The Natural History of Bicuspid Aortic Valve Disease

Abstract

The bicuspid aortic valve (BAV) is the most common congenital heart defect with an estimated prevalence of between 0.5% and 2% in the United States, representing up to 6.5 million individuals. Most individuals with BAV will develop valvular and/or aortic complications related to their BAV. How these various complications relate to one another and why they arise remain elusive. Yet, astute observations have yielded relevant classification systems that leverage valvular morphology, aortic shape, and genetic alteration patterns. Emerging evidence supports the existence of BAV phenotypes that may have different patterns of disease presentation, rates of progression, and risk of secondary complications. We review the natural history of BAV in light of known classification systems to illustrate a framework through which future hemodynamic, cell biologic, and other studies can better correlate with clinical endpoints. Consistent utilization of valvular, aortic, and genetic classification systems in the management and study of BAV may facilitate insight into the patterns of the disease, with prognostic and therapeutic significance for individuals who experience this common structural heart disease.

Introduction

The earliest description of the bicuspid aortic valve (BAV) is attributed to Leonardo da Vinci who sketched the valve’s cusp geometry over 500 years ago as part of his extensive exploration of blood flow through the human heart.¹ Today BAV is recognized as the most common congenital heart defect (CHD) with a prevalence of 0.5% to 2% in the United States, representing up to 6.5 million individuals with an estimated annual incidence of 0.7%.2, 3, 4, 5 BAV is increasingly recognized as a syndrome that has valvular and aortic manifestations. Over 50% of BAV cases are associated with dilatation of the proximal ascending aorta, referred to as aortopathy.^6,7 Aortopathy predisposes individuals with BAV to aortic aneurysm and consequent aortic dissection. Notably, nearly 50% of people who undergo ascending aortic (AA) replacement have BAV.⁸

The valvular pathology commonly seen in BAV includes aortic valve stenosis, regurgitation, and infective endocarditis (IE).9, 10, 11, 12 BAV confers an estimated 50% lifetime risk of requiring valve replacement.¹³ In total, the procedures required for BAV management account for more morbidity and mortality than all other CHDs combined.^13,14 Clinical presentations of BAV exhibit significant heterogeneity, with some individuals encountering critical valvular lesions or aortic catastrophes early while others may never encounter these. Several valvular and aortopathy classification systems exist although there is no uniform acceptance of any one system, leading to inconsistent classification of both valvular and aortopathy phenotypes.15, 16, 17, 18, 19, 20, 21 This review details the natural history of BAV through the lens of emerging data, which employs valvular, aortic, and genetic classification. We visit the embryology of the aortic valve first in order to illustrate how the mechanisms of aortic valve maldevelopment lie upstream of many of the BAV phenotypes observed clinically. We then revisit the anatomy of the aortic valve before the review of the valvular, aortic, and genetic classification schemes that have been developed. We then review the natural history of BAV complications and their treatment in light of these classification systems.

Embryology

Between embryonic days 23 and 25 of human development, the heart tube is formed, and although it already produces 1-way anterograde flow, it will soon require valves to ensure that flows meet the metabolic demands of the growing embryo.²² Aortic and pulmonary valve development begins between days 31 and 35 of embryogenesis.²³ By this time, the heart tube has developed into a U-shaped structure, referred to as the looped heart tube, with 1 arm of the “U” giving rise to the outflow tract (OFT) and the other arm becoming the atrioventricular canal (Figure 1). Both regions include endocardial cushions, which ultimately become the semilunar and atrioventricular valves.

Figure 1.

“U”-shaped heart tube. “U”-shaped heart tube with limbs composed of the outflow tract (OFT) and the atrioventricular canal.

Formation of the OFT Endocardial Cushions

The OFT endocardial cushions are swellings of tissue that protrude into the OFT lumen. Similar to a forming blister, they are initially composed of 3 structures: a base, a cap, and a matrix filling the intervening space (Figure 2). The base is composed of cardiac progenitor cells found in the pharyngeal mesoderm that comprises the second heart field. In addition, forming the base of the OFT endocardial cushion, these derived myocardial cells ultimately make up the OFT conotruncus, until it branches into the developing aortic arch.²⁴ The “cap” is formed by a layer of endothelial cells that are derived from cardiac neural crest cells. The matrix between these 2 structures is composed of a range of substances that will facilitate the critical process of endothelial-mesenchymal transition (EndMT).²⁵ The substances composing this matrix include glycosaminoglycans (GAGs; hyaluronan and chondroitin sulfate), glycoproteins (fibronectin, laminin, vitronectin, cytotactin, fibulin, and thrombospondin), collagens (I, III, and IV), and various proteoglycans (PGs), all of which are secreted from the myocardial base underlying the endothelial cap. The secretion of these substances causes the overlying cap to swell, forming a gelatinous endocardial cushion.^25,26 Although the responsible genes for this process remain unknown, collagen and PG overproduction within this cushion matrix appear to be common features of BAV.²⁷

Figure 2.

Formation of endocardial cushions.

