Facioscapulohumeral muscular dystrophy 1 patients participating in the UK FSHD registry can be subdivided into 4 patterns of self-reported symptoms

Highlights

• Facioscapulohumeral muscular dystrophy 1 (FSHD1) has 4 independent presentations.

• Multiparity associates with later muscle weakness in FSHD1.

• Paternal inheritance associates with earlier foot dorsiflexor weakness in FSHD1.

Abstract

Facioscapulohumeral muscular dystrophy (FSHD) is an autosomal dominant incurable skeletal muscle disease. FSHD1 constitutes 95% of cases and is linked to truncation of the D4Z4 macrosatellite at 4q35. In most cases the condition initially presents with facial and proximal weakness of the upper limbs, but over the course of the disease involves lower limb and truncal muscles. Weakness is progressive and frequently asymmetric, which is a hallmark of the disease. Here we performed an analysis of 643 FSHD1 patients in the UK FSHD patient registry, investigating factors affecting rate of onset of 5 major FSHD symptoms: facial, periscapular, foot dorsiflexor, hip girdle weakness, and hearing loss. We found shorter D4Z4 repeat length associated with accelerated onset of each symptom. Furthermore, paternal inheritance of the pathogenic allele was associated with accelerated onset of foot dorsiflexor weakness, while pregnancy and carrying multiple children to term was associated with slower onset of all muscle symptoms. Lastly, we performed clustering analysis on age of onset of the 4 muscle symptoms across 222 patients. We identified 4 clinical presentations of FSHD1. A classical presentation (74%) and 3 facial sparing phenotypes: a mild presentation (5%) with later facial and periscapular involvement, an early shoulder presentation (10%) with accelerated periscapular weakness and an early foot presentation (9%) with accelerated foot dorsiflexor weakness. The mild presentation was associated with longer D4Z4 repeat lengths, while the early foot presentation had a female bias. We note, however that symptom progression differs significantly in these 4 clinical presentations independently of D4Z4 repeat length and gender, motivating investigation of further modifiers of FSHD1 severity.

Introduction

Facioscapulohumeral muscular dystrophy (FSHD) is an autosomal dominant skeletal muscle disease with an estimated prevalence of 12/100,000 [1] for which there is currently no cure. The condition is linked to epigenetic derepression of the subtelomeric D4Z4 macrosatellite region at chromosome 4q35, alongside a permissive 4qA haplotype in cis encoding a polyadenylation signal [2]. Epigenetic deregulation occurs either via truncation of the D4Z4 region to between 1 and 10 repeats (FSHD1) [3] or by mutation in chromatin remodelling genes (FSHD2), mainly SMCHD1 [4], with DNMT3B [5] also recently identified. Each 3.3 kb D4Z4 repeat unit encodes a retrotransposed open reading frame for the transcription factor DUX4, and the FSHD genotype permits expression of stable DUX4 transcripts from the distal-most repeat [3]. DUX4 expression is reported to drive FSHD pathology, via mechanisms that include induction of apoptotic target genes [6,7] and interference with myogenic transcription factors [8], [9], [10].

Clinically, FSHD classically presents with weakness of the facial muscles of the second branchial arch, the shoulder girdle and upper arm muscles, before affecting the lower limb musculature, especially the foot dorsiflexors and hip girdle muscles [11]. Unusually for muscular dystrophies, there is a clear left/right asymmetry in muscle groups affected within an individual [2]. In addition to muscular dystrophy, FSHD is associated with sensorineural hearing loss [12], retinal telangiectasia [13], subclinical cardiac arrhythmias [14,15] and respiratory insufficiency [16].

While the symptoms of FSHD are well described, the presentation is highly heterogeneous. Family members, including even monozygotic twins, often present with diverse clinical courses [17]. Genetic determinants of FSHD1 heterogeneity are well studied and the length of the D4Z4 macrosatellite is inversely related to disease severity [18]. Other genetic modifiers of FSHD1 have been identified, such as SMCHD1 [19,20] and DNMT3B [5] mutations, as well as D4Z4 hypomethylation [21], all of which associate with a more severe disease. While enlightening however, these genetic variants cannot explain the full clinical heterogeneity of FSHD1 [18]. Part of this heterogeneity may be related to gender: females have been reported to display a later penetrance and less severe disease than males in Italian and Brazilian populations [18,22], but an inverse association has been described in a Korean population [23] (though the authors noted a paucity of large D4Z4 repeat length patients in this study). Moreover, whilst asymptomatic FSHD1 carriers display a greater proportion of females [22,24], a Chinese study has recently reported an increased prevalence of female patients among the most severely affected [25]. This gender difference has led to the study of oestrogen exposure in explaining FSHD heterogeneity, and though oestrogens have shown phenotypic rescues in vitro [26], a recent clinical analysis has shown no association between endogenous oestrogen exposure and disease severity [27]. Some heterogeneity in FSHD1 may also stem from patient inheritance pattern, anticipation has been considered a possibility [28], [29], [30], however, it was called into question by a study of a large FSHD1 family from Utah [31]. It has also been suggested that severe cases of FSHD1 tend to be more maternally than paternally inherited, implying a role for imprinting in contributing to disease heterogeneity, however, this finding has not been demonstrated to statistical significance [24]. Finally, somatic mosaicism is well described in FSHD1, occurring in up to 40% of de novo cases [32]. Mosaic FSHD1 patients display a milder phenotype and often go undetected, however male mosaic patients are more likely to be affected than females [32,33].

It has been proposed that FSHD, rather than showing a spectrum of differences in clinical presentation, may comprise a number of distinct clinical phenotypes. Indeed, many studies have reported a facial sparing variant of FSHD as a common atypical presentation which associates with longer D4Z4 repeat length [34], [35], [36]. Others have reported a phenotype manifesting more in the lower limb musculature, with milder involvement of the facial muscles [37]. Importantly, an Italian study developed a tool, the Comprehensive Clinical Evaluation Form (CCEF), to divide the heterogeneity of FSHD into 4 major categories with 9 subdivisions based on clinical presentation, which has demonstrated strong inter-observer agreeablity [38]. More recently the CCEF has been utilised on the UK FSHD Registry to subdivide FSHD1 patients into two groups, one displaying classical FSHD symptoms (68.1% of patients) and a second displaying a facial sparing phenotype (24.1% of patients) [39].

Here we present an analysis of 643 FSHD1 patients participating in the UK FSHD registry, we investigate factors associated with early onset of the FSHD symptoms: facial, periscapular shoulder, foot dorsiflexor and hip girdle weakness, as well as hearing loss. Finally we perform a clustering analysis on age of symptom onset for each of the muscle weakness symptoms described, identifying 4 distinct clinical presentations of FSHD1.