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

Outcomes of TAVR in BAV Patients

Early Experience of TAVR in BAV

The first TAVR series in BAV patients came out in 2010 with the first generation of BE valve (n=11),[92] followed by a series with the second generation of SE and BE valves in 2014 (n=139).[93] In these earlier series, the rate of periprocedural complications was high with 13% to 34% equal or greater than moderate PVR, 13% to 43% of permanent pacemaker implantation (PPI), and a 1-year mortality between 4% and 18% (Table 1). In 2017, Yoon et al compared the outcomes of propensity score–matched cohorts of 561 BAV patients and 4546 TAV patients treated by old and new generations of THV.[43] The 2-year mortality was similar in the 2 groups (17.2% versus 19.4%; P=0.28), but procedural success was lower in the BAV cohorts because of more surgical conversion (2% versus 0.2%; P=0.006), fewer implantation successes (85.3% versus 91.4%; P=0.002), and higher rates of equal or greater than moderate PVR (10.4% versus 6.8%; P=0.04). There was no difference in the rate of PPI (14.7% versus 13.7%; P=0.72).[43] A recent large observational study from the STS/ACC TVT registry including all types and generations of valves compared 5412 BAV patients to 165 547 TAV patients and reported that STS was relatively low at 3.8% (2.3–6.1) and patients were younger in the BAV group than in the TAV group (74.0 years [65.0–81.0] versus 82.0 years [76.0–87.0]; P<0.001). The 1-year unadjusted risk of mortality was slightly lower in BAV patients (hazard ratio [HR], 0.88 [0.78–0.99]).[52] The risk of conversion to open heart surgery was similar in both groups but a slightly lower procedural success (96.0% versus 96.7%; P=0.004) and higher incidence of equal or greater than moderate PVR (4.7% versus 3.5%; P<0.001) were observed in BAV patients. There was no difference in the adjusted 1-year hazard of stroke (HR, 1.14 [95% CI, 0.94–1.39]).

Impact of Refining TAVR Technology

Major improvements of the third generation THV combined a smaller delivery sheath and an outer sealing skirt for SAPIEN-3, or a pericardial outer wrap and recapturable capabilities for Evolut-R/Pro. The systematic use of CT scan for aortic annulus and root sizing also contributed to an increase in procedural success and dramatically decreased the rate of significant PVR (Table 1). Perlman et al first described a cohort of 51 BAV SAPIEN 3 recipients without any equal or greater than moderate PVR.[94] A study comparing 102 BAV patients treated with older generation THV versus 199 BAV patients treated with third generation THV demonstrated a sharp decrease in the rate of equal or greater than moderate PVR (8.5% versus 0%; P=0.002), and an increase in device success implantation rate (80.9% versus 92.2%; P=0.01).[95] While low and decreasing compared with the first valve generation, Yoon et al observed that the rate of conversion to surgery in BAV patients remained numerically higher with the third generation valve (1.3% versus 0.0%; P=0.25) than in TAV patients.[43] In the STS/ACC TVT registry (n=3714; 73.4% SAPIEN-3), the change for the third generation valve was associated with a global decrease in the rate of adverse events in BAV patients while remaining slightly higher than in TAV patients with a higher rate of equal or greater than moderate PVR (2.7% versus 2.1%; P=0.006) and a lower rate of device success (96.3% versus 97.4%; P<0.001).[52] The study by Makkar et al, also based on the STS/ACC TVT registry, compared 2691 BAV patients only treated with SAPIEN 3 with the same number of propensity score–matched TAV patients (Table 1). There was no difference in the 30-day (2.6% versus 2.5%; P=0.82) or in the 1-year mortality (12.9% versus 14.1%; P=0.74). However, a greater risk of stroke (2.5% versus 1.6%; P=0.02), PPI (9.1% versus 7.5%; P=0.03), and conversion to open chest surgery (0.9% versus 0.4%; P=0.03) at 30 days were observed in BAV patients than in TAV patients. At 1 year, the rate of stroke (3.4 versus 3.1%; P=0.16), equal or greater than moderate PVR (3.2% versus 2.5%; P=0.47), and AV reintervention was similar in the 2 groups.[96] Forrest et al reported the outcomes of 929 BAV patients propensity score–matched with 929 TAV patients in the STS/ACC TVT registry study with Evolut-R/Pro. There was no difference between the 2 groups in the 30-day or 1-year mortality (10.4% versus 12.4%; P=0.63), or for device success (96.5% versus 96.4%; P=0.88), conversion to open chest surgery (0.6 versus 0.2%; P=0.29), or PPI rate (15.4% versus 13.7%; P=0.30). The rate of stroke at 1 year was similar in both groups (3.9% versus 4.4%; P=0.63), but the need for AV reintervention at 1 year was greater in BAV patients than in TAV patients (1.7% versus 0.3%; P=0.01). AR was more prevalent in BAV than in TAV (5.6% versus 2.1%; P<0.001) at 30 days but was similar at 1 year. In a multicentric registry of 301 patients, the need for PPI was high but similar between BAV and TAV patients (13.1% versus 16.7%; P=0.40).[95] Recently, Elbadawi et al compared, after propensity score matching, the in-hospital outcomes of TAVR in BAV versus TAV patients (n=1035 in each arm). Mortality (2.9% versus 3.4%; P=0.76) and other procedural-related complications were similar between the 2 groups, including the rate of stroke (1.9% in both groups) and PPI (14% in BAV versus 12.1% in TAV; P=0.51). However, the study had several limitations related to the methodology (data obtained from the National Inpatient Sample database, which is administrative and does not have detailed information about the procedures) and did not provide granularity on THV design and BAV subtypes.[25] Overall, the short-term outcomes improved dramatically with the newest generation of THV and approximated the excellent outcomes achieved in TAV patients.