Formation of the Aortic and Pulmonary Valves

In the early stages of arterial valve development, once the endocardial cushions have formed, myocardial cells, comprising the cushion base, initiate signals that cause adjacent endocardial cells from the cushion cap to undergo EndMT and invade the intervening matrix (Figure 2).²³ EndMT is a complex biological process in which endothelial cells undergo internal regulatory changes that transition their endothelial phenotype, typically characterized by cell polarity and cell-cell junctions, to a mesenchymal phenotype, here characterized by the expression of mesenchymal proteins and increased secretory functions.²⁸ Impairment of specific genes critical to EndMT, like Notch1 and Nos3, is associated with BAV.29, 30, 31 Invasion of mesenchymal cells into the primitive endocardial cushion matrix results in a relatively bulky and cellularized endocardial cushion.²⁷ In the OFT, there are 2 such cushions, the conotruncal and intercalated cushions, each of which is composed of 2 halves that oppose one another (Figure 3). The halves of the conotruncal cushion are referred to as the superior septal cushion and the inferior septal cushion, while the 2 halves of the intercalated cushion are referred to as the right-posterior and left-anterior intercalated cushions (Figure 3). Aortic and pulmonary valve cusps develop from these 4 structures, with the conotruncal cushions (superior septal cushion and inferior septal cushion) giving rise to the right and left cusps of each semilunar valve. For the aortic valve, this corresponds to the right and left coronary cusps. The right-posterior and left-anterior intercalated cushions become the posterior aortic cusp (noncoronary cusp) and the anterior pulmonic cusp, respectively.²⁴ The absence of several genes has been identified to reproducibly affect fusion and partitioning of specific valve cusps, predictably forming certain BAV subtypes.^31,32 One gene that exemplifies this paradigm is Nos3, a gene which encodes a nitrogen oxide synthase critical to EndMT of neural crest cells. The absence of the Nos3 gene is strongly associated with the BAV subtype resulting from the fusion or failure of separation of the right and noncoronary cusps.³¹ These mechanisms deserve a greater study, and many important questions remain. For instance, it is not yet understood why, in spite of their common embryologic origin, there is a poor correlation between BAV and bicuspid pulmonary valves.³³

Figure 3.

Endocardial cushion origins of aortic cusps.

Abbreviations: ISC, inferior septal cushion; LAIC, left-anterior intercalated cushions; OFT, outflow tract; RPIC, right-posterior intercalated cushions; SSC, superior septal cushion.

Valvulogenesis continues with elongation and thinning of the endocardial cushions by means of cusp remodeling and stratification into 3 tissue layers: the collagen-rich fibrosa, the PG-rich spongiosa, and the elastin-rich ventricularis. These layers are filled with valvular interstitial cells (VICs) and lined by endothelial cells on the exposed surface (endothelium).²⁷ This process of cusp remodeling continues into the postnatal life.³⁴ Grewal et al.³⁵ performed a review of aortic valve and ascending aorta development and suggested a framework for classifying genes relevant to the pathology. This system organizes genes based on the embryologic cell type in which they are expressed, associated BAV leaflet morphology, and association with or without aortopathy (Figure 4).

Figure 4.

Grewal genetic classification of the bicuspid aortic valve.

Abbreviations: LCC, left coronary cusp; NCC, noncoronary cusp; RCC, right coronary cusp.

Anatomy

Understanding the natural history of the BAV requires a grounding, and a common language, when discussing the annular, valvular, and aortic anatomy of the BAV.

Annular

The fibrous structures that compose the aortoventricular junction and the root of the aorta constitute the center of the cardiac skeleton. Although the fibrous portion of the aortoventricular junction is often referred to as the aortic annulus and depicted as a distinct plane (Figure 5), in reality, there is no anatomic plane that separates the ventricle and aorta. Often called the “surgical annulus,” this structure is a virtual ring that lies in the plane determined by the nadir of the basal attachments of the 3 aortic valve cusps (the green ring in Figure 5). In practice, this virtual structure is used to determine the appropriate size for valve replacement; it is easily identified both intraoperatively, to guide surgical aortic valve replacement (SAVR), and by means of preoperative cross-sectional imaging, in advance of transcatheter aortic valve replacement (TAVR).³⁶ In contrast, the “anatomic annulus” is the coronet-shaped structure formed where the cusps attach to the walls of the sinuses of Valsalva and sinotubular junction (the red crown-like ring in Figure 5). These same tissue relationships describe both BAV and tricuspid aortic valve (TAV) anatomic annuli.³⁷ Despite similar anatomic annuli, the surgical annuli differ between BAV and TAV. Notably, BAV annuli are generally larger than TAV annuli.^38,39 Philip et al.³⁸ reported the aortic annular area for BAV to be 5.21 ± 2.1 cm² compared to the 4.63 ± 2.0 cm² for TAV in a cohort of 400 patients.

Figure 5.

Valvular attachments of the normal tricuspid aortic valve.

Valvular

The aortic root refers to the entire valvular apparatus which includes 4 distinct components that each play a major role in valvular function: the (anatomic) annulus, cusps, sinus segments, and sinotubular junction (Figure 6). Immediately above the annulus lie the sinuses of Valsalva, which have wider diameters on average in BAV than in TAV.³⁸ Depending on the cusp morphology, BAV may have 2 or 3 aortic sinuses. Cusps attach to the sinuses of Valsalva and find their peak at their terminal point of attachment referred to as the commissures. Just above the commissures is the sinotubular junction (Figure 6). When cusps fuse or fail to separate, they typically exhibit a thickened ridge of fibrous tissue, referred to as raphe, that emanates from an incompletely developed commissure, which is itself referred to as a false commissure (Figure 7). Raphe occurs in varying locations and to varying degrees, forming the basis for the different BAV subtypes. The leading (free) edge of the cusps refers to the mobile edge of the cusps that stretches from commissure to commissure. These anatomic elements have been used to develop classification systems of BAV subtypes.

Figure 6.

Aortic root tissue landmarks. The aortic root tissue landmarks of a typical tricuspid aortic valve are depicted here.

Figure 7.

Bicuspid aortic valve (BAV) with raphe. (a) A Sievers 1 left-right BAV defined by the presence of a raphe between the left and right coronary cusps. There is a corresponding false commissure where the raphe meets the sinotubular junction. (b) The excised cusps of a Sievers 1 left-right BAV demonstrating a raphe between the left and right coronary cusps. The boundary of the basilar cusp incision includes the false commissure, present at the point at which the raphe met the sinotubular junction.

Abbreviations: L, left coronary cusp; N, noncoronary cusp; R, right coronary cusp.

