Chapter 92 Muscular Dystrophies
Introduction
Muscular dystrophies are progressive, inherited skeletal muscle disorders resulting in muscle degeneration and loss of strength. This heterogeneous group of disorders has been further characterized at the clinical and molecular level since the 1980s, giving rise to a complex classification based on phenotype and molecular genetic correlates. The onset of clinical symptoms ranges from the neonatal period to late adulthood. The distribution of predominant muscle weakness helps identify six major phenotypes [Emery, 2002]:
The mode of inheritance in muscular dystrophies can be autosomal-dominant, autosomal-recessive, or X-linked. Spontaneous mutation is responsible for many cases of muscular dystrophy usually transmitted by autosomal-dominant and X-linked inheritance. The underlying molecular defects responsible for these disorders are diverse, including extracellular matrix proteins (laminin-2, collagen VI), transmembrane- and sarcolemma-associated proteins (dystrophin, sarcoglycans, caveolin-3, α7 integrin, dysferlin), cytoplasmatic proteases (calpain-3), cytoplasmatic proteins associated with organelles and sarcomeres (titin, fukutin, telethonin), and nuclear membrane proteins (lamin, emerin). The molecular characterization of the different muscular dystrophies has advanced rapidly since the 1990s and has increased our understanding of sarcolemmal organization and underlying muscle biology responsible for the maintenance of muscle cells in health and disease. Figure 92-1 is a cartoon of the muscle membrane. Binding relationships among proteins are intricate, and the specific relation of these proteins, as well as the presence of phosphorylation and glycosylation sites, provides the basis for the pathophysiology of many of these disorders. Diseases caused by mutations of these proteins are listed in Table 92-1.
In view of the variable manifestation of the muscular dystrophies, the broad spectrum of other organs that might be involved in addition to muscle (i.e., myotonic dystrophies, syndromic congenital muscular dystrophies), and the diverse groups of proteins that cause these disorders, it may seem that the major uniting feature is the characteristic pathologic changes in muscle that allow characterization of these disorders as a “dystrophy.” Neuromuscular specialists have become increasingly dependent on molecular genetic testing to diagnose and counsel patients with muscular dystrophy and their families. DNA testing is expensive and access to this diagnostic tool is uneven. An invaluable resource is the GENETESTS website maintained by the University of Washington. This site offers excellent updated review articles on many genetic disorders (including the muscular dystrophies), as well as current information regarding molecular genetic testing available commercially or through research centers (http://www.genetests.org).
A molecular-based classification of the muscular dystrophies is presented in Table 92-2. Several of these disorders first manifest clinically in adulthood and are not covered in detail in this chapter. The most common muscular dystrophy in children is Duchenne muscular dystrophy, and it is the main focus of this review. Extensive investigations of this disorder have shed light on the pathophysiology of other muscular dystrophies, especially those related to the dystrophin-associated glycoprotein complex (DGC). The principles of clinical delineation of phenotype, categorization of functional deficits, and supportive therapies utilized for the Duchenne patient are applicable for nearly all the disorders discussed in this chapter.
Dystrophinopathies (Duchenne and Becker Muscular Dystrophies and Clinical Variants)
Dystrophinopathies are a group of muscular dystrophies resulting from mutations in the dystrophin gene, located in the Xp21 region [Baumbach et al., 1989]. Of these, Duchenne muscular dystrophy (DMD) is the most common dystrophinopathy, resulting from complete absence of the dystrophin gene product: the subsarcolemmal protein, dystrophin. It affects 1 in 3500 live male births and respects no ethnic or racial boundaries worldwide. Its allelic variant, Becker muscular dystrophy (BMD), is rarer. BMD usually results from in-frame mutations of the dystrophin gene that cause decreased quantity and decreased quality of the dystrophin protein, giving rise to a disease with varied severity and time of onset, ranging from childhood to adulthood. Whereas DMD typically manifests before age 4 years, BMD patients can be asymptomatic for decades.
Duchenne Muscular Dystrophy
History
In a communication to the Royal Medical and Chirurgical Society of London in December 1851, Edward Meryon, an English physician, described in detail a disease affecting eight males in three families. He noted the predilection for male family members, the familial nature of the disease, and the fact that the progressive muscle wasting and weakness resulted from a disease of muscle and not the central nervous system. He appears to have been the first physician to make a detailed clinical, genetic, and pathologic study of the disorder, which he accomplished several years before the brilliant work of Guillaume-Benjamin Duchenne [Emery, 1993]. Duchenne, a French neurologist and pioneer of electrical stimulation of muscle, described a syndrome characterized by muscular paralysis associated with muscle hypertrophy that resulted from accumulation of large amounts of fat and connective tissue in muscle. Duchenne described the first instrument for procuring muscle for biopsy and analysis. The disease he described as pseudohypertrophic, X-linked muscular dystrophy bears his name.
The discovery of the molecular defect in this disorder constituted the first example of reverse cloning. Restriction fragment length polymorphisms were used to track the co-segregation of the disease phenotype with DNA markers within a family. The initial cloning of the gene responsible for DMD occurred in 1985 and 1986 [Kunkel et al., 1985; Monaco et al., 1985; Ray et al., 1985], followed by isolation of the DNA clones complementary to the Duchenne dystrophy gene [Burghes et al., 1987; Cross et al., 1987; Monaco et al., 1986]. These complementary DNA clones, produced from RNA transcripts from the DMD locus, contain the coding exon sequences of the Duchenne gene. Subsequently, the protein product of the human DMD locus was identified through the use of polyclonal antibodies directed against fusion proteins containing two distinct regions of the DMD locus’s complementary DNA. This protein was named dystrophin [Hoffman et al., 1987a].
Molecular Pathogenesis
DMD is a relentlessly progressive skeletal muscle disorder, caused by a mutation in the X-linked dystrophin gene, resulting in absence of a critical protein, dystrophin [Hoffman et al., 1987a, 1988; Koenig et al., 1987]. The dystrophin gene includes 86 exons (including seven promoters linked to unique first exons), which make up only 0.6 percent of the gene; the rest consists of introns. The gene spans a distance of more than 2.5 million basepairs and is the largest human gene isolated to date [Burmeister and Lehrach, 1986]. In more than 90 percent of males with the DMD genotype, there is an absence of dystrophin resulting from an “out-of-frame” mutation that disrupts normal dystrophin transcription [Gillard et al., 1989]. These mutations include deletions, duplications, point mutations and splice site mutations, which cause early termination of messenger RNA transcription. As a result, an unstable RNA is produced, undergoes rapid decay, and then leads to production of absent or very low concentrations of abnormal dystrophin. If the mutation maintains translational reading as seen with an “in-frame” deletion, the BMD phenotype, with variably decreased amounts of abnormal molecular weight dystrophin, results [Hoffman et al., 1988]. The reading frame rule holds true for over 90 percent of afflicted individuals and is used both for diagnostic confirmation of dystrophinopathies and to assist in differentiating DMD from BMD; this distinction is occasionally difficult due to the spectrum of clinical involvement in these allelic disorders.
