Centronuclear myopathy--morphological relation to developing human skeletal muscle : a clinicopathological evaluation.
Centronuclear myopathy (CNM), an uncommon condition, is one of the congenital myopathies. It is believed to arise as a result of maturational arrest, with persistence of myotubes postnatally. However, denervation being the basic disease process and its possible influence on central nervous system causing defect in nuclear migration has also been postulated. Keeping in view these existing controversies, we have studied 17 cases of CNM (neonatal - 1, childhood - 13, adulthood - 3) during the last twelve and a half years. Diagnosis was based on histological and enzyme histochemical findings of muscle biopsy along with clinical data. Ultrastructural characterstics of muscle have been studied in 10 cases. The affected muscle fibres showed a central nucleus (40-99%) with perinuclear halo. Type I fibre predominance with hypoplasia was consistently seen. Fibre type disproportion was noticed in 7 cases. The neonatal form revealed dense oxidative enzyme reaction product in the centre. The morphological features of CNM were compared with foetal skeletal muscles obtained at gestational ages ranging from 9 weeks - 36 weeks (n = 18). In the severe neonatal form th myofibres resembled the foetal myotubes. In the less severe childhood and adult form of CNM, aberrant organization of cytoskeletal network might have played a pathogenetic role in causing the disease.
Centronuclear myopathy (CNM) is a primary congenital muscle disorder, pathomorphologically characterised by the occurrence of skeletal muscle fibres with centrally placed sarcolemmal nuclei in contrast to normal subsarcolemmal location. This is a heterogeneous disease with diverse clinical and genetic features and clinically poses diagnostic problems. The gene for X-linked recessive form of centronuclear myopathy, a severe neonatal disorder has recently been mapped to the long arm of sex chromosome Xq 28, while the genetic locus for the autosomal recessive/dominant forms is not known. The first description of this entity was given by Spiro et al, who found that 40-50% of the skeletal muscle fibres in the biopsy had material rich in oxidative enzymes and/or sarcolemmal nucleus in the centre. Based on the central position of the nucleus, resembling foetal muscle during development, the authors named the entity as myotubular myopathy. Following this, several cases have been reported, proposing different pathogenetic mechanisms for this condition. While Kinoshita and Cadman suggested arrest of development at myotube stage, Engel et al proposed that the presence of centronuclear myofibres resembling myotubes could be secondary to functional deficiency of maturational factors from the motor nerve, or inability of the muscle fibres to utilise them. Serratrice et al attributed the disease to malfunctioning of a factor from CNS controlling the nuclear migration. Later it was suggested that the centronuclear myopathy could be secondary to postnatal migration of the nucleus to the central region, as in other muscular dystrophies., The mechanism proposed was the arrest in maturation of the muscle fibres beyond the 20th week of foetal gestation, due to lack of trophic factors from the motor nerve., Thus, the underlying pathogenetic mechanism of centronuclear myopathy still remains unresolved, though majority of the authors favour deranged development as the mechanism for central nucleation.
In view of the unresolved issues, we carried out histological, histochemical and ultrastructural study of human foetal skeletal muscle from 9 weeks - 36 weeks of gestation and compared with the morphological features observed in the biopsied skeletal muscle from patients of CNM, to examine the hypothesis of arrested developmental maturation as the pathogenetic mechanism.
Clinical material: Seventeen cases of centronuclear myopathy, diagnosed during January 1985 to June 1998, were taken for the study. Among them, 7 cases were examined in detail at our centre by the neurologists, while 10 cases were referred to us with muscle biopsy, for diagnostic consultation. All the clinical records were reviewed by a neurologist. The final diagnosis for selecting the cases of CNM was based on morphological features in the muscle biopsy, along with clinical findings of myopathy. The muscle biopsy for the diagnosis was collected from quadriceps (12 cases) and biceps (5 cases).
Foetal tissues: Human foetuses (n=18) with gestation age ranging from 9 weeks to 36 weeks (9wks-1, 10 wks-1, 14 wks-1, 15 wks-1, 17 wks-2, 18 wks-2, 19 wks-1, 20 wks-2, 22 wks-1, 24 wks-1, 28 wks-2, 29 wks-1, 31 wks-1, 36 wks-1) were collected fresh, following medical termination of pregnancy, with informed consent of the parents and the clinician. Clearance was obtained from the ethics committee for biomedical research of NIMHANS to use human foetal tissues for study. They were transported on ice to the pathology laboratory for dissection. The interparietal distance, crown-rump and crown-heel length and the weight of the foetuses were recorded. The gestational age was calculated from the clinical records and the above measurements. Quadriceps, gastrocnemius, biceps, triceps and deltoid muscles of foetuses were sampled.