Aorta

BAV is associated with dilatation of the aorta in a majority of cases, referred to as aortopathy.^6,40 Enlargement of the tubular ascending aorta in BAV often has an asymmetric configuration with prominent dilatation along the convexity of the anterolateral aortic wall.⁶ The etiology of this asymmetry is still unknown although genetic, pathobiologic, hemodynamic, and biomechanical elements likely contribute.41, 42, 43, 44, 45, 46, 47 Indeed, multiple biomolecular investigations have found asymmetric spatial patterns of extracellular matrix (ECM) protein expression along with smooth muscle cell changes which occur in the dilated portions of the aorta in patients with BAV.48, 49, 50, 51 Although these studies do not demonstrate causality, they support the notion that BAV may be associated with altered microarchitecture of the aortic tunics contributing to aortopathy. Concurrently, many studies have demonstrated altered hemodynamics within the ascending aorta of patients with BAV which correlate increased wall sheer stress (WSS) with regions that commonly experience dilatation.52, 53, 54, 55 Again without demonstrating causality, these abnormal flow patterns have been implicated in the development and progression of aortopathy, either directly, through increased WSS, or indirectly, through the induction of matrix metalloproteinase (MMP) gene expression.⁵⁶

Classification Systems

Valvular Classification

“Bicuspid aortic valve” describes a range of valve morphologies that each exhibit 2 valve cusps. A number of classification schemes have been developed to parse these various morphologies into characteristic subtypes.^{3,15,17,18,20} The 2 most utilized systems have been developed by Schaefer et al.¹⁷ and by Sievers et al.¹⁸ The Schaefer system is based on the orientation of the resulting BAV orifice. In this tiered system, BAVs are classified first based on the orientation of the opposing cusps. Subclasses are then determined by the presence or absence of a raphe (Figure 8). By contrast, the Sievers system first determines the number of raphes present in a BAV and then forms subclasses based on the location of the raphe (Figure 9).

Figure 8.

Schaefer classification system.

Abbreviations: LMCA, left main coronary artery; LCC, left coronary cusp; NCC, noncoronary cusp; RCA, right coroncary artery; RCC, right coronary cusp.

Figure 9.

Sievers classification system. Sievers type 2 subtypes have been extended since the initial report by Sievers et al.¹⁸ to reflect all valve morphologies that have been reported.

Abbreviations: A/P, anterior-posterior; Lat, lateral; L-R, Left-right coronary cusp fusion; N-L, noncoronary-left coronary cusp fusion; R-N, right-noncoronary cusp fusion.

The Sievers classification system is the most widely used one and describes 3 types of BAV (Figure 9). A Sievers type 0 BAV has no raphe and only 2 true commissures. A Sievers type 0 BAV may exist in 2 orientations relative to the coronary arteries, anterior-posterior and lateral. A Sievers type 1 BAV has 1 raphe that may be present in any of 3 locations, among the left (L), right (R), and noncoronary (N) cusps. These 3 subtypes are referred to as type 1 L-R, type 1 R-N, and type 1 L-N. A Sievers type 2 BAV has 2 raphes present; although this allows for 3 possible orientations, only 1 of these was observed by Sievers in their original report. Subsequent reports have observed the remaining orientations.^57,58

Although the Sievers classification system has seen the widest adoption, there continues to be inconsistent use of the valvular classification systems in the management and reporting of BAV. Recently, an international consensus statement on the classification of BAV was published and endorsed by a range of surgical organizations.⁵⁹ A nuanced valvular classification, included in this statement, documents BAV fusion, raphe calcification, cusp size/shape, as well as symmetry and represents a gold standard for valvular classification in BAV.

As with all classifications, a balance exists between the number of elements classified and ease of implementation. Although it captures fewer elements, the Sievers classification system may represent a bridging solution for the broader rollout of BAV classification due to its relative ease of implementation. The Sievers system is a robust framework with similar prevalence of BAV subtypes across multiple studies (Table 1), and although it was derived from surgical pathology specimens, this system has been successfully applied to multiple noninvasive evaluations of BAV, including echocardiography and cardiac magnetic resonance imaging.60, 61, 62

Table 1.

Prevalence of Sievers subclasses across reports

Report	Total patients	Sievers 0 (%)	AP	Lat	Sievers 1 (%)	L-R	R-N	N-L	Sievers 2 (%)	(L-R/R-N)	(R-N/N-L)	(N-L/L-R)
Mylotte57	120	32 (26.7)	-	-	82 (68.3)	60 (50.0)	15 (12.5)	7 (5.8)	6 (5.0)	-	6 (5.0)	-
Hayashida58	21	0 (0)	-	-	18 (85.7)	16 (76.2)	1 (5)	1 (5)	3 (14.3)	-	-	3 (14.3)
Jackson9	159	17 (10.7)	-	-	142 (89.3)	119 (74.8)	23 (14.5)	0 (0)	0 (0)	0 (0)	-	-
Schaefer17	191	57 (29.8)	39 (20.4)	18 (9.4)	133 (69.6)	113 (59.2)	20 (10.5)	1 (0.5)	0 (0)	0 (0)	-	-
Sievers18	304	21 (6.9)	7 (2.3)	13 (4.3)	269 (88.5)	216 (71.1)	45 (14.8)	8 (2.6)	14 (4.6)	14 (4.6)	-	-
Niaz63	1010	-	-	-	1010 (100)	715 (70.8)	280 (27.7)	15 (1.5)	-	-	-	-

AP, anterior-posterior; Lat, lateral; N-L, noncoronary-left coronary cusp fusion; L-R, left-right coronary cusp fusion; R-N, right-noncoronary cusp fusion.

