Rhabdomyolysis in pontocerebellar hypoplasia type 2 (PCH-2)

1. Introduction 

The name pontocerebellar hypoplasias (PCH) is generally applied to a group of autosomal recessive neurodegenerative disorders with fetal onset and fatal outcome in infancy or childhood. The main representatives are type 1 (MIM 607596) [1], [2] with spinal anterior horn involvement, type 2 (MIM 277470) with progressive microcephaly, severe developmental delay and dyskinesia/dystonia [3], [4] and olivopontocerebellar hypoplasia (OPCH, MIM 225753) [5], a more severe and early lethal variant, presented as type 4 in a newly proposed classification [6]. Typical neuroradiological features are cerebellar hypoplasia, which predominantly affects the hemispheres, a flat profile of the pons and variable atrophy of the cerebral cortex. The macroscopic infratentorial characteristics are shared with disorders of N and O-glycosylation (CDG1A, dystroglycanopathies) [7], [8] that can be excluded by appropriate laboratory procedures. A combination of neuropathological findings in PCH-2 (the splitting up of the cerebellar dentate nucleus in clusters, the severe loss of neurons in the ventral pons and the preservation of spinal anterior horn cells) provide a solid base for the diagnosis [3]. The genes associated with PCH-2 and allied disorders have not yet been identified. Neuromuscular symptoms or laboratory findings compatible with muscular involvement in PCH-2 have not been reported previously. We here report two non-related patients with PCH-2 who underwent an episode of rhabdomyolysis, which was fatal in one. A muscle biopsy performed in one of these two patients before the episode showed focal necrosis. The association of two rare conditions, rhabdomyolysis and PCH-2, prompted further search for myopathic involvement in PCH-2. Retrospective analysis was performed on three muscle samples from patients with PCH-2 known to the investigators and creatine kinase was analyzed in other PCH-2 patients. None of the patients studied retrospectively had clinical symptoms of neuromuscular involvement.

2. Material and methods 

Muscle biopsies from three patients with PCH-2 and autopsied muscle from another patient with PCH-2 were reviewed. Snap frozen samples and samples prepared for electron microscopy were originally taken to exclude mitochondrial disease and/or PCH-1. Data on the patients are given in Table 1.

Table 1. Microscopic findings in five patients with pontocerebellar hypoplasia type 2

	Pat. No., sex, age at histological examination	Muscle sample obtained by	Clinical signs of neuromuscular disorder	Light microscopy	Electron microscopy	Cerebrospinal findings on autopsy
	1. Female 3yearsa	Autopsy sampling of quadriceps muscle and diaphragm	Rhabdomyolysis	No abnormalitiesb	Z-disk irregularities Interpretation uncertain because of preceding rhabdomyolysis	3years confirmation PCH-2. No spinal anterior horn involvement
	2. Female 6months	Biopsy quadriceps muscle	Rhabdomyolysis at 11months	No abnormalitiesb,d	Focal myocyte necrosis and concentric laminated bodies	Alive
	3. Male 9monthsa	Biopsy gastrocnemius muscle	No	Groups of small fibres, hyperactive on NADH oxidaseb	Focal sarcomeric disruption, duplication of Z-band, small rods	15years confirmation PCH-2. No spinal anterior horn involvement
	4. Male 1montha	Biopsy quadriceps muscle	No	No abnormalitiesb,c,e	Focal sarcomeric disruption, Z-disk streaming, duplicated Z-disks, concentric laminated bodies	2years confirmation PCH-2. No spinal anterior horn involvement
	5. Male 22years	Autopsy	No	Large number of scattered very small fibres, in part hyperactive on NADH oxidase, positive for myosin HC, some positive for CD68	Not done	22years confirmation PCH-2. No spinal anterior horn involvement

a Families related by common ancestry. b Routine frozen sections: NADH oxidase, myosin ATPase 4.3, PAS, ORO, Gomori trichrome. c Immunostaining for dystrophin (three epitopes), utrophin, merosinα, β-dystroglycan, α, β, γ, δ-sarcoglycan. d Immunostaining for dystrophin (three epitopes), α, β, γ sarcoglycan, sarcospan, α2 laminin (merosin), β-dystroglycan, caveolin-3, and dysferlin. e Myosin HC, CD68, HLA ABC, HLA-DR, desmin.

