A number sign (#) is used with this entry because Loeys-Dietz syndrome-1 (LDS1) is caused by heterozygous mutation in the TGFBR1 gene (190181) on chromosome 9q22.

The Loeys-Dietz syndrome (LDS) is an autosomal dominant aortic aneurysm syndrome with widespread systemic involvement. As defined by Loeys et al. (2006), the disorder is characterized by the triad of arterial tortuosity and aneurysms, hypertelorism, and bifid uvula or cleft palate. Some patients have craniofacial involvement consisting of cleft palate, craniosynostosis, or hypertelorism. Bifid uvula may also be present. The natural history is characterized by aggressive arterial aneurysms and a high rate of pregnancy-related complications.

LDS is also associated with immunologic-related disorders: approximately one-third of affected individuals exhibit food allergies, in contrast to a prevalence of 6 to 8% in the general population, and LDS patients have an increased prevalence of asthma, rhinitis, and eczema (summary by MacCarrick et al., 2014).

Nomenclature

In initial reports, LDS patients, defined as those with mutations in TGFBR1 or TGFBR2, were stratified into 2 types, depending on severity of craniofacial features (type 1) or cutaneous features (type 2) (MacCarrick et al., 2014). Given that vascular disease is the major concern in LDS irrespective of the severity of systemic features, a revised nosology was proposed with sequential numbering corresponding to the gene mutant in each group (see below).

Genetic Heterogeneity of Loeys-Dietz Syndrome

LDS1 is caused by mutation in the TGFBR1 gene. LDS2 (610168) is caused by mutation in the TGFBR2 gene (190182). LDS3 (613795), which is associated with early-onset osteoarthritis, is caused by mutation in the SMAD3 gene (603109). LDS4 (614816) is caused by mutation in the TGFB2 gene (190220). LDS5 (615582) is caused by mutation in the TGFB3 gene (190230). LDS6 (619656) is caused by mutation in the SMAD2 gene (601366).

Reviews

MacCarrick et al. (2014) provided a review of LDS, stating that there are no specific clinical criteria for the diagnosis, which is confirmed by molecular testing. They proposed that mutation in any of the 4 genes, TGFBR1, TGFBR2, SMAD3, or TGFB2, in combination with arterial aneurysm or dissection or a family history of documented LDS, should be sufficient to establish the diagnosis. The authors noted that rapidly progressive aortic aneurysmal disease is a distinct feature of LDS, and they discussed management strategies for cardiovascular issues as well as other complications of LDS.

Schepers et al. (2018) reviewed the clinical manifestations of LDS associated with mutation in the TRFBR1/2, TGFB1/2, and SMAD2/3 genes, concluding that the LDS phenotype represents a broad spectrum that is emerging as more patients are reported. They suggested genetic testing of the LDS genes be performed in the following scenarios: patients with the typical clinical trial of hypertelorism, cleft palate/bifid uvula and arterial tortuosity/aneurysm; early-onset aortic aneurysm with variable combination of other features including arachnodactyly, camptodactyly, clubfeet, any type of craniosynostosis, blue sclerae, thin skin with atrophic scars, easy bruising, joint hypermobility, bicuspid aortic valve, patent ductus arteriosus, atrial and ventricular septal defects; sporadic young probands with aortic root dilatation/dissection; families with autosomal dominant thoracic aortic aneurysms, especially those families with early-onset aortic/arterial dissection or aortic disease beyond the aortic root, including cerebral arteries; patients with a Marfan syndrome (MFS; 154700)-like phenotype, especially those without ectopia lentis, but with aortic and skeletal features not fulfilling the MFS diagnostic criteria; and patients with clinical features reminiscent of vascular Ehlers-Danlos syndrome (130050), including thin skin with atrophic scars, easy bruising, and joint hypermobility, who have normal type III collagen biochemistry and/or normal COL3A1 (120180) genetic testing.

