Experimental Therapeutics in Hereditary Neuropathies: The Past, the Present, and

[Guest guest]

Experimental Therapeutics in Hereditary Neuropathies: The Past, the

Present, and the Future

N. Herrmann

Volume 5, Issue 4, Pages 507-515 (October 2008)

Neurotheraputics (FULL TEXT ARTICLE)

Summary

Hereditary neuropathies represent approximately 40% of undiagnosed

neuropathies in a tertiary clinic setting. The Charcot–Marie–Tooth

neuropathies (CMT) are the most common. Mutations in more than 40

genes have been identified to date in CMT. Approximately 50% of CMT

cases are accounted for by CMT type 1A, due to a duplication within

the peripheral myelin protein 22 gene (PMP22). Mutations in the gap

junction beta 1 gene (GJB1), the myelin protein zero gene (MPZ), and

the mitofusin 2 gene (MFN2) account for a substantial proportion of

other genetically definable CMT. Some 15% of demyelinating CMT and

70% of axonal CMT await genetic clarification. Other hereditary

neuropathies include the hereditary sensory and autonomic

neuropathies, the familial amyloid polyneuropathies, and multisystem

disorders (e.g., lipid storage diseases and inherited ataxias) that

have peripheral neuropathy as a major or minor component. This review

surveys the challenges of developing effective therapies for

hereditary neuropathies in terms of past, present, and future

experimental therapeutics in CMT.

Key Words: Charcot–Marie–Tooth neuropathies, peripheral neuropathy,

hereditary neuropathy, outcome measure, clinical trial design, therapy

Article Outline

• Summary

• Introduction

• Review of Current Treatment Options for CMT and Past Therapeutic

Trials

• Development of Novel Therapies for CMT: Pathogenetic Considerations

• Mechanism-Based Therapies Currently Under Study in Human CMT Trials

• Ascorbic acid (CMT1A)

• Coenzyme Q10

• Neurotrophin-3

• Therapies in Preclinical Testing for Forms of CMT

• Future Therapeutic Strategies for CMT: RNA and Gene-Based Therapies

• Future Therapeutic Strategies for CMT: Targeting of Cytoplasmic

Aggregates of Mutant or Misfolded Protein

• Challenges to Developing Therapies in Inherited Neuropathies

• Experimental Therapeutics in CMT Will Require Use of Novel Outcome

Measures and Trial Designs

• Economic Implications of Designer Therapies for Rare Forms of CMT

• Ethical Considerations in Development of Future Therapies for

Inherited Neuropathies

• Summary Comments

• References

• Copyright

Introduction

Hereditary neuropathies represent approximately 40% of undiagnosed

neuropathies in a tertiary clinic setting.1 The most common

hereditary neuropathies are the Charcot–Marie–Tooth neuropathies

(CMT). Mutations in more than 40 genes have been identified to date

in CMT (see, for example, http://www.molgen.ua.ac.be/CMTMutations).

Among CMT, approximately 50% are accounted for by CMT type 1A

(CMT1A), due to a duplication within the peripheral myelin protein 22

gene (PMP22).2, 3 Mutations in the gap junction beta 1 gene (GJB1),

the myelin protein zero gene (MPZ), and the mitofusin 2 gene (MFN2)

account for a substantial proportion of other genetically-definable

CMT.2, 3 Some 15% of demyelinating CMT and 70% of axonal CMT await

genetic clarification. Other hereditary neuropathies include the

hereditary sensory and autonomic neuropathies, the familial amyloid

polyneuropathies, and multisystem disorders (e.g., lipid storage

diseases and inherited ataxias) that have peripheral neuropathy as a

major or minor component. Major categories of hereditary neuropathies

and any known associated gene defects are detailed in an online

database resource (http://www.molgen.ua.ac.be/CMTMutations). The

present review illustrates the challenges of developing effective

therapies for hereditary neuropathies, via a focus on past, present,

and future experimental therapeutics in CMT.

Review of Current Treatment Options for CMT and Past Therapeutic

Trials

Most forms of CMT are characterized by slowly progressive symmetric

distal limb weakness, atrophy, foot deformity, and sensory loss.2, 3,

4 Particular genetic causes of CMT may also lead to hearing,

respiratory, vocal cord, or autonomic impairment. The age of symptom

onset and the degree of associated disability range widely, from a

childhood requirement for a wheelchair to mild impairment of gait and

balance in later life.

No disease-modifying therapies have yet been discovered for CMT.

