG) or surgical intervention on the ascending aorta or another heart valve, demonstrable symptoms or sustained fall in blood pressure (> 20 mmHg) on exercise testing (IIa B-C), and/or procedural low-risk plus a risk parameter (very severe aortic stenosis, severe valve calcification and peak aortic valve velocity progression ≥0.3 m/sec/year, markedly elevated brain natriuretic peptide levels) [13].

Once indication to valve replacement is defined, the choice between surgical and transcatheter intervention lies on age, surgical hazard, previous cardiac surgery, a concomitant cardiac condition requiring intervention, technical parameters, comorbidities, and frailty. These parameters should be evaluated by a multidisciplinary heart team, whose role is predominant, especially in moderate-risk patients, in which cases guidelines provide less indications [13]. Aspects favoring SAVR are younger age (typically <75 years), low surgical risk, no previous thoracic surgery, coronary or heart valve disease requiring intervention, and nonrelevant comorbidities, while older age (≥ 75 years), high surgical risk, previous thoracic surgery, and comorbidities favor TAVR. The scores that are commonly used in the definition of the surgical risk are Society of Thoracic Surgeons Mortality (STS) score and EuroSCORE II. Although, these scores were born and developed for stratifying risk in patients undergoing cardiac surgery and not for those who are scheduled for transcatheter therapy. Moreover, they provide just only low correlation with 30-day mortality [21]. In a recent multicenter study performed on patients assigned to TAVI, STS score and EuroSCORE II demonstrated just a moderate correlation and a low accuracy for inhospital adverse events and for 30-day and medium-term mortality, pointing out the necessity of dedicated scores [22].

Technical aspects will be discussed in a separate section (see Anatomical assessment).

2.3 Futility

Transcatheter aortic valve implantation was developed to improve prognosis and has revolutionized the treatment of elderly patients affected by severe aortic stenosis. The expansion in indication and the spread among centers determined the increase of its demand. Consequently, adequate patients selection has become fundamental to avoid wasted resources. Currently, TAVI represents a highly expensive intervention and a relevant issue in a health system where economic resources are limited. However, cost/efficacy analysis had demonstrated a non-inferiority of TAVI respective to SAVR in the long run; in particular, Cohen et al. [23] demonstrated how, despite a higher procedural cost, TAVI allows significantly reduced follow-up costs, compared to SAVR. According to the 2017 American College of Cardiology (ACC) consensus, avoiding intervention on patients who are not going to benefit in survival or quality of life is appropriate. In particular, futility is defined for patients with a life expectancy inferior to 1 year and for those with expected survival with benefit of <25% at 2 years, as evaluated with NYHA class and/or Canadian Cardiovascular Society (CCS) angina grade improvement [24].

In the recent frailty in older adults undergoing aortic valve replacement (FRAILTY-AVR; NCT01845207) study, 646 TAVI patients have been stratified with several frailty scores. The one that had the major correlation with prognosis was the essential frailty toolset (ETF) score. This score is composed of 4 items: mobility (assessed by the time necessary to get up from a chair), cognitive function, hemoglobin value, and serum albumin value. It is interesting to notice that the highest (i.e. the worst) score (5) was associated with a mortality rate of 63% and a major disability rate of 16%. The persistence of a high ETF score value after interventions focused on its reduction could be a futility marker in this kind of patient [25].