Diagnosis

Schepers et al. (2018) noted that no formal diagnostic criteria had been developed for LDS, but stated that the diagnosis is established in individuals with a heterozygous pathogenic variant in SMAD2, SMAD3, TGFB2, TGFB3, TGFBR1, or TGFBR2 who exhibit either aortic root enlargement (defined as an aortic root z-score greater than or equal to 2.0) or type A dissection, or compatible systemic features including characteristic craniofacial, skeletal, cutaneous, and/or vascular manifestations found in combination. Special emphasis should be given to arterial tortuosity, particularly of the head and neck vessels, and to aneurysms or dissections involving medium-to-large muscular arteries throughout the arterial tree.

Furlong et al. (1987) described a male of normal height and intelligence with dolichocephaly, a high palate, dolichostenomelia, a mild scoliosis, L4-5 spondylolisthesis, long thorax, prominent pectus carinatum, camptodactyly, pes planus, mitral valve insufficiency, bilateral recurrent inguinal hernias, myopia with normal lenses, mitral valve prolapse, and a dilated aortic root which subsequently dissected at age 18 years. Features atypical for Marfan syndrome (154700) included multisutural craniosynostosis, ptosis, hypertelorism, and hypospadias.

Nicod et al. (1989) described a family in which 9 members over 2 generations had aortic dissecting aneurysm or aortic or arterial dilatation at a young age. Three died of ruptured aortic dissecting aneurysms at ages 14, 18, and 24 years, respectively. A fourth member of the family died suddenly at age 48 years, a few years after aortic repair for aneurysmal dilatation. One member underwent surgical repair of an ascending aortic dissecting aneurysm at age 18 years and was still living at the time of report. Histologic examination of the aortic wall in 3 patients showed a loss of elastic fibers, deposition of mucopolysaccharide-like material in the media, and cystic medial changes--the typical findings of Erdheim cystic medial necrosis. Collagen of types I (see 120150) and III (see 120180) from cultured fibroblasts appeared to be normal on gel electrophoresis.

The patient of Lacombe and Battin (1993) had sagittal craniosynostosis, marfanoid habitus, cleft palate and micrognathia, and aortic root dilatation but normal height and intelligence. Pronounced craniofacial dysmorphism (scaphocephaly, facial asymmetry, unilateral ptosis, orbital dystopia, downslanting palpebral fissures) was reminiscent of that seen in Shprintzen-Goldberg syndrome (SGS; 182212). Other features included arachnodactyly, camptodactyly, kyphoscoliosis, and normal lenses.

Megarbane and Hokayem (1998) described the osseous findings in a 16-year-old male with marfanoid habitus and craniosynostosis. Observations included dolichocephaly, malar hypoplasia, low set, posteriorly rotated ears, ptosis, downslanting palpebral fissures, and microretrognathia. Prominent pectus carinatum, kyphoscoliosis, and limited mobility at the elbows were present, as well as camptodactyly of the fourth and fifth fingers and fifth finger clinodactyly. Flat feet, long toes, and hallux valgus were observed. Echocardiography documented aortic root dilatation. Although early psychomotor development was delayed, intelligence was normal. The authors described radiologic abnormalities that included atlantooccipital joint dislocation, biconvex vertebral bodies, a fusion defect of the dorsal arches of L4 and L5, hypoplasia of the posterior arches of L5 and S1, and elongated diaphyses. Megarbane and Hokayem (1998) considered the clinical findings of their propositus to be similar to those reported by Furlong et al. (1987).

Lacombe and Battin (1993) compared their case with that of Furlong et al. (1987) and 3 other cases. They suggested that the 5 patients might represent variable expressivity of the same syndrome with mental retardation as an inconstant feature, or that there may be 2 distinct syndromes, one with mental retardation (Shprintzen-Goldberg syndrome; 182212) and one without (Furlong syndrome). Megarbane and Hokayem (1998) noted similarities between Shprintzen-Goldberg syndrome and Furlong syndrome and proposed dividing craniosynostosis with marfanoid habitus into 2 types, nominating type 1 as Shprintzen-Goldberg syndrome and type 2 as those with normal intelligence, aortic root abnormalities, and mild skeletal dysplasia.

