Focus on the unique mechanisms involved in thoracic aortic aneurysm formation in bicuspid aortic valve versus tricuspid aortic valve patients: clinical implications of a pilot study

Abstract

OBJECTIVES

The involvement of different factors in the onset of thoracic aortic aneurysm (TAA) in patients with a bicuspid aortic valve (BAV) vs those with a tricuspid aortic valve (TAV) is well recognized. However, the molecular, genetic and cellular mechanisms driving TAA remain unclear. The aim of this study was to identify the different mechanisms involved in TAA development in patients with BAV vs TAV.

METHODS

Aorta specimens and DNA samples were collected from 24 BAV (18 men and 6 women; mean age: 54.2 ± 14.39 years) and 110 TAV (79 men and 31 women, mean age: 66 ± 9.8 years) patients. A control group of 128 subjects (61 men and 67 woman, mean age: 61.1 ± 5.8 years) was also enrolled. Histopathological and immunoistochemical analyses were performed, as well as genotyping of 10 polymorphisms.

RESULTS

In BAV-associated ascending aortas, significant severe plurifocal apoptosis of smooth muscle cells and matrix metalloproteinase-9 (MMP-9) amounts were detected. In contrast, TAV-associated ascending aortas were characterized by a significant severity of elastic fragmentation, cystic medial necrosis, medial fibrosis and inflammation. In addition, in BAV cases, the −1562TMMP-9 and −735TMMP-2 alleles represent independent risk factors for TAA. The effects of these genotypes combined with hypertension and smoking in BAV cases result in an increase in both the apoptosis (P = 0.0001) and levels of MMP-9 (P = 0.001). In TAV cases, the D angiotensin-converting enzyme and +896A Toll-like receptor-4 alleles seem to be the predictive factors for TAA risk. They, combined with hypertension and age, significantly increase both the microscopic lesions and inflammation.

CONCLUSIONS

Our data seem to suggest that TAA in BAV and TAV patients arises from different molecular, cellular and genetic mechanisms. They might help to identify the potential molecular and genetic biomarkers that are useful to detect BAV subjects at high TAA risk, to monitor and treat them differently from those with TAV, with approaches such as the complete removal of the ascending aorta, including the aortic root with or without dilatation.

INTRODUCTION

Congenital bicuspid aortic valve (BAV) is the most common cardiac malformation, occurring in 1–2% of the population, and it affects males four times more often than females [1]. It may evolve in serious complications, which occur in about 33% of patients. Among these, aortic dilatation is the most common consequence of BAV, which determines the onset of thoracic aortic aneurysms (TAA)s [1, 2]. Currently, it is assumed that the principal causes of the increased TAA incidence in BAV patients are genetic abnormalities (i.e. genetic defects in the neural crest-origin cells) and higher haemodynamic stress on the ascending aortic wall determined by turbulent blood flow over the malformed valve.

TAA onset is a very complex process, in which both cellular and extracellular mechanisms are involved. These mechanisms converge on multiple signaling pathways and result in the maladaptive remodelling of the vascular extracellular matrix [2–4]. The presence of molecular, cellular and genetic profiles in TAA samples from patients with BAV, which are different in patients with both idiopathic medial degenerative disease and tricuspid aortic valve (TAV), has been demonstrated [3–8]. These differential profiles seem to suggest that both aneurysm development and its progression are mediated, in each valve group, by unique mechanisms associated with distinct signaling pathways. As result, aortic lesions associated with BAV seem to differ from those correlated with TAV in terms of histopathology, anatomical configuration, natural history and genetic profiles [3–8]. However, the distinct mechanisms involved in TAA development in the two valve patient groups remain unclear, even if there is a growing number of studies.

In a previous, recent study, we evidenced in BAV patients, typical histological abnormalities and a genetic profile characterized by an over-expression of some polymorphisms [−786T/C endothelial nitric oxide synthase enzyme (eNOs), D/I angiotensin-converting enzyme (ACE), +896A/G Toll-like receptor-4 (TLR-4), −1562C/T metalloproteinase-9 (MMP-9), −735C/T MMP-2 polymorphisms]. We also demonstrated that these features predispose them to a major risk of an enlargement of the ascending aorta and severe complications, such as aortic rupture or dissection.

