Clinical, genetic, and histological features of centronuclear myopathy in the Netherlands

Abstract Centronuclear myopathy (CNM) is a genetically heterogeneous congenital myopathy characterized by muscle weakness, atrophy, and variable degrees of cardiorespiratory involvement. The clinical severity is largely explained by genotype (DNM2, MTM1, RYR1, BIN1, TTN, and other rarer genetic backgrounds), specific mutation(s), and age of the patient. The histopathological hallmark of CNM is the presence of internal centralized nuclei on muscle biopsy. Information on the phenotypical spectrum, subtype prevalence, and phenotype–genotype correlations is limited. To characterize CNM more comprehensively, we retrospectively assessed a national cohort of 48 CNM patients (mean age = 32 ± 24 years, range 0–80, 54% males) from the Netherlands clinically, histologically, and genetically. All information was extracted from entries in the patient's medical records, between 2000 and 2020. Frequent clinical features in addition to muscle weakness and hypotonia were fatigue and exercise intolerance in more mildly affected cases. Genetic analysis showed variants in four genes (18 DNM2, 14 MTM1, 9 RYR1, and 7 BIN1), including 16 novel variants. In addition to central nuclei, histologic examination revealed a large variability of myopathic features in the different genotypes. The identification and characterization of these patients contribute to trial readiness.

data concerning the different genetic subtypes.
To date, CNM epidemiological reports provide limited incidence and prevalence data. A recent integrated model utilizing available literature has been proposed to obtain a better estimate of overall CNM patient numbers by age, causative gene, severity, and geographic region. 18 This model calculated a CNM incidence higher than the current estimates. Therefore, knowledge on the actual prevalence in a geographically defined region is essential. Our aim therefore was to obtain epidemiological information regarding the Dutch CNM cohort, and to report their clinical, genetic, and histological features.
This could also facilitate CNM trial recruitment in the future.

| METHODS
This retrospective, cross-sectional study was conducted at the Radboudumc Neuromuscular Centre, Nijmegen, in collaboration with the Dutch Neuromuscular Centre. All CNM patients had been referred to our center between 2000 and 2020. The study was approved by the local ethics committee (Protocol 2017-4022), and all participants or, as appropriate, their parents provided informed consent.

| Patients
Inclusion criteria were (1) a (likely) pathogeneous mutation in one of the genes associated with CNM: MTM1, DNM2, BIN1, and RYR1, and a clinical phenotype of a myopathy; or (2) a clinical or histopathological diagnosis of CNM and genetic confirmation of first degree affected family member. In the first group, histological confirmation was not required since histopathological confirmation is not performed in all cases anymore. We included subjects without genetic confirmation in the second group since some XL-MTM patients had passed away before the diagnostic availability of genetic testing. Patients were divided into groups per genotype, including a distinction between male patients and female manifesting carriers with an MTM1 variant.

| Data collection
CNM patients of all ages were identified through four routes: (1) the (Paediatric) Neurology Outpatient Clinic at the Radboudumc; (2) the Genetics Department at the Radboudumc; (3) the Dutch Patient Organization Spierziekten Nederland; and (4) (Paediatric) Neurologists of the Dutch Neuromuscular Centre. This is estimated to provide a high coverage (>80%) since the Radboudumc is the national referral center for congenital myopathies and is acknowledged as such by the other Dutch Neuromuscular Centres. Hence, CNM patients are generally referred to the Radboudumc at least once as part of clinical management, for registration and for participation in studies. Clinical data from the patients were stored in our electronic patient file system, and systematically extracted by the researchers (S.R. and D.Z.). Data were pseudonymized and stored in a Castor database.

| Data collection, clinical features
We collected information regarding family history, medical history, clinical features, and ancillary investigations. Clinical features were grouped into motor symptoms (signs of delayed gross motor development, muscle weakness, muscle atrophy, and hypotonia), myalgia and cramps (myalgia, muscle cramps, and stiffness), facial and bulbar symptoms (facial weakness, abnormal ocular movements, dysphagia, and dysarthria), respiratory symptoms and cardiac involvement. Age at onset was retrieved from the medical file or estimated based on the history (congenital: 0 years; early childhood: ± 3 years; childhood: ±5 years). Age at diagnosis was determined by the time point where either a histological or suspected genetic diagnosis of CNM was made.
Reference values for CK (in IU/L) used in our medical center are ≤710 for neonates, ≤295 for infants, ≤230 for children, ≤270 and ≤ 123 for male and female adolescents, ≤170 and ≤ 145 for men and women.

