Bicuspid aortic valve (BAV) is a common congenital valvular defect, affecting about 0.5–1% of the general population [1, 2]. Nearly half of the BAV affected patients would undergo medical or surgical management during their lifetime due to the BAV-related valvular and/or aortic complications (i.e. aortic stenosis, aortic insufficiency, aortic aneurysm, and aortic dissection et al.) [3]. Patients with BAV had a high risk of development of the ascending aorta dilatation (to a size above 4.0 cm) [4]. Of them, about 15% of patients have a dilated aortic root frequently with aortic insufficiency at a young age [5]. According to a large community cohort, the risk of aneurysm formation is significantly higher in BAV patients than the general population, and about a quarter of the patients with BAV may ultimately underwent aortic surgery after BAV diagnosis. The risk of acute aortic emergencies, most commonly aortic dissection, is eightfold in patients with BAV disease compared with the general population [6].

The underlying mechanisms of BAV-associated aortopathy are widely discussed. There are two potentially competing or more likely complementary perspectives of the cause of BAV aortopathy, i.e. the hemorheological reason and genetic factors. The former argues that the fused-leaflet of BAV alters the blood flow and increases the shear stress in the tubular ascending aorta [7]. From the genetic view, the genetic defect may increase the risk of BAV aortopathy [8]. However, which kind of influence would play a more important role is still controversial.

The aortic valve and ascending aorta share some common embryological origin [9]. It is reasonable to hypothesize that rare variants of BAV associated genes are likely to be risk factors predisposing to BAV aortopathy, especially in the aortic root. Human BAV presenting together with aortopathy has been proved to be associated with rare variants in NOTCH1, TGFBR2, FBN1, and SMAD6 et al. [5, 10].

However, our knowledges of the inherited components affecting the clinical phenotype of BAV aortopathy are still limited. It could be argued that different types of BAV aortopathy may have different rare variants spectrum. By targeted next generation sequencing (NGS) of genes associated with BAV, we aimed to investigate distribution of rare variants in a large well-phenotyped BAV cohort.

In the present study, we reported on targeted resequencing detection of the rare variants associated with BAV and related aortopathy. With a 13-gene panel, an overall rare variants detection rate of 26.0% in 96 patients was achieved. To the best of our knowledge, this study was the largest cohort by targeted resequencing accessing the pathogenic variants associated with sporadic BAV disease in the Eastern Asian population. According to our data, the BAV patients with aortopathy displayed more complex genetic heterogeneity than BAV patients with normal aorta. The root type aortopathy was strongly associated with the rare variants of BAV genes.

The flourishing availability of NGS technologies provided us the convenience to rapidly detect multiple variants in multiple genes, by which, a more comprehensive understanding of the heritability of the BAV and associated aortopathy would be reached. Bonachea et al. conducted a targeted sequencing assay of 78 unrelated BAV patients [17]. They identified 31 putative disease-causing variants in 16 individuals. Dargis et al. used a panel of 9 genes associated with BAV to study 48 patients [18], they identified 19 potentially pathogenic variants. Of them, 7 variants in NOTCH1 took the highest burden of deleterious variants. Some researchers also reported the NGS results of the BAV aortopathy. Girdauskas et al. used a 20 genes panel to investigate 63 BAV patients with root aortopathy [19]. A total of 64 rare variants in 15 candidate genes were identified, of them 24 were potentially pathogenic/likely pathogenic variants. Gillis used a panel of 22 genes to study a large BAV/ thoracic aortic aneurysm cohort [10]. They identified 112 variants classified as pathogenic in 441 patients. In comparison with these previous studies, our cohort reaches a similar rare variants detection rate, suggests that the BAV might be a combination as a polygenic trait, many discrete genes being responsible for the disease.

Some studies have suggested that aortic dilatation of BAV is an inherent attribute of the disease. In a long-term follow-up research, the accelerated aortic dilatation in BAV was observed even independent of severe aortic stenosis or regurgitation and aortic coarctation [4]. Although the mechanism of aortic dilation has been long debated, the genetic theory was continuously advocated and investigated. Based on genome-wide single nucleotide polymorphism array, the early onset of thoracic aortic disease was demonstrated frequently in BAV patients, and associated with recurrent rare copy number variations [20]. These findings suggest that BAV and aortopathy might share some common molecular mechanism, the genetic defect may hurt the cardiac or vascular development, leading to the early onset of disease.

