Clinical Description
Facioscapulohumeral muscular dystrophy (FSHD) is characterized by progressive muscle weakness involving the face, scapular stabilizers, upper arm, lower leg (peroneal muscles), and hip girdle [Wang & Tawil 2016]. Asymmetry of facial, limb, and shoulder weakness is common [Kilmer et al 1995]. Typically, individuals with FSHD become symptomatic in their teens, but age of onset is variable. More than 50% of individuals with FSHD demonstrate findings by age 20 years. Individuals with severe infantile FSHD have muscle weakness at birth. In contrast, some individuals remain asymptomatic throughout their lives. Progression is usually slow; however, many affected individuals describe a stuttering course with periods of disease inactivity followed by periods of rapid deterioration. Eventually 20% of affected individuals require a wheelchair.
Scapular winging is the most common initial finding; preferential weakness of the lower trapezius muscle results in characteristic upward movement of the scapula when attempting to flex or abduct the arms. The shoulders tend to slope forward with straight clavicles and pectoral muscle atrophy.
Affected individuals show facial weakness, with symptoms more pronounced in the lower facial muscles than the upper. Some affected individuals recall having facial weakness before the onset of shoulder weakness. Earliest signs are often difficulty whistling or sleeping with eyes partially open in childhood. Individuals with FSHD are often unable to purse their lips, turn up the corners of their mouth when smiling, or bury their eyelashes when attempting to close their eyelids tightly. Extraocular, eyelid, and bulbar muscles are spared.
The deltoids remain minimally affected until late in the disease; however, the biceps and triceps are selectively involved, resulting in atrophy of the upper arm and sparing of the forearm muscles. The latter results in the appearance of "Popeye arms." In more severely affected individuals, distal upper extremity weakness typically involves the wrist and finger extensors.
Abdominal muscle weakness results in protuberance of the abdomen and exaggerated lumbar lordosis. The lower abdominal muscles are selectively involved, resulting in a Beevor's sign (upward displacement of the umbilicus upon flexion of the neck in a supine position).
The legs are variably involved, with peroneal muscle weakness with or without weakness of the hip girdle muscles, resulting in foot drop.
Sensation is preserved; reflexes are often diminished when the reflex involves weak muscles.
Respiratory dysfunction is relatively uncommon. Individuals who had pulmonary function testing by spirometry showed a restrictive lung disease pattern in 38% [Moreira et al 2017], which was likely a result of expiratory weakness. Respiratory support with noninvasive ventilation is uncommon (1%-3%) [Santos et al 2015].
Other manifestations. Retinal vasculopathy characterized by failure of vascularization of the peripheral retina, telangiectatic blood vessels, and microaneurysms can be demonstrated by fluorescein angiography in 40%-60% of affected individuals [Padberg et al 1995]. Vision is usually unaffected by this particular vascular malformation, but an exudative retinopathy clinically indistinguishable from Coats disease that can result in retinal detachment and vision loss has also been described. Bindoff et al [2006] reported two sisters with infantile-onset FSHD who had tortuous retinal vessels, small aneurysms, and yellow exudates.
Approximately 15% of individuals with FSHD have an abnormal audiogram. An abnormal audiogram was identified in up to 32% of individuals with a large pathogenic contraction of D4Z4 (D4Z4 fragments <20 kb) [Lutz et al 2013].
Both the exudative retinopathy and symptomatic sensorineural hearing loss are seen almost exclusively in individuals with a large pathogenic contraction of D4Z4 (1-3) repeats) or in individuals with early-onset disease [Lutz et al 2013, Statland et al 2013].
A predilection for atrial tachyarrhythmias has been reported in about 5% of cases, but symptoms are rarely experienced [Laforêt et al 1998, Galetta et al 2005, Trevisan et al 2006].
Chronic pain is likely underrecognized in affected individuals, with a prevalence as high as 77% [van der Kooi et al 2007].