Fresh frozen (fixed in isopentane and chilled in liquid nitrogen) cryostat cut sections of the skeletal muscles from the clinical biopsy in 14 cases and all the human foetuses were stained histochemically for the following enzymes: nicotinamide adenine dinucleotide tetrazolium reductase (NADH-TR), succinic dehydrogenase (SDH), adenosine triphosphatase (ATPase) at pH 9.4, 4.6, 4.3, to evaluate the fibre types and any other abnormality in the muscle fibres. The remaining muscle fixed in formalin was processed for paraffin embedding, and stained with haematoxylin and eosin to note the changes in myofibre, and Masson's trichrome stain to assess the connective tissue component. For ultrastructural study, tiny fragments of fresh tissue from all the 18 foetal skeletal muscles and 10 clinical biopsies were immediately transferred to 3% buffered gluteraldehyde and processed for araldite embedding. The ultrathin sections stained with uranyl acetate and lead citrate were viewed under JEOL 1OOCX II at 6OKV. Muscle biopsies were received from outstation in formalin, in three cases. They were processed for paraffin sectioning. Ultrastructural study and histochemistry could not be undertaken on them.
The histological, histochemical and ultrastructural features were compared with those observed in foetal skeletal muscle. The morphological features were also compared with those noticed in dystrophy and neurogenic atrophy at our centre to understand the evolution of the disease.
Clinical Features: In the present series of patients, there were 10 males and 7 females. The age at the onset of the symptoms was at birth (one patient), early childhood (13 cases) and adulthood (3 patients). The presenting symptoms were proximal muscle weakness (13 cases) and delayed motor milestones (9 cases).
The disease was slowly progressive. Salient clinical signs were: ptosis and ophthalmoplegia (8 cases), facial weakness (8 cases) and hyporeflexia (7 cases). Somatic abnormalities noted were in the form of high arched palate (4 cases), thin dysmorphic facies (4 cases) and pescavus (1 case). External ophthalmoplegia and somatic abnormalities were seen in patients with onset during childhood. The rather unusual clinical features were the presence of bilateral calf muscle hypertrophy and ankle contractures in one patient each, similar to muscular dystrophy. Family history of similar illness was noticed in 5 cases. The other affected members included paternal great grand aunt (1 case), other siblings (3 cases) and first cousin (1 case). In one family, the father and all the siblings were affected. Creatine kinase was estimated in nine patients. The values were normal (20-170) in four, moderately elevated (2-3 times the upper limit of normal) in two and markedly elevated (8-41 times the normal upper limit) in three patients. Electromyography done in 9 patients revealed features suggestive of myopathy in 8 and neurogenic in one. Clinical diagnoses were centronuclear myopathy (3), possibly congenital myopathy (7), mitochondrial myopathy (3), muscular dystrophy (2) and neurogenic atrophy (2) [Table I].
Pathology: In the neonatal form (one case) of CNM, the size of the muscle fibres was smaller than the age matched normal controls. The fibres were of uniform diameter, 70% of them having centrally placed large, vesicular nucleus with a halo around. Loose but normal mature peri and endomysial connective tissue was present [Figure. 1a]. Oxidative enzyme reaction showed condensed reaction product[Figure. 1b] similar to that seen in the foetal muscle during the 9-10th week of gestation. ATPase reaction revealed checkerboard pattern, type I fibres being smaller than type II fibres. In the early childhood form (13 cases), variation in the diameter of the fibres was noted. The striking feature was the presence of central nuclei with perinuclear halo involving 40-99% fibres [Figure. 2a]. Most fibres had a single central nucleus, though occasional ones had 2-3 nuclei randomly dispersed close to the centre. Longitudinal sections showed small rows of nuclei along the centre. In all the cases, a few hypoplastic fibres similar to foetal myotubes, as observed in the neonatal form of the disease were seen. Oxidative enzyme reaction (NADH-TR) showed fibres with central region devoid of enzyme activity (site of nucleus) and/or accumulation of excessive reaction product. A radiating pattern of the staining reaction giving 'spokes' like appearance was seen in some of the fibres [Figure. 2b]. The central nucleus was seen mostly in type I fibres, and sometimes in type II fibres. In addition to the presence of central nuclei, one of the cases revealed multicores (multiple small circumscribed regions devoid of oxidative enzyme activity). There was predominance and hypoplasia of type I fibres [Figure. 3a]. Six out of fourteen cases showed evidence of fibre size disproportion. Type I fibre hypertrophy and smaller diameter of type II in 2 cases while in other three the reverse i.e. hypertrophy of type II fibres and small type I fibres [Figure. 3b], was noted. In one case, hypertrophic fibres were of both types (types I and II) while the smaller fibres were type I histochemically [Table II]. In the adult form of CNM (3 cases), the histological features were similar to childhood form, the central nucleus being found in 40-90% of fibres. Fibre size disproportion observed in one case revealed type I fibre hypertrophy [Table II].
