Bethlem Myopathy Phenotypes and Follow Up: Description of 8 Patients at the Mildest End of the Spectrum

Abstract

The classical phenotypes of collagen VI-associated myopathies are well described. Little is known, however, about the progression of patients at the mildest end of the clinical spectrum. In this report, we describe the clinical findings and the results of MRI, muscle biopsy, collagen VI expression in cultured skin fibroblasts and genetic tests of a series of patients with Bethlem myopathy. Our series highlights the existence of mild presentations of this disorder that progresses only slightly and can easily be overlooked. Analysis of the genetic studies suggests that missense mutations can be associated to a milder clinical presentation. Muscle MRI is extremely useful as it shows a pathognomonic pattern in most patients, especially those with some degree of muscle weakness.

INTRODUCTION

Collagen VI myopathies are caused by mutations in the genes encoding the collagen VI α-chains: COL6A1, COL6A2 and COL6A3. Most authors recognize the existence of a phenotypic spectrum in which classical Ullrich congenital muscular dystrophy (UCMD) represents the most severe end while Bethlem myopathy (BM) includes a high diversity of milder forms [1, 2]. UCMD shows a congenital onset by definition and, in most cases, ambulation is never acquired or it is lost before adulthood. In BM, however, the age of onset varies and while some patients have a congenital presentation, others present the disorder later in adulthood. Nevertheless, symptoms usually appear in early childhood, with a mild motor developmental delay, difficulties running or doing sports, and proximal weakness or contractures. Classically, BM patients present a combination of proximal muscle weakness and flexion contractures of elbows, wrists, fingers and ankles [3, 4]. Two phenotypic variations of BM have recently been described: a limb-girdle muscular dystrophy (LGMD) presentation [5], in which muscle weakness predominates and contractures are either absent or minimal, and a “myosclerosis” phenotype [6] which presents with severe contractures without significant weakness.

While the course of disease is well established in UCMD, it is less extensively studied in BM, particularly regarding the more benign pauci-symptomatic phenotypes. Progression in BM is usually very slow but about two-thirds of patients may require some ambulatory aid after the sixth decade of life [7].

In this report, we describe the clinical features and progression in seven BM patients with a benign phenotype and compare them to a patient with classical BM.

METHODS

Patients and data

The study group consisted of 8 patients with collagen VI-associated myopathy followed-up at the Neuromuscular Disorders’ Unit of Hospital de la Santa Creu i Sant Pau in Barcelona. The expression pattern of COLVI in cultured fibroblasts and molecular diagnosis of the patients were obtained at the Pitié-Salpêtrière Hospital in Paris. Clinical records were retrospectively reviewed for demographic data, such as age and gender, as well as for clinical features and other disease-related elements: age at onset, follow-up duration, early and late clinical features (namely muscle weakness and its pattern, contractures or hyperlaxity and their location, scoliosis and skin changes). All patients signed an informed consent and the study was approved by the Ethics Committee.

Age and clinical examination including muscle balance and presence of contractures at last visit is reported for all patients. The modified Gardner-Medwin and Walton clinical severity score was calculated at every visit: grade 0 = hyperCKaemia, all activities normal; grade 1 = normal gait, unable to run freely, myalgia, atrophy; grade 2 = unable to walk on tiptoes, waddling gait; grade 3 = evident muscle weakness, stepping gait, climbing stairs with banister; grade 4 = difficulty in rising from the floor, presence of Gowers’ sign; grade 5 = unable to rise from the floor; grade 6 = unable to climb stairs; grade 7 = unable to rise from a chair; grade 8 = unable to walk without assistance; grade 9 = unable to eat, drink or sit unassisted.

Muscle MRI

We used a 1.5T Philips Achieva XR MRI to study our patients. Axial sections (10-mm slices) were obtained in a T1-weighted spin-echo sequence (repetition time 600–700 msec, echo time 30 msec). MRI findings were considered typical if concentric involvement of vastus lateralis and central high signal in rectus femoris were present [8, 9].

Muscle biopsy

Histological analysis included immunohistochemical staining for collagen VI as well as the following regular stainings: hematoxilin and eosin, modified Gömöri trichrome, nicotinamide adenine dinucleotide (NADH), succinic dehydrogenase (SDH) and adenosine triphosphate catalytic enzymes (ATPase). Histological features were deemed dystrophic if there was increased connective tissue and fatty replacement, with or without necrotic fibres.

Collagen VI Immunolabeling on cultured skin fibroblasts

Collagen VI and fibronectin co-immunolabeling was carried out on confluent cells cultured in the presence of 0.25 mM ascorbic acid as previously reported by Hicks et al. (2008), using triton X-100 to permeabilize the cells in order to detect intracellular retention of COLVI [10]. Slides were observed with an Axioplan 2 microscope (Zeiss) equipped with a HBO 100 mercury lamp (Zeiss). Images were captured using the Metavue software (Molecular Devices).

