The clinical, myopathological, and molecular characteristics of 26 Chinese patients with dysferlinopathy: a high proportion of misdiagnosis and novel variants

Dysferlinopathy is muscular dystrophy caused by the deficiency of dysferlin protein coding by the DYSF gene. Pathologic variants in DYSF lead to different clinical phenotypes, mainly including LGMD R2 and MM [20]. LGMD R2 mainly affects the proximal lower extremity muscle tissue in the youth; as the disease progresses, the scapular girdle and upper extremity muscles may also be affected, but the symptoms are mild; the neck and hand muscles are generally spared [21]. This disease is the second most common LGMD in Europe and Japan but is underdiagnosed in China previously [22]. MM is an adult-onset disorder characterized by early-onset gastrocnemius weakness, which is also accompanied by an increase in serum CK concentration [23]. But the onset of MM was found to be earlier than that of LGMD R2 in the Italian population [24]. Other phenotypes associated with dysferlin deficiency have also been identified, including distal anterior myopathy (DACM) (also known as distal tibial onset distal myopathy), and proximal-distal phenotype (PD) (this phenotype may be a proximally rapidly progressive MM) [25, 26] and asymptomatic hyperCKemia. We didn’t observe DACM in our cohort with the highest proportion of LGMD R2, which was consistent with the domestic sample [22] and foreign studies [25].

LGMD R2 is easily misdiagnosed as inflammatory myopathies, especially polymyositis (PM), which is very similar to LGMD R2 in clinical manifestation and muscle pathology. Both LGMD R2 and PM exhibit proximal muscle weakness and significantly elevated muscle enzymes and may show infiltration of immune cells in muscle pathology, but the treatment is different between them [21]. PM is an immune disease that responds well to hormone therapy [27], but glucocorticoids have been reported to exacerbate muscle weakness in LGMD R2 patients, and the damage to the muscle is irreversible [28]. We recommend that it is important to rule out dysferlinopathy before starting corticosteroid courses. Studies have shown that injection of glucocorticoids into the patient’s muscle cell membrane can damage the membrane stability [29], which may lead to an increase in the CK value, and this instability also exacerbates the lack of fibrillin repair capacity [29]. For LGMD R2, however, the focus is on early symptomatic treatment and appropriate exercise, which can slow disease progression and improve motor function. Therefore, the early diagnosis of LGMD R2 is closely related to the prognosis of patients. In our cohort, 5 patients showed inflammatory cell infiltration in muscle pathology. Twelve LGMD R2 patients (66.7%) in this group were misdiagnosed as polymyositis before biopsy, and 10 of them had received corticosteroid therapy, which may affect the level of CK. The CK level of patient 7 still repeatedly increased after corticosteroid therapy; in addition to corticosteroids, Patient 1 also took traditional Chinese medicine, but the CK level stayed at a high level (35.3 times the normal scope). Besides, 2 MM patients (50%) were misdiagnosed as polymyositis and had previously received corticosteroid therapy before the biopsy. The two diseases can be differentiated by analyzing the expression of dysferlin in muscle tissues by IHC. For patients with LGMD R2, IHC analysis showed a lack of dysferlin in the involved muscle fibers, and MHC-I results were negative or low [30].

Besides inflammatory myopathies, dysferlinopathy patients with a history of exercise intolerance or asymptomatic hyperCKemia [25, 31] may be misdiagnosed as metabolic myopathy. CK levels fluctuated in patients with metabolic myopathy but usually stayed at a high level in patients with dysferlinopathy, except in the late stage of the disease because of muscle wasting. In primary hospitals in China, patients with asymptomatic hyperCKemia at first tended to visit the department of cardiology or general medicine. Three patients of our cohort with asymptomatic hyperCKemia (75%) were misdiagnosed as viral myocarditis. Patients with viral myocarditis usually presented with chest pain, shortness of breath, fever, fainting, and palpitations. CK-MB is one of the diagnostic indicators of viral myocarditis. It has been reported that the CK-MB levels of children with viral myocarditis in the acute phase are about 3 times that of the normal control group [32] and CK levels can slightly increase, while in dysferlinopathy the CK but not the CK-MB levels usually increased significantly. Muscle damage in dysferlinopathy also results in an elevated level of liver enzymes (for example, ALT and AST) which may be confused with liver disease. Our study found that patients with asymptomatic hyperCKemia were easily misdiagnosed as myocarditis (75%) and liver disease (25%), indicating insufficient recognition of this disease in primary hospitals in China, especially for doctors of internal medicine.