Exceptions to the reading frame rule occur in approximately 10 percent of patients. Out-of-frame deletions affecting exons 3–7, 5–7, 3–6, or downstream at exons 51, 49–50, 47–52, 44, or 45, can result in a milder BMD phenotype. A likely explanation for the production of at least some dystrophin in these patients is “exon skipping,” which occurs through alternative splicing [Nicholson et al., 1992; Patria et al., 1996]. In this form of BMD, the carboxy-terminus is always preserved [Arahata et al., 1991]. Exon skipping is also the probable underlying mechanism for revertant fibers (muscle fibers exhibiting dystrophin immunostaining in muscle biopsies), evident in about 50 percent of males with DMD [Winnard et al., 1995]. Limited expression of dystrophin may result in a slower progression of muscle weakness than in the usual Duchenne phenotype [Arahata et al., 1989; Baumbach et al., 1989; Koenig et al., 1989; Malhotra et al., 1988]. A new therapeutic strategy in DMD is being developed, utilizing antisense oligonucleotides to induce exon skipping in patients to restore the dystrophin reading frame and allow the production of enough dystrophin to ameliorate the phenotype (see Therapeutic Approaches) [Hoffman, 2007; Kinali et al., 2009; van Deutekom et al., 2007].
Exceptions to the reading frame rule may occur with large in-frame deletions in the 5′ end of the dystrophin gene, extending to the middle of the rod domain (i.e., deletions of exons 3–31, 3–25, 4–41, and 4–18) [Nevo et al., 2003], as well as with small in-frame deletions of exons 3–13 that cause the severe DMD phenotype [Muntoni et al., 1994]. These small and large mutations affect another actin-binding domain in the N-terminus of dystrophin and therefore may have significant functional consequences, resulting in the more severe phenotype. Other mechanisms, such as an unexpected effect of the deletion on splicing behavior, might also be implicated in determining the phenotypic outcome [Muntoni et al., 1994]. After the complete dystrophin complementary DNA (cDNA) was isolated [Koenig et al., 1987], it became evident that about 60 percent of patients with Duchenne and Becker muscular dystrophies manifest structural rearrangements due to deletion mutations [Darras et al., 1988; den Dunnen et al., 1987; Forrest et al., 1987; Koenig et al., 1987; Kunkel, 1986]. Deletions and, more rarely, duplications can arise almost anywhere in the dystrophin gene; however, two deletion hot spots are known. The most commonly mutated region includes deletions in exons 45 to 55 with genomic breakpoints (i.e., the end points of where the deletions actually occur) lying within intron 44. The second hot spot is located toward the 5′ end and includes exons 2 to 19 with genomic breakpoints commonly found in introns 2 and 7 [Beggs et al., 1990; den Dunnen et al., 1989; Nobile et al., 1995; Oudet et al., 1992]. Forty percent of patients have dystrophinopathy that results from small mutations (point mutations resulting in frameshift or nonsense mutations) or duplications.
As noted, the incidence of DMD is approximately 1 per 3500 male births [Brooks and Emery, 1977; Jeppesen et al., 2003]. The most common mode of inheritance is X-linked recessive, with inheritance from a mother who is a known or unidentified carrier. However, DMD is associated with a high spontaneous mutation rate, which accounts for approximately 30 percent of cases [Brooks and Emery, 1977; Moser, 1984; Scheuerbrandt et al., 1986; van Essen et al., 1992]. Therefore, almost one-third of males with DMD have no family history of muscular dystrophy. This mutation rate is estimated to be 10 times higher than for any other genetic disorder [Hoffman et al., 1992], due at least in part to the extremely large size of the dystrophin gene [Hoffman and Kunkel, 1989]. The 2.5 million basepairs constituting the gene (a full 1 percent of the X chromosome) provide a large target for random mutational events. Nonfamilial DMD patients might be the product of germinal mosaicism on the X chromosome (a mutation occurring before the birth of the mother) [Bakker et al., 1989; Bunyan et al., 1994; Lanman et al., 1987; Smith et al., 1999], in which case the mother is a DMD carrier but no other family member is affected with DMD. Another possibility is maternal or paternal gonadal mosaicism: a new mutation in maternal or paternal germ cells in an unaffected father or mother, whose other cell lines (i.e., muscle, lymphocytes) are normal [Darras et al., 1988; Hurko et al., 1989; Lanman et al., 1987]. This mother neither is an obligate carrier, nor does she have the neuromuscular characteristics of a DMD carrier. The remaining new mutations are partial gene duplications [Hu et al., 1989, 1990].
As the genetic defect is an X-linked recessive trait, dystrophinopathies are expressed primarily in boys and young men. However, females may manifest signs and symptoms of DMD if they also exhibit skewed X-inactivation, wherein the abnormal X chromosome is expressed in an excessive proportion of cells [Kinoshita et al., 1990; Lesca et al., 2003; Pena et al., 1987; Yoshioka et al., 1986]. Females with both DMD and Turner’s syndrome have been described. This rare combination arises when the single X chromosome of the Turner patient contains a mutant gene [Bjerglund Nielsen and Nielsen, 1984; Ferrier et al., 1965; Lescaut et al., 2004].
Genotype/phenotype correlation
There is no simple relationship between the size of a dystrophin gene deletion and the resultant clinical phenotype. Large deletions, which may involve nearly 50 percent of the gene, have been described in patients with BMD [England et al., 1990], whereas deletions of small exons, such as exon 44, typically result in a severe DMD phenotype. The central and distal rod domains of the dystrophin protein seem to be nonessential; some deletions in this region are associated with a syndrome characterized only by myalgia and muscle cramps or by an isolated increase in creatine kinase [Beggs et al., 1991; Gospe et al., 1989]. This finding has been reported in patients with in-frame deletions in exons 32–44, 48–51, or 48–53, all of whom had normal or near-normal dystrophin expression [Melis et al., 1998]. The effect of the genetic mutations on the phenotype depends on whether or not it disrupts the reading frame and on specific essential signaling or binding sites in the dystrophin protein that might be affected by the mutation. Good correlation is generally found between the severity of the phenotype and the effect of the deletion on the reading frame.
