Our study describes an in-depth effort to compare the ascending aortic wall of patients with MFS (non-dilated), BAV (non- and dilated), and TAV (non- and dilated) in one study. A specific aortic wall architecture associated with aortic wall pathology in both MFS and BAV is an extensively discussed, yet controversial, subject. Previous studies compared the aortic wall of MFS and BAV histologically and found that the media were characterized by cytolytic necrosis and elastic fiber degeneration in both diseases [24]. An increased MMP activity and decreased fibrillin-1 in the aortic wall, without comparable reduction in matrix components elastin and collagen, were further noticed [4, 7, 8]. Similarities have thus been noted, though without getting grip on a possible common defect. In this study, we attempted to shed light on intrinsic defects of the aortic wall leading to the observed similarities by comparing the ascending aortic wall of both diseases with each other and with patients with TAV without MFS. As recently several publications focused on a difference in dilation progress after aortic valve repair dependent on whether the diseased aortic valve was stenotic or regurgitant with concomitant root dilation [25, 26]; in the latter, we also paid additional attention to the underlying aortic valve pathology besides structural histopathologic features in the patient groups. In the literature, it is argued that the aortic dilation in the stenotic type is a functional hemodynamic-induced problem, while the aortic wall problem in the root phenotype is genetically determined. Therefore, an isolated aortic valve replacement in stenotic BAV patients is believed to halt the aortic wall dilation. Girdauskas et al. further state that BAV patients with root dilation require a more aggressive surgical approach, being a genetic, connective tissue disorder-like form of aortic disease which is independent of transvalvular flow perturbations [25]. We, however, do not agree on this differentiation into two main groups for a several reasons. Firstly, we postulate that all BAVs are intrinsically a congenital malformation (Grewal et al.) and that an early developmental concomitant defect during embryogenesis contributes to the abnormal maturation of the vascular wall of the ascending aorta independent of a subsequent stenotic or a root phenotype [27]. Secondly, as shown in Table 2, in our study population, aortic valve pathology in non- and dilated BAV and dilated TAV was highly variable showing either stenosis or no stenosis, with or without regurgitation. The dilated BAV and MFS groups even consisted of patients without valve pathology, which indicates that even in the absence of specific aortic valve pathology, aortic complications can occur. We can however not completely refute the influence of hemodynamics on the development of aortic wall complications, as we do not have follow-up data of the operated patients. This is a limitation of our study and an important aspect which should be taken into account in future research.

Histopathologically, in our study, patients in the BAV group did not show marked degenerative features in the aortic media [5], which has also been reported earlier [28, 29]. The media of the ascending aorta in MFS, however, showed resemblance with the TAD with significant cytolytic necrosis, VSMC apoptosis, and degradation of the elastic lamellae, which could predispose the aortic wall for dissection. The MFS group thus showed characteristics of a weak aortic wall, although strikingly the pathologic aorta was not even markedly dilated in the MFS study population as compared to the TAD (Table 2). A role for increased MMP9 production in the occurrence of cytolytic necrosis and extracellular matrix production responsible for increased inter-elastic lamellar distance in TAD and MFS was not supported by our studies. The only marked increase of MMP9 was observed in the non-dilated, susceptible BAV with no marked cytolytic necrosis. So further studies are needed into the role for MMP9 in aortic wall pathology.

To further elucidate the high incidence of aortic wall pathology in MFS and BAV, we studied the level of fibrillin-1 expression. The immature aortic wall of BAV and MFS shared a marked resemblance in the level of expression of fibrillin-1. Decreased fibrillin-1 mRNA and fibrillin-1 protein have been demonstrated before in BAV individuals [8] and can lead to dissociation of VSMCs from medial matrix components [30]. The diagnosis of MFS is however clinical and relies on a set of defined clinical criteria (the Ghent nosology) [2]. The new diagnostic criteria emphasize cardiovascular manifestations of the disorder, in which aortic root aneurysm is one of the prime features. Recently, Pepe et al. [31] identified two fibrillin-1 mutations in a population of eight BAV patients. These mutations had never been detected in MFS patients before. Moreover, these BAV patients did not meet the clinical MFS criteria; therefore, they were not diagnosed with MFS [31]. We can conclude that in the present study and previous studies, a decrease in fibrillin-1 has been reported in four conditions: patients with BAV without a fibrillin-1 mutation [8] (current study), patients with BAV and a fibrillin-1 mutation with [9] or without [31] clinical MFS features, and finally patients with clinically MFS and a fibrillin-1 mutation but no BAV [32].