Of consequence, we speculated that, for BAV patients, the diameter of the aneurysm and the growing rate do not represent definite parameters for operation. The surgical strategy should probably consider the structure of the complete ascending aortic wall (the grade of medial degeneration) and genetic risk factors. In particular, we suggested that the use of a surgical procedure based on the complete removal of the ascending aorta including the aortic root with or without dilatation might be reasonable in BAV patients with a high risk of rupture and dissection, particularly in BAV patients with a thin aortic root wall at the time of operation and a particular genetic risk profile (an over-expression of −786T/C eNOs, D/I ACE, +896A/G TLR4, −1562C/T MMP-9 and −735C/T MMP-2 polymorphisms) [9].

In order to contribute to the identification of the molecular and genetic mechanisms associated with TAA development in BAV vs TAV patients, we assessed the molecular mechanisms and genetic risk factors of TAA in 24 BAV vs 110 TAV patients. The identification of the unique mechanisms involved in aneurysm formation in each valve group might help to recognize and to predict subjects at high TAA risk. In the case of BAV patients, this strategy might permit monitoring and treating them differently from those with TAV, preferably using, in BAV patients with a high risk of rupture and dissection, the surgical procedure suggested in our previous study.

MATERIALS AND METHODS

Subjects

Our study received approval from the local ethic committees, and all participants gave their informed consent. Data were encoded to ensure the anonymity of patients in the study and control groups. All measurements were performed without knowledge of the nature of the study.

The study included 24 BAV (18 men and 6 women; mean age: 54.2 ± 14.39 years) and 110 TAV patients from Western Sicily, undergoing TAA surgical repair at the Cardiac Surgery Unit of Palermo University Hospital. We selected BAV and TAV patients with TAA complication. For their selection, histopathological analyses and exclusion criteria for arteriosclerosis, connective tissue disorders and inflammatory diseases were utilized.

The BAV (stenotic or incontinent) was fibrocalcific in 13 (54%) cases and prolapsed in 1 (4%) (Table 1). In contrast, the TAV was fibrocalcific in 10 (8%) cases and prolapsed in 20 (18%).

The evaluation of aorta diameters in both BAV and TAV patients was made both preoperatively and in the operating room by using transthoracic echocardiography and transesophageal echocardiography. The estimations were performed as follows: by assessing the dimensions of the aortic annulus, sinuses of Valsalva and proximal ascending aorta (above 2.5 cm of the sinotubular junction) from the parasternal long axis and by evaluating those of the aortic arch from the suprasternal view. The mean sizes of Sinus of Valsalva were 33 ± 3.8 and 28.2 ± 3.3 mm in BAV and TAV patients, respectively. The mean sizes of the aneurysmatic ascending aorta were 52.3 ± 9.4 and 50.4 ± 6.9 mm in BAV and TAV patients, respectively (Table 1). Colour Doppler was used to assess the presence and severity of aortic regurgitation and stenosis. Furthermore, diameter sizes of the aortic root and ascending aorta were carried out using helical computed tomography image analysis techniques.

Relevant medical histories regarding aortic disease were obtained from patients' medical records. Thus, demographic and clinical features, and comorbidity conditions were collected (Table 1). In all BAV and TAV cases, hypertension was controlled by using beta-blockers.

To detect histopathological abnormalities, control ascending aortas were obtained from 30 individuals (20 men and 10 women; mean age: 55 ± 11.57 years) who died from causes unrelated to aortic disease and having no sepsis at the time of death, as confirmed by autopsy.