Loeys et al. (2005) described 10 families with a previously undescribed aortic aneurysm syndrome characterized by hypertelorism, bifid uvula and/or cleft palate, and generalized arterial tortuosity with ascending aortic aneurysm and dissection. The syndrome showed autosomal dominant inheritance and variable clinical expression. Other findings in multiple systems included craniosynostosis, structural brain abnormalities, mental retardation, congenital heart disease, and aneurysms with dissection throughout the arterial tree.

In view of the phenotypic overlap between the Loeys-Dietz syndrome (LDS) and vascular Ehlers-Danlos syndrome (EDS; 130050), Loeys et al. (2006) screened the TGFBR1 and TGFBR2 genes in 40 probands who had previously received a provisional diagnosis of vascular EDS by a medical geneticist, but in whom the diagnosis had been ruled out by studies of type III collagen (see 120180) biosynthesis. Twelve of these patients carried a heterozygous mutation in one of these genes and were assigned to the LDS type 2 phenotypic category (lacking craniofacial features). Physical findings in these patients included prominent joint laxity, easy bruising, wide and atrophic scars, velvety and translucent skin with easily visible veins, spontaneous rupture of the spleen or bowel, diffuse arterial aneurysms and dissections, and catastrophic complications of pregnancy, including rupture of the gravid uterus and the arteries, either during pregnancy or in the immediate postpartum period. None of these patients had cleft palate, hypertelorism, or craniosynostosis. Three patients had bifid uvula, and one had a family history of cleft palate. Mean age at the first major vascular event was 29.8 years, versus 24.5 years for LDS type 1 (with craniofacial involvement) patients. The extent of vascular and skin involvement was similar in the patients with true vascular EDS and in those with LDS type 2; only joint laxity was significantly more prevalent in those with LDS type 2 (12 of 12 vs 18 of 28, p = 0.03).

Loeys et al. (2006) presented the clinical characteristics of a series of 40 probands who exhibited a phenotype consistent with LDS type 1, including 10 previously described patients (Loeys et al., 2005). Besides the triad of hypertelorism, cleft palate or bifid uvula, and arterial tortuosity with aneurysms, patients in this group had additional cardiovascular, skeletal, and cutaneous findings. Neurocognitive signs included delayed development in 6 patients, hydrocephalus in 6 patients, and Arnold-Chiari malformation in 4 patients. When present, delayed development was not always associated with craniosynostosis or hydrocephalus, suggesting that learning disability is a rare primary manifestation. No patient had ectopia lentis, and few patients (18%) had dolichostenomelia.

Loeys et al. (2006) noted that the natural history of both types of LDS among 52 probands was characterized by aggressive arterial aneurysms (mean age at death, 26.0 years) and a high incidence of pregnancy-related complications (in 6 of 12 women). Patients with Loeys-Dietz syndrome type 1, as compared with those with type 2, underwent cardiovascular surgery earlier (mean age, 16.9 years vs 26.9 years) and died earlier (22.6 years vs 31.8 years). There were 59 vascular surgeries in the cohort, with one death during the procedure. This low rate of intraoperative mortality distinguishes the Loeys-Dietz syndrome from vascular EDS.

Loeys et al. (2006) pointed out that in both LDS and vascular EDS, dissection can occur without marked arterial dilatation. However, incidence of fatal complications during or immediately after vascular surgery is about 45% in vascular EDS but only 4.8% in LDS type 2, and only 1.7% in LDS overall. Thus, genotyping is beneficial in patients who present with features of vascular EDS.

Matyas et al. (2006) screened a cohort of 70 individuals with phenotypes related to Marfan syndrome (154700) without mutations in the FBN1 gene (134797) for mutations in TGFBR1. In a 43-year-old patient with thoracic aortic aneurysm and dissection they found a heterozygous missense mutation (190181.0006). No abnormality of the skeletal system or eyes was described, and family history was negative.