Supportive care includes exercise, physical and occupational therapy,

and orthotics designed to address foot drop and stabilize gait.4

Surgery intended to maintain function of primarily the distal lower

limbs is available. Analgesic agents are used for neuropathic and

mechanical pain.4

What is the evidence that supports these interventions? A Cochrane

Collaboration analysis of randomized trials of surgical, physical

therapy, and orthotic interventions for foot drop in neuromuscular

disorders found one study of CMT that met the review criteria.5 In

this study of 29 CMT subjects, a 24-week muscle strengthening and

exercise program resulted in improvement in a 6-meter timed walk,

with trends favoring the treatment in other outcomes including

quality of life (QOL), and stair climbing.5 A randomized, single-

blind, cross-over study of nocturnal ankle splinting in CMT1A (n =

14) showed no effect on ankle range of motion or strength.6

Systematic studies on the effect of orthotics on gait and falls in

CMT are relatively lacking, despite widespread use. Retrospective

studies of arthrodesis surgery have reached divergent conclusions in

CMT.7, 8 In one study of 16 subjects with an average of 21 years

follow-up, 14 of 30 arthrodesis sites had a poor outcome; the authors

concluded that triple arthrodesis should be used only as salvage

therapy for advanced CMT and severe rigid deformity.7 In contrast, in

a study of 57 subjects (with 25- and 40-year follow-up), 95% of

subjects were satisfied with their surgical interventions.8

Prospective studies comparing outcomes of conservative care versus

early surgery for foot drop in CMT are lacking. There is also a

paucity of controlled trials of pain management in CMT. Fatigue is

also a significant symptom in CMT, but no randomized, double-blind

controlled trials (RCT) of therapies for fatigue have been performed

for CMT. In a single open-label study, four subjects receiving

modafinil therapy reported improvements in disabling daytime fatigue.9

Before the availability of contemporary genetic approaches, few

trials aimed to modify the biology of CMT. Intramuscular Cronassial

(a ganglioside mixture) (Fidia Pharmaceutical Corporation,

Washington, DC) was ineffective in 30 CMT subjects treated in a

RCT.10 A small RCT of linoleic and & #947;-linoleic essential fatty acids,

which were added to a placebo containing vitamin E, was conducted in

CMT1 (n = 20) and controls. No effect of essential fatty acids was

noted at 12 months; however, after a lead-in period of 3 months of

vitamin E containing placebo, improvements in neurologic disability

measures were seen, which were sustained for 1 year.11 It was unclear

whether this improvement reflected a sustained placebo effect, or

therapeutic benefit of vitamin E.11 A single, double-blind, placebo-

controlled cross-over study of low-dose coenzyme Q10 (CoQ10) (100

mg/day) enrolled patients with several forms of muscular dystrophy,

and also clinically defined CMT; benefit was reported among the three

CMT subjects included.12

Development of Novel Therapies for CMT: Pathogenetic Considerations

Mechanisms of peripheral nerve damage in CMT are uncertain, but

unraveling of the genetic causes of CMT has led to novel therapeutic

targets. CMT1A stems from a duplication of a 1.5-Mb domain of the

PMP22 gene, with PMP22 overexpression in peripheral nerves.2, 3, 4

PMP22 overexpressing transgenic mice develop a demyelinating

neuropathy, and the animal phenotype depends on the degree of PMP22

expression.13 A toxic gain of function mediated by PMP22

overexpression is inferred in CMT1A. PMP22 is important in Schwann

cell (SC) differentiation and regulation and in the integrity of

compact myelin.2, 3, 4 It influences axon–glial cell interactions and

axonal integrity, loss of which is thought to be critical in the

progression of CMT1A.2, 3, 4 PMP22 overexpression produces

ubiquinated PM22 aggregates, impaired proteosomal function, and

autophagic profiles in rodent SC cytoplasm.14 Correction of PMP22

levels in animals leads to improvement of the neuropathy, and is a

rational strategy for CMT1A.15

Hereditary neuropathy with liability to pressure palsies (HNPP) is

usually due to a PMP22 gene deletion and haploinsufficiency.2, 3, 4

Therapeutic downregulation of PMP22 in CMT1A, if excessive, could

provoke a HNPP phenotype, whereas upregulation may be effective for

HNPP.3 An alternative approach for CMT1A, which requires preclinical

testing, would be one that targets PMP22 aggregates, on the

hypothesis that the aggregates cause proteosomal impairment and SC

toxicity.3, 14 Trophic factor deficits (e.g., ciliary neurotrophic

factor) have been found in rodent models and CMT1A nerves, suggesting

that trophic factor replacement may be efficacious.16 Animal models

of CMT1A suggest that inflammatory responses influence demyelination.

Small subsets of patients with CMT1A develop acute worsening, with T-

cell infiltrates in nerve biopsies, and respond to immunotherapy.17,

18, 19 It has not been shown, however, that anti-inflammatory

therapies influence animal models or the course of most patients with

CMT1A.20

CMTX is usually due to mutations of the GJB1 gene, which encodes the

connexin 32 (Cx32) gap junction protein, located in noncompact

myelin, and accounts for 15% of CMT.2, 3, 4, 20, 21, 22 More than 300

pathogenic GJB1 mutations have been described

(http://www.molgen.ua.ac.be/CMTMutations). Inheritance is X-linked

dominant. Most males develop CMT with intermediate range conduction

velocity slowing on nerve conduction studies, whereas females have

normal or minimally reduced conduction velocities consistent with

axonal CMT. Pathologic findings vary, but are primarily axonal in

type.2, 3, 4, 20, 21, 22 Cx32 has multiple functions, including

serving as a conduit for diffusion of ions and molecules between the

perinuclear and adaxonal SC cytoplasm, regulating SC gene expression,

and playing a role in early axon regeneration.3, 21, 22

It is disputed whether a relationship exists between the GJB1

genotype and the human phenotype.21, 22 For example, a mutation in

which the mutant Cx32 failed to reach the SC membrane was found to

impair axon regeneration in human sural nerve xenografts and to be

nonfunctional in vitro. This mutation produced a more severe human

phenotype than did a mutation that inserted within the SC membrane

and showed only subtle abnormalities in channel function.21 In

contrast, a study assessing 73 patients with 28 GJB1 mutations

revealed no mutations that were associated with a phenotype more

severe than missense mutations, suggesting a loss of function

mechanism.22

Animal transfection with a GJB1 gene that expresses in myelinating

SCs ameliorates the neuropathy, indicating that replacement of

deficient wild-type Cx32 channels may be effective for the human

disease.22 An inflammatory component is suspected in Cx32 null mice.

Occasional CMTX patients with accelerated weakness have inflammation

on biopsy, suggesting superimposed immune neuropathy.18 Segmental

demyelination is not a feature of most descriptions of CMTX, and thus

the contribution of inflammation to the human condition is unclear.