Atypical presentations. Clinical variants of typical FSHD in individuals with a pathogenic contraction of the D4Z4 locus in the subtelomeric region of chromosome 4q35 include the following:
Mosaicism for FSHD-associated alleles. Approximately half of de novo cases of FSHD (i.e., affected offspring of unaffected parents) show a mosaic distribution of D4Z4 repeat array lengths in peripheral blood. This mosaicism likely results from a postzygotic array contraction during the first few cell divisions in embryogenesis. In such cases, a proportion of cells have two normal-sized D4Z4 alleles, while the remaining cells have one normal-sized D4Z4 allele and one pathogenic contracted D4Z4 allele [Lemmers et al 2004]. Depending on when in embryogenesis the pathogenic contraction occurs at the D4Z4 locus and the proportion of cells with the contracted D4Z4 repeat, individuals with mosaicism can be affected or asymptomatic. FSHD with somatic mosaicism of D4Z4 array lengths is more penetrant in males than in females [van der Maarel et al 2000].
Genotype-Phenotype Correlations
D4Z4 repeat array contraction size. Evidence-based guidelines published in 2015 recommend that a large pathogenic contraction of D4Z4 (D4Z4 fragments of 10-20 kb) should alert clinicians to the increased likelihood of significant disability, earlier onset of symptoms, and increased likelihood of extramuscular manifestations [Tawil et al 2015].
Allele size explains roughly 10% of variability in phenotype [Mul et al 2018]. A correlation has been reported between the degree of the pathogenic contraction of the D4Z4 locus and the age at onset of symptoms [Zatz et al 1995], age at loss of ambulation [Lunt & Harper 1991], and muscle strength as measured by quantitative isometric myometry [Tawil et al 1996], particularly in affected females [Tonini et al 2004a]. Individuals with a large contraction of D4Z4 (1-3 repeats) have a higher probability of earlier-onset disease and more rapid progression than those with smaller contractions of the D4Z4 locus [Bindoff et al 2006, Hobson-Webb & Caress 2006, Klinge et al 2006, Nikolic et al 2016, Goselink et al 2019]. However, significant variation exists even with small repeats, and others have not been able to confirm a correlation between disease severity and degree of D4Z4 pathogenic contractions [Butz et al 2003].
A study of Italy's National Registry concluded that 76% of early-onset (age <10 years) disease was a result of de novo pathogenic variants. However, neither de novo pathogenic variants nor earlier disease onset were associated with a more severe phenotype [Nikolic et al 2016], contrasting with other studies showing that earlier onset is associated with more severe symptoms [Mah et al 2018, Goselink et al 2019]. Caution must be noted as this correlation may represent an ascertainment bias, where more mild forms of FSHD are detected when inheritance of a known pathogenic variant in a family is suspected.
Mosaicism. The phenotypic severity of individuals with mosaic distributions of one or more array sizes, which is typically less than that of individuals without mosaicism, may reflect the proportion of cells carrying the pathogenic contracted D4Z4 locus in addition to the degree of the contraction of the D4Z4 locus in those cells.
Compound heterozygosity. Two unrelated affected individuals homozygous for a D4Z4 pathogenic contraction were reported by Wohlgemuth et al [2003], suggesting that the presence of two FSHD-associated alleles can be compatible with life. However, both families demonstrated reduced penetrance for FSHD, leaving open the possibility that in other genetic/environmental settings, compound heterozygosity could be a lethal condition. In support of this possibility, the authors report a phenotypic dosage effect in both of the compound heterozygotes, compared to other family members.
Homozygosity.
Tonini et al [2004b] reported an individual homozygous for the contraction on two D4Z4 4A alleles whose clinical phenotype is not more severe than those of some of his heterozygous relatives. Within the same family, the authors also observed a large number of asymptomatic or minimally affected heterozygotes, reflecting the wide range of clinical variability that can occur in a given kindred.