Transcatheter Aortic Valve Replacement in Bicuspid Aortic Valve Stenosis

Flavien Vincent, MD, PhD; Julien Ternacle, MD, PhD*; Tom Denimal, MD; Mylène Shen, MSc; Bjorn Redfors, MD, PhD; Cédric Delhaye, MD; Matheus Simonato, MD; Nicolas Debry, MD; Basile Verdier, MD; Bahira Shahim, MD, PhD; Thibault Pamart, MD; Hugues Spillemaeker, MD; Guillaume Schurtz, MD; François Pontana, MD, PhD; Vinod H. Thourani, MD; Philippe Pibarot, DVM, PhD; Eric Van Belle, MD, PhD


Circulation. 2021;143(10):1043-1061. 

In This Article

Management of TAVR in BAV Patients

Patient Selection

Surgery remains the preferred option for BAV patients, especially in young patients or patients at low risk for surgery. TAVR may be an alternative in elderly patients (≥75 years of age) with a favorable AV complex anatomy but without significant aortic dilatation (<45 mm or <50 mm in high surgical-risk patients). According to the majority of published studies, the referral of BAV patients to TAVR should be considered at intermediate or greater risk for SAVR based on clinical assessments by a multidisciplinary heart team or an STS score >4% (Table 1). When TAVR is considered, a comprehensive CT scan assessment of BAV anatomic features is mandatory to identify suitable patients and to select the THV design and size based on patients' characteristics (Figure 5).

Figure 5.

Key screening steps when considering transcatheter aortic valve replacement in a patient with bicuspid aortic valve.
*<50 mm for high-surgical risk patients; **<683 mm2 according to instructions for use of SAPIEN 3 29 mm and <94.2 mm according to instructions for use of Evolut-R/Pro 34 mm; however, successful implantation with additional balloon volume until 904 mm2 has been reported with SAPIEN-3.106 CT indicates computed tomography; STS, Society of Thoracic Surgeons; and THV, transcatheter heart valve.

Which BAV Anatomy is Favorable for TAVR?

Procedural and clinical outcomes reported in observational studies can be applied only to patients with a BAV anatomy considered favorable by operators, and therefore, to a highly selected population. Data are lacking to clearly identify BAV patients who are the most favorable for TAVR, but we can delineate some markers of procedural complexity and potentially worse outcomes (Table 2). The major challenge is to predict the behavior of the AV complex (particularly the raphe) during and after THV deployment. Sievers type 1 BAV is preponderant and most patients implanted with TAVR had this anatomy. The characteristics of the raphe are critical, and a heavily calcified raphe is generally considered hostile to the proper expansion of the THV. In the study of 561 patients, all cases of annulus rupture (1.6%) occurred in type 1 BAV with a calcified raphe and treated with BE valve.[43] Yoon and Makkar reported the largest CT-based BAV cohort with core laboratory analysis of 1034 patients (89.7% type 1, 10.3% type 0) mainly treated with SAPIEN 3 (71.6%). They identified the presence of calcified raphe and excess leaflet calcifications (defined as more than median calcium volume of 382 mm[3]) as an independent morphological features associated with all-cause mortality.[20] Both calcified raphe plus excess leaflet calcification were found in 26.0% of patients and they had higher rates of procedural complications compared with those with either or none of these features (aortic root injury, 4.5% versus 0.7% versus 0.9%; P<0.001; equal or greater than moderate PVR, 6.5% versus 2.5% versus 1.6%; P=0.002) and higher 2-year mortality (25.7% versus 9.5% versus 5.9%; log-rank P<0.001).

Type 0 Sievers BAV has been associated with a favorable outcome, especially a low rate of PVR, but its prevalence is low in TAVR series (≈10%).[20,93] Type 2 is rare in the general population and could be associated with very poor expansion of the stent frame. Tricommissural BAV with incomplete and noncalcified raphe (ie, incomplete type 1 of Sievers) could be the most favorable to achieve a circular orifice and proper sealing of the THV. Beyond the BAV type, an extreme elliptic shape of the AV complex or a large discrepancy between annulus and sinus morphology may complicate the sizing and THV selection. Even if THV can be safely overexpanded, dimensions of the landing zone should fit with the available THV size[106] (Figure 5). Importantly, the burden and topography of the calcification should be carefully evaluated through the entire AV complex. The presence of bulky, heavy, and asymmetrical calcification patterns extending toward the left ventricle outflow track could be associated with a higher risk of PPI, PVR, and annulus rupture via direct damage of the calcified tissue or via a contrecoup mechanism.[107] Calcified leaflet tips could also be associated with a higher risk of coronary occlusion or embolization. Conversely, a low calcium burden, as observed in BAV-related AR or mixed AV disease, may increase the risk of THV migration and embolization. The presence of abnormal coronary artery implantation could also be considered because THV may further increase the difficulties of coronary access. Overall, new tools are clearly needed to stratify TAVR feasibility in BAV patients.