Drera et al. (2008) reported a 45-year-old Italian man with LDS confirmed by genetic analysis (190181.0008). He had a prominent and narrow nose, thin lips, bifid uvula and cleft palate, hypermobility of small joints, and soft skin. He also had a history of dissection of both internal iliac arteries and the right femoral artery. There was no aortic root dilatation or tortuosity of the great vessels. The phenotype, which was classified as type 2, was reminiscent of vascular EDS.

Ades (2008) described the evolution of craniofacial features in 7 patients with LDS type 2 and proven mutations in the TGFBR1 or TGFBR2 genes. Most patients had dolichocephaly, a tall broad forehead, frontal bossing, high anterior hairline, hypoplastic supraorbital margins, a 'jowly' appearance in the first 3 years of life, translucent and redundant facial skin that was most pronounced in the periorbital area, prominent upper central incisors in late childhood/adulthood, and an open-mouthed myopathic face. The adult faces appeared prematurely aged. Although not exclusive to the LDS type 2 phenotype, Ades (2008) suggested that recognition of these facial features and their evolution might assist in the differentiation of some cases of LDS type 2 from related clinical entities.

In 30 patients with Loeys-Dietz syndrome, 6 with a mutation in TGFBR1 and 24 with a mutation in TGFBR2, Sheikhzadeh et al. (2014) analyzed imaging findings for the presence of dural ectasia and compared them to 60 age- and sex-matched patients with Marfan syndrome (MFS) and mutations in FBN1. The authors observed a similar frequency and severity of dural ectasia in LDS and MFS, and suggested that it was a highly sensitive but not specific sign of both diseases. Analysis of other documented features in these patients corroborated that arterial tortuosity, aneurysms of nonaortic arterial vessels, patent ductus arteriosus, bifid uvula, and increased craniofacial severity indices are seen only in LDS patients, whereas ectopia lentis and myopia greater than 3 diopters are seen only in MFS patients.

Loeys et al. (2005) noted that some individuals with LDS had phenotypes that overlapped to some extent with that of Marfan syndrome (MFS; 154700), but none met the diagnostic criteria for MFS (De Paepe et al., 1996). All individuals with LDS had manifestations in multiple organ systems that are not associated with MFS. In these individuals, aneurysms tended to be particularly aggressive and to rupture at an early age or to be of a size not associated with high risk in MFS. From a management perspective, the distinction from MFS is neither ambiguous nor unimportant.

Loeys et al. (2006) stated that using 3-dimensional reconstruction of images from the head to the pelvis obtained by computed tomography with intravenous contrast material or magnetic resonance angiography, they identified aneurysms distant from the aortic root in 53% of their patients with LDS type 1; these aneurysms were not detected with the use of echocardiography. Most of these lesions were amenable to surgical repair. This imaging technique also detected arterial tortuosity, a finding of diagnostic importance.

Jondeau et al. (2016) analyzed 176 patients with mutations in TGFBR1 and 265 with mutations in TGFBR2 and found similar survival rates, aortic risks, and prevalence of extraaortic features between the 2 groups. The authors noted that the observed survival rate, 80% at age 60 years, was much better than previously reported. However, within the TGFBR1 group, males had a greater risk of aortic events than females. In addition, 10% of patients in either group who underwent surgical repair of an aortic root aneurysm subsequently presented with dissections of the ascending aorta, suggesting that aortic replacement should include the entire ascending aorta when possible. Aortic dissection was observed in 1.6% of pregnancies.

The transmission pattern of LDS1 in family 10 reported by Loeys et al. (2005) was consistent with autosomal dominant inheritance. The heterozygous mutations in the TGFBR1 gene that were identified in patients with LDS1 by Ades et al. (2006) occurred de novo.