One important structural change that would facilitate the broader use of BAV classification would be to call upon our large valve registries to require subtype data entry. For example, the vast Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database and the STS/American College of Cardiology (ACC) transcather valve therapy (TVT) registry each document the presence of BAV, but neither include a field to record BAV subtype according to any classification system. Although a balance must exist, other large registries provide insight into the impacts of narrow data-collection requirements; the FRANCE transcatheter aortic valve implantation registry does not require delineation of BAV vs. TAV, precluding important subgroup analyses. Collecting BAV subtype information in present and future databases may accelerate ongoing efforts to understand the relationships between the BAV valvular anatomy and outcomes, including rates of procedural, postoperative, and other secondary complications.64, 65, 66, 67

Aortic Classification

BAV is commonly associated with aortic dilatation; the cause of such dilatation is thought to be related to abnormal tissue properties of BAV aortas as well as abnormal hemodynamics related to the BAV itself.43, 44, 45^,52,54 Although studies have demonstrated several consistent flow patterns observed in BAV but not in TAV that may correlate with certain valve subtypes, these associations, and their impacts on aortopathy, are inconsistent across reports and will require a further analysis of well-classified data sets to resolve.^52,54 Wojnarski et al.⁶⁸ analyzed the 3D computed tomography scans of 656 BAV patients undergoing an AA surgery and correlated valve cusp morphologies with specific phenotypes of aortopathy using machine learning techniques. Three aortopathy phenotypes emerged, allowing for the creation of a BAV aortopathy classification system, illustrated best by Verma and Siu (Figure 10).⁶ The ascending phenotype refers to dilation of the anterolateral convexity of the ascending aorta from the sinotubular junction to the brachiocephalic branch takeoff and was the most common phenotype observed, accounting for approximately 70% of cases.^69,70 The arch or extended phenotype describes dilation beginning from the sinotubular junction that continues past the brachiocephalic trunk to involve the aortic arch. Lasty, the root phenotype describes dilation at the level of the aortic sinuses with normal dimensions of the tubular ascending aorta and aortic arch and accounts for approximately 20% of cases.⁷⁰ The international consensus statement on the classification of BAV classifies aortopathy in a similar manner although focusses primarily on the ascending and root phenotypes as discrete entities, with the arch or extended phenotype representing an aorta that originated with the root phenotype which has progressively dilated beyond the root segment.⁵⁹

Figure 10.

Wojnarski classification system of bicuspid aortic valve aortopathy. Printed with permission from the Massachusetts Medical Society.

Integration of aortopathy and valvular classification systems has revealed discernable presentations of BAV that had not previously been recognized. The ascending phenotype is typically associated with a later age of BAV diagnosis (>50 years), BAV aortic stenosis (AS), and Sievers type 1 L-R.^{17,41,55,71,72} The arch or extended phenotype has been most commonly associated with a Sievers type 1 R-N valve morphology. The root phenotype is notably rarely associated with the Sievers type 1 R-N valve morphology, but it is primarily associated with the Sievers type 1 L-R valve, a younger age of diagnosis (<40 years), male sex, and aortic regurgitation (AR).^41,73, 74, 75 Correlating classification systems with clinical events (e.g., dissection), time to procedure, and outcomes has not been conclusively demonstrated to date; this task is urgent as current monitoring and procedures are likely to prolong the lives of people with BAV, allowing time for secondary issues to emerge.

Coarctation accounts for 4% to 6% of all CHD cases.⁷⁶ BAV is commonly associated with cases of coarctation, being present in up to 70% of cases.^2,77 Conversely, coarctation is present in only 22% to 36% of BAV cases and typically has a younger age of diagnosis, often diagnosed before the age of 30 years.^61,78,79 Given that the classification of coarctation among the medical and surgical communities has been well established and routinely implemented, coarctation classification will not be discussed here.

Genetic Classification

The intricate temporal and spatial signaling that occurs throughout valvulogenesis provides many points at which BAV may form. Although BAV is classically associated with several connective tissue syndromes (CTS) with known genetic bases (Marfan syndrome, Loeys-Dietz syndrome, and vascular Ehlers-Danlos syndrome), in more than 95% of cases, it is found in individuals without these conditions.⁸⁰ These majority cases, previously thought to occur sporadically, have been shown to primarily follow an autosomal dominant inheritance pattern with reduced penetrance and variable expressivity.⁸¹ In spite of this finding, given that no gene has been clearly implicated or found to be prognostically significant, routine genetic testing of individuals with BAV is not currently recommended unless there is a striking family history of BAV or other left-sided CHD.⁸² Importantly, in order to evaluate for familial associations, society guidelines recommend a 1 time transthoracic echocardiography (TTE) in all first-degree relatives of BAV patients.⁸³

The many genes associated with BAV compels further investigation of the genetic contributions to the heterogeneous clinical presentations of BAV. If gene sequencing is pursued for a given patient, genetic classification of BAV may better (1) inform potential therapeutic targets and (2) direct the timing and approach of surgical interventions. The Grewal framework, by using regionalized cell types (Figure 4), sets a clever structure for classifying BAV genetics in a manner that appreciates drivers of both valvular disease and aortopathy.

Natural History of BAV Disease

The currently perceived BAV prevalence of 0.5% to 2% is likely to be underestimation of the actual prevalence of BAV; the widely quoted studies that derived these rates are limited in their size and often have some degree of selection bias due to populations that underwent imaging studies for other reasons and were incidentally found to have BAV.84, 85, 86 Our current understanding of the prevalence of BAV in the general population may be underestimated by virtue of the fact that many BAV patients are asymptomatic.

With the introduction of transcatheter therapies, it may be even more important to accurately classify BAV morphotype as we gain experience and recognize certain patient-device limitations that may be inherent to specific BAV subtypes.^39,87,88 Different cusp geometries impact TAVR device placement, expansion, and forces exerted on surrounding structures differently compared to TAVs⁸⁹; these differences may negatively impact long-term outcomes of BAV TAVR as compared to TAV TAVR or SAVR.