Deletions that disrupt the reading frame result in a severe phenotype, whereas in-frame deletions are associated with a milder disease course [Gillard et al., 1989]. These assumptions hold true for about 90 percent of the cases in which mutations are in the rod domain of the dystrophin protein. Rare exceptions to this rule exist, mainly resulting from frameshift mutations in the 5′ region of the gene (in particular, deletions involving exons 3–7), which are associated with a milder phenotype than expected [Malhotra et al., 1988]. When mutations arise in the 5′ region of the gene, especially in the first 13 exons of the dystrophin gene, one-third of patients have a phenotype different from that predicted [Muntoni et al., 1994]. Affected patients can have a severe clinical phenotype despite the presence of a small in-frame deletion, or they can have a mild phenotype with an out-of-frame deletion. In the first case, a Duchenne phenotype is associated with in-frame deletions of exon 5, of exon 3, and of exons 3–13. In the second case, an intermediate Duchenne or Becker phenotype is seen, with out-of-frame deletions involving not only the usual exons 3–7, but also 5–7 and 3–6 [Muntoni et al., 1994]. Therefore a proportion of patients with a deletion in the 5′ end of the gene have a phenotype that is not predictable on the basis of the effect of the deletion on the reading frame.
Very large chromosomal deletions, such as the Xp21 chromosomal microdeletion syndrome, cause multisystem disorders along with the DMD phenotype. Such contiguous deletions may result in a boy manifesting congenital adrenal hypoplasia, glycerate kinase deficiency, chronic granulomatous disease and/or McCleod’s syndrome [Francke et al., 1985]. The Leiden database (http://www.dmd.nl/) is a useful resource for phenotype/genotype correlation in situations in which genetic testing and phenotype do not appear to be clearly correlated.
Pathophysiology
Dystrophin protein
The normal dystrophin gene creates a 14-kilobase dystrophin messenger RNA that encodes 3685 amino acids, producing a 427-kD protein called dystrophin. Dystrophin localizes to the subsarcolemmal region in skeletal and cardiac muscle, and constitutes 0.002 percent of total muscle protein [Hoffman et al., 1987a, 1987b; Knudson et al., 1988]. Dystrophin binds to cytoskeletal actin and to the cytoplasmic tail of the transmembrane DGC protein, β-dystroglycan, and thus forms a link from the cytoskeleton to the extracellular matrix (see Figure 92-1).
Dystrophin contains four domains: the N-terminus, the rod domain, a cysteine-rich area, and the C-terminus [Koenig et al., 1987]. The N-terminus consists of the first 240 amino acids and provides an F-actin binding site at three distinct regions with α-actinin homology. The rod region is a succession of 25 triple-helical spectrinlike repeats and contains about 3000 residues, including four proline-rich regions that act as hinges. Another F-actin binding site is located near the middle of the dystrophin rod domain but with significantly lower affinity than the N-terminus site [Rybakova et al., 1996]. A cysteine-rich area with amino acids 3080–3360 has a WW domain, a protein domain with two highly conserved tryptophans that binds proline-rich peptide motifs, which binds to the protein β-dystroglycan, an essential protein through which dystrophin links to other integral membrane components of the DGC [Ozawa, 1995; Ozawa et al., 1995]. The site at which dystrophin binds to β-dystroglycan extends to the first half of the carboxy-terminal domain (C-terminus) [Watkins et al., 2000]. The C-terminus comprises the last 420 amino acids and has homology to utrophin and dystrobrevin. This area contains many potential phosphorylation sites and also binds to syntrophins at exon 74, and possibly to dystrophin-associated glycoproteins [Rybakova et al., 1996].
Besides the full dystrophin protein just described, seven different dystrophin transcripts can be identified. These transcripts have different promoter regions and initial exons. The N-terminus region is excluded in smaller dystrophins, but the C-terminus region is found in all seven dystrophin transcripts. The transcripts are specific to cell types. Three full-length isoforms have the same number of exons but are derived from three independent promoters in brain, muscle, and Purkinje cerebellar neurons. The best characterized of these transcripts is the previously described muscle protein, expressed not only in skeletal muscle but also in cardiac muscle, smooth muscle, and retina. A cortical 427-kDa transcript is found in cortical postsynaptic densities, retina, and skeletal muscle [Boyce et al., 1991]. A Purkinje cell 427-kDa transcript is found in the cerebellum [Gorecki et al., 1992]. There is also a retinal 260-kDa transcript, a brain and kidney 140-kDa transcript, and a Schwann cell 116-kDa protein resulting from transcription beginning on exon 56 [Byers et al., 1993]. A glial 71-kDa transcript resulting from transcription beginning at exon 63 can be found in glia, viscera, and cardiac muscle [Rapaport et al., 1992]. The presence or absence of these seven dystrophin transcripts in DMD depends on the location and size of deletion in the dystrophin gene, which determines phenotypic expression.