Mutations in MPZ occur in 5% of CMT. The gene product MPZ, a member

of the immunoglobulin superfamily, is located in compact myelin and

accounts for 50% of PNS myelin protein.2, 3, 4 MPZ facilitates myelin

compaction. More than 120 disease-causing mutations, located

primarily in the coding portion of the gene, have been identified

(http://www.molgen.ua.ac.be/CMTMutations/). MPZ mutations can

manifest as several phenotypes: an autosomal dominant CMT, with an

early onset severe (congenital hypomyelinating [CHN] or Dejerine

Sottas [DSS]) phenotype, a childhood-onset demyelinating phenotype,

or a milder adult onset phenotype that shows intra- and interfamily

variability in its severity and its degree of nerve conduction

velocity slowing.23 Phenotype–genotype correlations suggest that a

severe early-onset phenotype may be produced by mutations adding a

charged amino acid, by mutations that alter a cysteine residue in the

extracellular domain, or by mutations truncating the cytoplasmic

domain.24

Another important determinant of severity is the system of nonsense-

mediated decay of mutant MPZ mRNA.23 Upstream premature stop codon

mutations (e.g., nonsense or frameshift mutations) result in

degradation of mutant mRNA via nonsense-mediated decay, with a

typically mild MPZ haploinsufficiency phenotype.24 Downstream

mutations predict a truncated protein that escapes nonsense-mediated

decay. This results in translation of a mutant MPZ protein, which

generally produces severe disease, likely via a toxic gain of

function. In this situation, mutant MPZ protein is retained in the

endoplasmic reticulum as aggregates of misfolded protein.24 This may

trigger an unfolded protein response, which is a chaperone system

that mediates proteasome degradation of aggregates. MPZ protein

aggregates have in preclinical studies been associated with

apoptosis.24 Therapies that target misfolding or aggregation of

protein mutants in severe MPZ neuropathies may ameliorate the

phenotype. Although augmentation of the unfolded protein response–

proteasome pathway could be considered as a therapeutic target,

overactivation of this pathway can lead to apoptosis.

Axonal CMT account for approximately 30% of autosomal dominant CMT.2,

3, 4 Of these, 20% are due to mutations in the guanosine

triphosphatase outer mitochondrial membrane fusion protein, mitofusin-

2 (MFN-2).25, 26 More than 50 mutations in MFN2 have been associated

with CMT2A. Some kindreds with hereditary motor sensory neuropathy 5

(HMSN 5) (axonal CMT with pyramidal tract signs) and HMSN6 (axonal

CMT and optic atrophy) harbor a mitofusin mutation

(http://www.molgen.ua.ac.be/CMTMutations/). Cyclic fusion and fission

is vital to the maintenance of mitochondria.

Studies suggest diverse mechanisms whereby MFN2 mutations contribute

to axonal degeneration. Dorsal root ganglion cells transfected with

each of six MFN2 mutants demonstrated a marked defect in

mitochondrial transport, with abnormal clustering of fragmented

mitochondria in the soma and proximal axon.26 In this study, ATP

production, oxidative enzyme activity, and mitochondrial membrane

potential were normal.26 It was hypothesized that a paucity of

mitochondria in distal axons resulted in local energy failure and

thus in and dying-back degeneration.26 A further study analyzed

cultured skin fibroblasts in four subjects with CMT-2A, and three

novel MFN-2 mutations.25 These mutations impaired oxidative

phosphorylation, but mitochondrial morphology, enzyme activity, and

DNA content remained normal.25 Additional proteins associated with

mitochondria include kinesin, which is important in mitochondrial

transport, and GDAP-1, a glutathione transferase protein, which has

been associated with recessive demyelinating CMT4A and recessive

axonal CMT2.4 Therapies that address mitochondrial dysfunction may be

viable for CMT2A, and for additional rare forms of CMT.

Mechanism-Based Therapies Currently Under Study in Human CMT Trials

Ascorbic acid (CMT1A)

Ascorbic acid (AA) is essential for myelination in Schwann cell–

dorsal root ganglion (SC-DRG) cultures, where it is necessary for SC

basal lamina formation.27 Passage et al.28 studied AA versus placebo

in a C22 mouse model of CMT1A. After 1 month of therapy, all AA-

treated mice had arrest of motor decline, or improvement that was

sustained while on treatment. The AA-treated mice showed a marked

survival benefit and partial correction of the myelination defect.28

Ascorbic acid appeared to exert its effect by downregulation of PMP22

mRNA expression.28 Subsequent studies have suggested that high-dose

AA reduces PMP22 mRNA expression in a dose-dependent manner, via

decreasing adenylate cyclase activity and intracellular cAMP levels,

thereby inhibiting cAMP stimulation of PMP22 expression.29

Although these preclinical data have yet to be replicated, trials of

AA in CMT1A have been initiated. A study in North America aims to

enroll 120 subjects (age 13 to 70 years) in a futility design

protocol (see below) of 24 months of AA (2 g orally, twice daily) or

placebo to determine the impact of AA on progression of CMT1A

compared with historical controls and a cohort of placebo-treated

subjects (NCT00484510 at http://www.clinicaltrials.gov). A secondary

aim is to assess whether AA administration affects PMP22 mRNA levels

of myelinated nerve fibers in the skin. The primary outcome is a

composite clinical and nerve conduction study (NCS) measure, the CMT

Neuropathy Score (CMTNS). The neuropathy impairment score (NIS),

upper extremity NCS, PMP22 mRNA expression in forearm skin biopsies,

and the 36-item Short Form Quality of Life questionnaire (SF36) scale

are secondary outcomes.

Another multicenter, parallel-group design RCT in CMT1A of 1.5 g of

AA or placebo daily for 2 years is being conducted in Italy.30 This

study uses the CMTNS as a primary outcome, and maximal voluntary

isometric contraction, overall neuropathy limitations scale, timed 10-

meter walk test, nine-hole peg test, visual analog scale for pain and

fatigue, SF-36, NCS, and changes in skin biopsy PMP-22 expression as

secondary outcomes.30 The study design, with 220 subjects, assumes

that placebo-treated subjects will worsen by an average of 1 point on

the CMTNS over 2 years and that AA-treated subjects will improve by

0.5 points, to achieve 80% power to detect an effect of AA.30

Coenzyme Q10

Coenzyme Q10 is an electron acceptor in the mitochondrial electron

transport chain with antioxidant properties; it has shown a possible

beneficial effect in early studies in Parkinson's disease and

Huntington's disease. A single small study of low-dose CoQ10 reported

benefit among three clinically defined CMT subjects.12 A double-

blind, randomized, placebo-controlled trial of CoQ10 300 mg orally

twice daily administered for 60 weeks is currently enrolling, with a

target of 46 CMT subjects (NCT00541164 at

http://www.clinicaltrials.gov). However, this study does not focus

enrollment on those forms of CMT in which mitochondrial dysfunction

appears to be a fundamental mechanism, nor did the prior small pilot

study.

