Response to Pyeritz et al.

To the Editor: Subsequent to the meeting of the “Montalcino Consortium,” the Marfan Foundation convened a more inclusive gathering in Boston of all invested parties to discuss the important issue of nosology for inherited presentations of thoracic aortic aneurysm. In addition to representatives from the Montalcino Consortium, participants included leaders from many additional centers of excellence in the care of thoracic aortic disease and patients and representatives from patient advocacy groups including the US and Canadian Marfan foundations and the US Loeys–Dietz Syndrome Foundation. Virtually all of the comments and concerns raised by Pyeritz et al.¹ in their letter, “Loeys–Dietz Syndrome Is a Specific Phenotype and Not a Concomitant of Any Mutation in a Gene Involved in TGF-β Signaling,” were discussed in detail. We welcome this opportunity to more thoroughly present the prevailing themes of this meeting.

There was uniform consensus that knowledge about the gene or specific mutation underlying a disease can serve as a proxy for the biochemical and cellular events that drive disease initiation or progression and that influence response to therapy. This knowledge often provides predictive value with regard to when and where disease will manifest; the character of the disease (e.g., mild or severe, and more or less predictable than average); the best diagnostic, follow-up, and therapeutic protocols to apply; and the spectrum of risks for family members (present and future). The ability to “bin” patients into a specific predisposition class (based on commonality of phenotype/genotype/mechanism) increases the potential for anticipatory counseling and management while minimizing bias or other confounding factors. Such principles underlie current practices in the diagnosis and management of Marfan syndrome (MFS), in which the presence of a pathogenic FBN1 mutation or an unequivocally affected family member, in combination with aortic root enlargement, is sufficient for the diagnosis of MFS irrespective of the presence and/or severity of other systemic findings, ranging from catastrophic infantile presentations that often associate with striking dysmorphism to exceedingly mild adult presentations in individuals who do not show outward features of MFS. This alerts caregivers with variable familiarity with the intricacies of diagnosis to the spectrum of possibilities and informs patient management (e.g., the need for ophthalmologic evaluations, proper imaging modalities, and surgical thresholds). A description of these widely endorsed and applied priorities and nosologic practices included the leadership of the Montalcino Consortium as authors, and similar practices are in place (and productive) for other presentations of aneurysm, prominently including vascular Ehlers–Danlos syndrome.^2,3

The term “Loeys–Dietz syndrome” (LDS) was initially applied to describe a condition, caused by mutations in either of the genes that encode subunits of the transforming growth factor (TGF)-β receptor (TGFBR1 or TGFBR2), that associates many features of MFS (arachnodactyly, pectus deformity, scoliosis, dural ectasia, and aortic root aneurysm) with other discriminating features in the craniofacial (hypertelorism, cleft palate, bifid uvula, and craniosynostosis), skeletal (cervical spine malformation and/or instability and clubfoot deformity), and cutaneous (translucent skin, easy bruising, and dystrophic scars) systems.⁴ Most importantly, patients with LDS often show a widespread and aggressive vasculopathy (arterial tortuosity, aneurysms throughout the arterial tree, dissections at young ages and at relatively small vascular dimensions that do not infer risk or provoke surgery in MFS, and many other conditions). Within 1 year (in 2006), a second publication expanded the phenotypic spectrum of LDS to include individuals with TGFBR1 or TGFBR2 mutations with a similarly diffuse and aggressive vascular phenotype but only subtle or even absent craniofacial and/or skeletal manifestations.⁵ Around the same time, others proposed use of the diagnosis of familial thoracic aortic aneurysm and dissection for such patients, a term previously applied for patients with mutations in genes encoding components of the vascular smooth muscle cell contractile apparatus with predominant or exclusive aneurysms of the thoracic aorta that tended to be well served by imaging and surgical practices that had previously been developed for MFS.⁶ This practice might have merit if the diagnosis adequately captured the extent of predisposition for such patients, if true phenotype–genotype correlations exist, and if phenotypes bred true in families; none of these suppositions proved valid. Indeed, in the published literature and in many examples shared by multiple groups at the Boston meeting, such patients commonly showed aneurysms distant from the thoracic aorta, dissections at small dimensions as compared with those of MFS, and family members with classic craniofacial, skeletal, cutaneous, and cardiovascular manifestations of LDS. Furthermore, data were presented showing that virtually every familial thoracic aortic aneurysm and dissection–associated TGFBR1/TGFBR2 mutation had been observed in other individuals and/or families with classic and unequivocal LDS.

