Clinical Description
The FLNB-related disorders can be divided into two groups of conditions caused by loss of function or gain of function of filamin-B. Biallelic loss-of-function pathogenic variants in FLNB cause spondylocarpotarsal synostosis syndrome (FLNB-SCT). Monoallelic gain-of-function pathogenic variants in FLNB cause a spectrum of phenotypic severity ranging from apparently isolated clubfoot to Larsen syndrome (FLNB-LS), atelosteogenesis type 3 (FLNB-AO3), and atelosteogenesis type 1 (FLNB-AO1), which is perinatal lethal. For the purposes of this GeneReview, the previously described entities Piepkorn dysplasia, boomerang dysplasia, and spondylohumerofemoral (giant cell) dysplasia are subsumed under the FLNB-AO1 spectrum. To date, more than 100 individuals with pathogenic variants in FLNB have been identified [Daniel et al 2012, Girisha et al 2016, Yang et al 2016, Salian et al 2018, Stenson et al 2020, Quiggle et al 2022]. The following description of the phenotypic features associated with FLNB-related conditions is based on these reports.
FLNB-Related Larsen Syndrome (FLNB-LS)
FLNB-LS is compatible with survival into adulthood [Bicknell et al 2007, Girisha et al 2016]. Intelligence is normal.
Intrafamilial variation in FLNB-LS can be remarkable. In a large family segregating one of the recurring pathogenic variants leading to FLNB-LS, some individuals had cleft palate and multiple large joint dislocations, whereas others who had no major anomalies had short stature and very mild clinical and radiographic features, such as short distal phalanges and supernumerary carpal and tarsal bones [Bicknell et al 2007]. These mild phenotypes may be analogous to reports that isolated clubfoot may be a minimal manifestation of this spectrum of conditions [Yang et al 2016, Quiggle et al 2022]. Clinical variability can also result from the presence of somatic mosaicism for a causative pathogenic variant in a mildly affected parent and the presence of a germline pathogenic variant in more severely affected offspring [Bernkopf et al 2017].
In their study of 20 unrelated families with a total of 52 affected individuals, Bicknell et al [2007] determined that all probands had dislocations or subluxations of the large joints (80% hip, 80% knee, and 65% elbow). Delivery of an affected infant has the potential to be complicated by extended breech presentation due to dislocation of the hips and knees. The most mildly affected proband had subluxation of the shoulders as her only large joint manifestation. Clubfoot was present in 75% of affected individuals.
Stature is mildly affected. In 14 of 20 probands height was below the tenth centile; height was rarely below the first centile and in one individual was above the 97th centile [Bicknell et al 2007].
Spinal abnormalities were observed on radiographs in 16 (84%) of 19 probands. Cervical kyphosis was noted in 50%, usually from subluxation or fusion of the bodies of C2, C3, and C4, which was commonly associated with posterior vertebral arch dysraphism (i.e., dysplasia of the vertebral laminae and hypoplasia of the lateral processes of all cervical vertebrae). Individuals with FLNB-LS and cervical spine dysplasia are at significant risk for cervical cord myelopathy and secondary tetraparesis [Bicknell et al 2007]. The incidence of myelopathy is at least 15%. Evidence suggests that preemptive posterior stabilization of the cervical spine in individuals with FLNB-LS with cervical spine dysplasia may prevent this complication and that combined anterior and posterior stabilization can lead to clinical improvement in individuals with evidence of myelopathy [Sakaura et al 2007].
Craniofacial anomalies are found in all individuals with FLNB-LS. These include a prominent forehead, depressed nasal bridge, malar flattening, and widely spaced eyes. Cleft palate occurs in 15% of affected individuals.
Deafness is common [Herrmann et al 1981, Stanley et al 1988, Maack & Muntz 1991]. Conductive deafness, often with malformation of the ossicles of the middle ear, was observed in four (21%) of 19 probands [Bicknell et al 2007].
Although laryngotracheomalacia has been reported in association with FLNB-LS, few individuals with FLNB-LS and a documented FLNB pathogenic variant are severely affected.
Short, broad, spatulate distal phalanges, particularly of the thumb, are a common but not invariable manifestation (67%) of FLNB-LS [Bicknell et al 2007].
Genotype-Phenotype Correlations
FLNB-SCT. Homozygosity or compound heterozygosity for pathogenic frameshift or nonsense variants in FLNB causes FLNB-SCT [Krakow et al 2004]. Pathogenic variants associated with FLNB-SCT are associated with loss of protein expression and hence constitute true null alleles [Farrington-Rock et al 2006]. Consequently, no genotype-phenotype association has been described.
FLNB-LS, FLNB-AO1, and FLNB-AO3. The FLNB pathogenic variants associated with LS, AO1, and AO3 are either missense variants or small in-frame deletions and are predicted to encode full-length filamin-B protein.
The large majority of
FLNB pathogenic variants reported in individuals with
FLNB-AO1 are in exons 2-5 [
Bicknell et al 2005,
Daniel et al 2012]. Three individuals with Piepkorn dysplasia have had pathogenic variants in exons 28 and 29.
In some instances, the same FLNB pathogenic variant is associated with different phenotypes (e.g., c.502G>A [p.Gly168Ser] is associated with both AO1 and AO3).
Recurrent
FLNB pathogenic variants:
Nomenclature
Larsen syndrome. Some authors have described what they termed "autosomal recessive Larsen syndrome" [Clayton-Smith & Donnai 1988, Bonaventure et al 1992, Laville et al 1994, Yamaguchi et al 1996]; however, these disorders are phenotypically heterogeneous and are not typical of the FLNB-related phenotype [Topley et al 1994]. The conditions (which likely include a variety of chondrodysplasias with multiple joint dislocations) include: the "Reunion Island form of Larsen syndrome" [Bonaventure et al 1992, Laville et al 1994], which is clinically and radiographically distinct from FLNB-LS and is caused by pathogenic variants in B4GALT7 [Cartault et al 2015]; CHST3-related chondrodysplasia with congenital joint dislocations; and two different forms of Desbuquois dysplasia caused by pathogenic variants in CANT1 and XYLT1.
Atelosteogenesis types 1 and 3 were so named because the major manifestation was disordered and incomplete ossification of the skeleton [Maroteaux et al 1982, Sillence et al 1982, Stern et al 1990].
Note: Atelosteogenesis type 2, one of the sulfate transporter-related osteochondrodysplasias caused by pathogenic variants in SLC26A2 (DTDST), is genetically distinct from FLNB-AO1 and FLNB-AO3 (see SLC26A2-Related Atelosteogenesis).
Piepkorn dysplasia, formerly considered to be the same disorder as boomerang dysplasia, has been readdressed by Rehder et al [2018], who described a phenotype that was more severe but with significant overlap with FLNB-AO1. This severe phenotype included flipper-like limbs, a characteristic form of synpolydactyly, and completely absent ossification of many skeletal elements at mid-gestation.
Boomerang dysplasia was so named due to the distinctive bowing of the femora in the original descriptions. The definition of its pathogenesis has confirmed that this condition is best subsumed under the FLNB-AO1 spectrum.