In addition to the futility issue, the TAVI’s outcome still has several possibilities of improvement; after implantation, there is a 30-days mortality rate of 7.8% and 2.2%, with old and new devices, respectively [26]. Moreover, considering 5 years of follow-up derived from major trials, the mortality rates exponentially increase. In the Core Valve US pivotal extreme and High-Risk trial (NCT01240902), a prospective, multicenter, and single-arm clinical trial of TAVI enrolling 639 patients with severe aortic stenosis at extreme surgical risk, with a mean age of 82.8 ± 8.4 years, a 5-year mortality rate of 71.6% was observed (with futility of 50.8%) [27]. The same behavior was confirmed in the PARTNER I and II trials, enrolling, respectively, inoperable and surgical intermediate-risk patients randomly assigned to TAVI or SAVR, reporting a 5-year mortality of 71.8% vs. 93.6% (p < 0.0001; PARTNER I) and of 47.9% and 43.4% (p = 0.21; PARTNER II) [6, 7]. It is interesting to notice that a rapid increase in mortality is observed after about 30 days from intervention [28]. Several studies have demonstrated that in patients undergoing TAVI, after an early phase of high cardiovascular risk, this drastically reduces but, in elderly, non-valvular heart failure and noncardiac diseases represent the main causes of death. According to Chen et al. metanalysis, main etiologies are: infections/sepsis (14%), cancer (7%), renal insufficiency (4%), multi-organ failure (3%), and other causes (23%) [29]. This scenario is easily understandable considering that TAVI patients are generally older and have more comorbidities, compared to patients addressed to conventional surgery, and despite valvular disease correction, advanced age and comorbidities still represent a heavy frailty burden [30, 31]. Although, age is not automatically a synonymous of frailty, the latter has demonstrated to significantly affect the outcome of patients undergoing TAVI even in the 90-years-old population [32].

Several factors are associated with increased morbidity and mortality at 1 year after aortic valve replacement among cardiac conditions, atrial fibrillation (AF), left ventricular systolic dysfunction, mitral regurgitation, pre-capillary pulmonary hypertension, and right ventricular dysfunction. Among extracardiac conditions, COPD and restrictive lung diseases, chronic kidney disease, cancer, advanced age, and frailty are the most impacting from a prognostic point of view [33]. In this context, a fundamental question concerns if aortic valve replacement may improve or resolve symptoms and associated conditions affecting prognosis. Indeed, left ventricular systolic dysfunction, when other potential causes are excluded, improves in about two-thirds of the patients from 48 hours to 1-year post aortic valve replacement [34], and mitral regurgitation, especially if functional, may be positively affected after TAVI [35]. Conversely, COPD (with poor exercise tolerance, oxygen-dependency or use of noninvasive ventilation), precapillary pulmonary hypertension (especially with systolic pulmonary artery pressure > 60 mmHg), primary mitral valve regurgitation, active cancer, and cognitive impairment are unlikely to get better after aortic valve replacement and are thus associated with a worse prognosis [36, 37, 38].

Considering the multidimensional phenotype and the discordance among the various tests and scores used in clinical practice, quantifying the impact of frailty could be challenging. Its assessment is however essential in patients’ selection in order to improve extracardiac diseases and avoid vain invasive procedures.

2.3.1 Balloon aortic valvuloplasty (BAV)

Widely used in high surgical risk patients since it was first introduced by Cribier et al. in 1985, BAV is progressively gaining significance in patients’ stratification, clinical stabilization, and forecasting the results of a definitive correction of the valve disease, as a “bridge to decision (medical therapy/TAVI/SAVR)” therapy. In particular, the evaluation of BAV’s results provides prognostic information and is capable of identifying patients who are going to take advantage of aortic valve disease correction [39, 40]. In addition, performing BAV could improve mobility and general status of frailty patients, helping them to bear intensive rehabilitation courses that could be fundamental to face up to aortic valve intervention with the lowest frailty degree. In the end, in frailty patients, in which a judgment of futility has been made (especially for poor life expectancy), BAV may be used as a temporary palliative treatment, as a “destination therapy” [24]. A major limitation of BAV has always been the risk of vascular complications, as the most widespread vascular access site is the femoral venous and arterial access. This site is associated with a rate of major and minor vascular complications of, respectively, 2.7% and 6.6%, even with the use of advanced hemostasis systems (Angio-Seal and ProGlide) [41]. In recent years, there was an important effort in researching techniques to minimize periprocedural complications and, in this context, the Safety and Feasibility of Transradial Mini-invasive Balloon Aortic Valvuloplasty (SOFTLY; NCT03087552) study showed the feasibility and safety of a mini-invasive approach combining radial artery access and LV pacing through the wire (without implantation of a temporary pacemaker through venous access) [42]. The possibility of a mini-invasive approach able to significantly reduce access-related complications could be a great incentive for the use of BAV in order to improve frailty situations before an aortic valve disease definitive correction is performed.

3. Anatomical assessment

3.1 Echocardiography

Echocardiography is fundamental to diagnosis and to assess aortic stenosis severity, valve calcifications, LV systolic and diastolic function, and other cardiac pathologies. Current ESC guidelines underline the importance of echocardiographic evaluation when blood pressure is well controlled to reduce confounding flow effects of increased afterload [13].