In the initial and subsequent reports describing SMAD3 mutations in individuals with a syndromic presentation of aneurysms, it was clear that virtually all craniofacial, skeletal, cutaneous, and cardiovascular manifestations of LDS were commonly observed, prominently including diffuse and aggressive vascular disease.⁷ The impression of this group at that time was that a predisposition for early and severe osteoarthritis was a predominant distinguishing feature, prompting the proposal of use of the designation aneurysms–osteoarthritis syndrome. However, with further experience, they have refined this position to accommodate the finding of osteoarthritis in many etiologies of LDS (and many other connective tissue disorders) and the defining aggressive vascular course reminiscent of LDS, leading to their adoption of the designation LDS-type 3 for this condition, as expressed at the Boston meeting. Likewise, our initial description of mutations in the gene encoding the TGF-β2 ligand (TGFB2) described frequent LDS-associated features including hypertelorism, cleft palate, bifid uvula, clubfoot deformity, arterial tortuosity, and widespread aneurysms with a propensity for early dissection.⁸ Given this substantial clinical overlap with LDS caused by other etiologies, we proposed the designation LDS-type 4, an observation and conclusion now shared by others (Leutermann et al.⁹). Although other reports have described patients with familial thoracic aortic aneurysm and dissection with or without systemic features reminiscent of MFS in patients with mutations in SMAD3 and TGFB2, phenotypic descriptions were often vague, clinical pictures were not provided, and aneurysms and dissections distant from the thoracic aorta were also described. Importantly, there were no distinguishing characteristics regarding the location, nature, or cellular consequence of the mutations, with absolute consensus that heterozygous mutations in TGFBR1, TGFBR2, SMAD3, and TGFB2 impair the performance of positive effectors of TGF-β signaling in cell culture systems but lead to paradoxically enhanced TGF-β signaling in the vessel wall in affected patients and mouse models. This is in striking contrast to Camurati–Engelmann disease or Myhre syndrome, in which site-specific mutations in the TGFB1 or SMAD4 genes lead to tissue-specific (predominantly skeletal and not vascular) phenotypic features through a mechanism that is not well defined and to multiple self-healing epitheliomas (Ferguson–Smith syndrome) and in which a germ-line null allele of TGFBR1 is paired with somatic loss of the opposite allele. There are many situations in which the conditions caused by genes in a common pathway with a common molecular and predominant phenotypic consequence share a single designation—with Noonan syndrome caused by mutations in genes encoding effectors and regulators of the RAS–RAF–MEK–ERK axis serving as a prime example. Use of a numbering system that designates both the condition and locus (i.e., LDS-1, LDS-2, LDS-3, and LDS-4 for the conditions caused by mutations in TGFBR1, TGFBR2, SMAD3, and TGFB2, respectively), as we have proposed, would alert clinicians to the commonalities of phenotype while accommodating current and future recognition of potentially important locus-specific differences.

In our proposed guidelines for the management of LDS we do not advocate a “one-shoe-fits-all” approach. Existing evidence from all groups would strongly support widespread cardiovascular imaging on an intermittent basis and imaging of the cervical spine at diagnosis. We have proposed that surgical repair of aortic root aneurysms be “considered” with measurements around 4 cm, but we stress that other factors such as family history, historical knowledge about the specific genotype (gene and/or mutation), severity of systemic findings, and the patient’s personal assessment of risk versus benefit should influence this decision. The decision regarding the use of medications such as angiotensin receptor blockers and/or β-blockers can and should be influenced by each center’s analysis of available data. Given the data regarding the underlying mechanism of disease, the ability of angiotensin receptor blockers to suppress TGF-β signaling in the aortic wall, the evidence that angiotensin receptor blockers can attenuate the progression of aortic root enlargement in LDS mouse models and in both mouse models and people with MFS (another TGF-β vasculopathy), the excellent tolerance profile for this class of medications, and the aggressive, diffuse, and often catastrophic course of untreated LDS, we believe that the use of these agents for LDS remains a responsible choice in patients who are fully informed regarding what is known and what remains to be learned. The selection of dose will require additional study, but we are compelled by the evidence for both safety and superior efficacy of ultrahigh dose angiotensin receptor blocker regimens in the management of chronic kidney disease and other TGF-β–related pathologies. Although the proposed management guidelines remain descriptive rather than prescriptive, we believe that they make a genuine attempt to integrate the prevailing tenets of both evidence-based and individualized medicine, with the sole and unwavering priority of best serving our patients and their families. Refinement will undoubtedly occur with further research and knowledge.

Disclosure

The authors declare no conflict of interest.