Primary and secondary (downstream) events
Muscle cell death in the muscular dystrophies (by apoptosis and necrosis) is conditional and reflects a propensity that varies between muscles and changes with age [Rando, 2001b]. The fact that adjacent muscle groups in DMD can be completely normal and others are undergoing active necrosis not only supports this concept but also argues against the concept of inevitability. If endogenous biochemical mechanisms alter the ability of a muscle cell to live or die while the genetic and biochemical defects remain constant, then pharmacologic modulation of these pathways might result in successful therapies for DMD and other muscular dystrophies [Rando, 2001b]. Although dystrophin deficiency is the primary cause of DMD, multiple secondary pathways are responsible for the progression of muscle necrosis, abnormal fibrosis, and failure of regeneration that results in progressive clinical deterioration [Tidball and Wehling-Henrichs, 2004]. Since the mid-1990s, hypothesis-driven research has dissected several abnormal pathways involved in muscular dystrophy progression. There is ample literature establishing evidence of oxidative radical damage to myofibers [Baker and Austin, 1989; Haycock et al., 1996a, 1996b; Murphy and Kehrer, 1986; Rando et al., 1998], inflammation [Cai et al., 2000; Kissel et al., 1993; Lagrota-Candido et al., 2002; McDouall et al., 1990; Nahirney et al., 1997; Porter et al., 2002; Shaw et al., 1996; Spencer et al., 2000; Spencer and Tidball, 2001], abnormal calcium homeostasis [Baker and Austin, 1989; De Luca et al., 2002; Leijendekker et al., 1996; Murphy and Kehrer, 1986; Pulido et al., 1998; Ruegg et al., 2002; Wrogemann and Pena, 1976], myonuclear apoptosis [Adams et al., 2001; Rando, 2001b; Sandri et al., 1995, 1997, 1998a, 1998b, 2001; Smith et al., 2000; Spencer et al., 1997; Tews, 2002; Tews and Goebel, 1997; Tindall et al., 1987], and abnormal fibrosis and failure of regeneration [Bernasconi et al., 1995, 1999; D’Amore et al., 1994; Iannaccone et al., 1995; Luz et al., 2002; Melone et al., 2000; Morrison et al., 2000; Murakami et al., 1999; Passerini et al., 2002; Rando, 2001b; Yamazaki et al., 1994]. This evidence has been validated by cross-sectional genome-wide approaches that allow an overall analysis of multiple defective mechanisms in DMD [Chen et al., 2000; Porter et al., 2002]. The increasing understanding of these events has led to the discovery of potential pharmacologic targets designed to reverse, stop, or slow down muscle damage and progression of disease, even if the primary genetic defect cannot be repaired at present. Thus, it is important to understand these mechanisms as the basis of present and future therapies for this otherwise fatal disease.
Mechanical Membrane Fragility
Dystrophin is a link between the intracellular cytoskeleton and the extracellular matrix. All evidence supports a structural function for dystrophin and the DGC, establishing the physical link for force transduction between the actin skeleton and the extracellular matrix. The carboxy-terminal of dystrophin is attached to the sarcolemma, the surface membrane of striated muscle cells [Arahata et al., 1988; Bonilla et al., 1988; Watkins et al., 1988; Zubrzycka-Gaarn et al., 1988], binding to β-dystroglycan [Jung et al., 1995] and other dystrophin-associated glycoproteins [Ervasti and Campbell, 1991]. When dystrophin is lost, disconnection of the link between contractile proteins to β-dystroglycan results in loss of β-dystroglycan and DGC from the sarcolemma, expressed as reduced sarcoglycan complex immunoreactivity. The location of dystrophin supports the concept of fragile membranes in DMD and BMD, with membrane gaps that allow extracellular leakage of cytoplasmic components, such as creatine kinase, and the influx of excessive Ca2+. Initially, it was thought that the main function of dystrophin was to provide mechanical reinforcement to the sarcolemma and thereby protect it from the mechanical stress of muscle contraction [Petrof et al., 1993]. Indeed, dystrophin is a load-bearing element, and deficiency in dystrophin leads to muscle membrane fragility and aberrant mechanotransduction [Kumar et al., 2004]. In the absence of dystrophin, there is disruption to normal force transmission and greater stress is placed on myofibrillar and membrane proteins, leading to muscle damage [Lynch, 2004], especially during lengthening (eccentric) contractions, during which a muscle contracts against a mechanical force pulling in the opposite direction [Lynch et al., 2000]. The consequent muscle membrane fragility and abnormal permeability characteristics allow increased intracellular Ca2+ influx [Mallouk and Allard, 2000; Mallouk et al., 2000] and initiate the pathologic cascade of events, resulting in muscle necrosis and fibrosis [Ruegg and Gillis, 1999; Ruegg et al., 2002]. However, it is likely that Ca2+ enters the cell through both “broken” membrane and physiologic Ca2+ channels.
Abnormal Permeability to Calcium and Chronic Increase of Intracellular Calcium
Although the theory of increased membrane microdisruptions with subsequent increased Ca2+ permeability cannot be refuted, there is accumulating evidence that abnormal Ca2+ handling may be related to direct dystrophin regulation of mechanosensitive transient receptor potential channels [Vandebrouck et al., 2002b; Yeung and Allen, 2004], as well as abnormal Ca2+ intracellular cycling [Doran et al., 2004; Dowling et al., 2004; Woods et al., 2004]. Several nifedipine-insensitive, voltage-independent Ca2+ channels might be involved in the initial abnormal Ca2+ entry. This evidence is extremely important because modulation of these channels could be a new therapeutic target. Although there is concordant evidence of increased Ca2+ influx to the cell, there is no consensus on an overall increase in resting intracellular Ca2+ concentration [Gillis, 1996]. One line of evidence involves abnormal function of “stretch-activated” and “stretch-inactivated” Ca2+ channels, a subfamily of the transient receptor potential channels [Clapham, 2003]. These stretch-activated channels are abnormally active under mechanical stimulation in myotubes of the mdx mouse, a murine model of DMD, and result in an increase in intracellular Ca2+ [Mallouk and Allard, 2000; Vandebrouck et al., 2002a, 2002b]. In addition, there is evidence that a large number of these channels, when subjected to mild mechanical stress, irreversibly shift to a stretch-inactivated channel form [Franco and Lansman, 1990] that remains open at rest and allows a chronic increase of calcium currents [Franco-Obregon and Lansman, 2002]. Accumulation of abnormally active Ca2+ leak channels over time results in a gradual loss of Ca2+ homeostasis and eventual cell death [Alderton and Steinhardt, 2000]. This increase of Ca2+ current might also be initiated by other types of Ca2+-permeable, growth factor-activated channels. The latter are normally localized in the cytoplasm of skeletal muscle, and translocate to the plasma membrane only under insulin-like growth factor 1 and possibly other growth factor stimuli [Iwata et al., 2003]. In dystrophic muscle, these channels are abnormally increased in the sarcolemma and seem to be involved directly with the increase in intracellular Ca2+ and abnormal creatine kinase efflux seen during mechanical stress of dystrophic muscle membrane [Iwata et al., 2003].
Of practical importance is the fact that blockade of these stretch-activated channels has been achieved by the nonselective transient receptor potential blockers gadolinium and mycin, as well as by the selective cationic transient receptor potential blocker GsMTx4 (spider venom toxin), and has resulted in normalization of intracellular calcium and muscle force generation ex vivo in mdx mouse muscle [Yeung et al., 2003, 2005]. Furthermore, treatment of mdx mice with oral streptomycin resulted in decreased muscle necrosis [Yeung et al., 2005]. Forced expression of the full-length and mini-dystrophin protein in dystrophin-deficient So18 skeletal myotubes also rectified steady-state levels of subcellular concentrations of Ca2+ and of Ca2+ transients [Marchand et al., 2001, 2004].