| Data collection, genetic findings
Results of genetic testing previously performed as part of the diagnostic procedure were retrieved from the medical files. In most patients, Sanger sequencing was performed until the introduction of wholeexome sequencing with muscle panel analysis in 2013. 19 The variants were classified as pathogenic, likely pathogenic, variant of uncertain significance (VOUS), likely benign or benign, according to the ACMG classification. 20

| Histologic features
Results of muscle biopsy were retrieved from the medical files. Muscle biopsy samples were frozen and stored at À80 C, specimens were processed for routine histological procedures. The majority of samples were processed with several enzyme histochemical staining, including hematoxylin and phloxin (HPhlox), nicotinamide adenosine dinucleotide (NADH), succinate dehydrogenase (SDH), cytochrome C oxidase (COX), Gömöri trichrome, and ATPase 4.2, ATPase 4.6, and ATPase 10.3. All available muscle biopsy slides were reviewed by the pathologist at our center (B.K.) to confirm and/or amend the findings described in the clinical report. We paid particular attention to the following histological features: increased fiber size variability, type I fiber predominance, increased internal and central nuclei (>5%), fatty or connective tissue, nuclear clumps, and radial sarcoplasmic strands (RSS).

| Statistical analysis
Data were analyzed using IBM SPSS Statistics software (version 25. Armonk, NY: IBM Corp.). Descriptive statistics used were mean with SD (n ± SD) and frequencies with percentages (n[%]).

| Patients
We identified 50 patients with a CNM diagnosis in the Netherlands.
Two patients with an additional diagnosis of nemaline myopathy were excluded. We retained two patients with a DNM2 variant and a mixed myopathy-neuropathy phenotype (neurophysiologically and histologically classified).
Seven of the 10 male XL-MTM patients had passed away (30% survival, mean age at death was 7 ± 15 years). Five out of seven XL-MTM patients died shortly after birth because of respiratory failure. One patient died at the age of 11 due to respiratory failure after recurrent respiratory tract infections and pneumothorax. The oldest XL-MTM patient (41 years) passed away because of sudden cardiac failure, this patient was reported previously (III.3). 21 Survival in the other genotypes was 100%. Of the DNM2-CNM patients, 61% had family members with a diagnosed myopathy; in MTM1 and RYR1 patients this was the case in 50% and 44%, respectively. In 13% of the patients stillbirths in the family were reported; these were mainly MTM1 families. Percentages per genotype are given in Table 1.

| Clinical features
Patient characteristics are summarized in Table 1. Overall mean age was 32 ± 24 years, ranging from 0 to 80 years. All male MTM1 patients had congenital onset, most RYR1 and DNM2 patients had onset in childhood. Age at onset was highly variable for BIN1 patients.
Age at onset in the two DNM2 patients with a mixed phenotype was 17 and 46 years, respectively. The delay between age at onset and age at diagnosis was only short for male MTM1 patients (2 ± 4 years) and longest for female MTM1 carriers (29 ± 24 years). Patient age, including the age at onset and diagnosis, is depicted in Figure 1C.
Clinical features for each genotype are illustrated in Figure 2, more detailed information is listed in Table 2. Ambulatory status was highly variable ( Figure 1B). None of the male MTM1 patients achieved independent ambulation, and none of the BIN1 patients were dependent on assistance or a wheelchair. Two patients used disease modifying medicine (nutritional supplements); one BIN1 patient and one MTM1 patient used pyridostigmine. In both patients, this had no effect. Two RYR1 patients used acetylcysteine, but without significant effects. All CNM patients had at least one motor symptom (signs of delayed gross motor development, muscle weakness, muscle atrophy, and hypotonia), except for two BIN1 patients whose main clinical features were myalgia and muscle cramps. Myalgia and cramps were reported by many BIN1 patients and female MTM1 carriers, while fatigue and exercise intolerance were common in all groups of CNM.

| Creatine kinase
Creatine kinase (CK) levels were available in 29 Table 3, and more detail is listed in Table S1. Ten out of eleven DNM2 variants were missense variants and one in-frame deletion was reported. Two DNM2 variants were identified as de novo after segregation analysis (c.596G > A and c.1105C > T) and

| Genetic findings
are associated with a more severe phenotype than the other DNM2 variants in this study. One DNM2-CNM patient had a somatic mosaicism   Typical histopathological features seen in our cohort are displayed in Table 4 and shown in Figure 3.  24 Dutch policy has a lot of influence on this.

| DISCUSSION
Because of the routine ultrasound examination around 20 weeks of pregnancy, termination of pregnancy is more common. 25 In addition, euthanasia for neonates has been allowed in the Netherlands since 2005. 26 Moreover, we have performed a retrospective study most likely confounded by selection bias. Therefore, we might not have reported the totality of prevalent patients but rather the majority of them, reaching an estimate of the prevalence in the Netherlands.
We will discuss the most important findings for each genotype in the following section.