In addition to the typical dystrophic features, 2 patients in this dysferlinopathy group presented with several ragged red fibers (RRF) seen on histopathological MGT staining. Previous reports have also documented mitochondrial abnormalities in some patients with dysferlinopathy, in which there is an accumulation of subsarcolemmal mitochondria in muscle fibers, including one patient with RRF and paracrystalline mitochondrial inclusions [33, 34]. The mechanisms for the formation of mitochondria abnormalities observed in muscle pathology are undefined. Previous research showed that dysferlin has a ferlin Ca²⁺ domain with a variable affinity for Ca²⁺ and helps regulate the cytoplasmic Ca²⁺ [33], which becomes abnormally high in the absence of dysferlin. Doug M. Turnbull [35] et al. suggested that dysferlin gene variants increased the concentration of cytoplasmic Ca²⁺, leading to mitochondrial aberrations. However, not only do mitochondria regulate cytoplasmic Ca²⁺ levels, but the abnormal elevation of Ca²⁺ would also affect mitochondria, and calcium influx into the cytoplasm would lead to fragmentation of the mitochondrial network and increase mitochondrial fission [36]. Further studies are needed to investigate the mechanisms which may explore potential therapeutic strategies for dysferlinopathy.

Decreased expression of dysferlin supports the diagnosis of dysferlinopathy, but it should be noticed that the expression of dysferlin may also decrease secondary to deficiency of other related genes, such as CAPN3 (causative gene for LGMD R1), so genetic analysis remains the definitive diagnostic criterion for dysferlinopathy. A wide range of DYSF variants has been identified, including missense, nonsense, frameshift deletions/insertions, splice variants and large exonic deletions [9]. Missense variants accounted for nearly half of the study in this cohort, and a comparison of the variant spectrum with a large French cohort [37] suggested a possible difference, with exonicframeshifting (18% vs. 30%) and splice (10% vs. 16%) being lower in our cohort, while missense (42% vs. 34%) and nonsense (25% vs. 20%) were more common in our cohort which also had 2 exon duplication variants. The top three outcomes in the world patients dataset were missense (42.3%), splicing (13.7%), and frameshift (11.1%) [19]. Chinese patients showed a similar pattern of variant sequence distribution as patients worldwide[17‐25, 39] (Table 2). Most reported pathogenic variants for dysferlinopathy are single nucleotide variants and small insert/deletions [37], but large exonic deletions and duplications have also been described [38]. Pathogenic variants identified in this study consist of 4 (4/27) canonical-splice, 10 (10/27) nonsense, 6 (6/27) missense, and 7 (7/27) frameshift variants. Most of the variant types are single nucleotide variants consistent with previous reports. Two variants were identified previously in Chinese patients: c. 937 + 1G > A6, splicing, and c.3112 C > T (p. Arg1038Ter) [20] were also retrieved in our study. The c.3112 C > T (p. Arg1038Ter) was identified in more than one patient, and the patients who had the same variant came from different provinces, suggesting the variant may be recurrent in China. The c.2997 G > T (p. Trp999Cys) variant was the most common variant in the LGMD group in the previous study [39], but no c.2997 G > T (p. Trp999Cys) variant was observed in our cohort. Previous studies involving other genotypes have shown no observed relationship between reduced expression levels and the severity of clinical symptoms. Therefore, the effect of genotype on protein levels, and thus on phenotype, should be further investigated for each variant [39]. In addition, we also identified 15 novel variants, expanding the molecular spectrum of dysferlinopathy, and highlighted the high proportion of novel variants in Chinese patients with dysferlinopathy.

In summary, we reviewed the clinal and molecular characteristics of 26 Chinese patients with dysferlinopathy. This study showed clinical and genetic heterogeneity of dysferlinopathy and a high proportion of novel variants in the Chinese population. We also found a high rate of misdiagnosis of dysferlinopathy in primary hospitals, suggesting more attention should be paid to improving the knowledge and awareness of this disease.