Toward TAVR for Low-risk BAV Patients

Waksman et al conducted a small (61 patients) multicenter prospective trial in low-risk patients and reported excellent 30-day outcomes (no death and no disabling stroke) in a highly selected BAV population mainly treated with SAPIEN 3 (74%).[97] The rate of PPI was at 13% and that of equal or greater than moderate or equal or greater than mild PVR was at 1.6% and 35.3%, respectively. The Low Risk Bicuspid Study was a prospective single-arm study evaluating Evolut-R/Pro in 150 BAV low-risk patients with neither aortopathy nor prohibitive calcification in the left ventricle outflow track. STS score was 1.4±0.6%, and 90.7% of patients had type 1 BAV (Sievers), while the remaining had type 0. The early results at 30 days showed a high device success rate of 95.3%, 1 case of coronary obstruction, and no annulus rupture. The number of deaths or disabling stroke was low (1 event for each). The rate of PPI was 15.1%, which was high but comparable to the 17.4% rate observed in low-risk TAV patients.[5] No patient had moderate or worse AR, whereas 40.4% had mild AR.[98] Preliminary data (n=75) from another prospective cohort (BIVOLUT-X [Bicuspid Aortic Stenosis with Evolut Platform International Experience]; NCT03495050) with Evolut-R/Pro confirmed promising early results with no equal or greater than moderate PVR and excellent hemodynamic outcomes. However, an unusually high risk of disabling stroke was observed (5.3%) that could be attributable to a high calcium burden, frequent predilation maneuvers, and valve repositioning required during implantation.[34] The PARTNER 3 trial (Placement of Aortic Transcatheter Valves 3) bicuspid registry with SAPIEN 3 that enrolled 75 low-risk patients will be presented this year. We can anticipate that these single-arm registries will confirm the feasibility and safety of the TAVR procedure on low-risk patients for both THV design. However, their limited sample size (75–150 patients) and the enrollment of highly selected patients without bulky eccentric calcification, left ventricle outflow track significant calcifications, or anomalous coronary arteries (selected both by clinical sites investigators and then by eligibility committee of the trials) limit the incremental clinical value of these cohorts. The anatomy of patients not included in these registries would be very informative. Overall, results of TAVR in low-risk BAV patients are encouraging but larger cohorts and long-term data are needed to determine the impact of PPI and mild PVR as it could contribute to a late superiority (beyond 2 years) of SAVR over TAVR similar to the one observed in intermediate-risk TAV patients with second generation valve.[99]

THV Durability in BAV Patients

Data on THV durability in BAV versus TAV are scarce and limited to valve hemodynamic performance. In the BAVARD (Bicuspid Aortic Valve Anatomy and Relationship with Devices) registry, 30-day post-TAVR mean gradient (9.4±4.9 mm Hg versus 10.7±4.9 mm Hg; P=0.15) and effective orifice area (2.1±0.5 cm2 versus 1.9±0.6 cm2; P=0.07) were comparable between BAV and TAV patients. The indexed effective orifice area was smaller (1.17±0.4 cm2/m2 versus 1.33±0.37 cm2/m2; P<0.01) in BAV patients than in TAV patients, but the incidence of severe prosthesis–patient mismatch was similar in both groups.[11] In the STS/ACC TVT registry, 1-year mean gradient (13.1±8.1 mm Hg versus 13.0±6.2 mm Hg; P=0.86) and equal or greater than moderate PVR (3.2% versus 2.5%; P=0.47) were similar between BAV and TAV patients treated with SAPIEN-3.[96] The longer follow-up by echocardiography reported stable mean transvalvular gradients of 14.5±9.6 mm Hg at 15 months after TAVR in a cohort of 79 BAV patients.[100] The BAVARD registry was also reassuring on the overall THV stent circularity after TAVR with BE, SE, or ME valves between BAV (n=101) and TAV (n=88) patients (ellipticity index at the annulus level, 1.18±0.14 versus 1.18±0.11; P=0.5). However, BE valves, having a higher radial force than SE valves,[101] achieved a lower ellipticity index. All THVs were also more constrained and underexpanded in BAV patients compared with TAV patients.[11] As prosthesis expansion could impact the THV durability and promote leaflet thrombosis, these findings raised concerns of a potential accelerated structural valve degeneration process in BAV patients.[102] In a recent study, a 30-day CT scan was systematically performed after TAVR in 60 BAV patients and found hypoattenuated leaflet thickening[97] in 10% of patients, which is comparable to the findings of previous studies in TAV.[103] Thus, the short-term durability of THV seems acceptable although long-term data are required.[103]

Outcomes of SAVR for THV Explant

The increasing rate of THV implantation in BAV patients, especially in a young population, may also lead to an increased need for surgical intervention for structural valve degeneration or aortopathy during a patient's lifetime. Recent data suggest that TAVR explant could be challenging. A study of 17 patients undergoing TAVR explantation reported neoendothelialization of the THV into the aortic wall (both BE and SE valves), requiring intense aortic endarterectomy in all 5 THVs older than 1 year and 2 required unplanned aortic root repair.[104] A recent registry on 123 patients implanted with an early generation of THV requiring SAVR for early valve dysfunction (median time, 2.5 months; interquartile range, 0.7–13) reported worse outcomes than expected by the STS risk model for redo-SAVR (in-hospital mortality of 17.1%).[105] Further research is needed to optimize surgical techniques, evaluate the risk of these operations, and understand the difference between SE and BE valve explant. These gaps of knowledge are of paramount importance as successive TAVR and SAVR could be considered in the lifetime management of BAV patients.