There was no evidence of rounding, hyalinisation, regeneration or sarcoplasmic degeneration with myophagocytosis. The endomysial connective tissue was marginally increased in 2 cases, while fatty infiltration was prominent in childhood and adult forms of CNM. Inspite of variation in fibre size and type predominance, there were no overt features of muscular atrophy except for rare, type I angulated fibres [Figure. 3b].
Electron microscopic observations in 10 cases revealed myofibres of varying diameter in the childhood and the adult form, while the fibres were of uniform diameter in the neonatal form. The central regions were occupied by nuclei unlike the usual subsarcolemmal position [Figure. 4a]. Perinuclear noncontractile area rich in glycogen and mitochondria was noticed corresponding to the halo observed at light microscopy. Accumulation of normal looking mitochondria was seen in the center of some of the fibres (Figure. 4b). Lipofuscin was observed in the center specially at the nuclear poles and in the subsarcolemmal region. The spokes like appearance seen on light microscope on oxidative enzyme stain corresponded to the radial patterned widening of inter myofibrillar spaces. The mitochondria were distributed within these spaces, thus giving a radiating pattern [Figure. 4a]. Structural alterations in the form of mild disarray of myofilamentous pattern and streaming of the Z-band was observed in a few fibres in all the cases examined. However, the features were not comparable to the disorganisation noted in the muscular dystrophies. No redundant basal lamina around the muscle fibres suggestive of atrophy was noted. No abnormal forms of mitochondria with paracrystalline inclusions suggestive of mitochondrial myopathy were noted in the sarcoplasm in any of the cases.
In the neonatal form, the fibres did not reveal radiating pattern in their myofibrillar organisation. The central regions in some of the fibres was occupied by normal mitochondria and in a few with glycogen granules, similar to that seen in the foetal myotubes. The basement membrane was well formed. The nucleus was vesicular with prominent nucleoli.
Foetal muscle: Sections from normal human foetal skeletal muscle revealed myotubes during 9-10 weeks of foetal life. These were round or oval cells, with centrally placed nuclei, scattered amidst loose myxoid embryonal mesenchyme. The nucleus was large, vesicular with prominent nucleoli and perinuclear halo was noticed [Figure. 5a]. Histochemical stain for oxidative enzyme (NADH-TR) revealed condensed reaction product at the centre [Figure. 5b]. There was no histochemical differentiation into different fibre types. At ultrastructural level, glycogen granules were noticed in the perinuclear region. Myofilaments were seen beneath the sarcolemma and were organised into sarcomeres. Migration of centrally placed nuclei to the subsarcolemmal position started at the 14th week of embryonic life, though histochemical distinction into fibre type was not yet evident. At ultrastructural level, mature myotubes consisted of 2-3 cells enclosed within a common basement membrane. The cells revealed variable morphology in their myofibrillar organisation. One of the cell types showed well organised myofibrils with nuclei in the subsarcolemmal region. The other cell type was rich in rough endoplalsmic reticulum and showed a few myofibrils and mitochondria, which probably were the undifferentiated cells. Membrane densities present at a few points on the plasma membrane of these cells probably represented sites of fusion. During 22-36 weeks of gestation, the fascicular architecture was well formed. Groups of uniform sized fibres, round to polygonal in shape with peripherally placed nuclei, were seen in each of the fascicles. Histochemical differentiation into major fibre types, namely type I and II was noticed by the 24th week of gestation, with predominance of type II fibres. By the 28th week of foetal life distinct checker board pattern was observed. The morphology of the fibres was similar to the post natal myofibres, consisting of polygonal shaped fibres with peripherally placed nuclei of uniform diameter. The classical ultrastructural features of muscle with well organised 'Z' band and myofibrillar architecture was noted at 36 weeks. At no stage of the foetal muscle development, either histochemical or ultrastructural features of 'radiating spoke' like pattern seen in CNM was observed [Figure. 5c].