Immunohistochemistry on the muscle of patients 4 and 5 was performed using the following antibodies: HLA-ABC from Dako, Glostrup, Denmark, clone W6/32; HLA DR from Sigma USA, clone Tal1b5; CD68 mouse monoclonal ab against monocytes and macrophages, from DAKO clone PG-M1; myosin Heavy Chain, from Novocastra, clone WB-MHCn, 1:400. Antibodies against dystrophins 1, 2, 3, sarcoglycans alpha, beta, gamma and delta, merosin, alpha- and beta-dystroglycan, and desmin were obtained from Novocastra.

3. Results 

Patients 1, 3, and 4 belong to different families. Their parents are consanguineous and they share a common ancestry, linked to the Dutch community of Volendam in which families with PCH-2 are clustered [3].

3.2. Case 2 

This patient is of Australian-English origin. She is the second child of unrelated parents with a healthy older sibling. She was born at term with a birth-weight of 3.8kg. She was well as a neonate, but by four months of age developed continuous lower extremity pedalling movements, episodic opisthotonos and limb hypertonia with axial hypotonia. Her head circumference at four months was two standard deviations below the mean. She was not dysmorphic. Cerebral magnetic resonance imaging showed cerebellar hemispheric hypoplasia and a flat profile of the pons. No abnormalities were identified on screening for inborn errors of metabolism (blood and CSF lactate, urinary organic acids, urinary amino acids, and urinary glycosaminoglycans), and the serum transferrin isoelectric focussing pattern was normal. A muscle biopsy, performed to exclude PCH type 1 was normal on light microscopy, but on electron microscopy showed a single focus of necrotic material (Fig. 1) and concentric subsarcolemmal lamellated bodies (Fig. 2). Muscle enzyme- and immunohistochemistry was normal (Table 1). An assay of respiratory chain enzyme activity (complexes I, II, II and III, IV) in skeletal muscle homogenate was normal. At 11months of age, after a two day history of upper respiratory tract infection, she presented with lethargy, shock, cyanosis and rhabdomyolysis (maximum CK 258200U/l) with myoglobinuria. Her highest temperature was 40.6°C during the acute phase. Peak values for serum sodium, 143mmol/l, and serum potassium, 5.7mmol/l, were reached on the day of admission. She developed acute renal failure requiring dialysis. Her serum carnitine and fibroblast acylcarnitine profiles were normal. She was weaned from dialysis and made a full recovery. Her CK returned to normal. Three months later an assay of respiratory chain enzyme activity in liver homogenate was normal. A repeat muscle biopsy was normal on both light and electron microscopy.

View Large Image Download to PowerPoint

Fig. 1. Case 2: Quadriceps femoris muscle. A necrotic cell (arrow head) containing copious necrotic debris is partially surrounded by a satellite cell (asterisk), both lying within the basal lamina of a normal looking muscle fibre. Bar 0.5μm.

View Large Image Download to PowerPoint

Fig. 2. Case 2: Quadriceps femoris muscle. A subsarcolemmal group of concentric laminated bodies in a muscle fibre. Bar 0.5μm.

3.3. PCH-2 cases without symptoms of neuromuscular disease 

Slides from two muscle biopsies taken earlier for diagnostic evaluation in 1984 (case 3) and 1997 (case 4) and one muscle sampled at autopsy in 2001 (case 5) were reviewed (Table 1). Frozen blocks from the autopsied muscle were available for additional staining. In case 3, a biopsy of gastrocnemius muscle at age nine months showed groups of small fibres, hyperactive on NADH-oxidoreductase with coarsened intermyofibrillar staining (Fig. 3). Electron microscopy showed small fibres with sarcomeric disruption, duplication of Z-bands and small rods. Subsarcolemmal and intermyofibrillar mitochondria were morphologically normal (Fig. 4).