To perform genotyping analyses, 128 controls [61 (47%) men and 67 (53%) women; mean age: 61.08 ± 5.83 years] were also enrolled. They were in good health according to their clinical history and blood tests (complete blood cell count, erythrocyte sedimentation rate, glucose, urea nitrogen, creatinine, electrolytes, C-reactive protein, liver function tests, iron and proteins). Concerning the major cardiac risk factors, smoking (17%), hypertension (31%), cardiovascular ischaemic familiarity (8%), dislipidemy (6%) and diabetes mellitus (3%) were evidenced (Table 1). Furthermore, echocardiography imaging examinations confirmed the absence of any aorta wall abnormalities in all controls. We selected a very homogenous population. Patients and controls belonged to the same ethnic group, since their parents and grandparents were born in Western Sicily.

Aortic specimens, histopathological and immunohistochemical assays, TUNEL testing and genotyping

BAV patients and controls, BAV aorta specimens, control aorta tissues, DNA samples of the BAV cases and 128 controls were enrolled, obtained, analysed and genotyped as described in the study of Pisano et al.^. [9]. The data previously obtained were compared in this present study with those of TAV patients.

In addition, the TAV aorta samples, as well as their blood and DNA samples, were collected, obtained and analysed in this study, utilizing the same procedures reported by Pisano et al. [9]. Their DNA samples were also genotyped for the same single-nucleotide polymorphisms (SNPs) selected in our previous study [9]. Furthermore, procedures, criteria, definitions and grading systems for tissue sample collection, staining, histopathological and immunohistochemical assessment, Terminal deoxynucleotidyl transferase dUTP Nick End Labeling (TUNEL) testing, semi-quantitative MMP-9 evaluation and genotyping were performed as described in detail in our previous study [9].

Statistical analysis

All analyses were performed with R and EXCEL software. The χ² test was utilized to compare all demographic and clinical features, comorbidity conditions between BAV and TAV patients, and between cases and controls. The same test was also utilized to verify the hypothesis of an association between aorta lesions and valvular dysfunctions. To verify the hypothesis of a relationship between the severity of aorta lesions and size of the aorta dilatation in BAV and TAV patients, a non-parametrical Spearman correlation test was done. It was performed for other correlations. To analyse significantly the relationships between quantitative variables, the Wilcoxon rank-sum test was employed.

Allele and genotype frequencies were evaluated by gene count. Data were tested to find out the consistency between observed and expected genotype frequencies, according to Hardy–Weinberg equilibrium, by χ² tests. Significant differences in frequencies among groups were calculated by using the χ² test and appropriate tables. Furthermore, odds ratios (ORs) with 95% confidence intervals (CIs) and their significances were calculated.

The significant relationship between genetic variables and pathology of TAA risk was analysed using quasi-likelihood binomial models.

RESULTS

Patient and control characteristics

All patient and control features are summarized in Table 1. A significant difference was first observed in age between BAV and TAV patients. BAV patients were significantly younger than TAV cases (54.2 ± 14.3 vs 66 ± 9.8 years, respectively, P < 0.00001). No differences were evidenced in the number of males and females among the two groups. No significant differences were also observed in the size of aorta dilatation between BAV and TAV cases. In contrast, a higher sinus Valsalva size was found in BAV than in TAV patients (33 ± 3.8 and 28.2 ± 3.3, respectively, P = 0.0001). Among the TAA risk factors, a higher number of BAV cases (33%) were smokers than in the other two cohorts (33 BAV vs 11 TAV cases and 17% controls, P = 0.003 and 0.005, respectively). In addition, significant differences were observed for hypertension between cases and controls (75 BAV and TAV cases vs 31% controls, P = 0.00005 and 0.001, respectively). Furthermore, BAV (stenotic or incontinent) was fibrocalcific in 13 (54%) cases and prolapsed in 1 (4%) (Table 1). In contrast, TAV was fibrocalcific in 10 (8%) cases and prolapsed in 20 (18%). Comparisons between these TAV and BAV values evidenced significant differences (P = 1.1e−7 and 0.001, respectively). Other significant differences were found comparing BAV and TAV cases for aortic valve dysfunctions. Higher faint incontinence and moderate and severe stenosis were, indeed, found in BAV than in TAV cases (38 vs 5% and 38 vs 16%, P = 0.000001 and 0.01, respectively).