Aortic stenosis severity assessment lies on the measurement of mean pressure transvalvular gradient, peak transvalvular velocity (Vmax), and aortic valve area (AVA). Based on these parameters, three categories of severe aortic stenosis may be identified and could benefit from aortic valve replacement [13]:

• High-gradient AS: characterized by mean gradient ≥40 mmHg, Vmax ≥4 m/s, AVA ≤ 1 cm2 (or AVAi ≤0.6 cm2/m2);

• “Classical” low-flow, low-gradient AS (LF-LG AS): characterized by mean gradient <40 mmHg, AVA ≤ 1 cm2 (or AVAi ≤0.6 cm2/m2), LVEF <50%, and as an additional variable, indexed stroke volume (SVi) ≤ 35 ml/m2; in these cases, dobutamine stress echocardiogram is recommended to identify true classical LF-LG AS, which presents increasing mean pressure (≥ 40 mmHg) and could benefit from AVR, form pseudo-severe AS. In particular, patients with true classical LF-LG AS have developed functional improvement one year after TAVI, but no significant LV function improvement [43].

• “Paradoxical” low-flow, low-gradient AS: characterized by mean gradient <40 mmHg, AVA ≤ 1 cm2 (or AVAi ≤0.6 cm2/m2), LVEF ≥50% and SVi ≤ 35 ml/m2. This condition is typical of patients with profound concentric LV hypertrophy with small cavities that are not able to generate enough SV to effectively open the aortic valve [44]. In this context, a computerized tomography (CT) assessment of valve calcification’s degree helps to define the probability of true severe AS (highly likely with Agatston units >3000 for men and > 1600 for women).

In peculiar cases, especially with patients with poor echocardiographic transthoracic windows, transesophageal echocardiography could be a valid alternative (Figure 3).

Figure 3.

Integrated assessment of patients with aortic valve stenosis [13].

3.2 Cardiac catheterization

Despite the evaluation of aortic valve stenosis is mainly based on echocardiography, there is a not negligible discrepancy between effective aortic valve area (AVA) derived from Doppler and from cardiac catheterization. According to Minners et al., there are inconsistencies in grading aortic valve stenosis in patients with normal LV function, in particular with respect to AVA, while mean pressure gradient seems to be a more robust parameter [45]. In a prospective study on assessment of aortic stenosis severity between echocardiography and cardiac catheterization, AVA correlated poorly between the two techniques, with an average AVA difference of 0.25 cm2 (range 0–1.59) [46]. That is due to the fact transvalvular pressure gradient is maximal at the level of the vena contracta, the point in a fluid stream where the diameter of the stream is the least and fluid velocity is at its maximum, which occurs where all the layers of the stream converge, slightly downstream of anatomic aortic valve area. After the vena contracta, part of the jet kinetic energy is recovered in pressure but, during this process, there is some energetic dispersion as a result of flow separation and vortex formation. Echocardiography, measuring transvalvular pressure gradient at the vena contracta (where it is maximal), tends to overestimate pressure gradient and, therefore, underestimate aortic orifice area. Cardiac catheterization, instead, tends to measure a lower transvalvular pressure gradient because it samples it at some distance downstream to vena contracta, where conversely catheter would have trouble maintaining the position of the pressure sensor due to the instabilities secondary to flow-jet turbulence [46]. As assessed by Garcia et al., effective orifice area calculated by catheterism (EOA_cath) may therefore be larger than the one calculated by echocardiography (EOA_echo). This overestimation becomes relevant as the ascending aorta diameter decreases, mostly when sino-tubular junction diameter is ≤30 mm [47]. Moreover, echocardiography could also overestimate EOA because of poor alignment of the ultrasound beam with the stenotic jet [48]. In the end, cardiac catheterization provides data about pulmonary pressures and resistances that, if elevated, could identify an advanced pathology grade that may not benefit from valve correction [37]. Nevertheless, current ESC/EACTS Guidelines for the management of valvular heart disease recommend LV catheterization only when there is a severe aortic stenosis clinic and noninvasive assessment is inconclusive [13]. Criteria for defining aortic valve stenosis severity and its prognosis are derived from catheter measurements, and nowadays the invasive assessment could be a valid ally in an accurate definition of aortic stenosis severity, although a proper selection is mandatory to limit unavoidable complications related to its invasiveness.