Aortic Stenosis

The natural history of AS in BAV is accelerated as compared to TAV. This appears to be due to differences in the valve biology, which may be exacerbated by disrupted hemodynamics. Acquired AS involves the sclerosis and calcification of valve cusps characterized by ECM remodeling, angiogenesis, inflammation, and calcium deposition (Figure 11).⁹⁰ Much of our understanding of the progression and natural history of AS is derived from studies investigating TAV AS.91, 92, 93 It remains unknown whether BAV stenosis presents as an accelerated progression of the same mechanisms inherent to TAV AS or whether there are distinct pathophysiologic mechanisms of BAV AS.⁹⁴

Figure 11.

Sievers 1 right-left bicuspid aortic valve (BAV) demonstrating heavy nodular calcification. Sievers 1 right-left BAV demonstrating heavy nodular calcification that has resulted from extracellular matrix remodeling, angiogenesis, inflammation, and calcium deposition.

Abbreviations: L, left coronary cusp; N, noncoronary cusp; R, right coronary cusp.

In BAV, the typical 3-layer cusp architecture and proper distribution of VICs are each disrupted.²⁷ There is also increased volume of PGs, GAGs, and ECM within BAV cusps.²⁷ Increased PG and GAG contents are prominent features observed in calcific aortic valve disease.⁹⁵ Moreover, VIC signaling is recognized as paramount to the homeostasis of heart valve biology.⁹⁶ VIC disorganization, present in BAV, correlates with higher levels of MMPs in BAV cusps, specifically MMP-2 and -9, both of which have proteolytic activity toward type IV collagen and elastin.⁹⁶ The increased levels and lytic activity of these enzymes, and the ECM disorganization that results, are pathways unique to BAV that contribute to BAV AS and influence its natural history.⁹⁷ Consequently, even in those without valve dysfunction, calcification is typically present by 40 years of age.⁹⁸

Unlike typical calcific aortic valve disease, the progression of BAV AS is accelerated, often requiring a valve replacement much earlier in life.¹⁹ Multiple studies have estimated the disease prevalence of either moderate or severe BAV AS to be between 17% and 49% of all BAV cases.^84,99, 100, 101, 102 Michelena et al.⁸⁵ found that among 212 asymptomatic BAV patients studied over a 15-year period, 26 individuals (12.3%) required AVR for severe AS at a median age of 49 years.⁸⁵ In comparison to TAV, BAV requires AVR at higher rates and at a younger age. Shen et al.¹⁰³ prospectively studied the progression of AS between BAV and TAV patients, finding that when adjusted for age and comorbidities, BAVs have faster hemodynamic progression of AS than TAVs. Specifically, via Doppler echocardiography, they found BAV to have a 2-year progression in peak aortic velocity (V_peak) of 0.16 m/s, an increase in the mean gradient of 1.8 mmHg, and a valve area reduction of 0.08 cm².

Although the mechanisms driving the rates of AS progression by Sievers classification have not yet been established, a meta-analysis demonstrates differences in aortic valvular dysfunction by BAV subtype¹⁰⁴; specifically, a higher incidence of AS has been demonstrated in Sievers type 1 R-N while Sievers type 1 L-R demonstrates a higher incidence of AR. The finding that disease progression may vary among Sievers subclasses underpins data showing the variance in the prevalence of AS among Sievers subclasses. Kong et al.¹⁰¹ found a 9:1 ratio of Sievers type 1 to Sievers type 0 BAV among 2118 echocardiograms and 1.6-fold greater number of Sievers type 1 BAVs with severe AS (38.7% vs. 24.1%). In their original report, investigating 304 pathology specimens following AVR, Sievers and Schmidtke reported a marked predominance of severe AS with type 1 BAV compared to type 0 BAV: 137 (45.1%) vs. 12 patients (3.9%).¹⁸ A routine classification of BAV will allow researchers to better determine whether certain BAV subclasses are more at risk of progression to severe AS, which would dramatically improve the prognostication of our current screening guidelines.

Aortic Regurgitation

Due to alterations in cusp and aortic wall composition, cusp prolapse and immobility commonly occur in BAV, resulting in AR.⁴⁹ Among BAV patients, the prevalence of AR is between 26.7% and 47.2%.^{13,85,100,101} Significant isolated AR in BAV, characterized by New York Heart Association (NYHA) functional class III or greater symptoms, tends to present at an earlier age than AS, with a mean age near 45 years, compared to 54 years for severe BAV AS.¹⁰⁰ In the young, AR typically develops due to prolapsing cusps, endocarditis, or following balloon valvuloplasty.¹⁰⁵ As patients age, AR may also develop due to dilation of the ascending aorta in the sinus segments or sinotubular junction, impacting cusp coaptation (Figure 12a).¹⁰⁶ Similar to AS, Sievers type 1 BAV patients present more frequently with significant AR than those with type 0 BAV.^18,101 AR may be associated with different pathophysiologic processes within the aortic wall; Roberts et al.¹⁰⁷ reported that among the BAVs with isolated AR, 50% of patients have significant aortic-medial-elastic-fiber loss compared to only 10% of patients with BAV AS displaying similar findings.

Figure 12.

The effects of sinotubular dilation on cusp coaptation in BAV. (a) A Sievers 0 anterior-posterior BAV which has developed aortic regurgitation in the setting of sinotubular dilation. Dilation has significantly impacted cusp coaptation evidenced by the wide space between leading edges of the anterior and posterior cusps. (b) The same valve from (a) has undergone aortic valve repair and concomitant aortic root replacement with significantly improved cusp coaptation.

Abbreviation: BAV, bicuspid aortic valve.

