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

TAVR Challenges in BAV Anatomy

Aortic Valve Anatomy

BAV patients were generally excluded from RCTs evaluating TAVR because of several anatomic features that could increase the risk of TAVR procedural complications or failure. There remain some distinct differences between bicuspid and tricuspid AS. First, the dimensions of all components of the AV complex are generally larger in BAV than in TAV, increasing the likelihood of having an annulus size outside of the range covered by currently available THV (annulus area, 521±21 mm2 versus 463±20 mm2; ascending aorta, 36.7±5.4 mm versus 30.4±3.4 mm; P<0.01).[11,31,32] Second, several studies on CT or transesophageal echocardiography for BAV have described a similar circular geometry of the aortic annulus as that found in TAV.[31,33–35] However, the supra-annular geometry, especially at the level of the sinus of Valsalva, is more elliptical in BAV patients.[33]

The AV complex shape (from the annulus to the leaflet tips) is nontubular (flared or tapered) in two-thirds of BAV patients[11,36] (Figure 3). It has been suggested that in this latter configuration (14%–40.1% of patients[11,36,37]) the narrowest dimensions and point of highest resistance of the AV complex may be located above the annulus, at the commissural level,[11,38,39] especially in those with a tricommissural raphe-type anatomy[36] (Figure 3), which may contribute to a less circular deployment,[40] THV oversizing, and subsequent risk of THV underexpansion or annulus injury if sized according to the annulus dimensions.

Figure 3.

Computed tomography scan sizing strategy and transcatheter heart valve design choice in bicuspid aortic valve.
Top, A, Aortic valve sizing at different level of a tapered bicuspid aortic valve type 1 left/right bicuspid aortic valve (A) annulus level, (B) 4 mm above annulus, (C) 8 mm above annulus, (D) simulated Evolut-R/Pro 26 mm (dashed blue and pink circle). Bottom, B, Aortic valve sizing at different level of a flared bicuspid aortic valve type 1 left/right bicuspid aortic valve (A) annulus level, (B) 4 mm above annulus, (C) 8 mm above annulus, (D) simulated SAPIEN 3 26 mm (blue circle). BE indicates balloon-expandable; and SE, self-expandable.

Third, BAVs are more heavily calcified than TAVs, as demonstrated by the higher weight in excised BAVs than in TAVs (men, 3.61 g versus 2.31 g; women, 2.62 g versus 1.64 g),[13] and higher calcium score (5143±4730 AU versus 3038±3671 AU; P=0.02).[11] The BAV's calcification burden is also more eccentric and asymmetrical. Fourth, the presence of a calcified raphe, cumulated with the heterogeneous distribution of the calcium load and asymmetrical morphology of the AV complex, might prevent the optimal expansion of the THV stent frame.[31,41,42] When localized between the right coronary cusp and left coronary cusp, the raphe may also indirectly induce conduction disturbances through an increase of the compression forces applied on the contralateral stent frame toward the noncoronary cusp area, which is very close to the His bundle.[43] Fifth, coronary anomaly may be more frequent in BAV than in TAV,[32,44] particularly a left dominance and separate ostia for the left main and circumflex artery. Moreover, 1 or both of the coronary ostia may lie more frequently in close proximity to the commissure, unlike in TAV in which both ostia are in the middle of the sinuses. This, along with a longer and heavier calcified leaflet particularly at the tip, could potentially increase the risk of coronary obstruction. However, the height of the coronary ostia has been reported to be similar or higher for BAV than for TAV, which is reassuring about this risk.[31,34,45] Sixth, the aorta also has a more horizontal orientation exposing it to difficulties in accurate positioning of the THV within a vertical annulus.[35,46] Lastly, aortopathy frequently coexists with BAV and cannot be addressed by TAVR.

BAV Patients With Pure Aortic Regurgitation or Mixed Aortic Regurgitation/AS Disease

BAV patients, especially the type 1 R-N of the Sievers classification, have faster hemodynamic valve deterioration compared with TAV patients.[47,48] Moderate-to-severe calcific AS is the most common presentation between the fifth and sixth decades of life, while pure isolated aortic regurgitation (AR) accounts for only 10% to 15% of young BAV patients.[49] Some degree of AR is frequently associated with AS (mixed AV disease) and may worsen the hemodynamic burden imposed on the left ventricle. Indeed, it was previously demonstrated that patients with moderate mixed AV disease had similar outcomes as patients with isolated severe AS,[50] while AVR was associated with better outcomes.[51] Hence, both BAV and TAV patients with mixed AV disease may be considered for TAVR. However, there is a paucity of data on this subset of patients because the prevalence of mixed AV disease is rarely reported.[4,5] In the recent Society of Thoracic Surgeons (STS)/American College of Cardiology (ACC) transcatheter valve therapy (TVT) studies focusing on BAV patients, moderate-to-severe AR was observed in 3% to 15% of patients treated by TAVR.[52,53] Unfortunately, none of these studies reported subgroup analysis in patients with mixed AV disease. However, it can be assumed that for those with a predominant stenotic lesion, TAVR can be considered, as is performed for BAV patients with pure AS. BAV patients with isolated pure AR are typically young with a low calcification burden and frequently present root aortopathy. These patients should be treated with surgical replacement or repair, as the results are excellent.[54,55] TAVR may be considered for very high-risk patients with pure AR, but this is associated with a low device success (81%)[56] and data on BAV anatomy are scarce.