3.3 Computerized tomography scan

Electrocardiogram-gated CT scan has a central role in the pre-procedural planning for TAVI. First of all, it is fundamental to evaluate annular valvular area and perimeter (essential to guide the choice of prosthesis’ size), extent and distribution of calcifications, aortic root anatomy, and height of coronary ostia from aortic annulus and LV outflow tract dimension (Figure 4). All this information is pivotal to define prosthesis implantation. For example, an overestimation of the aortic annulus dimensions poses a significant risk for aortic root lesions or disruption during prosthesis release. On the other hand, underestimation increases the risk of paravalvular aortic regurgitation [49, 50]. Considering that aortic annulus dimensions vary throughout the cardiac cycle, they should be measured during systole, i.e., when they are larger.

Figure 4.

Computed tomography evaluation for TAVI procedural planning. Aortic annulus measure (A) and calcium distribution (B). Coronary distance from virtual basal ring (C, D). Aorta and peripheral artery evaluation for a transfemoral access (E-G).

Another main scope of CT scan concerns the planning of vascular access through imaging of aorta and iliofemoral vasculature. This assessment has become increasingly important and has led to a significant decrease of pre- and post-procedural major and minor vascular complications in TAVR patients [51].

4. Device

Transcatheter therapies for the treatment of aortic stenosis have seen a fast and progressive development in technology. Many platforms are nowadays available; it is possible to categorize the devices according to the deployment mechanism: balloon-expandable valves (BEV) and self-expandable valves (SEV). The third category of devices, mechanically expandable, is less widespread. They also differ in the leaflets’ position and their relationship with the annular plane (Figure 5).

Figure 5.

Principal TAVI platforms and technical characteristics.

4.1 Balloon-expandable valves

The SAPIEN platform (Edwards Lifesciences, Irvine, USA) is one of the most diffuse BEVs. It is an intra-annular device, with bovine pericardial leaflets mounted on a cobalt-chromium balloon-expandable frame. SAPIEN valves have a flexible delivery system that allows adapting the implantation in angulated aorta; the balloon expansion allows volumetric modification according to annular sizes, although it is not recapturable during the implantation. The fourth-generation SAPIEN 3 Ultra features an increased outer seal cuff to reduce paravalvular leak (PVL). There are 4 currently available sizes: 20, 23, 26, and 29 mm. The SAPIEN family valves have a lower stent frame profile, which makes easier the coronary catheterization after TAVI [52].

4.2 Self-expanding valves

The Evolut PRO+ (Medtronic, Minneapolis, USA) is the last generation prosthesis of the Evolut family of SEV. They have a supra-annular design and consists of three porcine pericardial leaflets attached to a self-expanding nitinol stent. The stent is a diamond-shaped cell and the valve has an hourglass shape, with a larger circumference at the proximal and distal anchoring points. The delivery system allows the device to recapture after partial deployment and repositioning. Four valve sizes are available (23, 26, 29 and 34 mm) [52].

The ACURATE Neo 2 valve (Boston Scientific, MA, USA) is a SEV with supra-annular design and porcine pericardial leaflets. Its design includes stabilizing arches to facilitate correct positioning. Its top-down deployment, with or without the need for ventricular pacing, does not allow any recapture. It has less radial force, so pre-dilatation is mandatory. The open-cell design and the short-stent body should ease coronary access after implantation. Furthermore, it has a superior crown designed to keep the native cusps away from the coronary ostia [53].

The Portico valve (Abbott Vascular, Santa Clara, CA, USA) comprises a bioprosthetic bovine pericardial aortic valve mounted upon a self-expandable nonflared nitinol frame. The leaflets are located at the annular level, ensuring valve function immediately upon deployment.

Allegra (Biosensors International, Morges, Switzerland) and HYDRA (SMT, Wakhariawadi, India) are two self-expanding nitinol frame valves with bovine pericardial leaflets. Their use is limited to high surgical risk patients and the evidence of safety and efficacy are quietly poor.