Neurotrophin-3

Studies have suggested a deficit of neurotrophin-3 (NT-3) in CMT1A.31

Human mutant SCs harboring PMP-22 duplications fail to support axonal

regeneration of nude mice axons in an injury model. Addition of NT-3,

but not brain-derived neurotrophic factor (BDNF) partially corrects

the defect in axonal regeneration in a human CMT1A xenograft in nude

mice.31 NT-3 is known to be important for SC survival and axonal

regeneration. A pilot, double-blind, RCT of 150 & #956;g/kg of r-metHuNT-3

three times a week for 24 weeks or placebo in CMT1A, showed

encouraging results, with significant improvement in sensory and

reflex subsets (but not motor subsets) of the NIS, and increased

density of regenerative clusters and single myelinated nerve fibers

on pre- and post-treatment nerve biopsies, compared with placebo-

treated subjects.31 NT-3 was well tolerated.

Therapies in Preclinical Testing for Forms of CMT

Curcumin, a widely available turmeric extract, has shown antioxidant,

antiproliferative, anticancer, anti-inflammatory protein folding–

misfolding modulatory effects in vitro and in animal studies.24, 32

Curcumin appears well tolerated in humans in doses as high as 12

g/day, and protects against apoptosis in cell cultures expressing MPZ

or PMP22 mutations that ordinarily lead to intracellular aggregates

and apoptosis.24, 32 Newborn Trembler-J mice with missense Pmp22

mutations that are associated with severe human CMT1 show marked

improvement in their clinical and pathologic phenotype on curcumin

therapy.32 Studies in this model indicate that curcumin does not

affect PMP22 levels, but rather likely alleviates mutant PMP22-

induced SC apoptosis.32 Precisely how curcumin is active in these

models is unclear; its effects may variously occur via release of

mutant protein from the endoplasmic reticulum, by alleviating

sequestration of endoplasmic reticulum chaperones, through reversal

of the impairment in cellular protein synthesis and trafficking

associated with an unregulated unfolded protein response, or by

correction of misfolding of mutant proteins.24, 32 Curcumin is thus a

potential therapy for forms of CMT that are associated with

endoplasmic reticulum aggregate formation and protein misfolding.

There are potential barriers to use of curcumin, including concern

about bioavailability. Studies in Trembler-J mice suggest, however,

that (at least in this model) curcumin crosses the blood–brain

barrier, and, although measurable plasma levels are not achieved, it

is taken up by peripheral nerve.32 Pharmacokinetic modifications of

curcumin are being sought to develop a form that maximizes its

therapeutic effects while optimizing its bioavailability. Another

concern is that, in the Trembler-J model, curcumin ameliorates the

CMT phenotype only in newborn mice, but not when treatment is

initiated in adulthood.32 Explanations for the difference include

that higher concentrations of curcumin are achieved in peripheral

nerve in the newborn mice, and that curcumin may be most active

during brisk myelination and prior to axon loss.32 If the early

treatment effect applies to human CMT, it would limit the therapeutic

potential of curcumin.

Progesterone is produced by SCs and neurons, and influences myelin

gene expression (PMP22, MPZ, and their transcription factors).33, 34

Studies in an animal model of CMT1A (5-day-old transgenic mice) have

shown that progesterone worsens the clinical and pathologic phenotype

and increases sciatic nerve PMP22 mRNA by approximately 30%.33 Wild-

type mice also show a 25% increase in sciatic nerve PMP22 mRNA

expression. Treatment of 5-day-old CMT1A mice for 7 weeks with the

pure progesterone receptor antagonist onapristone improved the

clinical and pathological phenotype, and reduced sciatic nerve PMP22

mRNA levels by 15%.33 In a subsequent study, adult CMT1A-type mice

treated with onapristone showed improvement in the clinical,

pathological, and electrophysiological phenotype, with a 20%

reduction in PMP22 mRNA expression in sciatic nerve and skin

biopsies.34 Onapristone mediated its effects in these studies via an

increase in axon density, although there was no effect upon myelin

pathology.34

These studies provide proof of principle for consideration of

progesterone antagonists in CMT1A, and also indicate that the effects

of SC PMP22 overexpression on axon integrity are divorced (at least

in part) from myelinopathy.34 The results are important, because

progesterone antagonists address axon loss, which correlates with

disability in CMT1A. There are barriers to use of progesterone

antagonists in human CMT1A clinical trials. Onapristone is unsafe in

humans (liver toxicity). Mifepristone (RU486), an oral progesterone

antagonist and partial agonist, has a better long-term safety record

in both males and females, and has been used in meningiomas and

leiomyomas for its antiprogesterone effects and has been in trial for

Alzheimer's disease. There are no reports on RU486 in animal models

of CMT1A, but this may be worthy of investigation. It is unclear,

however, whether partial agonists will abrogate any beneficial

effects of progesterone antagonism. Additionally, assuming that

lifelong therapy would be required, effects of RU486 on endometrial

hyperplasia and possibly endometrial carcinogenesis may be limiting.

A role for progesterone antagonists in CMT1A awaits development of

safer preparations, and then human clinical trials.

Progesterone agonists, for which several safe preparations are in

wide human use, may conversely have therapeutic potential for some

forms of hereditary neuropathies. In HNPP, with its PMP22 haplotype

insufficiency, and potentially also for MPZ mutations associated with

a haplotype insufficiency mechanism (in which the mutant MPZ allele

undergoes mRNA nonsense-mediated decay), upregulation of the normal

allele could conceptually improve haploinsufficiency.33 To date,

preclinical studies are lacking to provide proof of concept for

consideration of progestins in these neuropathies.

Future Therapeutic Strategies for CMT: RNA and Gene-Based Therapies

Several evolving gene-based technologies hold therapeutic promise in

hereditary neuropathies and require study in preclinical CMT models.