Centronuclear myopathy constituted 30% of the congenital myopathies studied at our centre. Clinically, 17 cases of centronuclear myopathy in our study could be subdivided into three groups along the lines of classification proposed by Banker.
(1) The neonatal form with X-linked recessive inheritance presents with severe hypotonia and weakness at birth. Swallowing difficulties may be present. Extraocular involvement is less common than the childhood variety. The condition is often fatal. One baby in the present study had these features. He underwent biopsy at the age of five months.
(2) In the late infantile or early childhood form of autosomal recessive inheritance, attainment of motor milestones, especially walking is delayed. There is ptosis and ophththalmoplegia, with a long, narrow face and slender body; high arched palate may be present. In the present series, thirteen children had these features. The clinical diagnosis of congenital myopathy was entertained in 4, centronuclear myopathy in 3, muscular dystrophy in one, SMA in one and mitochondrial myopathy in 3, because of varied clinical manifestations. Since progressive external ophthalmoplegia is one of the manifestations of mitochondrial myopathy, clinical differentiation between the two conditions, namely CNM and mitochondrial myopathy, may be difficult.
(3) The late childhood or adult type of CNM is of autosomal dominant inheritance. Ocular muscle involvement and dysmorphic features do not usually occur and it resembles limb girdle syndrome with mild proximal muscle weakness. There were three patients with these clinical features. Calf muscle hypertrophy, seen in two of our patients, prompted a clinical diagnosis of muscular dystrophy. Harriman and Haleem described a case of CNM in a 67 year old lady with marked hypertrophy of calves. Matsushita et al reported a case of CNM with generalised muscle hypertrophy who presented with muscle weakness since infancy, progressing to the second decade. A ten year old boy in the present study had only proximal muscle weakness and was diagnosed as a case of spinal muscular atrophy. In the absence of ptosis and ophthalmoplegia in this group of cases, clinical diagnosis of centronuclear myopathy can be difficult.
Histologically, internalisation of nuclei in neuromuscular disorders is not uncommon. It is often detected in primary and secondary disorders of muscle and during the course of development of human skeletal muscle as a normal event. The diagnostic criteria for centronuclear myopathy is a combination of a number of pathological features - a high proportion of mature muscle fibres with central nuclei, the presence of perinuclear halo (better appreciated on paraffin sections), predominance of type I fibre and relative absence of other identifiable muscle alterations indicative of either atrophy or dystrophy. In the present study, all the 17 cases on light microscopic observation revealed the aforementioned pathological changes and hence a diagnosis of CNM was considered. The number of fibres with central nuclei however varied in each case (40-99%). Predominance of type I fibres and presence of hypoplastic type I fibres, as observed in most of our cases, are the usual findings reported in the literature.,,,, An interesting observation made in one of our cases is the presence of type II fibre hypoplasia. Histochemically, very large hypertrophic fibres belonging to either type I or II or both were seen in patients with childhood type (6 cases) and adult form (1 case) of CNM. There was no difference in the percentage of fibres with central nuclei in the three clinical forms. In a comparative study with foetal skeletal muscle, myotubes were noticed during the 9th-10th week gestational age and by the 14th week the nuclei were seen in the subsarcolemmal region in a few myofibres. In human ontogenesis, polyneural innervation is noticed from the 10-14th week gestation. Whether the peripheral migration of the nucleus is governed by the appearance of single motor unit, the numerical density of the myofilaments, formation of basement membrane or some other factors is not known. Fibres with central nuclei seen in infantile form of CNM and the hypoplastic fibres in the other forms are identical morphologically to the foetal myotubes at 9-10 weeks gestation, in having central large vesicular nucleus with prominent nucleoli and perinuclear halo. This suggests arrest in maturation at an early stage (9-10 weeks) as in the neonatal form, or aberrant formation of cytoskeletal frame with delayed maturation as in late infantile and adult form of CNM, deranging the normal migration of nuclei to the periphery. However, the mechanism by which this translates into weakness affecting different muscle groups and presenting at various ages is not known.