4.3 Device choice

So far, there is insufficient evidence to claim the superiority of a prosthesis or another. Each TAVI device has a unique design, and certain elements may slightly favor one or another prosthesis. Among the factors to consider when choosing a valve for TAVI, those that may favor BEV are short or narrow sinus of Valsalva, the presence of conduction disturbances (right bundle branch block or 1st degree AV block), and the anticipated need for future coronary re-access and a horizontal aorta. In small annuli and in case of severe LV outflow tract calcification, SEV may be preferred [52, 54].

The intra-annular design is associated with higher trans-prosthetic gradients and more frequent patient-prosthesis mismatch (an effective orifice area too small in comparison to patient’s body surface area) [55]. Patient-prosthesis mismatch is associated with a worse prognosis in surgical prosthesis; however, the clinical relevance of TAVI remains uncertain [56].

Only few randomized trials directly compared different TAVI devices. Direct comparisons are difficult because the small number of events makes necessary the use of composite endpoints. Furthermore, data from early generations TAVI devices cannot be automatically extrapolated to current-generation prosthesis. The Comparison of Transcatheter Heart Valves in High-Risk Patients With Severe Aortic Stenosis (CHOICE; NCT01645202) trial, which randomized high-risk patients to receive a BEV (Sapien XT) or a SEV (Core Valve), showed a greater rate of device success with early generation BEV. The greater device success of BEV in comparison to SEV (95.9% vs. 77.5%; relative risk, 1.24; 95% CI, 1.12–1-37; p < 0.001) was driven by a significantly lower frequency of significant aortic regurgitation and less frequent need for the implant of a second valve. Placement of a new permanent pacemaker was less frequent in the BEV group (17.3% vs. 37.6%, P = 0.001). A randomized trial compared the SAPIEN 3 valve with the ACURATE Neo valve (Safety and efficacy of the Symetis ACURATE Neo/TF Compared to the Edwards SAPIEN 3 Bioprosthesis - SCOPE 1; NCT03011346) [57]. The non-inferiority of the ACURATE Neo was not met in a composite endpoint. In the SCOPE 2 trial (NCT03192813), the ACURATE Neo valve did not meet the non-inferiority criteria in comparison with Core Valve Evolut prosthesis. Nevertheless, the ACURATE neo showed a significative reduction in permanent pacemaker implantation (PPI) (10.5% vs. 18%) [58]. Data from observational analysis seem to favor BEV [59, 60], but should be interpreted with caution.

5. Minimizing complications

5.1 Paravalvular leaks

Paravalvular leak (PVL) consists of a residual gap between the native calcified aortic valve, aortic annulus, and the prosthesis. PVL can be identified during the TAVI procedure using invasive hemodynamics and cine-angiography, while echocardiography is the most diffusely used technique to detect, grade, and follow PVL [61].

The hemodynamic effects of a significant residual regurgitation have a negative clinical impact. Moderate to severe PVL are independent predictors of short-term and long-term mortality, while the impact of mild PVL is unclear [62].

Calcification of the aortic valve, leaflet asymmetry, prosthesis malposition and under-sizing, and the use of SEV is associated with the development of PVL [63]. SEV is more influenced by the calcium burden, as they exert less radial force than BEV, therefore they are more often under-expended or eccentrically shaped. On the other side, the higher radial force exerted by BEV could lead to annular rupture [63, 64].

The first-generation devices had a 30-day incidence of moderate to severe PVL of 9.0% and 11.8% (respectively for SEV and BEV) in high-risk patients [65, 66]. Newer-generation devices were designed with features aimed to reduce PVL, such as external skirt of the SAPIEN 3 Valve and the external sealing system of the Evolut PRO Valve. In the more recent PARTNER III trial, the SAPIEN 3 valve had a decreased incidence of moderate to severe PVL in low-risk patients (0.6%) and of note that was similar to residual PVL of SAVR (0.5%) [8]. In the Evolut Low-Risk Trial, there was a greater incidence of moderate to severe PVL: 3,5% in TAVI vs. 0,5% in SAVR [10]. This discrepancy is consistent with the different designs of prosthesis.

A significant PVL may benefit from several treatment options, which includes balloon post-dilatation of the prosthesis, percutaneous closure with plugs, and TAVI-in-TAVI to exert a superior radial force against the PVL, and surgical intervention.