RNA interference (RNAi) is a naturally occurring process in which

small, double-stranded RNA strands (short interfering RNA) target

undesired homologous mRNA sequences for destruction, or prevent their

translation.35, 36 This regulates gene and protein expression, and

has led to use of exogenous viral-based and non-viral-based RNAi to

silence genes in cell lines and in animal models. Potential

therapeutic applications of RNAi include silencing of mutant alleles,

which lead to a toxic gain of function, or of alternatively spliced m-

RNA to regulate expression of protein isoforms.35

Antisense oligonucleotides (ASO) are short, single-stranded RNA or

DNA sequences that bind to mRNA via hydrogen bonds and inhibit

translation or cause degradation of target mRNA.35, 36, 37 ASOs can

also be used to modify pre-mRNAs in several ways: through blocking of

cryptic pre-mRNA splice sites; via blockade of normal splice sites to

induce specific exon skipping; or through targeting of splicing

inhibitors, to facilitate inclusion of specific exons.35

Morpholino phosporodiamidate ASOs (morpholinos), which substitute a

morpholine for the ribose in the oligonucleotide backbone, and

phosporodiamidate to link nucleotides instead of phosphodiesters,

hold particular promise because they are quite resistant to

nucleases.36, 37 Morpholinos injected into the tibialis anterior

muscle, and directed toward Cl-c1 exon 7a, ameliorated the chloride

channelopathy and myotonia in that muscle in a myotonic dystrophy

mouse model, and, in a proof-of-concept study, injection of an ASO

(PR0051) into the tibialis anterior muscle of four boys with Duchenne

muscular dystrophy did safely and successfully produce skipping of

exon 51 and reading frame restoration, allowing for translation of a

partially functional dystrophin protein.38, 39, 40 Fomivirsen, a CMV

retinitis therapy, is the first ASO drug to receive FDA approval;

others are in various stages of clinical evaluation in oncology and

virology studies.

Deoxribozymes (catalytic DNA, or DNAzymes) are deoxynucleotide

sequences that target mRNA via –Crick base-pair binding, and

catalyze breakdown of that mRNA.36 DNAzymes targeting vascular

endothelial growth factor (VEGF) mRNA have been used in animal models

of cancer, and to inhibit neovascularization in models of macular

degeneration. They can be synthesized more economically than some

other gene-silencing tools, and can be modified to be relatively

resistant to degradation. DNAzymes could be considered for silencing

mutant RNA for CMT genes that lead to toxic gain of function.36

The technique of therapeutic RNA trans-splicing allows for repair of

defective m-RNA.41 RNA-trans-splicing is a rare naturally occurring

splicing process, which occurs between two separately transcribed pre-

mRNAs producing a composite m-RNA.35, 41 Spliceosome- or ribozyme-

based therapeutic trans-splicing approaches can be used to remove a

mutant sequence in pre-mRNA and replace it with an exogenously

constructed sequence to permit production of corrected mature mRNA,

and translation of a normal protein.35, 41

A further therapeutic approach to dominant negative effects in PMP22

overexpression disorders includes downregulation of the promoter

sequences for PMP22. The PMP22 gene has two promoters, P1 and P2. P1

is of particular interest, because it exerts its effect mainly in

myelinating SCs.42 Downregulation of the activity of P1 may therefore

have therapeutic value with specificity for PMP22 overexpression.

Development of sequence triplex=forming oligonucleotides that bind

competitively to the P1 promoter in a sequence-specific manner has

been achieved.42 The triplex-forming oligonucleotides exert their

effects in the nucleus. This may be an advantage, in that they are

less subject to nucleases than are strategies targeting mature mRNA

(e.g., conventional ASOs, RNAi), and also if multiple copies of the

mRNA exist.42

In summary, knockdown of mutant alleles using RNAi, ASOs, or DNAzymes

may be effective for dominant forms of CMT associated with a toxic

gain of function.3, 4 ASO therapy may also have a role for hereditary

neuropathies with abnormal pre-mRNA splicing.43 Repair of mutant pre-

mRNA using therapeutic trans-splicing could be applied to address

various mechanisms in hereditary neuropathies including

haploinsufficiency and dominant negative disorders. Traditional gene

replacement or stem cell approaches could also be considered for

haploinsufficiency in hereditary neuropathies.4 These strategies have

varying advantages and disadvantages in terms of cost, barriers to

effective delivery to target tissue, stability, and potential for off-

target effects, but hold future therapeutic potential for various

forms of CMT.35 Although major challenges lie ahead in harnessing

these technologies for therapeutic advantage in CMT, in vitro and

animal model work for hereditary neuropathies should now proceed.

Gene and RNA-based therapies would usher in an era of therapy

customized to the mutations of individuals and families, raising also

the vexing questions of how to manage the attendant costs, clinical

testing, and regulatory approval.40

Future Therapeutic Strategies for CMT: Targeting of Cytoplasmic

Aggregates of Mutant or Misfolded Protein

Although curcumin shows preclinical promise for forms of CMT mediated

via cytoplasmic aggregates of mutant protein, other approaches that

break down protein aggregates, or correct protein misfolding, could

be valuable in CMTs to convert the phenotype from a more severe toxic

gain of function to a milder haploinsufficiency. Protein aggregation

and misfolding is believed to be a key event in neurodegenerative

disorders, including Alzheimer's disease, prionopathies, Huntington's

disease, and Parkinson's disease.44 A number of strategies are being

considered in these disorders, including targeting proteases

important in aggregate formation, and disruption of aggregates with

antibodies and other compounds, such as Congo red analogs.45 A better

understanding of mechanisms of protein aggregation and misfolding in

particular forms of CMT is required to develop such therapies.3

Challenges to Developing Therapies in Inherited Neuropathies

The translation of the expanding range of therapeutic targets into

therapies for CMT will be a major challenge. Apart from diabetic

neuropathy and neuropathic pain, the pharmaceutical industry has

devoted relatively limited effort toward therapies for neuropathies.

Limited resources and funding sources are available to clinical and

research scientists and academic institutions for independent

development of therapies for hereditary neuropathies, and

collaborations across institutions and continents and among clinician

investigators interested in CMT have been few. Patient advocacy in

hereditary neuropathies has not been of the scale seen with other

neurodegenerative disorders. The initiation of clinical trials of AA

in CMT1A in North America and Europe has led to improved

infrastructure for conduct of treatment trials. Nonetheless,

partnerships among the pharmaceutical industry, clinical

investigators, patient advocacy groups, and private foundations—in

addition to enhanced levels of governmental funding of therapeutic

trials in peripheral neuropathies—will be necessary for successful

therapy development for hereditary neuropathies.