The L-type, voltage-gated Ca2+ channels appear to be abnormal in the absence of dystrophin; it was demonstrated that the Ca2+ currents in response to an action potential were much smaller in mdx mice than in normal control subjects. A disrupted direct or indirect linkage of dystrophin with these channels may be crucial for proper excitation–contraction coupling to initiate Ca2+ release from the sarcoplasmic reticulum. This linkage seems to be restored fully in the presence of mini-dystrophin [Friedrich et al., 2004].
Calcium cycling within the dystrophic muscle cell is also altered in mdx mice. Although the subsarcolemmal Ca2+ concentration might be elevated [Mallouk et al., 2000], release of Ca2+ by the sarcoplasmic reticulum in response to an action potential is much decreased in this model [Plant and Lynch, 2003; Woods et al., 2004] and might contribute to abnormal activation–contraction coupling and subsequent muscle weakness. The Ca2+ buffering system in dystrophin myotubes is equally affected. A key luminal Ca2+-binding protein, sarcalumenin, is affected in mdx skeletal muscle, which results in an abnormal shuttling of Ca2+ between the Ca2+-uptake sarco-endoplasmic reticulum Ca2+-adenosine triphosphatase and calsequestrin, and might amplify indirectly the Ca2+ leak channel-induced increase in cytosolic Ca2+ levels [Dowling et al., 2004]. Abnormal intracellular Ca2+ levels result in abnormal activation of Ca2+-activated proteases (e.g., calpain), with subsequent abnormal degradation of intracellular proteins that likely contribute to the abnormal functioning of the leak channels [Turner et al., 1993]. There is evidence that exercise worsens the abnormalities in calcium homeostasis in mdx mice [Fraysse et al., 2004]. This supports the clinical observation that excessive exercise by DMD patients may be deleterious and exacerbate muscle weakness [Allen, 2004; Ansved, 2003].
Abnormal Immunologic Response
In normal skeletal muscle, contraction-induced damage is followed by an inflammatory response involving multiple cell types that subsides over several days. This transient inflammatory response is a normal homeostatic reaction to muscle damage. In contrast, a persistent inflammatory response is observed in dystrophic skeletal muscle, which leads to an altered extracellular environment that includes an increased presence of inflammatory cells (i.e., macrophages) and elevated levels of various inflammatory cytokines (i.e., tumor necrosis factor-α, transforming growth factor-β). Therefore, the signals that lead to successful muscle repair in healthy muscle may promote muscle wasting and fibrosis in dystrophic muscle. Evidence supporting the role of the immune system in promoting muscle pathology in DMD, as well as the active role of cytotoxic T cells and myeloid cells in the pathophysiology of progressive muscle necrosis and fibrosis in DMD and mdx mice, has accumulated since the mid-1990s [Spencer and Tidball, 2001]. More recently, genome-wide approaches investigating the gene expression profile of DMD and mdx mouse muscles have revealed that an abnormal immunologic response is induced very early and is maintained at a low level in DMD patients and mdx mice from early in the neonatal period [Chen et al., 2000; Porter et al., 2003a]. Microarray experiments conducted longitudinally in mdx mice have allowed identification of abnormally increased cytokines and their receptors within the muscle cells, which implies that the dystrophic muscle might contribute to abnormal chemotaxis [Porter et al., 2002, 2003a]. These cytokine pathways could be new therapeutic targets.
Several pathways are involved in abnormal immune response in DMD. First, invasion of antigen-presenting cells activates cytotoxic and helper T cells [Banchereau and Steinman, 1998; Hart, 1997]. These release interleukin-2, which initiates polyclonal expansion of cytotoxic T cells [Gussoni et al., 1994; Spencer et al., 1997, 2001]; this, in turn, mediates muscle necrosis by perforin-mediated [Cai et al., 2000; Spencer et al., 1997, 2001] and cytokine-mediated (tumor necrosis factor-α and transforming growth factor-β) killing mechanisms [Isenberg et al., 1986; Lundberg et al., 1995; Morrison et al., 2000; Spencer et al., 2000; Tews and Goebel, 1996]. Another pathway is through circulating dermal dendritic cells, which are abundant in DMD muscle [Chen et al., 2000], and through CD4+ cells, which in turn activate (through interleukin-2) CD8+ cells. Another immune pathway starts with the invasion of muscle by myeloid cells, including macrophages, neutrophils, eosinophils, and mast cells [Arahata and Engel, 1988; Cai et al., 2000; Gorospe et al., 1994b; McDouall et al., 1990]. Most of these can kill target cells by liberating cytokines (tumor necrosis factor-α; macrophage inflammatory protein-1αβ, RANTES), and by generating high concentrations of oxidative radicals. Mast cells can also promote muscle fibrosis [Gorospe et al., 1994a; Granchelli et al., 1996]. Muscle cells might become autoreactive as they liberate tumor necrosis factor-α and could be an additional source of antigen-presenting cells [Behrens et al., 1998]. Many of these pathways are known to be blocked by prednisone, including induction of the transcription of I B (inhibitor), which keeps nuclear factor κβ in the inactive state; decreased production of proinflammatory cytokines; and induction of genes that inhibit cyclo-oxygenase-2, adhesion molecules, and other inflammatory mediators.
Abnormal Signaling Functions
Although the mechanical theory of muscle membrane disruption remains important and well documented [Kumar et al., 2004], there is accumulating evidence that the DGC also has important muscle cell-signaling functions, and its integrity is essential for muscle cell viability [Rando, 2001a]. These functions include transmembrane signaling (through β-dystroglycan), docking of signal transduction molecules (e.g., caveolin-3), and interaction with or regulation of other transmembrane complexes (e.g., integrins) [Rando, 2001a]. The absence of dystrophin causes extensive abnormalities and disruption of the DGC, in which sarcoglycans, neuronal nitric oxide synthetase [Rando, 2001b], and other members of the DGC lose their association with the sarcolemma. Many of the putative signaling cascades and molecules associated with this complex are known to regulate the balance between pro- and anti-apoptotic pathways [Rando, 2001a]. In addition to direct regulation of apoptosis, the defective DGC signaling might alter the metabolic pathways necessary to modulate cell susceptibility to injury. Of the downstream pathways involved in DMD pathogenesis leading to cell death, those involving oxidative radical metabolism [Baker and Austin, 1989; Haycock et al., 1996b; Kumar and Boriek, 2003; Ragusa et al., 1997; Rando, 2002; Spencer et al., 2001] and intracellular Ca2+ regulation are of particular interest because they could be targets of pharmacologic modulation. In addition, mechanical stretch has been demonstrated to activate the classic nuclear factor κβ pathway abnormally (in a Ca2+– independent manner) in the mdx mouse, resulting in increased expression of inflammatory cytokines interleukin-1β and tumor necrosis factor-α which precedes the onset of symptomatic myopathy [Kumar and Boriek, 2003].