Histopathological and immunohistochemical observations: comparisons between BAV and TAV cases

As reported in Table 2, we evidenced significant differences comparing microscopic aorta lesions in ascending aortas from 24 BAV and 110 TAV patients. In particular, higher severity for elastic fragmentation, medionecrosis, cystic medial necrosis, medial fibrosis and inflammation was found in TAV aortas than in BAV aortas (P = 0.0001, 0.0009, 0.006, 1e−8 and 0.000008, respectively). However, these lesions characterized only the aorta segments of TAV aortas with dilatation.

In contrast, more significant amounts of MMP-9 and severe plurifocal apoptosis were detected in BAV aorta samples than in those associated with TAV (P = 0.0001 and 1e−8, respectively). Interestingly, they were observed in the different aorta segments with or without dilatation (precisely, of the aortic root without dilatation, and the aneurysmatic ascending aorta portion) of the BAV aortas.

Furthermore, we compared the severity of aortic wall abnormalities with the aortic diameter in BAV and TAV cases. Positive correlations were detected (r = 4.1, P = 0.02 and r = 3.4, P = 0.03, respectively, by the non-parametrical Spearman correlation test; data not shown). The severity of aortic wall changes was also dependent on age and hypertension only in TAV patients (r = 5.2, P = 0.009 and r = 3.6, P = 0.005, respectively, by the non-parametrical Spearman correlation test; data not shown).

Genotype distributions and allele frequencies of the 10 SNPs

As previously reported in the study of Pisano et al. [9] and in Table 3, significant differences both in genotype distributions and allele frequencies were observed for −786T/C eNOs, D/I ACE, −1562C/T MMP-9 and −735C/T MMP-2 SNPs comparing BAV cases with the 128 controls. No significant differences between BAV cases and controls were found both in genotype distributions and allele frequencies for the two TLR4 SNPs, CC-chemokine receptor 5 (CCR-5) WT/Δ32 SNP and other SNPs of MMP-2 and eNOs genes selected.

Comparing the genotype distributions and allele frequencies of the 10 SNPs between TAV cases and controls, significant differences were obtained. In particular, the genotype distributions of −786T/C eNOs, D/I ACE, −1562C/T MMP-9, −735C/T MMP-2, +896A/G TLR4 and CCR5WT/Δ32 SNPs were significantly different between TAV cases and controls (P = 0.002, 0.0001, 0.01, 0.01, 0.01, 0.01, respectively). Accordingly, significant differences were observed in the allele frequencies of these SNPs (Tables 3 and 4). In contrast, no significant differences between TAV cases and controls were observed both in genotype distribution and allele frequencies for the other SNPs analysed. Interestingly, compared with controls, both TAV (P = 0.03, 0.0001, 0.04 and 0.001, respectively) and BAV (P = 0.01, 0.02, 0.00001 and 0.0001) cases showed very greater frequencies of −786T eNOS, D ACE, −1562TMMP-9 and −735TMMP-2 alleles associated with a very significantly increased TAA risk (Tables 3 and 4). However, higher significant values of OR for these alleles were detected in BAV than in TAV cases (Table 3). Furthermore, using quasi-hood binomial statistical models, we evidenced that the −1562C/T MMP-9 and −735C/T MMP-2 genotypes represent independent risk factors for TAA in BAV patients (P = 0.001 and 0.0002, respectively). The effects of these genotypes combined with hypertension and smoking in BAV cases resulted in an increase in both the apoptosis (P = 0.0001) and levels of MMP-9 (P = 0.001).

Using the same quasi-hood binomial statistical models, the allele D ACE and the +896A TLR4 allele result in being predictive factors for TAA in TAV patients (P = 0.01 and 0.0001, respectively). A synergistic effect between the D ACE allele and hypertension was also found in TAV cases for both an increased aortic diameter (P = 0.002) and the severity of aortic lesions (P < 0.001). The same effect was found in TAV cases between the +896A TLR4 allele and hypertension and age for the increased inflammation (P = 0.002) and the severity of microscopic lesions (P = 0.0001).