Role of the Preoperative CT Scan

Beyond conventional measures (annulus size measurements with modification for bicommissural phenotype, sinus of Valsalva width, and height of coronary ostia), a thorough assessment of the number of cusps, presence of a raphe, extension and topography of calcifications, and dimensions of the aortic root is essential, as defined per guidelines.[108] This helps to plan the procedure, anticipate the THV expansion, and avoid potential complications such as PVR or annulus injury[109,110] (Figure 3). Yoon et al illustrated in detail how a CT scan before TAVR can improve the procedural results in BAV patients. A CT scan was performed in 100% of patients treated with the third generation THV (versus 78.9% with the second generation), and its use was associated with a greater procedural success rate (92.2% versus 80.9%; P=0.01) and lower rates of annulus rupture (2.0% versus 1.0%; P=0.01) and equal or greater than moderate PVR (0% versus 8.5%; P=0.002).[95]

THV sizing should be primarily based on the annulus size which has been shown to be a reliable and safe approach in BAV patients.[36,37] For the minority of patients with tapered or funnel anatomy, a supra-annular sizing[38] has been suggested to be of incremental value[11] to select a smaller THV than suggested by annulus dimensions and avoid the oversizing-related risks[39] (Figure 3); however, supra-annular techniques of measurements are not standardized, have inferior reproducibility than annulus measurements,[37] and would result in a beneficial modification of THV size in a minority of cases. Hence, the clinical evidence supporting this approach are scarce.

Overall, BAV sizing should rely on a multilevel (from the annulus to the aortic root) and multiparameter (area, perimeter, supra-annular or intercommissural distance/area, and median/minimum/maximum diameter) integrative approach to determine the optimal THV size and implant position (Figures 3 and 5). A CT scan is also useful to localize and quantify the calcification burden through Agatston or calcium volume scoring.[41,107] New tools such as patient-specific computer simulation have been recently developed to better identify patients who are suitable for TAVR and optimize sizing and procedural planification.[111]

Procedural Characteristics

TAVR procedure may be more challenging in BAV anatomy. The asymmetrical and heavily calcified valve orifice, presence of the raphe, and horizontal aorta may lead to guidewire crossing difficulties and lateralization of the stiff guidewire path through the valve. The risk of coronary obstruction should also be carefully evaluated beforehand. Innovative strategies of leaflet laceration by electrified guidewire have been proposed to prevent coronary obstruction or to create tricuspidization and allow better prosthesis expansion, but further validation studies are required.[112,113] Predilation of the valve is more frequently performed in BAV than in TAV to facilitate the crossing of the delivery system and ensure appropriate expansion of the THV, as well as to assess the response of the supra-annular anatomy.[114] Balloon valvuloplasty should be sized according to the minor axis of the AV complex, and the appearance of a waist may indicate a calcified or resistant raphe and a risk of THV underexpansion or annulus injury, thus influencing the THV type (SE valve preferred) and size selection. In cases of BE valve implantation, a low degree of oversizing is recommended (<10%) with respect to the risk of annulus rupture. In combination with contrast injection, balloon valvuloplasty may also be valuable to judge the leaflets behavior in relation to coronary ostia. In some instances, there may be a higher-than-normal AR after predilation and the implanting team should be facile in moving forward quickly with the TAVR if necessary. In a tapered anatomy, a supra-annular positioning of the THV may be considered to achieve optimal sealing while minimizing the risk of PPI. Recapturable features of SE and ME valves may be advantageous for deploying the THV at the optimal landing zone. Postdilation should also be more frequently considered in BAV than in TAV because raphe and calcifications burden may lead to valve underexpansion and insufficient sealing in the pericommissural zones, particularly with SE valves that have a lower radial force.[115] All of these technical and anatomic parameters might explain the potential higher risk of stroke and of new ischemic lesions[115] a noted in some observational exploratory studies. The benefit of using a cerebral embolic protection device remains unclear in TAVR. However, a recent pooled analysis suggested that the use of dual-filter protection decreases the rate of periprocedural stroke.[116] This benefit may be greater in populations at high risk of stroke such as BAV patients (Figure 5), but there are no data to support this hypothesis.

Which THV Design is Suitable for BAV Anatomy?