Differentiation into fibre types was noticed by the 22nd week and normal checker board pattern was seen at the 28th week of gestation. Predominance of type I fibres seen in our series can be ascribed to lack of differentiation of type II fibres during the course of development. However, acquisition of fibre type differentiation indicates that neural innervation is not wholly defective. Commitment to a particular myogenic cell lineage is probably inherited during the period of myoblast proliferation that precedes myotube formation. Therefore, it is the myoblasts that carry the genetic information required to form the diverse types of embryonic muscle fibres. Although physiological properties of innervating alpha motoneurons and target muscle fibres appear to be closely matched, it remains difficult to identify histochemically fast and slow firing motoneurons, because of overlapping anatomical and biochemical properties. Postmortem examination of central and peripheral nervous system in patients with centronuclear myopathy, has not revealed any abnormalities.,
Even though investigations of foetal and cultured muscle and human myogenesis, indicate that developing myoblasts are transformed into myotubes independent of innervation, neural influences are essential for contraction and in controlling the ultimate fibre size, myofibril organisation and peripheral placement of nucleus. The defective neural influence at an early stage of foetal life before nuclear migration and checker board pattern probably is the cause for nucleus remaining in the centre and marked variation in the diameter of the fibres and predominance of fibre type. Immature appearance of myoneural junctions in CNM,, lends support to the view that the muscle cell abnormality is due to defective innervation. In a comparative study of postsynaptic zones from muscle of an infant with CNM and normal foetal muscle, Fidzianska et al observed that neuromuscular junctions (NMJ) in muscle fibres from CNM resembled foetal NMJ and expressed neural cell adhesion molecules (N-CAM). These findings prompted the authors to speculate abnormal nerve-muscle cell interactions induced by motor end plate immaturity leading to arrest of myofibre. The idea of arrested or aberrant morphogenesis of myofibre is further strengthened by the presence of foetal cytoskeletal proteins namely vimentin and desmin, and intracytoplasmic distribution of dystrophin. High concentration of vimentin and desmin in the postnatal life in CNM is due to genetically determined failure of these proteins, which are probably responsible for the central nucleation without causing generalised arrest in other maturational processes to diminish. However, there is no persistence of foetal myosin in the muscle fibres of CNM.
Maturation of foetal muscle involves changes in the internal architecture. Mitochondria are redistributed from the central core of the cytoplasm to the intermyofibrillar spaces. In CNM, the distribution of mitochondria in a radiating pattern observed in this study could represent an aberrant pattern of redistribution of the sarcoplasmic reticulum and myofibrillar orientation due to persistence of the central nucleus, unlike secondary nuclear migration in dystrophies. This can be further validated only by repeated examination of biopsies during the temporal evolution of the disease. Presence of myofibrillar disarray, Z-band streaming and lipofuscin material suggest simultaneous breakdown and cellular degeneration. Association of multicores in one of the cases of CNM suggests a common factor influencing the migration of nuclei and structural reorganisation of myofibres. The temporal evolution and switch over from the foetal form to adult form of cytoskeletal protein framework and its organisation appear to have a role in morphological expression of the disease.
The relative reduction in number of type II fibres, predominance of type I fibre and structural disarray, as in other myopathies could be responsible for the weakness. The functional activity of the patients was hampered to a variable degree, as seen by clinical presentation. Lack of information about gene frequency and mutation rate as well as the temporal evolution of the severity of the disease point to the need for accurate diagnosis.
The variable clinical and morphological features in centronuclear myopathy at different ages of presentation, appear to represent a continuum in evolution of the disease. In severe neonatal form, the muscle fibres resemble the embryonal myotubes suggesting a probable arrest in further maturation. The acquisition of fibre type differentiation and further growth in the muscle fibre, especially in type II, as noted in majority of our cases, suggests that neural innervation is probably not wholly defective. In less severe childhood and adult forms of CNM, aberrant organisation of cytoskeletal network may play a role in the pathogenesis of the disease.