Several recent studies focused on the genic origin of BAV related aortopathy. Pepe et al. suggested FBN1 was an important cause of the aortic root dilatation in BAV [21]. Gillis et al. [10] considered variants in SMAD6 has a significant contribution to BAV aortopathy, with a variant burden of 2.5%. In our study, there is a wide spectrum of rare genetic variants in patients with aortic dilatation, no gene plays an outstanding role.

NOTCH1 is the first gene proved to be associated with both familial and sporadic BAV cases. Altered Notch signaling causes malformations of the left ventricular outflow tract including BAV and facilitates the progress of aortic valve disease [22]. Controversy also exists in whether the NOTCH1 variant is a causal of BAV aortopathy. Girdauskas et al. found a wide spectrum of the variants in 19 of 63 patients with BAV root aortopathy, and NOTCH1 was the leading causal gene [19]. In contrast, Kent et. al investigated BAV related aortopathy by sequencing the NOTCH1 gene in 13 BAV families. They failed to identified causal variant in the patients and concluded that the NOTCH1 variant contributed little to the noncalcified bicuspid aortic valve combined with ascending aortic aneurysm [23]. Ambitiously, in another multiple center sequencing study, NOTCH1 was suggested as a protective factor for BAV related aortic aneurysm [10]. In our cohort, 9 nonsynonymous rare variants in NOTCH1 were identified in unrelated patients, with the detection rate of 9.4%. The phenotypes of rare variant carriers were distinct from each other. Generally, the NOTCH1 variants were more common in the phenotype of valve dysfunction with normal aortic shape than the phenotype of aortopathy. Our findings support the argument that NOTCH1-dependent mechanism has a heterogeneity and low penetrance of aortic aneurysm in the BAV patient. Still, NOTCH1 plays the important role in the pathogenesis of BAV and in its complications.

According to the previous study and our data, the incidence of root aortopathy is ranged from 7.3 to 13.5% in patients with BAV [24]. This type of aortopathy has a tendency of rapid progress and high risk of adverse aortic events [24, 25]. Since the root aortopathy presents a marfanoid-like aortic root morphology, it is reasonable to suggest that the genetic factors may determine a defect of the aortic wall, leading to the aortic root dilation. In our study, the pathogenic variants in TGFBR1, TGFBR2, FLNA, and FBN1 are responsible for the aortic root dilatation. Indeed, it is a compelling finding that the pathogenic variants accounted for 57.1% of root phenotype in this consecutive BAV cohort.

The FBN1 variant has been associated with syndromic or no syndromic bicuspid aortic aneurysm [21, 26]. Meanwhile, the FLNA defects have been demonstrated to cause various developmental malformations involving the brain, skin, bone, and cardiovascular. According to Chen et al. [27], the BAV is one of the most common cardiac malformations occurred in 5.3% of the filaminopathy patients. In our group, the majority of root group patients are lack of typical systemically connective tissue disease presentations. The explanation may be the reduced penetrance of the genetic variants.

Since the phenotypes are overlapped and with low penetrance of some variants, it is challenging to make differential diagnoses with BAV related aortopathy such as fibrillinopathy and filaminopathy by only clinical manifestation. The insight of BAV genetic background helps to improve the understanding of the pathology of these diseases. In view of the fact that rare variants in some special genes act as potential modulator of BAV and associated aortopathy, the molecular diagnosis can refine the accuracy of clinical diagnosis, and thereby provides optimal guidance on patient’s management and helps to develop more tailored therapeutic strategies.

Firstly, our participants were all recruited from the surgical department. There would be a potential bias of assessing genetic effect to represent the general population. Secondly, our study method focused on the impacts of the rare variants of known BAV genes on the disease. Whole-genome sequencing in a larger cohort might be helpful to broaden genetic horizon of BAV and relative complications. The variants would also need further functional studies to evaluate the relevance. Long-term follow-up is also warranted for assessing the potential pathogenicity of the rare genetic variants.

In conclusion, this targeted NGS approach revealed a wide and specific gene variant spectrum in a large cohort of patients with BAV. From the insight of the genetic view, it is reasonable to classified the aortopathy as tubular and root type. The root phenotype presented high pathogenic variant burden among BAV patients. Patients with rare nonsynonymous coding variants had an accelerated disease progression and underwent surgical management at an early age. The molecular evaluation of BAV disease could help to accurate the diagnosis and was worth further researched.