Specifically considering CMT, given that CMT1A accounts for 50% of

all cases, achieving effective therapy for PMP gene dosage disorders

would represent a major milestone. Traditional experimental

therapeutic models appear feasible in CMT1A, because all patients

harbor the same mutation. For other forms of CMT (e.g., Cx32 or MPZ

disorders), in which hundreds of mutations exist, individualizing

therapies and assessing efficacy for mutations that affect only a

handful of kindreds worldwide will be especially challenging. This

underscores the need to focus therapy development in such forms of

CMT to be effective (where possible) for a variety of mutations. For

example, therapies for MPZ mutations with toxic accumulation of

mutant proteins might ideally target a common mechanism in the

accumulation of diverse MPZ mutants.

Beyond the hurdle of developing approaches for multiple mutations,

there is wide intra- and interfamilial variability in the age of

onset, phenotype, and severity of CMT, such that a unitary approach

may not be appropriate, not even for two members of the same family.

For some patients with very mild, late-onset phenotypes, intervention

with costly, designer therapies may not be justifiable. Because

patients with CMT demonstrate disability primarily in relation to

axon loss,4 early therapy before significant axon loss has occurred

may be more effective. Once axonal loss has taken place,

reinnervation of fibrotic muscle or skin that has lost its SCs may be

more difficult to achieve. Research that identifies predictors of a

severe course for a given mutation will be critical for selection of

patients who need disease-modifying therapy, and for timing of

interventions.

Experimental Therapeutics in CMT Will Require Use of Novel Outcome

Measures and Trial Designs

For CMT1A, there are sufficient numbers of patients to permit

multicenter, parallel-group RCTs to assess efficacy.46 There are,

however, only scarce data on optimal outcome measures for such

studies.46, 47, 48 Current outcome measures include the NIS, a

quantified neurological examination, or the CMTNS, a composite

measure consisting of nine elements (sensory symptoms, motor symptoms

[lower limb, upper limb], strength [lower limb, upper limb],

vibration and pin sensation, ulnar or median sensory nerve action

potential amplitude, and an ulnar or median compound muscle action

potential amplitude). Limited natural history data for CMT1A suggest

that there is slight, yet measurable worsening of a mean of 1.37 NIS

points per year, or 0.69 points per year on the CMTNS.48 Given the

available natural history data, at least 280 subjects would be

required to demonstrate a 50% slowing of progression of CMT1A

relative to placebo in a parallel-group design over a 2-year

period.48 Because CMT is a lifelong progressive disorder, even more

modest effects on progression may translate, over decades, into a

clinically meaningful effect. Lengthy treatment trials to identify

small effects would be difficult to achieve for more than a handful

of prospective therapies in CMT1A.

One approach to streamlining screening of potential therapies in

CMT1A is the futility study design. It is considered futile to

proceed with further larger efficacy trials if the treatment group

does not worsen by at least a prespecified amount less than expected

based on natural history data. This approach reduces the required

sample size by comparing changes in the outcome in study subjects to

changes expected based on natural history data, and also uses single-

tailed hypothesis testing. Futility studies are nonetheless a major

undertaking in CMT1A, because they require the same length of trial

as a traditional efficacy study. This study design also depends on

reliable natural history data and on selection of an appropriate

prespecified amount of slowing of progression relative to natural

history.

Development of surrogate measures for hereditary neuropathies that

rapidly and reliably predict a beneficial clinical effect is

therefore critical. One promising biomarker involves analysis of

myelinated dermal nerve bundles in punch skin biopsies.43, 49 Skin

biopsies can be obtained to assess SCs and myelinated and

unmyelinated innervation, and to track expression of myelin

components at the mRNA and protein levels. These biopsies have been

used to reveal differences in PMP22 expression and localization among

controls, CMT1A, and HNPP, and to explore molecular mechanisms of

novel CMT mutations.49 Skin biopsies are under pilot trial as a

secondary outcome in two ongoing trials of AA in CMT1A.4, 30 In these

studies, PMP22 mRNA levels are being followed, with the hypothesis

that downregulation of PMP22 mRNA levels will predict clinical

slowing of progression or improvement of CMT1A, and will provide a

suitable surrogate measure for future studies in CMT.4, 30 Skin

biopsies may also be useful to monitor mutant mRNA or protein levels,

or aggregate formation in other forms of CMT. If skin biopsies are to

be used in this manner, the reliability and reproducibility of these

assessments will need to be established.

For rare forms of CMT, in which particular mutations occur in only a

few kindreds worldwide, conduct of traditional efficacy trials will

not be possible. Here, single-patient (N-of-1) double-blind,

randomized, placebo controlled, crossover trial designs could be

considered.50, 51 In N-of-1 trials, an active therapy and placebo are

administered to an individual patient in a multiple cross-over

manner, and the outcomes for each treatment period are compared to

assess for a treatment effect. Results of multiple N-of-1 clinical

trials can be combined to draw population inferences regarding

treatment effects.51 N-of-1 clinical trials are inherently most

suitable for assessing therapies for chronic disorders such as

neuropathic pain, diabetes, and hypertension. In diseases such as

these, independent measurement of the effects of interventions can be

determined during successive periods, because they do not cause

contamination between periods, nor require lengthy washout periods.

For CMT, the N-of-1 design is problematic. Any effective treatment

might take a long period of time to show a clinical effect (e.g.,

axonal regeneration may be required before a clinical effect can be

measured). Likewise, it may take considerable time for washout of

therapeutic effects. Treatment periods may thus need to be lengthy.

This approach may be viable if sensitive surrogate measures can be

developed. For example, if a hypothetical RNAi therapy was developed

for a toxic gain-of-function MPZ mutation, the effects of blinded

cross-over application of the RNAi versus sham RNAi could be assessed

within a short time interval by measuring expression levels of the

mutant mRNA or protein in skin biopsies. Conventional clinical and

electrophysiological outcomes might be expected to respond too slowly

to an intervention to make them suitable for N-of-1 applications in

CMT. Careful consideration of the optimum manner in which to assess

efficacy of designer therapies in rare hereditary neuropathies is

therefore required. Proper dosing of prospective therapies and

measurements to verify that desired tissue concentrations and half-

life are achieved in the human subject are critical elements that

have often received only limited attention in prior failed

therapeutic trials in neuromuscular disorders.