Genetic analysis

Total RNA was extracted from confluent fibroblast cultures using the BioRobot EZ1 Workstation (Qiagen). After retrotranscription, mRNA of the COL6A1 (NM_001848.2), COL6A2 (NM_001849.3) and COL6A3 (NM_004369.3) genes was amplified in overlapping fragments and sequenced on the ABI3730 system (Applied Biosystems, Foster City, CA). Mutations detected at mRNA level were confirmed at the gDNA (extracted from blood) level. The pathogenicity of the non previously described mutations was assessed as previously described, depending on the affected domain in the protein, the nature of the mutation (nonsense or insertion/deletion vs missense), in silico analyses with the Alamut (Interactive Biosoftwares) software and cosegregation with the disease in available relatives [11]. Sup. Table 1 describes the genetics findings in our series of patients.

RESULTS

Clinical features and disease evolution

Description of a classical BM case

Patient 1 (mutation c.801+1del in the COL6A2 gene, p.Cys246_Lys267del), is a 41 year-old woman who presented decreased fetal movements and then a mild motor developmental delay. In early childhood she had frequent falls and some difficulty standing up from a chair or climbing up stairs. By the age of eight, significant elbow and heel contractures were noted. Both muscle weakness and joint contractures showed progressive worsening and, at the age of 30, she needed a cane to walk short distances and a wheelchair for long distances (Walton score: 8 (Fig. 1). Clinical examination at the last visit showed contractures in multiple joints, particularly severe in elbows, heels and fingers, as well as scoliosis (Fig. 2) and severe proximal weakness of the lower limbs (MRC was 3/5 for hip extension and flexion, and 4/5 for hip adduction and abduction). Mutation c.801+1del was not previously described. Whereas in silico analysis by the Alamut software predicted the use of another adjacent cryptic splice site, mRNA analysis in this patient showed skipping of exon 5 and the deletion of 22 amino-acids, including the beginning of the triple helical domain.

Description of the milder phenotype cases

Patient 2 (mutation c.6239G>A in the COL6A3 gene, p.Gly2080Asp) is a 52 year-old woman who presented with frequent falls in her early childhood. She later started developing muscle weakness, particularly in shoulder and pelvic girdles, but also in the anterior and lateral leg compartments. Weakness progressed very slowly and at last visit it mainly involved the psoas and the glutei (MRC 4/5). She had no significant functional disability and no joint contractures were identified (Walton score: 3), but we found keratosis pilaris involving the arms and the thighs.

Her daughter, patient 3, showed a mild motor developmental delay and later developed elbow and heel contractures. These findings did not cause significant limitations in her daily life (Walton score: 1) at age 23 when she was visited for the last time. Clinical examination at last visit showed no muscle weakness and only mild contractures in elbows and heels. Mutation c.6239G>A in the COL6A3 gene is located in exon 17, in the triple helical domain of the protein. It was previously described in a severe BM patient [12].

Patient 4 (mutation c.1076G>A in the COL6A1 gene, p.Gly359Asp) is a 59 year-old woman who mentioned some gait clumsiness during childhood. During the fourth decade she started noticing moderate muscle weakness which, at last visit, mainly affected both quadriceps (MRC 4/5), without significant gait impairment (Walton score = 3). There were no joint contractures other than a mild flexion contracture in fingers. Keratosis pillaris involving the thighs and the forehead was also observed in this case. Her daughter, patient 5, had a mild motor developmental delay and showed joint hyperlaxity during early childhood. At the age of 27, she had no muscle weakness and presented only a very mild heel contracture, remarkably retaining hyperlaxity in fingers. She had no significant limitations in her daily life activities (Walton score = 0). Mutation c.1076G>A in the COL6A1 gene was not previously reported. The Alamut software, predicted this new mutant to be pathogenic (see table in supplemental data). Moreover, this mutation affects a glycine codon from the triple helical domain, which is usually disease-causing in a dominant transmission pattern [13].

Patient 6 (mutation c.877G>A in the COL6A1 gene, p.Gly293Arg) is a 38 year-old woman who reported frequent falls and some weakness during childhood. The disease had a very slow progression and at the last visit her clinical examination showed weakness in psoas (MRC = 3/5) and in shoulder girdle and long finger flexors and extensors (MRC = 4/5). She also developed heel contractures, which preclude her heels from touching the floor during walk producing some difficulties to walk (Walton score: 4). Her daughter, patient 7, had similar presenting symptoms. At the age of 18, she showed evident heel contractures, which still allow a full plantar contact with the floor, and she had only a mild weakness in psoas (MRC 4+/5), without significant functional impairment (Walton score = 1). This mutation affecting a glycine codon of the triple helical domain of the α1(VI) chain has been previously identified in the literature [13] and in one family of the Pitié-Salpêtrière Laboratory (personal communication).

Patient 8 (mutation c.1002-2A>C in the COL6A1 gene, p.Gly335_Asp352del), is a 36 year-old woman who reported at first visit a mild motor developmental delay and frequent falls during early childhood. Later on she developed pelvic girdle muscle weakness and significant joint contractures, especially affecting elbows. At last visit, clinical examination showed weakness of psoas and glutei (MRC = 3/5) as well as mild weakness in limb girdle muscles and in long finger flexors and extensors (MRC = 4/5). She presented evident contractures in elbows and also in finger flexors –coexisting with hyperlaxity of fingers –and heels (Fig. 1). At last visit the patient could walk long distances without problems but she needed the bannister to climb up stairs (Walton score = 4). This 1003-2A>C mutation caused exclusion of exon 14 at mRNA level. This mutation was not previously reported, although another mutation at this locus was previously described (c.1003-2A>G) in ClinVar (ClinVar ID: 210745) and in our laboratory.