The “vascular” theory of DMD pathogenesis is supported by morphologic findings in DMD and mdx mice of muscle fiber group necrosis, which occurs very early in the disease, presumably secondary to ischemia. Recent findings indicate that the mislocalization and reduction of neuronal nitric oxide synthetase in dystrophic muscle affects smooth vessel vasodilatation in response to alpha-adrenergic stimuli during exercise [Thomas et al., 2003], and results in muscle ischemia [Sander et al., 2000]. Dystrophin-associated α-syntrophin appears to be essential for the membrane localization of neuronal nitric oxide synthetase [Thomas et al., 2003]. The phosphodiesterase-5 inhibitor, sildenafil, has been shown to be cardioprotective in mdx mice [Adamo et al., 2010], prompting a clinical trial in DMD boys [Wagner, 2010].
Abnormal Fibrosis and Muscle Regeneration
Fibrosis (excessive deposition of endomysial and perimysial extracellular matrix) is a phenomenon known to be secondary to chronic muscle inflammation and fiber degeneration in DMD [Porter et al., 2002]. However, the amount of fibrosis in DMD seems disproportionate in relation to the clinical severity in the earlier stages of the disease, raising the question that an abnormal fibrotic process might be related directly to the absence of dystrophin and might occur in parallel to (as well as after) muscle necrosis and degeneration. There is evidence that both enhanced fibrinogenesis and decreased fibrinolysis [von Moers et al., 2005] are implicated in the development of muscle fibrosis in DMD. There is an increase in expression of the fibrogenic cytokine transforming growth factor-β1 in muscle of DMD patients [Bernasconi et al., 1995] and in serum samples from DMD boys [Bernasconi et al., 1999]. Preclinical studies in mdx mice have revealed that transforming growth factor-β1 levels are also significantly elevated in the diaphragm, the muscle in this model that best correlates with human skeletal muscle pathology [Hartel et al., 2001; Iannaccone et al., 1995; Morrison et al., 2000; Passerini et al., 2002; Porter et al., 2002]. Similar findings of abnormal cytokine levels and fibrosis early in the disease process are found in the golden retriever model of DMD [Passerini et al., 2002]. The level of messenger RNA transcript for transforming growth factor-β was found to be increased in mononucleated cells around areas of fiber necrosis in 6- and 9-week-old mdx mice, but not in 12-week-old mdx mice [Gosselin et al., 2004]. These findings suggest a role of transforming growth factor-β1 during the early stages of fibrogenesis in dystrophic diaphragmatic muscle. Transforming growth factor-β1 induces organ fibrosis by increasing extracellular matrix synthesis [Gosselin et al., 2004] and by simultaneously inhibiting matrix-degradation proteases, such as matrix metalloprotease 1 [Herbst et al., 1997; von Moers et al., 2005]. Another abnormal fibrotic pathway in DMD relates to an increase in levels of platelet-derived growth factor and its receptors, which have an important modulating role in the active stage of tissue destruction, as well as initiation and promotion of muscle fibrosis [Zhao et al., 2003]. Of concern is the fact that fibroblasts of DMD patients appear to have a paracrine function, inhibiting satellite cell growth [Melone et al., 2000], in which case abnormal fibrosis would decrease muscle regeneration directly.
Temporal gene expression profiling studies in DMD have demonstrated that inflammatory and profibrinogenic pathways predominate in the presymptomatic stages of the disease, whereas the acute activation of transforming growth factor-β and failure of metabolic pathways occur later in the disease [Chen and Nagaraju, 2005]. In hind limb mdx muscle, inflammation, proteolysis, and extracellular matrix upregulation and fibrosis are initiated very early in the course of the disease, represent a substantial component of the transcriptional response at that early age, and are tightly coordinated [Porter et al., 2003b]. Pharmacologic blockade of these pathways may have promising therapeutic implications, especially in symptomatic subjects [De Luca et al., 2005; Porter et al., 2002].
Clinical Features
The disease is on-going in infancy, with muscle fiber necrosis demonstrable on muscle biopsy and high serum creatine kinase enzyme levels. However, clinical manifestations frequently are not recognized until the child is 3 years of age or older. This potential “presymptomatic therapeutic window” has been underemphasized previously; however, it lends itself to the development of early therapeutic interventions to prevent or delay the onset of symptoms secondary to more advanced muscle degeneration. Walking frequently begins later than in normal children; independent ambulation often is achieved after 15 months of age. Affected children often experience more falls than their peers. Gait abnormality, difficulty rising from the floor, and problems negotiating stairs usually become apparent at 3–4 years of age, prompting clinical evaluation. Muscle weakness is present initially in neck flexor muscles, with less than antigravity power. As a result, children need to turn on their side when getting up from a supine position on the floor, which is the initial component of the Gowers’ maneuver (Figure 92-2). Hypertrophy of calf muscles typically occurs, usually being very prominent by age 3 or 4 years (Figure 92-3). Hypertrophy of other muscles also may develop, especially the vastus lateralis, infraspinous, deltoid, tongue, and, less frequently, the gluteus maximus, triceps, and masseter muscles. Muscle mass is usually decreased in later stages of the disease in the pectoral, peroneal, and anterior tibial muscles. Hip-girdle muscles are affected earlier than shoulder-girdle muscles. Because of weakness of the hip extensor muscles that act as “shock absorbers” when weight is placed on one leg, children tend to rock from side to side when walking, producing a waddling gait. Increased lumbar lordosis ensues, which is necessary to keep the center of balance stable, with shoulders lined up over hips, knees, and ankles. The preschooler has difficulty rising from the floor, turning 45, then 90, and, finally, 180 degrees, and placing the hands on the floor to get up. As pelvic weakness increases, the complete Gowers’ sign (see Figure 92-2) is exhibited. The patient assumes a locked-knee, buttocks-first position, followed by pushing off the floor with the hands, literally pushing the trunk erect by bracing the arms against the anterior thighs. The maneuver is necessary because of pronounced weakness of the hip muscles – predominantly, the gluteus maximus – and is present by age 5 or 6 years. As muscle deterioration proceeds, climbing stairs becomes difficult, necessitating use of both hands on a railing or crawling on all four extremities. Distal muscles of the arms and legs exhibit weaken as the disease progresses.