BAV Aortopathy and Evolution After AVR

BAV patients have a dilated ascending aorta compared with healthy subjects or TAV patients with a similar degree of AS severity.[57] The prevalence of BAV aortopathy is difficult to estimate but ranges between 20% and 84% depending on the evaluation technique and aortic size thresholds used to define aortic dilatation.[58,59] However, most BAV patients undergoing AVR do not require aortic root replacement surgery as prophylactic replacement of the ascending aorta is performed in approximately 25% of BAV patients.[60,61] All segments from the aortic root to the proximal descending aorta can be dilated, but we commonly distinguish 3 different patterns according to the segment predominantly involved: isolated dilation of the tubular ascending aorta (the most common; ≈70%), isolated dilation of the aortic root (≈20%), or dilation of the entire ascending aorta.[58] All BAV phenotypes can be associated with each aortopathy pattern, but some associations are more frequent,[62] even though numerous inconsistencies exist between studies.[63] This association may be explained by different embryological entities and/or different distributions of the aorta wall shear stress related to the type of AV dysfunction (aortopathy is more preponderant in AS than in AR),[64] and the eccentric transvalvular systolic flow pattern according to the BAV phenotype.[65–67] Patients with BAV type 1 left-right coronary cusp fusion present mainly aortic root dilation.[68] Patients with BAV type 1 right-noncoronary cusp fusion are associated with a larger diameter of the tubular ascending aorta and aortic arch while sparing relatively the aortic root.[69] Patients with type 1 noncoronary-right coronary cusp fusion have mainly aortic root dilation. Data for BAV type 0 are scarce but isolated dilation of the tubular ascending aorta is frequently associated with AS, whereas isolated aortic root dilation is frequently associated with AR.[70,71] BAV type 2 is rare but frequently associated with diffuse aortopathy in almost 50% of patients. The BAV phenotype and severity of the valve dysfunction may also influence the progression of the aortopathy, but published data are inconsistent.[64,72–74] The risk conveyed by the aortopathy is also difficult to determine. Patients with BAV have an 8-fold increased relative risk of aortic events compared with the general population,[60] but the absolute risk remains low and considerably lower than in patients with Marfan syndrome.[75] Although the aortic dimension is associated with the risk of aortic dissection, it remains a poor predictor of this complication.[76,77] In a recent investigation, the mean diameter at dissection in BAV patients was 10 mm greater than that in TAV patients (66±15 mm versus 56±11 mm; P=0.0004).[78] The flow pattern generated by a BAV and assessed by 4-dimensional flow MRI is abnormal (higher wall shear stress level than control) and could participate in aortopathy progression.[79,80] To reduce the need for late reoperation, the European Society of Cardiology, American Heart Association, and American Association for Thoracic Surgery guidelines,[6,81–83] recommend performing a combined AVR and aortic surgery when the diameter is ≥45 mm regardless of the aortopathy pattern or the underlying BAV phenotypes (with respectively IIa/C, I/C, and IIa/B levels of recommendation and evidence). These recommendations are mainly based on single-center observational studies[30,84] that mixed up AS and AR with diverse aortopathy patterns. This aggressive surgical treatment strategy for all BAV-associated aortopathy remains debated.[76] A previous study comparing SAVR alone or associated with aortic root surgery in BAV patients with aortic dilation (including AR) failed to delineate any prognostic cut-off.[85] The abnormal flow pattern induced by BAV[79,80] can remain abnormal after AVR and could impact postoperative progression of the aortopathy.[86] Yasuda et al observed progressive aortic dilatation after isolated SAVR in BAV patients, but not in TAV patients, and Kim et al found a similar annual aortic dilatation rate after SAVR in both BAV and TAV patients.[87] Other studies found that BAV patients with mild-to-moderate (40–50 mm) ascending aorta dilation undergoing isolated SAVR for AS were at very low risk of adverse aortic events at long-term (15 years) follow-up,[88] which is comparable to patients with TAV stenosis.[28] Concerning TAVR, only 2 small studies described the postoperative evolution of aortopathy. A cohort of 67 patients with aortopathy showed a stabilization of aortic dilatation after a short follow-up of 12 months.[89] A CT scan study comparing 86 BAV patients with 122 TAV patients demonstrated a similar rate of progression of the aortic dilatation (0.2±0.8 mm/year versus 0.3±0.8 mm/year; P=0.59, respectively), but very few (3.5%) patients completed the 5-year follow-up.[90] Overall, aortopathy and BAV phenotypes are intertwined between genetic and mechanical hemodynamic factors.[91] These features should probably be considered when determining the need for concomitant aortic surgery. However, to date, no classification encompassing the aortopathy pattern, BAV phenotype, and type of AV dysfunction has demonstrated its prognostic significance. Furthermore, other factors should be considered such as age, concomitant hypertension, progression rate, diameter indexed to body surface area, and family history of acute aortic events. Given the uncertainties on the need and optimal timing for aortic replacement, TAVR could represent the best option in selected patients without severe dilatation of the proximal aorta.