View Large Image Download to PowerPoint

Fig. 3. Case 3: Gastrocnemius muscle biopsy at age 9months, stained for NADH-oxidoreductase, showing coarsely stained fibres (centre), with abnormal distribution of reaction product. Bar 100μm.

View Large Image Download to PowerPoint

Fig. 4. Case 3: Gastrocnemius muscle at age 9months, electron microscopy. The longitudinal section shows two fibres. The fibre on the right shows multiple small rods originating from Z-bands, that on the left showing duplication of a number of Z-bands. Bar 2.5μm.

In case 4, a biopsy from the quadriceps muscle at one month of age was normal on light microscopy but on electron microscopy showed normal mitochondria, focal sarcomeric disruption, Z-band streaming (Fig. 5), duplicated Z-bands and subsarcolemmal concentric laminated bodies, the latter similar to case 1 (Fig. 2). Frozen muscle was assayed for mitochondrial respiratory chain activities (complexes I–IV) with normal results.

View Large Image Download to PowerPoint

Fig. 5. Case 4: Quadriceps femoris muscle at 1month, electron microscopy. Two small foci of myofibrillar disarray and sarcomere disruption (upper part of picture) and Z-disk streaming. Bar 1μm.

Postmortem examination of muscle from an adult with PCH-2 who died from progressive anaemia and inanition (case 5) showed a subpopulation of very small fibres (Fig. 6A) with increased activity of NADH-oxidoreductase (Fig. 6B). A proportion of normal sized fibres had a peripheral rim of NADH-oxidoreductase activity, higher than the central parts. Desmin staining reflected the same uneven distribution as NADH-oxidoreductase. Ca ATPase 4.3 showed equal size of type 1 and type 2 fibres with preponderance of type 2 fibres. Activity was evenly distributed without core-like structures. The population of small fibres expressed both fibre types with ATPase 4.3. The smallest fibres were positive for myosin heavy chain (MHCn) (Fig. 6C), indicative of muscle regeneration. HLA-ABC was expressed in a proportion of small muscle fibres including tiny myocytes (Fig. 6D), the apparent result of a cycle of regeneration and decay. Three foci of muscle cell necrosis with a rim of activated CD68 positive macrophages were seen per cross section of one square cm (Fig. 6E). Antibodies to proteins associated with muscular dystrophies (Table 1) stained positive. These included merosin and alpha-dystroglycan, deficiencies of which may be associated with pontine hypoplasia.

View Large Image Download to PowerPoint

Fig. 6. Case 5: Quadriceps femoris muscle taken at autopsy. (A) HE. Increased diameter variability with some small sized and numerous extremely small fibres. (B) NADH-oxidoreductase. Some of the larger fibres are rimmed by a zone of increased activity. Numerous extremely small myocytes are interspersed between normal sized fibres. Some smaller fibres stain darkly (asterisks). (C) Myosin HC. Fetal myosin identifies early stage of differentiation in smallest fibres. (D) HLA-ABC. Expression is present in small fibre (arrow), extremely thin fibre with two nuclei (open arrow), and macrophage (arrowhead). (E) CD68. Necrotic muscle fibre surrounded by a rim of activated macrophages. Magnification bars: A, B, C: 50μm; D, E: 25μm.