Fig. 92-3 Pseudohypertrophy of the gastrocnemius muscles is usually profound.
The large muscle mass is secondary to replacement of normal muscle with adipose and collagen tissue.
Pulmonary involvement
Pulmonary function becomes compromised because of weakness of intercostal and diaphragmatic muscles. Scoliosis occurs later in the disease in nonambulatory patients. Respiratory failure is the primary cause of mortality in DMD. Prolonging ambulation and standing with corticosteroid treatment [Moxley et al., 2005] and the use of stationary standers can have a significant impact on the development of scoliosis and respiratory function [Galasko et al., 1995]. Use of noninvasive respiratory support and aggressive pulmonary and cardiac care have increased survival [Gomez-Merino and Bach, 2002], even in patients who have not been treated with corticosteroids [Eagle et al., 2002].
Muscle weakness affects all aspects of lung function, including mucociliary clearance, gas exchange at rest and during exercise, and respiratory control during wakefulness and sleep [Gozal, 2000]. DMD is often associated with sleep-disordered breathing, which may be asymptomatic or only mildly symptomatic. Patients can have normal waking oxygen saturation and capillary blood gas levels and normal or near-normal forced vital capacity. Overnight polysomnography is useful for detecting abnormalities [Kirk et al., 2000].
Cardiac involvement
Children with DMD are at risk for cardiomyopathy, especially if they have deletion of exons 48–53 [Nigro et al., 1994]. Early screening for cardiomyopathy at ages 5–6 years, followed by cardiac surveillance with electrocardiography and echocardiography, allows detection of impaired cardiac output before signs of heart failure are apparent. If heart failure occurs, rigorous intervention to improve cardiac function and output is necessary. Mild degrees of cardiac compromise in DMD may occur in up to 95 percent of affected males [Melacini et al., 1996]. Chronic heart failure may affect up to 50 percent [Melacini et al., 1996; Wahi et al., 1971]. Sudden cardiac failure can occur, especially during adolescence. In one series of 19 patients, postmortem studies revealed that 84 percent had demonstrable cardiac involvement [Leth and Wulff, 1976]. Characteristic electrocardiographic changes include sinus tachycardia; tall R1 waves in lead V1; and prominent Q waves in leads II, III, aVF, and V6. Autonomic dysfunction may contribute to tachycardia [Danzig et al., 2003; Finsterer and Stollberger, 2003]. Routine echocardiography may yield normal findings. However, early abnormalities in cardiac function can be detected with newer echocardiography techniques in children between the ages of 4 and 10 years [Giglio et al., 2003]. Some young children with DMD might exhibit regional wall motion abnormalities in areas of fibrosis [Melacini et al., 1996]. With development of cardiac fibrosis, left ventricular dysfunction and ventricular dysrhythmias can occur. In the late stages of the disease, systolic dysfunction may lead to heart failure and sudden death. Subclinical or clinical cardiac insufficiency is present in about 90 percent of patients with DMD and BMD, and is the cause of death in 20 percent of those with DMD and 50 percent of those with BMD [Finsterer and Stollberger, 2003; Melacini et al., 1996].
Neuropsychologic involvement
Males with DMD have an intelligence quotient (IQ) curve shifted to the left. The mean IQ score in one study was 83 (range, 46–134) [Ogasawara, 1989]. However, other investigators have not been able to demonstrate a decrement in full-scale IQ relative to normal males [Bushby et al., 1995; Felisari et al., 2000; Hinton et al., 2000; Roccella et al., 2003]. Some cognitive areas are more affected than others, especially verbal memory [Hinton et al., 2000]. Genotype/phenotype studies have revealed that deletions localized in central and 3′ parts of the gene are preferentially associated with mental impairment, some of these directly affecting the regulatory and coding sequences for the three central nervous system-specific carboxy-terminal isoforms [Giliberto et al., 2004]. Another study of 137 males with DMD/BMD confirmed these findings, demonstrating that all patients with deletions upstream of the 5′end of the gene were mentally normal, whereas all patients with mental retardation or autism had deletions containing the 3′ end of the gene. Some researchers have found average intelligence [Sollee et al., 1985] but difficulties with immediate memory [Wicksell et al., 2004]. DMD boys are more likely to exhibit poor performance on measures of story recall, digit span, and auditory comprehension, a profile suggesting selective involvement of verbal working memory [Hinton et al., 2001]. Morphologic changes in the central nervous system and brain imaging studies have been inconsistent. Although many investigators have found no pathologic or radiologic brain abnormalities in DMD, others, using positron emission tomographic scanning, have reported hypometabolism in the cerebral and cerebellar hemispheres [Bresolin et al., 1994; Rae et al., 1998]. Overall, there is no firm evidence that an abnormality of the brain, either gross or histologic, is common in patients with DMD, and a correlation between abnormality and intellectual impairment has yet to be clearly established [Anderson et al., 2002].
Diagnosis
Clinical laboratory tests
The serum creatine kinase level is the most valuable and universally used screening test for Duchenne dystrophinopathy. Levels of creatine kinase muscle isozyme are greatly elevated, typically from 10,000 to 30,000 International Units/L, early in the disease. A normal or minimally elevated creatine kinase level effectively excludes the possibility of DMD. Gaps in the sarcolemma allow efflux of creatine kinase into the circulation. Serum creatine kinase levels can vary greatly with activity and decrease as muscle mass is lost with disease progression. There is no correlation between the serum creatine kinase level and clinical severity, and the use of creatine kinase levels as a surrogate marker of treatment response is not well supported. Due to the leakage of intracellular muscle proteins, other muscle isoenzyme levels also increase in the circulation. These include lactate dehydrogenase (LDH), alanine aminotransferase (ALT), and, to a lesser degree, aspartate aminotransferase (AST), all tested routinely as markers of liver function. Physicians may perform extensive investigations, including liver biopsy, before it is realized that these “liver enzymes” are of muscle derivation. Similar diagnostic confusion may occur in presymptomatic BMD patients [Korones et al., 2001; Lin et al., 1999; Tay et al., 2000; Zamora et al., 1996]. Obtaining a truly “liver-specific” gamma glutamyl transferase (GGT) level eliminates diagnostic error.