All these patients had some typical skin changes: most showed areas of follicular hyperkeratosis and some had keloid scars.

Muscle MRI, muscle biopsy and study of dermal fibroblasts

Muscle MRI was performed in all patients after a mean disease duration of 30.5 years (between 13 and 50). Except for patient 5 who had no muscle weakness, all patients presented typical collagen VI-associated MRI findings (Fig. 2). Only two patients had a clear relative sparing of sartorius, gracilis and adductor longus, and three showed a high signal rim between soleus and gastrocnemius. We found considerable differences between patient 1 and the other 7 patients concerning the degree of fatty infiltration in the muscles. As shown in Fig. 1, the degree of fatty infiltration is clearly higher in patient 1 than in the other patients.

All five patients who had a muscle biopsy (patients 1, 2, 4, 6, and 8) displayed mild dystrophic changes. Collagen VI immunostaining was normal in these five cases. On the other hand, collagen VI secretion by dermal fibroblasts was abnormal in all the cases, with severe to complete absence of secretion, associated with intracellular retention detected upon permeabilization of fixed cells (Fig. 3).

DISCUSSION

We describe a series of 7 patients with a very mild phenotype of Bethlem myopathy. All these patients have been followed-up for a long period of time (8 to 20 years) and we have not observed a significant decline in overall motor function affecting their activities of daily living. We want to emphasize that patients 2 and 4, who were in their sixth decade of life at their last visit, continued to have only mild weakness without significant problems to walk. Previous reports of patients with Bethlem myopathy have described a significant functional decline occurring in the fourth and fifth decades of life [2, 14] and a high proportion of patients requiring ambulatory aid after the sixth decade [7]. However, it has also been recently reported that missense mutations in any of the COL6A1, COL6A2 and COL6A3 genes are associated with a more benign progression [14]. All patients in our series harboured a missense mutation, except patients 1 and 8, probably explaining their mild clinical presentation compared to classical patients with Bethlem myopathy. Patient 8, who also had a mild phenotype, harboured the 1003-2A>C mutation in the COL6A1 gene producing skipping of exon 14. Skipping of exon 14 has been associated with worse progression, but it has also been found in patients with a mild phenotype [10]. In contrast, Patient 1, who had a more severe clinical evolution, harboured the c.801+1del that skipped exon 5 of the COL6A2 gene, associated with a severe phenotype[11, 14].

Although most of our patients presented at least some of the typical clinical features of Bethlem myopathy, we observed considerable phenotypic heterogeneity, even within the same family (patient 2 had a LGMD phenotype while her daughter, patient 3, had a contracture predominant phenotype). We found several less often described clinical findings in our patients such as predominant weakness of quadriceps muscle, weakness of long finger flexors and extensors, and a striking finger hyperlaxity coexisting with contractures. Furthermore, in our group of patients we identified two cases with pure weakness without joint contractures which are known as “limb-girdle muscular dystrophy” phenotype [2, 5] and one patient who did not have any weakness but only some mild joint contractures, who could be classified as mild “myosclerosis” phenotype [2, 6]. These different clinical presentations confirm previous descriptions of high clinical variability across the collagen VI-associated myopathy phenotypic spectrum [1, 14].

Patient 5 is intriguing: she carries the same mutation that her symptomatic mother, but she can be considered a paucisymptomatic patient as she has only distal hyperlaxity and mild Achilles tendon contractures. Although we do not have a clear explanation about her normal muscle balance and muscle MRI, we would like to point out that her mother had not any clinical problem until the fourth decade of life. We should continue following-up this patient to observe whether muscle weakness do appear in that decade of life.

In our series, the muscle biopsy with collagen VI immunohistochemistry had a very poor diagnostic yield, concurring with previous reports [10]. In contrast, but in agreement with other authors [15], immunolabeling of skin fibroblasts displayed altered secretion in all patients tested, thereby confirming the pathogenic nature of all mutations. Regarding muscle MRI, in all our patients with muscle weakness we found the highly characteristic features of collagen VI-associated myopathy [8, 16], particularly a central area of high signal in rectus femoris, concentric involvement of vastus lateralis and relative sparing of sartorius, gracilis and adductor longus. In fact, muscle MRI seems to be an invaluable diagnostic tool for this disease before a compatible clinical phenotype because its pattern is pathognomonic and, compared with skin fibroblasts culture, it has the advantages of being non-invasive and easilyaccessible.

As concluding remarks, Bethlem myopathy displays a phenotypic spectrum with heterogeneous severity and sometimes atypical phenotypes. Our series describes the benign evolution of one end of this spectrum. Muscle MRI seems to be a very useful tool for the diagnosis of this disease.

CONFLICT OF INTEREST

The authors have no conflict of interest to report.