Aortopathy

There are large discrepancies in the reported prevalence of aortopathy among individuals with BAV. Depending on definitions, evaluation techniques, and study populations, the reported prevalence of BAV aortopathy has ranged between 20% and 84% of patients.5, 6, 7^{,70,73,98,108} For individuals with BAV, aortopathy often begins in childhood and is progressive. Notably, BAV patients have larger ascending aortas after matching for body size and sex. When matched for valvular lesion severity, BAV aortas experience a greater rate of increase in AA dimensions over TAV controls.^40,98

Regarding the contribution of hemodynamics to aortopathy, Barker et al. performed 4D MRI studies investigating the aortic hemodynamics of 15 BAV patients and 45 TAV controls that were age-, sex-, and comorbidity-matched.⁵⁵ The TAV controls were further split into 3 groups of 15 patients each based on aortic dimensions and the history of aortic, heart, or valve disease.⁵⁵ Compared to all TAV controls, BAV had significantly greater WSS values and grossly eccentric aortic jets. In spite of demonstrating elevated WSS values in all BAV subclasses studied, there were no correlation between BAV subclass and any specific dilatation pattern. Elevated WSS in BAV appears to occur in both stenotic and nonstenotic BAVs.^55,109

Regarding the contribution of genetics, the mechanism of medial matrix remodeling appears to be different between BAV and TAV aortas; Phillippi et al.⁴³ have reported ECM fiber content and organization differences between BAV and TAV aortas, with BAV aortas demonstrating elastin and collagen fibers that are more highly aligned. In later work, this group used BAV subclass as a discretizing characteristic to investigate genetic contributions to aortopathy; they found that region-specific regulation of super oxide dismutase isoforms (Sod 1-3) in ascending aorta samples was significantly different depending on the BAV Sievers class.¹¹⁰ These studies illustrate that characterizing BAV subclass can provide a valuable framework to understand differences between BAVs and their associated aortopathies. Given that an association between BAV subclass and aortopathy phenotype is not yet conclusively established by longitudinal studies, aortic classification, in addition to valvular classification, represents a launchpad for future research.

The correlation between aortic disease and concomitant valvular lesions in BAV is not well established. Davies et al.,¹¹¹ in an analysis of 514 patients with AA aneurysms, found that those with BAV and concomitant AS had faster growth rates of AA aneurysm but had similar rates of aortic rupture, dissection, and death over the 20 years of investigation reported. Notably in this study, those with BAV AA aneurysms had improved long-term survival over TAV controls, which is partially attributable to a younger age at presentation and having fewer comorbidities. Despite this, those with BAV have an age-adjusted relative risk of 86.2, compared to the general population, for the formation of an aortic aneurysm—the most common predisposing condition to aortic dissection.¹³

Among established AA aneurysms, the rate of dilatation appears to be similar among BAV and TAV patients; this is also true regarding the incidence of aneurysm-related complications.^111,112 However, the risk of aortic dissection among equally-sized BAV-associated aneurysms may be less than that of TAV-associated aneurysms^113,114; Michelena et al.,¹³ in a community-based study of 416 consecutive patients with BAV, confirmed a very low absolute rate of dissection, with only 2 patients (0.48%) in the entire cohort experiencing aortic dissection after a mean follow-up duration of 16 ± 7 years.

The incidence of aortic dissection among all patients with moderate AA dilatation (4.0 cm – 5.5 cm) is 0.1% per patient year and is not increased in BAV compared to TAV patients.¹¹⁴ The age-adjusted relative risk of experiencing aortic dissection in BAV compared to that in the general population was found to be 8.4 in a study of Olmsted County, Minnesota, and this elevated relative risk is reflective of the fact that AA dilation is common in BAVs compared to the general population; in spite of the very low absolute incidence of aortic dissection in BAV, the dramatically increased individual risk of developing AA aneurysm in BAV accounts for the higher incidence of dissection in BAV cases relative to the general population.¹³ Understanding aortic event rates according to the aortopathy pattern remains an important area for future studies. It has been demonstrated that patients with the root phenotype are at significant risk of aortic events following isolated AVR¹¹⁵; importantly, this finding is independent of preoperative aortic root dimensions or maximal root diameter, highlighting an area in which aortopathy classification may provide greater therapeutic resolution than current dimension-based guidelines. Regarding coarctation, it has been shown that concomitant coarctation in BAV is an independent predictor for AA complications (dissection, rupture, aneurysm >5.5 cm) with an odds ratio of 4.7 compared to noncoarctation BAV controls.⁷⁸

For a given individual, the rate of aortic events in BAV appears to be much lower than aortic event rates in other CTS. Among these, the best studied is Marfan syndrome which is associated with an aortic dissection event rate between 10% and 36% in several national retrospective studies.^116,117 Comparing these figures to aortic dissection rates in BAV, which approximate those of the general population, it is indicated that management considerations for CTS are unlikely to properly apply to BAV aortas.^{13,85,111,114}

Infective Endocarditis

Patients with BAV are prone to developing IE, with some estimates suggesting the lifetime risk of developing IE in BAV to be 23 times greater than that in TAV.¹¹⁸ BAV IE typically occurs in younger patients with relatively few comorbidities.¹¹⁹ Zegri-Reiriz et al.¹²⁰ found that among 3208 consecutive patients with IE, those with BAV had a median age of 43 year and were predominantly male. Pathogens responsible for BAV IE are most often viridans-group Streptococci, with the oral cavity identified as the most frequent portal of entry, and Staphylococci.^12,120 In their study of 50 patients with BAV IE, Lamas and Eykyn found that 30% of individuals had recently had dental work performed or had poor dentition and 74% of individuals were unaware that they had a BAV at their time of presentation for IE.¹² This highlights the importance of a timely diagnosis of BAV and the critical role of good dental hygiene.