Economic Implications of Designer Therapies for Rare Forms of CMT

Several of the therapeutic strategies discussed would likely require

chronic or intermittent administration. For some agents (e.g., AA),

compounds are widely available and inexpensive. For RNAi and various

gene-modifying approaches, costs of therapy development could be

considerable. This is exemplified by the development of enzyme

replacement therapy for Pompe's disease. Enzyme replacement therapy

for this disorder requires lifelong biweekly infusions, with a

potential cost of approximately $200,000 to $300,000 per year. Many

of the effective therapies of the future will likely need to be

customized for a diversity of genetic causes of hereditary

neuropathies, and so society will need to confront the challenge of

how to provide costly designer therapies to patients with rare

disorders, for lengthy periods of time.

Ethical Considerations in Development of Future Therapies for

Inherited Neuropathies

Early genetic characterization of CMT will be desirable for timely

therapeutic intervention. Preimplantation genetic diagnosis (PGD),

with a goal of avoiding transmission of hereditary disorders, is an

evolving discipline. PGD has been used in at-risk families to select

embryos without a PMP22 duplication for in vitro fertilization.4

Although PGD will become increasingly available, it probably will not

initially be universally accessible or accepted by families for

prevention of hereditary neuropathies. Severe congenital or early-

onset CMT implies a need for prenatal diagnosis for optimum timing of

future therapies. Those at risk for later onset CMT will similarly

require presymptomatic genetic screening.

Ethical, social, psychological and economic considerations

surrounding presymptomatic genetic screening have been confronted in

other neurodegenerative disorders, including some with devastating

consequences (e.g., Huntington's disease). The Huntington's disease

experience will be of value in prenatal or presymptomatic diagnosis

in CMT, although the highly variable phenotype of CMT raises

different ethical questions, among them the genetic labeling of

individuals who may or may not develop a severe disorder years or

decades later. Predictors of which individuals are at risk for severe

hereditary neuropathies phenotypes will be critical for informing the

decisions of patients and families regarding presymptomatic screening

and therapy.

Summary Comments

Rapid molecular and genetic advances have enabled genetic

characterization of many hereditary neuropathies syndromes. Isolation

of causative genes for CMT over the past decade has led to emerging

therapeutic targets, and advances in gene and mechanism-based

strategies provide hope that therapies will emerge to slow or partly

reverse several forms of hereditary neuropathies. The advent of

designer strategies for hereditary neuropathies will challenge

stakeholders in health care in the 21st century to develop ways to

make novel therapies available to those who need them.

References

1. 1Dyck PJ, Oviatt KF, Lambert EH. Intensive evaluation of referred

unclassified neuropathies yields improved diagnosis. Ann Neurol.

1981;10:222–226. MEDLINE | CrossRef

2. 2Nicholson GA. The dominantly inherited motor and sensory

neuropathies: clinical and molecular advances. Muscle Nerve.

2006;33:589–597. CrossRef

3. 3Saifi GM, Szigeti K, Snipes GJ, CA, Lupski JR. Molecular

mechanisms, diagnosis, and rational approaches to management of and

therapy for Charcot–Marie–Tooth disease and related peripheral

neuropathies. [Erratum in: J Investig Med 2004;52:50] J Investig Med.

2003;51:261–283. MEDLINE | CrossRef

4. 4Shy ME. Therapeutic strategies for the inherited neuropathies.

Neuromolecular Med. 2006;8:255–278. MEDLINE | CrossRef

5. 5Sackley C, Disler PB, -Stokes L, Wade DT. Rehabilitation

interventions for foot drop in neuromuscular disease. Cochrane

Database Syst Rev. 2007;(2):CD003908.

6. 6Refshauge KM, J, Nicholson G, van den Dolder PA. Night

splinting does not increase ankle range of motion in people with

Charcot–Marie–Tooth disease: a randomised, cross-over trial. Aust J

Physiother. 2006;52:193–199. MEDLINE

7. 7Wetmore RS, Drennan JC. Long-term results of triple arthrodesis

in Charcot–Marie–Tooth disease. J Bone Joint Surg Am. 1989;71:417–

422. MEDLINE

8. 8Saltzman CL, Fehrle MJ, RR, Spencer EC, Ponseti IV. Triple

arthrodesis: twenty-five and forty-four-year average follow-up of the

same patients. J Bone Joint Surg Am. 1999;81:1391–1402. MEDLINE

9. 9 GT, Han JJ, Mayadev A, Weiss MD. Modafinil reduces fatigue

in Charcot–Marie–Tooth disease type 1A: a case series. Am J Hosp

Palliat Care. 2006;23:412–416. MEDLINE | CrossRef

10. 10Bradley WG, Badger GJ, Tandan R, et al. Double-blind controlled

trials of Cronassial in chronic neuromuscular diseases and ataxia.

Neurology. 1988;38:1731–1739. MEDLINE

11. 11 LL, O'Dougherty MM, FS, Bobulski RJ, Horrocks

LA. Dietary essential fatty acids, vitamin E, and Charcot–Marie–Tooth

disease. Neurology. 1986;36;:1200–1205.

12. 12Folkers K, Simonsen R. Two successful double-blind trials with

coenzyme Q10 (vitamin Q10) on muscular dystrophies and neurogenic

atrophies. Biochim Biophys Acta. 1995;1271:281–286. MEDLINE

13. 13Huxley C, Passage E, on AM, et al. Correlation between

varying levels of PMP22 expression and the degree of demyelination

and reduction in nerve conduction velocity in transgenic mice. Hum

Mol Genet. 1998;7:449–458. MEDLINE | CrossRef

14. 14Fortun J, Go JC, Li J, Amici SA, Dunn WA, Notterpek L.

Alterations in degradative pathways and protein aggregation in a

neuropathy model based on PMP22 overexpression. Neurobiol Dis.