DISCUSSION

A very large number of BAV patients develop TAA [1, 2]. Different natural histories and pathophysiological events are well recognized as contributing to TAA onset in patients with BAV vs those with TAV [3–8, 10–15]. However, the molecular and cellular mechanisms driving TAA development in these two valve groups remain unclear. Of consequence, several research groups are attempting to discover the molecular, cellular and genetic putative factors involved in the TAA pathogenesis in BAV patients. Different speculations have been proposed, and different histological, genetic and proteomic comparisons between TAV and BAV cases have been executed [3–8, 10–15]. The goal of this increasing interest is focused on identifying and predicting BAV subjects at high TAA risk, to monitor and treat differently from those with TAV, i.e. with prophylactic treatment vs observation, valve repair vs replacement and ascending aortic replacement with smaller diameters vs observation. However, discordant and inconclusive data have been obtained [3–8, 10–15].

The results obtained in the present study demonstrated that BAV-associated ascending aortas are characterized by more severe plurifocal apoptosis and significant MMP-9 amounts than TAV-associated ascending aortas. This datum is in accordance with previous studies [10–13]. In addition, we observed a lower inflammatory grading and a medial degeneration prevalently without fibrosis and necrosis in BAV than in TAV case aortas. The severity of aortic wall abnormalities correlated with the aortic diameter both in BAV and TAV cases. However, only in TAV cases, it was dependent on age and hypertension.

Based on these data, we proposed some suggestions regarding a different TAA onset in BAV and TAV cases. In particular, we assumed that, in TAV cases, medial degeneration, prevalently induced by physiological ageing and hypertension, was probably the primary pathological mechanism for TAA complication. Although physiological ageing and hypertension are generally considered to be independent for TAA risk, they are able to mediate synergic effects on the aortic wall by inducing pathological changes, i.e. medial degeneration, resulting in the TAA development and progression [16, 17]. In our study, TAV cases enrolled were older than the BAV cases. Their mean age was 66 ± 9.8 years. In contrast, the BAV cases were younger, with a mean age of 54.2 ± 14.3 years. Thus, an early TAA onset characterizes the BAV cases when compared with TAV cases. However, the same percentage of hypertensive subjects characterized the two case groups in our study. In addition, we detected, in TAV-associated ascending aortas, very significant aortic microscopic lesions, i.e. elastic fragmentation, cystic medial necrosis and fibrosis and severe inflammation. These aorta alterations seem to validate the assumption suggested previously. On the other hand, a strong relationship exists between ageing and/or hypertension and medial degeneration associated with inflammation [18].

In addition, we also suggested that the early aortic TAA pathology in the BAV cases might probably be caused by defects in the cellular and molecular microenvironment, resulting from inherent, genetically determined abnormalities of the aortic wall [19]. Accordingly, we observed the same medial degenerative lesions in tissue samples of different aorta segments with or without dilatation (precisely, of the aortic root without dilatation, and the aneurysmatic ascending aorta portion).

In order to validate our suggestions, we researched genes deregulated in BAV and TAV patients, when compared with controls. In particular, we analysed 10 SNPs of following genes: CCR5 (NM-00579), TLR4 (NM-138554.1), MMP-9 (NM-004985), MMP-2 (NM-001121363.1), ACE (NM-152830.1), eNOs (NM-000594). Their selection was based on the literature data that demonstrate their capacity to modulate endothelial dysfunction, local and systemic inflammatory degree, tissue injury, hypertension and vascular matrix homeostasis [9]. Data obtained demonstrated that both TAV and BAV cases had very greater frequencies of −786T eNOS, D ACE, −1562TMMP-9 and −735TMMP-2 alleles associated with a very significant increased TAA risk. However, higher significant values of OR for these alleles were detected in BAV than in TAV cases. In addition, higher frequencies of +896A TLR4 and CCR5WT alleles were only observed in TAV cases and not in the two other cohorts.