IE in BAV often causes significant AR, the symptoms of which are often the cause for patient presentation. As seen in TAV, IE can destroy the BAV cusps or cause cusp perforation leading to severe AR.¹²¹ Zegri-Reiriz et al.¹²⁰ found that 64.8% of BAV IE patients had moderate or severe valve dysfunction at the time of presentation and IE diagnosis. Tribouilloy et al.,¹¹⁹ in their study of 50 consecutive patients with BAV IE, found that heart failure due to severe AR occurred in 32% of individuals and was independently predictive of in-hospital mortality. Cardiac surgery with AVR, including the Ross procedure, is required for BAV IE in 54% to 68% of cases.^102,119

Although BAV IE occurs in younger individuals with fewer comorbidities than TAV IE, few studies have investigated the natural history of BAV IE according to valvular, aortic, or genetic classification schemes. Using the International Bicuspid Aortic Valve Registry, which comprises data collected at 13 tertiary care centers, Kong et al. identified 125 BAV patients with IE¹²²; 84.8% of patients had Sievers type 1 BAV, 64.8% of these were type 1 L-R and 17.6% were type 1 R-N. Interestingly, all BAV type 1 L-R patients developed septic emboli and annular abscesses. The incidence of stroke was higher in Sievers type 1 than that in Sievers type 0 BAV (60.1% vs. 10.5%, p < 0.001) in this IE cohort. During the median follow-up duration of 10 years, 19 individuals (15.2%) died. The highest period of mortality occurred within the first 30 days of BAV IE diagnosis in which 16 patients (12.8%) died.

Treatment

The 2020 ACC/American Heart Association guidelines parse the management of BAV into screening, treatment of aortic lesions, and treatment of valvular lesions.⁸³ Although emerging evidence suggests that there are discernable BAV phenotypes, without a routine classification of valvular, aortic, and genetic subclasses, clear recommendations for subclass-specific management strategies are not yet available. There are currently no medical therapies that have shown benefit in BAV beyond the proper management of blood pressure, healthy diet, and regular exercise.^123,124 Although beta blockers have been proposed to mechanistically reduce aortopathy development by reducing the WSS and heart rate, this has been unfounded by studies demonstrating equivalent aortic WSS values in BAV patients on and off the beta blocker therapy.¹²⁵ Routine beta blockade is not recommended in BAV.

Current society guidelines recommend TTE imaging for all patients with BAV and consideration of a 1-time TTE evaluation for all first-degree relatives. For those with BAV, if the aorta is not adequately imaged with TTE, cardiac magnetic resonance or computed tomography angiography is recommended. The cadence of TTE screening is currently left undefined unless aortic dilatation >4.5 cm is found, in which case, annual screening is recommended. Other risk factors associated with aortic events (family history of dissection, growth rate >0.5 cm/y, coarctation), may be considered for more-frequent screening.^78,83 Although guidelines state that imaging is recommended “to evaluate valve morphology,” a specific valve morphologic classification system has not been endorsed. Additionally, classification of aortopathy phenotype is not currently mentioned as a goal of annual screening.

Current recommendations governing the replacement of the aorta in patients with BAV are based primarily on the aortic diameter and rate of expansion. In spite of guidelines, the case-by-case replacement of the BAV aorta continues to be an area of controversy. Better aortic classification in BAV may improve the precision of future recommendations. At present, our best data support the replacement of BAV aortas in which the diameter of the aortic sinus or ascending aorta are >5.5 cm.⁸³ This threshold appears to be a point at which the natural history of aortic phenotypes converges in their dissection risk and can be governed by a unilateral recommendation. Current guidelines support consideration of concomitant aortic replacement at the time of AVR for patients with aortic diameters >4.5 cm.⁸³ In aortas with a diameter under 5.5 cm, rather than a discrete cutoff for the consideration of isolated AVR vs. paired AVR and aortic replacement, evidence supports the consideration of a set of variables to guide replacement strategies; specifically, consideration of aortic phenotype, concomitant valvular lesions, patient age, and surgical risk are all critical in determining an appropriate aortic strategy at the time of AVR.^{111,115,126,127}

The timing of BAV valve replacement, either with SAVR or TAVR, has been primarily subsumed under the AVR criteria for TAV. There is evidence that ventricular remodeling may occur in BAV far below these echocardiographic thresholds, calling into question the appropriateness of using valvular metrics in determining the timing of AVR in BAV.¹²⁸ Echocardiographic assessment of the valve remains among the standard assessments for monitoring and determining the timing of AVR. Commonly performed in TAV, aortic valve repair is now recommended in BAV for those patients with severe AR meeting the criteria for AVR who can receive care at a comprehensive valve center (Figure 12b).⁸³

With the encouraging results of TAVR in the low-risk TAV populations and as indications for TAVR expand, we are likely to see continued and expanded use of TAVR for BAV.129, 130, 131, 132 Indeed, the concentration of BAV cases within low-risk TAVR populations has been observed; in a recent report from the STS/ACC TVT registry looking at TAVR in low-risk patients with BAV, they reported 3.2% of patients receiving TAVR in their registry had BAV.⁶⁶ Roberts et al. reported that roughly half of 932 patients in their study who received SAVR for AS had a congenitally bicuspid valve.¹⁹ This figure rose to 68% in those aged 40 to 60 years; this high percentage is reflective of the fact that the majority of young AS patients have AS because they have BAV.¹³³ It is estimated that among all TAVR cases, 10% occur in BAV¹³⁴; given the concentration of BAV in low-risk AS populations, increasing low-risk TAVR caseloads are likely to increase BAV TAVR in the coming years.¹³⁵