| DNM2-CNM
Most patients had symptom onset in childhood or adolescence onset and respiratory insufficiency, as reported in some other patients with de novo DNM2 variants. 27 Some DNM2 variants have been described to be more frequent in late-onset phenotypes, 23 however, this was not observed in our cohort. The delay between onset and diagnosis was long (14 ± 14 years), probably related to the late onset, relatively mild symptoms, and slow progression, 9 Table 3. DNM2 plays an important role in mutated cytoskeleton and membrane proteins, both involved in CNM and Charcot-Marie-Tooth disease (CMT). In the CNM phenotype it affects mainly skeletal muscles and in CMT it mainly affects peripheral nerves. 28 These mixed phenotypes, although not common, have been well-recognized in DNM2-CNM and might be considered for inclusion in future CNM trials. 29 Mosaicism detected in one patient has only been reported once, 30 also with a mild phenotype (no hypotonia, respiratory or feeding difficulties at birth). Two of 11 mutations were de novo, which is not uncommon in DNM2-CNM. 27,30 Connective tissue replacement was the most prevalent muscle biopsy finding in DNM2 patients. This could be an age effect, since four DNM2 patients had their muscle biopsy at ages >35 years. Variable degrees of radial sarcoplasmic strands were present in six patients; both features have been described previously. 9,31

| XL-MTM male patients and female carriers
Male XL-MTM patients mostly have congenital onset and are severely affected. Cardiac involvement was reported in one male patient in our cohort and not in female carriers, although cardiomyopathies have been previously also described in MTM1 carriers. 32 There is also one report of a mildly affected male with XL-MTM who developed a cardiomyopathy in early adulthood. 33 Seven patients had passed away during the retrospective study window of 20 years. The severe phenotype in males may be the reason for the only short delay between onset and diagnosis (2 ± 3 years). Fatigue and exercise intolerance were common in all groups of CNM, except for the male MTM1 patients (30%). This is probably biased by the early death of half of these patients, and also because of their severely reduced mobility. Female manifesting carriers generally have a later onset and a less severe phenotype. This, and the fact that disease manifestation in carriers has been neglected for a long time, has probably contributed to the long diagnostic delay (29 ± 24 years). Our cohort included four affected XL-MTM carriers, one of them previously published by Biancalana et al. 16,34 The female phenotype covers almost the full disease spectrum in males, with wheelchair dependency (50%), respiratory insufficiency (50%), and facial and bulbar symptoms (25%, including extraocular muscle involvement).
Nine variants in the MTM1 gene were identified, including one entire gene deletion which has been reported before. 35 Type I fiber predominance in muscle biopsies was not reported in female MTM1 carriers, and nuclear clumps did not occur in XL-MTM males. In one female MTM1 carrier, necklace fibers were observed in the biopsy. Central nuclei would be expected in all CNM muscle biopsies.

| RYR1-CNM
However, the muscle biopsy of several patients with a pathogenic variant in one of the CNM genes (n = 9) showed no centralized nuclei at all. This is likely to be related to our current sequence of ancillary investigations-first genetic testing and subsequently muscle biopsy 4,41 -in contrast to the historical diagnostic approach where genetic testing was mainly prompted by suggestive features on muscle biopsy. Fatty or connective tissue and nuclear clumps were less common features in our CNM cohort. Nuclear clumps can occur in long-standing neurogenic or myopathic conditions, but also in other genetic muscle disorders such as myotonic dystrophy type 2 (DM2). 42,43 Fibrosis and increases in fatty tissue have previously been reported in DNM2-and RYR1-CNM with varying frequency and severity. In our cohort, fibrosis was observed in 36% of our DNM2 patients. 30,41 Although RSS were described in previous studies focusing on CNM patients with mainly autosomal-dominant inheritance, we found this feature in only 10 of our biopsies, including six biopsies from patients with a DNM2 mutation. 41 The range of patient age when the muscle biopsy was performed was wide (0-66 years), with a likely effect on the observed variability of features, as some abnormalities in the muscle are not always visible at a younger age and may only develop over time. Other features, for example type I fiber predominance and the presence of connective tissue that are known to increase as part of the aging process, 44 may be more prominent in biopsies taken at an older age.
A limitation of this study is the retrospective study design. Data were collected by medical chart review, preventing a more detailed description of the phenotype of this cohort. Another constraint is the small size of the different genetic subgroups, resulting in difficulties in making comparisons between the different genotypes and with regards to the wider applicability of our findings. In addition, previously reported genotype-phenotype variability and intrafamilial variability have to be taken into account. 38,45,46 The next step will be to assess these patients prospectively to collect natural history data. A recent natural history study in Belgium and France focusing on CNM patients with a DNM2 mutation has provided reliable natural history data and sensitive outcome measures. 17  In conclusion, to our knowledge, this is the first detailed study in the Netherlands to report the complete identified Dutch CNM cohort.
The identification and characterization of these patients contributes to trial readiness.

ACKNOWLEDGMENTS
We would like to thank all neuromuscular neurologists and clinical geneticists who contributed patients to our study cohort.
Furthermore, we thank Spierziekten Nederland for dissemination of information about this study. We also thank Daan Zegers for his contribution in data collection, and Stéphanie Hoffmann, Chris Freitag, and Leen Thielemans from Dynacure Inc. for contributing to the discussion on future studies concerning CNM. Several authors of this paper are members of the Netherlands Neuromuscular Center (NL-NMD), and the European Reference Network for rare neuromuscular diseases (EURO-NMD).