Furthermore, in BAV cases, the −1562TMMP-9 and −735TMMP-2 alleles represent independent risk factors for TAA, whose effects combined with hypertension and smoking result in increasing apoptosis and MMP-9. In TAV cases, the D ACE and +896A TLR4 alleles seem to be predictive factors for TAA risk. In combination with hypertension and age, they significantly increase both the microscopic lesions and inflammation.

In the light of the data obtained, TAA complications in TAV cases seem to be the result of a combined effect of physiological ageing and hypertension with genetic variants, i.e. the D ACE and +896ATLR4 alleles are able to cause strong inflammation, hypertension and tissue injuries.

In the case of TAA complications in BAV patients, the data obtained lead us to assume that genetic factors (i.e. −786T eNOS, D ACE, −1562TMMP-9 and −735TMMP-2 alleles) combined with congenital defect or lifestyle factors (i.e. smoking) seem to be the principal cause. This is in accordance with the biological effects induced by the −786T eNOS, D ACE, −1562TMMP-9 and −735TMMP-2 alleles. In particular, the −786T eNOS, D ACE, −1562TMMP-9 and −735TMMP-2 alleles determine a reduction of eNOs tissue endothelium levels, lower NO production and increased release of MMP-2 and -9. As a consequence, endothelial dysfunction and activation of stretch and stress pathways with the release of different molecules, such as MMP-2 and -9, increased hypertension, increased apoptosis of SMC cells and weakening extracellular matrices are induced. These conditions finally lead to degeneration and dilatation of the aortic wall.

CONCLUSIONS

On the whole, our data seem to suggest that TAA in BAV and TAV patients arises from different molecular, cellular and genetic mechanisms. Further and larger studies are certainly necessary to validate these promising findings and our suppositions. The reproducibility of our results from a very large sample size enriched by clinical data might provide the possibility of identifying potential molecular and genetic biomarkers useful to detect BAV subjects at high TAA risk, to monitor and treat them differently from those with TAV. In particular, this could allow us to propose specific clinical recommendations on the surgical approach to use in the case of BAV patients with TAA and a normal aortic root. On the other hand, bypassing and clarifying the problem of the more suitable surgical procedure with or without composite aortic root replacement represent the major aims of recent studies, since no clear and comprehensive guidelines actually exist. However, contrasting data characterize the current literature. Some authors recommend composite aortic root replacement as the appropriate surgical strategy to treat and prevent these aortic diseases in BAV patients [20–23]. This surgical procedure has been advocated to prevent ascending aortic or root dilatation and to reduce the risk of rupture or dissection in a very large number of BAV patients [20–25]. Accordingly, we assumed in the recent study of Pisano et al. [9] that, in BAV cases, the diameter of the aneurysms and the growing rate do not represent a definite parameter for operation. Surgical strategy should also consider the structure of the complete ascending aortic wall (the grade of medial degeneration) and genetic risk factors. However, our suggestion is based only on histological and genetic features. Clinical data and a large BAV patient number are required to support it, as well as a longitudinal study of both procedures or a larger series on the both the techniques, as suggested above.

Limitations

Some limitations characterize our study, such as the limited sample size and the rarity of BAV pathology, as previously suggested in the study of Pisano et al. [9]. In order to support the specific clinical recommendations, clinical data are also necessary. Another limitation regards the SNPs selected in our study. As suggested for other pathological conditions, we also underline, for TAA, the concept that functional SNP effects depend on the presence of one or different environmental causes. Because of this relationship between genetic and environmental factors, a genetic variant might never manifest itself phenotypically, if the carrier is never exposed to a specific trigger. In addition, different alleles might respond differently to the same environmental condition. Consequently, each exposed individual's risk of a given disease phenotype is determined by the particular allele an individual carries. Thus, further and larger investigations are certainly needed. For example, genomic, transcriptomic and epigenomic investigations may eventually lead to a better understanding of the molecular and cellular mechanisms associated with TAA in BAV patients.

Funding

This work was supported by grants from Italian Ministry of Education, University and Research to Giovanni Ruvolo.

Conflict of interest: none declared.

REFERENCES