As fibrillin-1 is produced by VSMCs [33], a significant decrease (of structurally normal) fibrillin-1 is plausible in an aortic wall which constitutes less well-differentiated VSMCs, also without apparent fibrillin-1 mutations. As described in this study, in both MFS and BAV, the aortic wall is primarily less well differentiated which can explain the secondary decrease in the (structurally normal) expression of fibrillin-1. Besides a decreased amount of fibrillin-1 in the aortic wall, in line with earlier research, we found that the distribution and localization of fibrillin-1 was different in the MFS and BAV, with more accumulation within the VSMCs [7, 34, 35], whereas it was mostly seen extracellular in the TAVs. As earlier described by Nataatmadja et al., the few extracellular fibers observed in BAV and MFS were thick and short [7]. Hollister et al. also found a deficiency in the amount of microfibrillar fibers, analyzed immunohistochemically [35], comparable with our results.

As described above, cytolytic necrosis and VSMC apoptosis are observed in both MFS and TAV. However, VSMC apoptosis, which leads to cytolytic necrosis, seems to occur due to a different pathogenetic mechanism in MFS as compared to TAV. In TAV, aging, accompanied by an increased progerin expression, and atherosclerosis cause VSMC apoptosis [5, 36], whereas as we have seen in this study in MFS these features are not apparent. In MFS, the media of the aortic wall contain less well-differentiated VSMCs similar to BAV. Therefore, we hypothesize that there must be a different pathway leading to cytolytic necrosis in MFS, which is on the one hand not related to cardiovascular aging, but on the other hand can also not be associated directly to the immature state of the VSMCs as in BAV where cytolytic necrosis scarce [5, 28].

VSMC apoptosis and subsequent cytolytic necrosis in MFS have earlier been described to occur at a much younger age as compared to the TAV and as a direct consequence of the fibrillin-1 mutation which leads to an increased Angiotensin II (AT2) receptors signaling and subsequently induction of TGFβ signaling [37‐39]. We postulate that two different pathologic pathways might thus be distinguished leading to aortic dissections. For clarification, we have schematically provided a working hypothesis (Fig. 4) showing how the various differentiation pathways and the involved gene pathways could lead to the observed histopathological similarities and differences we have seen between the investigated groups. In MFS, there is thus a combination of immaturity of the aortic media and cytolytic necrosis, due to increased VSMC apoptosis, related to increased AT2 receptor signaling [31, 32], rendering the aortic wall very weak. These observations could explain why angiotensin-receptor-blockers (ARBs), as losartan, have been identified as a potentially therapeutic agent to prevent progressive dilation of the ascending aorta in MFS. ARBs reduce the signaling that occurs through both AT receptors, (AT1 and AT2 receptor). Many trials have emerged in recent years investigating the efficacy of ARB’s in human patients with MFS [12‐18]. The first results from these studies show reduced dilation of the aortic root in MFS with losartan [17].

A similar positive effect has not been reported in a population of BAV patients who were treated with an angiotensin inhibitor (ACE inhibitor) [19] which can be understood on the basis of our current results. In BAV, medical treatment is not effective because the aortic wall in BAVs is not characterized by apoptosis and cytolytic necrosis. In conclusion, comparison of the aortic wall samples in BAV, MFS, and TAV seems to demonstrate that aortic dissections in MFS might have a different pathogenesis as compared to the TAV, explaining the higher incidence and the younger age of occurrence. These findings could be relevant for understanding why medical treatment to inhibit angiotensin is seemingly successful in MFS and less effective in TAV and BAV patients also being prone for aortic wall pathology.