5.2 Coronary obstruction and coronary re-access

Almost half of the patients undergoing TAVI have coronary artery disease, and about a third of the patients are in a low-risk population [8, 67]. TAVI may influence coronary in two ways: the prosthetic valve struts may prevent the selective catheterization of coronaries during PCI and the prosthesis or the dislodged native leaflets may cause acute coronary obstruction.

The coronary re-access following TAVI is influenced by several anatomical factors (sino-tubular junction dimensions, sinus height, leaflet length and bulkiness, sinus of Valsalva width, and coronary ostial height) and device-related and procedural factors (commissural tab orientation, sealing skirt height, and valve implantation depth) [68]. Prosthesis with higher frame design hinders coronary re-access more than those with a lower frame due to the barrier of the stent frame in allowing coronary catheters to directly engage the coronary ostia (Figure 6). Therefore, selective coronary angiography after TAVI with some SEV could be more challenging than with BEV [68].

Figure 6.

Difference between TAVI device profile in coronary re-engagement. Device with low profile could theoretically guarantee a easier coronary cannulation.

Otherwise, some SEVs, such as ACURATE NEO (Boston Scientific, MA, USA), are characterized by lower stent frame, which could allow for easy coronary engagement.

The alignment of the TAVI valve commissures with the native aortic valve commissures is a promising modifiable factor to facilitate coronary re-access. TAVI differs from aortic valve replacement in the fact that the orientation of commissural posts relative to the coronary ostia is random. It has been shown that specific orientations of the Evolut and ACURATE neo at initial deployment could improve commissural alignment [69]. Of note, a commissural alignment is particularly helpful in high-frame SEV in avoiding coronary artery overlap; this may be fundamental in coronary artery access and redo TAVR.

Acute or delayed coronary obstruction after TAVI is a rare but life-threatening complication, with an incidence inferior to 1% [70, 71]. Coronary obstruction is usually caused by the displacement of the calcified native valve leaflet over the coronary ostium or by the direct occlusion of the coronary ostium by the covered skirt of the transcatheter aortic prosthesis. Anatomical factors associated with coronary obstruction are low coronary ostia height and shallow sinuses of Valsalva. Procedural-related elements include BEV and valve-in-valve (VIV) for surgical bioprosthesis [70]. To prevent this complication some coronary protection techniques may be used, such as preventive coronary wiring or positioning of an undeployed stent in high-risk patients. If the coronary blood flow is compromised during or after TAVI release, the stent is retracted and deployed to create a channel for coronary perfusion between the displaced leaflets and the aortic wall (chimney technique) [72, 73].

5.3 Pacemaker implantation

High-grade atrioventricular block requiring permanent pacemaker implantation (PPI) is one of the most common complications following TAVI, with an incidence ranging from 2 to 36%, depending on the patient population in exam and the prosthesis design [74]. Notably, the rate of PPI remains high even in recent trials with newer generation devices compared with previous trials [74, 75]. SEV is associated with higher risk of PPI than BEV, probably because of the increased radial force exerted on the left ventricle outflow tract (Figure 7). In the Core Valve High-Risk trial, PPI was significantly more frequent in the TAVI group than in the SAVR group (19.8% vs. 7.1%, p < 0.001) [65]. A more frequent occurrence of PPI in TAVI patients was also observed in the Evolut Low-Risk trial [10]. In the PARTNER III trial, the rate of PPI-associated TAVI was similar to that of surgical patients (6.6% vs. 4.1%, hazard ratio 1.65; 95% CI, 0.92 to 2.95), although the onset of a new left bundle block was more common after TAVI (22.0% vs. 8.0%; hazard ratio 3.17; 95% CI 2.13 to 4.72) [8].

Figure 7.

Relationship between transcatheter heart valve and conduction system.

The link between the occurrence of conduction disturbances and the TAVI procedure is explained by the proximity between the aortic valve and the structures of the cardiac conduction system. The atrioventricular node is situated in the right atrium, continues as the Bundle of His, and then splits into the left and the right bundle branches. The Bundle of His emerges at the level of the interventricular membranous septum, caudally to the commissure between the right and noncoronary cusp. The course of the Bundle of His may be within the right half of the membranous sept, within the left half, or under the endocardium; conduction disorders during TAVI are lower with the first anatomic variant [76, 77]. During TAVI, the conduction system can be injured by the insertion of guidewires, balloon pre-dilation, and valve deployment.