No head-to-head randomized comparison has ever been conducted in BAV patients, which precludes inferences being drawn on the superiority of 1 valve over another. The large majority of studies separately reported the results of each THV design, and the few studies reporting post-TAVR outcomes according to the THV design did not observe significant difference in outcomes on hard clinical endpoints such as mortality or stroke.[18,95] In a large STS/ACC registry, Halim et al reported the same device success rate between SE and BE valves (P=0.19) but observed a higher incidence of PVR after SE valve than after BE valve (odds ratio, 2.97, [1.93–5.59]; P<0.001) in the third generation THVs, as previously observed in TAV populations.[52,117] Conversely, a meta-analysis of 758 patients reported a comparable rate of equal or greater than moderate PVR (3.4% versus 3.4%; P=0.80) and postprocedural PPI (13.3% versus 11.7%; P=0.43) between BE (SAPIEN-3) and SE valves (Evolut-R).[118] Waksman et al, in the prospective, single-arm, low-risk trial including BAV patients mostly treated with SAPIEN 3 (74%), reported a higher rate of PPI in patients treated with SE valve (31.3%) than in those treated with BE valve (6.7%),[97] but in a very limited sample size of 61 patients. Yoon et al reported that SAPIEN XT recipients with BAV had more frequent aortic root injury than TAV patients (4.5% versus 0.0%; P=0.015), whereas no root injury was observed with Corevalve recipients.[43] Interestingly, the same team also reported that all cases of annulus rupture in type 1 BAV patients with calcified raphe occurred with BE valve (SAPIEN XT or SAPIEN-3).[95] Similarly, Makkar et al observed a low, but higher, risk of annulus rupture with SAPIEN 3 in BAV patients than in TAV patients (0.3% versus 0.0%; P=0.02) in an STS/ACC TVT registry.95 A meta-analysis reported a similar 30-day and 1-year mortality, rate of stroke, equal or greater than moderate PVR, coronary obstruction, and acute kidney injury in BAV patients treated with a BE versus SE valve, while the use of a SE valve was associated with a higher rate of second device and PPI but a lower rate of annulus rupture.[119] A first multicenter registry comparing the third generation THV design ([SAPIEN-3] n=242 versus [Evolut-R/Pro] n=111) observed similar device success but BE valve recipients experienced higher annular ruptures (1.7% versus 0%; P=0.173) and higher transvalvular gradient (11.5±4.3 9.7±4.9; P=0.026) and had less equal or greater than moderate PVR (0.8% versus 10.8%; P<0.001) than SE valve recipients. Only the PVR rate and mean gradient remained significantly different between the 2 designs after propensity score matching.[115] These findings suggest that a rapid balloon inflation at high pressure against a highly and asymmetrical calcified area could produce a contrecoup injury and annulus damages on the opposite and less-calcified side of the AV complex.

In the most hostile anatomy, such as a calcified raphe, a SE or ME valve could be preferred to mitigate the risk of annulus injury. THV with repositionability features (SE valve or ME valve) allows the operator to test different depths of implantation, which could be advantageous in tapered-shape anatomy. The ME valve had a high radial strength that could help in achieving optimal sealing in heavily calcified BAV patients.[120] A postmarket registry with ME valve on a very small sample size (n=31) reported excellent procedural success and no equal or greater than moderate PVR, but a very high rate of PPI in BAV patients compared with that in TAV patients (37.4% versus 22.2%; P=0.14). Overall, large multicentric prospective cohorts including all commercially available THVs with core laboratory CT and echocardiographic analysis and an independent adjudication committee are eagerly needed to define precisely the BAV anatomies and THV designs that are the most favorable for TAVR.

Shared Decision-making for BAV Treatment in a TAVR Era: A Lifetime Plan

Shared decision-making in AS management is an increasingly complex process given the pros and cons of both AVR modalities. Risk and benefits of these 2 techniques should be explained to the patients as well as the uncertainties of using a treatment not as well evaluated as in TAV anatomy. SAVR is an open-chest intervention but offers the option of a combined surgery (coronary and/or aorta), whereas TAVR can only treat AS but without sternotomy nor cardiopulmonary bypass. Hence, a lifetime plan of care should be discussed with the patient to put in perspective that both techniques are complementary and could be required during the patient's lifespan. The presence and severity of aortopathy will be at the center of the discussion in many situations. Some may speculate that a patient in their mid-60s with favorable BAV anatomy and mild or moderate aortopathy (<45 mm) could benefit from a TAVR followed by a low-risk SAVR combined—or not—with aorta intervention at the time of THV deterioration during their 70s. Then, a TAVR valve-in-valve could be safely performed when the patient is 80 years of age or older. If an aorta surgery is required during the patient's 70s, before THV deterioration, a TAV-in-TAV may be considered at a later time when the THV will be deteriorated.

This strategy could offer 3 AVRs using a bioprosthetic valve for only 1 sternotomy. However, given the many uncertainties on TAVR durability, aortopathy evolution, and surgical outcomes of SAVR after TAVR, it seems too early to suggest such a standardized plan for lifetime management. Overall, heart teams may be more important now than ever before to individualize and optimize the therapeutic strategy according to a patient's risk, age, anatomy, preference, life expectancy, and functional status.