In 4 families (families 7-10) with Loeys-Dietz syndrome who did not carry a mutation in TGFBR2 (190182), Loeys et al. (2005) identified a unique missense mutation in TGFBR1 (190181). Two mutations occurred in the kinase domain, one occurred at the junction of the glycine-serine-rich domain and kinase domain, and one occurred just past the kinase domain at the C terminus. Loeys et al. (2006) added observations on 30 new probands with a phenotype consistent with LDS type 1. Of the 30 newly identified probands, 9 had mutations in TGFBR1 and 21 had mutations in TGFBR2.

In 4 patients with LDS type 2, Loeys et al. (2006) detected heterozygosity for 4 different missense mutations in the TGFBR1 gene. Three of these involved the same codon and occurred at the same codon, arg487 at the C-terminal end of the kinase domain (see 190181.0004). The fourth mutation occurred in exon 4 of the TGFBR1 gene at codon lys232 at the N-terminal end of the kinase domain.

Ades et al. (2006) found that 2 patients with a phenotype resembling Furlong syndrome were heterozygous for the same de novo missense mutation in the TGFBR1 gene (190181.0005).

In 7 individuals referred with classic MFS in whom no mutation in fibrillin-1 (FBN1; 134797) was found, Loeys et al. (2005) sequenced the TGFBR1 and TGFBR2 genes and found no mutations.

At the extreme of clinical severity some individuals with LDS had phenotypes that overlapped considerably with the Shprintzen-Goldberg craniosynostosis syndrome (SGS; 182212), also known as the marfanoid craniosynostosis syndrome. However, SGS is not associated with cleft palate, arterial tortuosity, or risk of aneurysm or dissection other than at the aortic root, and most affected individuals demonstrate no vascular pathology. Loeys et al. (2005) found no mutations in the TGFBR1 and TGFBR2 genes in 5 individuals with classic SGS. Nevertheless, given the extent of phenotypic overlap between SGS, MFS, and selected individuals with mutations in either TGFBR1 or TGFBR2, Loeys et al. (2005) concluded that the pathogenesis of SGS probably relates to alteration in TGF-beta (190180) signaling.

Tran-Fadulu et al. (2009) analyzed the TGFBR1 gene in 150 unrelated families with thoracic aortic aneurysm (AAT) and identified heterozygous missense mutations in 4 families, including a 4-generation family originally described by Nicod et al. (1989) (190181.0007). Tran-Fadulu et al. (2009) compared the clinical features of 30 affected individuals from these 4 families with TGFBR1 mutations to those of 77 patients from 4 families previously reported with mutations in the TGFBR2 gene (Pannu et al., 2005) and found that the average age of onset of vascular disease was significantly younger in the TGFBR1 cohort compared to the TGFBR2 cohort (31.4 vs 45.6 years; p = 0.002). In addition, men in TGFBR1 families presented with vascular disease at a statistically significant younger age compared with affected women (23 vs 39 years; p = 0.019). Thoracic aortic aneurysm was the predominant vascular presentation in both cohorts of patients, but the TGFBR1 patients were twice as likely to present with vascular disease elsewhere (23% vs 8%, respectively; p = 0.039), and vascular disease presentation differed based on gender in the TGFBR1 families: all men but 1 presented with AAT, whereas half of the affected women presented with disease in other vascular beds, including abdominal aortic aneurysms and carotid and coronary artery dissections (p = 0.038). In a combined analysis of the families, there was no difference in overall survival; however, survival was significantly worse in men than in women in TGFBR1 families (p = 0.017) but not in TGFBR2 families. The data also suggested that individuals with TGFBR2 mutations were more likely to dissect at aortic diameters less than 5.0 cm than individuals with TGFBR1 mutations: 3 TGFBR2 patients had dissections with aortic diameters under 5.0 cm, whereas there were no dissections under 5.0 cm in TGFBR1 patients, who often had dramatically enlarged aortic diameters at dissection (6.5 cm to 14.0 cm) or repair (8.5 cm). One TGFBR1 patient who refused repair had been stable for 3 years with an aortic diameter of 5.6 cm.