The conduction disturbances after TAVI range from new-onset complete atrioventricular blockade to left bundle branch block and transient complete atrioventricular block. The presence of baseline right bundle branch block (RBBB) is the strongest predictor of need for PPI. Other predictors for PPI after TAVI are PR-interval prolongation, left anterior hemiblock, older age, presence of left ventricle outflow tract calcifications, severe mitral annular calcification, and the length of the membranous septum. Procedural predictors are the use of SEV, deeper valve implantation, balloon pre- and post-dilation, and prosthesis oversizing (Figure 8) [74, 78, 79, 80].

Figure 8.

Major factors associated with permanent pacemaker implantation after TAVI.

As per standard of care, PPI is recommended when the patient develops a persistent complete or high-grade atrioventricular block after TAVI. It is also recommended in case of new-onset alternating bundle branch block, while it may be considered in patients with pre-existing right bundle branch block who develop new post-procedure conduction disturbance. There is not yet consensus about the optimal strategy for patients with other conduction abnormalities [78].

PPI after TAVI has been associated with increased mortality and rehospitalization, as the need for RV pacing may lead to decreased LV function and heart failure, yet there is still conflicting evidence [78, 81]. Risk factors that should be assessed in the preoperative TAVI evaluation are preexisting conduction disturbances and LVOT calcification. There may be a trade-off between the reduction of PVL and the risk of PPI, as a greater radial force may reduce the regurgitation, but it may damage the conduction system [52]. A BEV may be preferred in patients with baseline conduction disorders. A higher implantation strategy may minimize the contact between the valve and membranous septum, reducing conduction defects after the implantation [82]. In this context, an angiographic view providing an accurate visualization of the implantation depth (the cusp overlap view, as the right coronary cusp and the noncoronary cusp appear overlapping) demonstrates to reduce the rate of PPI [83].

6. Particular cases

6.1 Pure aortic regurgitation

Moderate to severe aortic regurgitation has a prevalence of 0.5%. The course of chronic aortic regurgitation leads to left ventricular dilation and heart failure. Primary aortic regurgitation may be caused by infective endocarditis, rheumatic disease, or degenerative/calcific valve disease. Bicuspid aortic valve, while more commonly associated with stenosis, may cause pure aortic regurgitation or a mixed disease. Aortic regurgitation may also be secondary to marked dilation of the ascending aorta [84].

The gold standard treatment is surgery, with both aortic valve replacement or aortic valve-sparing root replacement. Currently, the role of TAVI is limited to selected patients with aortic regurgitation deemed ineligible for SAVR [13, 85].

The commercially available TAVI devices have been designed for the treatment of degenerative calcific aortic stenosis. The presence of a rigid frame of calcium in the annulus provides an anchoring point for device deployment. The lack of calcium poses thus a significant challenge, as there is increased risk of device malposition, dislodgment, and embolization. The lack of calcification may also lead to higher rates of PVL. Another issue is the risk of implanting undersized devices, as regurgitant aortic valves are more elastic than calcific stenotic valves and can expand to a greater degree during valve deployment. Furthermore, the concomitant presence of a certain degree of aortic disease with dilation and friable tissues poses a further degree of risk for the procedure [85, 86].

Registry data show that TAVI in pure aortic regurgitation has worse outcome than TAVI in aortic stenosis. A 331 patients registry showed a 3% rate of procedure-related death, a 3.6% conversion to open surgery, a 1.2% rate of coronary obstruction, a 1.5% of aortic root injury, and a 16.6% need for second valve implantation. Newer generations valves scored better, as device success went from 61.3% to 81.1% (p < 0.001) and moderate to severe aortic regurgitation decreased from 18.8 to 4.2% (p < 0.001). [85] In another registry, also including patients with failing bioprosthesis, device success was achieved in 85% of patients with new-generation devices [87].

New prosthesis specifically designed for aortic regurgitation are currently being investigated, such as the Trilogy Heart Valve (Trilogy; Jena Valve Technology), which features anchor rings to clasp the native aortic leaflets [88]. While TAVI may be an alternative for selected patients deemed at high risk for surgical aortic valve replacement, it is currently an off-label indication; randomized control trials and long-term data are still needed.