4. Discussion 

We report two children with pontocerebellar hypoplasia type 2 with acute rhabdomyolysis in the context of intercurrent illness, with associated hyperthermia and electrolyte inbalance in one and respiratory infection in the second. Neuromuscular involvement has not been described previously in any series of PCH-2 patients, including two larger series [3], [4]. The differential diagnosis of PCH type 2 includes PCH-1 with spinal anterior horn involvement phenotypically similar to SMA type 1 [1], [2], a mitochondrial form of PCH which presents in infancy with respiratory insufficiency and multiple contractures [9] and other genetic conditions causing both cerebellar hypoplasia and pontine hypoplasia that include the disorders of N- and O-glycosylation [7], [8]. These conditions were ruled out in both patients. Rhabdomyolysis may develop in individuals without a hereditary predisposition after stressors such as vigorous exercise or intercurrent illness, particularly in the face of exacerbating factors such as hyperthermia, hypernatraemia [10] and dystonia [11]. The fact that only a small proportion of individuals exposed to such stressors develop this complication, however, suggests that there may be unrecognised genetic factors conditioning individual susceptibility. Hyperthermia is a documented cause of death in PCH-2 [3]. It is unknown whether this was associated with rhabdomyolysis, because reported patients were not hospitalized at the time of the death. The two cases of rhabdomyolysis in PCH-2 led us to review available muscle samples previously taken from patients with PCH-2. Light microscopy was normal in three of five cases, including case 2, the patient who subsequently developed rhabdomyolysis, and the postmortem sample from the patient who died of rhabdomyolysis (case 1). Of two cases that were abnormal on light microscopy one (case 3) presented foci of small fibres, with coarsened intermyofibrillar network and increased NADH-oxidoreductase staining (Fig. 3). In the other case (case 5) autopsied frozen muscle showed foci of necrosis with macrophage activation (Fig. 6), together with numerous small and extremely small fibres. Positive fetal myosin (MHCn) staining identified these as regenerating fibres. Surprisingly, a proportion of these smaller fibres expressed HLA-ABC. Capillaries and small vessels appeared unaffected by the process, and no leucocyte perivascular cuffing was seen. This suggests an intrinsic process leading to HLA-ABC upregulation and subsequent macrophage attack.

Abnormalities seen on electron microscopy were subsarcolemmal laminated bodies (cases 2, 4; Fig. 2), foci of sarcomere disruption (loss of register between adjacent myofibrils, Z-disk streaming) (cases 3, 4) and necrosis of a presumptive satellite cell (case 2). The presence of areas of myofibrillar disruption without storage products (cases 3, 4) raise the question of a relation to central core and multicore disease. Sewry et al. [12] reported a broad spectrum of inter- and intrafamilial variability in dominant central core disease. Comparing our findings to this report the absence of clearly delimited cores, type 2 predominance (instead of type 1 uniformity) and the abundance of regenerating fibres in case 5 are at variance with this spectrum. The presence of regenerative phenomena is also at variance with the findings in minicore myopathy [13]. Concentric laminated bodies have been reported in various neuromuscular disorders, are not specific and generally accompany disorders that cause myofibrillar disarray and sarcomere disruption [14], [15], [16], [17], [18]. The association between the encephalopathy of PCH-2 and neuromuscular findings cannot be explained. Three of the five patients here described belong to a Dutch cluster of families in the community of Volendam. The two others, one originating from the Netherlands (case 5) and one from Australia with English ancestors (case 2) are not related to the Volendam cluster. Myopathic involvement in PCH-2 therefore is not restricted to a single pedigree, largely ruling out the possibility of a chance association between PCH-2 and a second genetic disorder which causes the muscular involvement. Elevation of serum creatine kinase (CK) in four of six randomly sampled cases of PCH-2, with only one from the Volendam cluster, also supports silent muscle involvement associated with PCH-2. The number of CK samples is still limited and has to be verified in a larger group, preferably also accounting for spontaneous muscular activity at the time of sampling. Anesthetic accidents resulting in malignant hyperthermia have not been documented in PCH-2, and also are not known to the authors by experience. Previous studies on PCH-2 do not refer to a myopathic component. One study on PCH-2 [3] mentions the occurrence of fatal hyperthermic crises. No previous data have appeared yet on CK elevation during such crises. Disseminated intravascular coagulation, which may relate to tissue thromboplastin release from injured muscle, often complicates severe rhabdomyolysis. Altered coagulation was described as part of a Reye-like complication in a patient with PCH-2 [19]. No mention of CK activity was made. We conclude to the existence of a subclinical myopathy in a significant proportion of patients with PCH-2 and an associated risk of rhabdomyolysis. Further studies are needed to assess the frequency and molecular biological characteristics of this association. The finding of a “myopathic” association with a condition previously considered to be exclusively neurodegenerative also bears on the ultimate choice of candidate genes for the linkage study presently in progress.