Despite an early concern that BAV geometry was not well suited for TAVR, continued innovation appears to have reduced the rates of initial limitations and associated complications, best illustrated by the steep reduction in rates of paravalvular leak after implantation with newer-generation devices.⁶⁶ Large studies have reported on the success of TAVR in BAV. For example, Halim et al.⁶⁶ found in their analysis of 170,959 TAVR cases from the STS/ACC TVT Registry that 5412 had BAV with a lower 1-year adjusted risk of mortality than those with TAV (hazard ratio, 0.88 [95% CI, 0.78-0.99]) and no difference in 1-year adjusted risk of stroke (hazard ratio, 1.14 [95% CI, 0.94-1.39]). Recent meta-analyses have challenged these results, finding a significantly higher 1-year incidence of stroke/transient ischemic attack (2.4% vs. 1.6%) and procedural annular rupture (0.3% vs. 0.02%) in BAV compared to characteristically matched TAV controls.¹³⁶ Furthermore, given the excitement of TAVR performance in low-risk patients, there have been several studies investigating TAVR in low-risk BAV cases.^129,137,138 These studies have demonstrated excellent performance, with rates of permanent pacemaker insertion, rates of PVR, and 30-day device success rates in BAV similar to those in TAV.^129,137, 138, 139, 140

Although encouraging, it is important to note these early studies have investigated highly selected BAV populations, excluding patients with bulky or eccentric valve calcification, significant left ventricular OFT calcification, and an anomalous coronary artery anatomy.^129,137, 138, 139^,141 These variables are of special relevance to BAV TAVR given that these procedures frequently require maneuvers (predilation, valve repositioning) and device settings that may increase embolization risks and damage to surrounding structures as calcium burdens increase.^{39,89,142,143} Understanding outcomes within well-classified unselected low-risk BAV populations will be critical in order to understand the future role of TAVR in BAV from a standpoint of procedural risk.

Another factor impacting the future of TAVR in BAV is device durability. Long-term data are distinctly lacking, and given the younger age of the BAV AS population, durable devices are paramount. Preliminary studies indicate that TAVR implantation and patient-device relationships in BAV AS are not equivalent to those in TAV AS. Tchetche et al. analyzed 101 patients from the Bicuspid Aortic Valve Anatomy and Relationship with Devices multicenter registry in which they found consistent underexpansion of TAVR in BAV across all cases.³⁹ There is indication that underexpansion may hamper device durability and even promote cusp thrombosis.^39,144 Understanding device relationships to the supravalvular anatomy is also important given that TAVR devices have not been designed to accommodate aortopathy. Feasibility of TAVR according to the aortopathy phenotype has not been established. Studies have demonstrated that the rate of aortic dilatation after TAVR is similar between BAV and TAV, and the rate of expansion following BAV TAVR is independently associated with the PVL grade.^145,146

Given concerns over device durability, more generally with TAVR, studies have investigated the feasibility of staged, hybrid procedures in which a bioprosthetic SAVR is followed by a valve-in-valve TAVR at the time of valve deterioration. This strategy is understudied but may expand the timeline of optimal device function, an important consideration for BAV patients given that life expectancy often exceeds the reported durability of devices available for AVR. Although existing studies have not investigated BAV specifically, metanalysis by Mahmoud et al.¹⁴⁷ has demonstrated high procedural success of valve-in-valve TAVR within bioprosthetic valves at 30 days (97%) with an incidence of all-cause mortality of 12% at 1 ​year and 29% at 2 ​years.

Further clinical trials for TAVR in BAV are warranted to understand the appropriate application of TAVR devices to the BAV anatomy.^135,148 To date, no randomized, controlled trials have investigated SAVR as compared to TAVR in BAV. Several observational studies have demonstrated similar outcomes between TAVR in TAV and TAVR in BAV, but important questions remain about the appropriate use of available devices in light of the defining characteristic of BAV, heterogeneity; BAV is heterogeneous in its development, in its valvular and aortic anatomy, and in its natural history. Understanding how TAVR devices perform in BAV must respect the span of conditions, referred to by the umbrella term “BAV”, through routine preprocedural classification.

Given that many uncertainties remain regarding short- and long-term outcomes of TAVR in BAV, consideration of an AVR strategy in BAV must incorporate the expertise of a heart team including cardiac surgeons, interventionalists cardiologists, imaging specialists, and others. This team-based expertise is needed to optimize the index approach for patients’ AVR while also considering lifetime interventional planning. At present, SAVR remains the first-line treatment for the majority of patients with BAV. Available evidence supports the consideration of newer-generation TAVR devices for those with BAV who have intermediate or high surgical risk. Further studies are needed to understand the safety and efficacy of staged, hybrid procedures in BAV, especially for those patients whose life expectancy exceeds the durability of currently available AVR technologies.

Conclusion

Most individuals with BAV will develop valvular and/or aortic complications related to their BAV. Despite the recognized association between BAV and these complications, our understanding of the development of BAV itself and the mechanisms of complicated disease is incomplete. Despite these knowledge deficits, astute observations have yielded relevant classification systems for BAV that leverage valvular morphology, aortic shape, and genetic alteration patterns that help to categorize different BAV phenotypes and thus allow for a more granular study of mechanisms of disease development. They may also have implications on predicting patterns of presentation and clinical progression. Codifying BAV according to valvular, aortic, and genetic classification systems provides a framework through which future hemodynamic, cell biologic, and other studies may be better correlated with clinical endpoints. Consistent utilization of valvular, aortic, and genetic classifications systems into both clinical practice and research may facilitate better insight into patterns of disease that have both prognostic and therapeutic significance.^35,73,94

Credit Authorship Contributions

All authors have participated in conception and design of the study or analysis and interpretation of the data; (b) drafting the article or revising it critically for important intellectual content; and (c) approval of the final version.

Funding

The authors have no funding to report.

Disclosure Statement

The authors report no conflict of interest. This manuscript has not been submitted to, nor is under review at, another journal or other publishing venue.