6.2 Bicuspid aortic valve stenosis

The bicuspid aortic valve is the most common congenital heart defect, with an incidence of around 1% [89]. Almost half of the patients undergoing isolated aortic valve replacement have a bicuspid aortic valve, with a higher incidence in younger patients [90]. In the contemporary practice, up to 10% of patients with bicuspid aortic valve stenosis are referred to TAVI [91].

Echocardiography often underestimates the prevalence of bicuspid valves in calcified aortic stenosis [91]. CT scan provides a more accurate diagnosis and visualization of the bicuspid morphology [92]. Bicuspid aortic valve encompasses a wide range of morphologies; the most common classification categorizes it according to the number of raphes [93].

Aortic annuli in patients with bicuspid valve tends to be larger than in patients with a tricuspid valve. The annulus size may be outside of the range for the currently available devices. Furthermore, the aortic valve complex may have a non-tubular geometry, such as tapered or funnel anatomy. This adds complexity to the selection of a compatible prosthesis [94].

Bicuspid valves have a higher calcific burden than tricuspid stenotic valves. The calcium involves the leaflets in an asymmetrical way and often extends to the LV outflow tract. The majority of the bicuspid valves have a fibrotic and calcified raphe. These anatomic elements hinder the optimal expansion of the valve during TAVI. The asymmetric expansion of the prosthesis increases the risk of PVL. The presence of a highly calcified raphe, if localized between right coronary cusp and non-coronary cusp, increases the risk of conduction disturbances. Calcified raphe and excess leaflet calcification have been found to predict all-cause mortality in TAVI, and when both were present patients had higher rates of aortic root injury and PVL [94, 95].

In addition, coronary anomalies are more frequent in patients with bicuspid aortic valve, and 20 to 30% of them have concomitant aortic disease [89, 94]. Many patients may need aortic root surgery in addition to the valve replacement.

Data about the outcome of TAVI in bicuspid valve anatomy are limited to observational studies, as it was an exclusion criterion in all the randomized trials confronting TAVI with SAVR. In patients at increased surgical risk included in the STS/ACC transcatheter valve therapy registry (STS/ACC TVT Registry; NCT01737528), TAVI for bicuspid aortic valve stenosis showed acceptable safety outcomes with low complications rates [96]. When current-generation devices were used, device success was higher (96.3 vs. 93.5; P = 0.001) and the incidence of moderate to severe PVL was lower (2.7% vs. 14.0%; P < 0.001) in comparison with older-generation devices. With current-generation devices, device success was slightly lower in the bicuspid valve group (96.3% vs. 97.4%; P = 0.07) in comparison with tricuspid stenosis, with a slightly higher incidence of residual moderate or severe PVL. A comparable 1-year mortality was observed, with no increase in the risk of stroke [97]. Results of TAVI in low-risk patients with bicuspid valve anatomy seem similar to those patients with tricuspid aortic valve. In the PARTNER 3 bicuspid registry, with a population of 169 patients, a propensity score matching with TAVI patients showed no difference in the primary endpoint and in the individual components (death, strokes, cardiovascular rehospitalization). Of note, almost half of the patients submitted (47%) were not treated, being excluded because of anatomic or clinical criteria [98]. Another small prospective trial showed good short-term outcomes in low-risk patients with bicuspid aortic valve [99]. Despite the good outcomes in selected patients with favorable anatomies those results cannot be inferred for all the patients with bicuspid aortic valve.

Technical recommendations for TAVI include more frequent balloon valvuloplasty and post-dilation, a low degree of oversizing, and the use of repositioning prosthesis. In tapered anatomies, a supra-annular positioning of the prosthesis has been suggested [94].

7. Conclusions

In recent years the treatment of severe aortic stenosis has been deeply transformed by the introduction of the transcatheter approach. We have reported an overview of the more relevant clinical and technical aspects of the TAVI procedure. As the indications extend to younger patients and with lower surgical risk, it is even more crucial to optimize the results and reduce the complication rate. Further improvements in both technologies and techniques are needed before expanding indications in aortic stenosis in bicuspid valve and in aortic regurgitation.