Genetic testing
A precise molecular genetic diagnosis is essential in all patients with DMD and BMD, even in those whose diagnosis has been confirmed by immunostaining for dystrophin on muscle biopsy. Genetic testing for DMD and BMD is widely available. Initial screening for deletions and duplications within the dystrophin gene confirms the diagnosis of dystrophinopathy in most patients. Direct sequencing of the entire dystrophin gene will define the vast majority of DMD/BMD patients not identified by deletion/duplication testing. The screening of only 19 exons by multiplex polymerase chain reaction identifies about 98 percent of all deletions [Beggs et al., 1990]. Southern blot analysis of these samples frequently can predict whether the deletion, when in the rod domain, will shift the reading frame, and thus is confirmatory for dystrophinopathy. This technique is very effective for the molecular diagnosis of common deletions (60 percent of patients); however, it cannot be used to identify duplications or to establish genotype in female carriers. Other diagnostic approaches, such as quantitative polymerase chain reaction [Abbs and Bobrow, 1992; Yau et al., 1996], multiplex amplifiable probe hybridization, and multiplex ligation-dependent probe amplification [White et al., 2002], can be used for detecting patients with either gene deletions or duplications, as well as carrier diagnosis. With more recent technology, it is possible to sequence the entire dystrophin gene for the specific molecular defects responsible in the other 30–40 percent of patients with DMD and BMD in whom a genetic abnormality has not been detected. These techniques include a protein truncation test [Tuffery-Giraud et al., 1999a, 1999b], single-condition amplification/internal primer sequencing [Mendell et al., 2001], and denaturing high-performance liquid chromatography followed by sequencing, which enables the detection of small mutations in nearly all the 79 exons [Bennett et al., 2001]. In a large series with single-condition amplification/internal primer sequencing, deletions of one or more exons were found in 66 percent of probands, which was consistent with the frequency of deletions in the previously reported literature. Point mutations were found in 18 percent, including premature stop codons in 13 percent and missense mutations in 4 percent of the total mutations. Frameshift mutations were found in 3 percent of patients. Fourteen subexonic mutations were also found, and these included nonsense mutations in 9 (64 percent), missense mutations in 3 (21 percent), and frameshift mutations in 2 (14 percent). Of the 45 deletions detected, 3 (7 percent) either were undetected or would have been undetected by that technique [Dent et al., 2005]. Duplications were detected in 6 percent of patients, and no disease-causing mutation was identified in 7 percent of the evaluated patients [Dent et al., 2005].
Several commercial laboratories offer gene sequencing for the diagnosis of suspected dystrophinopathy patients, as well as for carrier detection and prenatal diagnosis. (See GENETESTS website: www.genetests.org.)
Muscle biopsy
Histologic study reveals fiber size variation, degenerating and regenerating fibers, clusters of smaller fibers, endomysial fibrosis, and a few scattered lymphocytes. Large, opaque fibers, distinguished by intense staining with the modified Gomori trichrome, are prominent (Figure 92-4). As the disease progresses and degeneration exceeds regeneration, a decrease in the number of muscle fibers is apparent, with replacement of muscle with fat and connective tissue. Fiber typing of type 1 and type 2 muscle fibers with ATP histochemistry is less distinct than normal. Oxidative histochemistry shows that the intramyofibrillary network is maintained. On electron microscopic study, gaps in the sarcolemma with preservation of the basement lamina are seen in non-necrotic fibers (Figure 92-5 and Figure 92-6). Absence of immunoreactivity for dystrophin, with monoclonal antibodies against the C-terminal, rod domain, and N-terminal, is important for an accurate diagnosis of dystrophinopathy (Figure 92-7). However, quantitative dystrophin analysis by immunoblot correlates more closely with the diagnoses of DMD than immunostaining, with levels of dystrophin being less than 5 percent of normal in DMD patients.
Fig. 92-5 Electron micrograph of a specimen from a 6-year-old child with Duchenne muscular dystrophy.
Management
Comprehensive standard of care guidelines for the diagnosis and treatment of patients with DMD have been recently published [Bushby et al., 2010a, 2010b]. The needs of children and adults with DMD and BMD change substantially with disease progression. The information presented in this section, although largely similar to that presented by Bushby, reflects the bias of the authors of this chapter.
Pharmacologic treatment
In the era before widespread treatment with corticosteroids and assisted ventilation of DMD patients, death usually occurred late in the second decade of life (median age of 18 years), and only 25 percent of males with DMD lived beyond 21 years of age Gardner-Medwin, 1970]. This natural history has changed [Eagle et al., 2002], and in current studies, researchers are trying to characterize the impact of corticosteroid treatment and age at start of treatment, as well as noninvasive ventilation, rehabilitation, scoliosis surgery, and pulmonary care, on the quality of life and survival of DMD patients. Daily corticosteroid therapy stabilizes or improves the strength of children with DMD, and it is the only treatment proven to be effective for this disease [Moxley et al., 2005]. Drachman and associates [1974] first reported the beneficial effect of prednisone in an open trial of 2 mg/kg/day. This finding was duplicated in other open-design trials [Brooke et al., 1987; DeSilva et al., 1987] and subsequently, through the collaboration of the Clinical Investigation of Duchenne Dystrophy investigators, in double-blinded placebo-controlled trials [Fenichel et al., 1991; Griggs et al., 1993; Mendell et al., 1989]. To date, the most effective, well-studied prednisone dose is 0.75 mg/kg/day. There is a dose–response effect, the lowest proven effective dose being 0.3 mg/kg/day [Mendell et al., 1989]. Effects on strength are seen as soon as 10 days after treatment starts, with a peak at 3 months, followed by slowing of disease progression [Griggs et al., 1991]. In a 3-year follow-up study, improvement was maintained in children who were kept on doses of at least 0.5 and 0.6 mg/kg/day [Fenichel et al., 1991]. It is hoped that long-term monitoring of these patients will confirm continuous benefit in muscle strength and respiratory function. Children with DMD treated with prednisone from an early age often remain ambulatory into their teens, have a lower incidence of scoliosis [Alman et al., 2004; Biggar et al., 2004] and of contractures [Yilmaz et al., 2004], and maintain better respiratory muscle function [Angelini et al., 1994; Biggar et al., 2004; Campbell and Jacob, 2003; Merlini et al., 2003