2006;22:153–164. MEDLINE | CrossRef

15. 15Perea J, on A, Tolmachova T, et al. Induced myelination

and demyelination in a conditional mouse model of Charcot–Marie–Tooth

disease type 1A. Hum Mol Genet. 2001;10:1007–1018. MEDLINE | CrossRef

16. 16Vigo T, Nobbio L, Hummelen PV, et al. Experimental Charcot–

Marie–Tooth type 1A: a cDNA microarrays analysis. Mol Cell Neurosci.

2005;28:703–714. MEDLINE | CrossRef

17. 17Kobsar I, Hasenpusch-Theil K, Wessig C, Müller HW, i R.

Evidence for macrophage-mediated myelin disruption in an animal model

for Charcot–Marie–Tooth neuropathy type 1A. J Neurosci Res.

2005;81:857–864. MEDLINE | CrossRef

18. 18Ginsberg L, Malik O, Kenton AR, et al. Coexistent hereditary

and inflammatory neuropathy. Brain. 2004;127:193–202. MEDLINE |

CrossRef

19. 19Dyck PJ, Swanson CJ, Low PA, Bartleson JD, Lambert EH.

Prednisone-responsive hereditary motor and sensory neuropathy. Mayo

Clin Proc. 1982;57:239–246. MEDLINE

20. 20Nave KA, Sereda MW, Ehrenreich H. Mechanisms of disease:

inherited demyelinating neuropathies—from basic to clinical research.

Nat Clin Pract Neurol. 2007;3:453–464. CrossRef

21. 21Abrams CK, Freidin M, Bukauskas F, et al. Pathogenesis of X-

linked Charcot–Marie–Tooth disease: differential effects of two

mutations in connexin 32. J Neurosci. 2003;23:10548–10558.

22. 22Shy ME, Siskind C, Swan ER, et al. CMT1X phenotypes represent

loss of GJB1 gene function. Neurology. 2007;68:849–855. CrossRef

23. 23Shy ME, Jáni A, Krajewski K, et al. Phenotypic clustering in

MPZ mutations. Brain. 2004;127:371–384. MEDLINE | CrossRef

24. 24Khajavi M, Inoue K, Wiszniewski W, Ohyama T, Snipes GJ, Lupski

JR. Curcumin treatment abrogates endoplasmic reticulum retention and

aggregation-induced apoptosis associated with neuropathy-causing

myelin protein zero–truncating mutants. Am J Hum Genet. 2005;77:841–

850. MEDLINE | CrossRef

25. 25Loiseau D, Chevrollier A, Verny C, et al. Mitochondrial

coupling defect in Charcot–Marie–Tooth type 2A disease. Ann Neurol.

2007;61:315–323. MEDLINE | CrossRef

26. 26Baloh RH, Schmidt RE, Pestronk A, Milbrandt J. Altered axonal

mitochondrial transport in the pathogenesis of Charcot–Marie–Tooth

disease from mitofusin 2 mutations. J Neurosci. 2007;27:422–430.

CrossRef

27. 27Eldridge CF, Bunge MB, Bunge RP, Wood PM. Differentiation of

axon-related Schwann cells in vitro: I (Ascorbic acid regulates basal

lamina assembly and myelin formation). J Cell Biol. 1987;105:1023–

1034. MEDLINE | CrossRef

28. 28Passage E, Norreel JC, Noack-Fraissignes P, et al. Ascorbic

acid treatment corrects the phenotype of a mouse model of Charcot–

Marie–Tooth disease. Nat Med. 2004;10:396–401. MEDLINE | CrossRef

29. 29Kaya F, Belin S, Bourgeois P, Micaleff J, Blin O, Fontés M.

Ascorbic acid inhibits PMP22 expression by reducing cAMP levels.

Neuromuscul Disord. 2007;17:248–253. Abstract | Full Text | Full-Text

PDF (153 KB) | MEDLINE | CrossRef

30. 30Pareyson D, Schenone A, Fabrizi GMCMT-TRIAAL Group. A

multicenter, randomized, double-blind, placebo-controlled trial of

long-term ascorbic acid treatment in Charcot–Marie–Tooth disease type

1A (CMT-TRIAAL): the study protocol. [EudraCT no.: 2006-000032-27]

Pharmacol Res. 2006;54:436–441. MEDLINE | CrossRef

31. 31Sahenk Z, Nagaraja HN, McCracken BS, et al. NT-3 promotes nerve

regeneration and sensory improvement in CMT1A mouse models and in

patients. Neurology. 2005;65:681–689. CrossRef

32. 32Khajavi M, Shiga K, Wiszniewski W, et al. Oral curcumin

mitigates the clinical and neuropathologic phenotype of the Trembler-

J mouse: a potential therapy for inherited neuropathy. Am J Hum

Genet. 2007;81:438–453. CrossRef

33. 33Sereda MW, Meyer zu Hörste G, Suter U, Uzma N, Nave KA.

Therapeutic administration of progesterone antagonist in a model of

Charcot–Marie–Tooth disease (CMT-1A). Nat Med. 2003;9:1533–1537.

MEDLINE | CrossRef

34. 34Meyer zu Hörste G, Prukop T, Liebetanz D, Mobius W, Nave KA,

Sereda MW. Antiprogesterone therapy uncouples axonal loss from

demyelination in a transgenic rat model of CMT1A neuropathy. [Erratum

in: Ann Neurol 2007;61:282] Ann Neurol. 2007;61:61–72. MEDLINE |

CrossRef

35. 35Wood M, Yin H, McClorey G. Modulating the expression of disease

genes with RNA-based therapy. PLoS Genet. 2007;3:845–854.

36. 36Bhindi R, Fahmy RG, Lowe HC, et al. Brothers in arms: DNA

enzymes, short interfering RNA, and the emerging wave of small-

molecule nucleic acid-based gene-silencing strategies. Am J Pathol.

2007;171:1079–1088. CrossRef

37. 37Sumanas S, Larson JD. Morpholino phosphorodiamidate

oligonucleotides in zebrafish: a recipe for functional genomics?.

Brief Funct Genomic Proteomic. 2002;1:239–256. MEDLINE | CrossRef

38. 38Wheeler TM, Lueck JD, Swanson MS, Dirksen RT, Thornton CA.

Correction of ClC-1 splicing eliminates chloride channelopathy and

myotonia in mouse models of myotonic dystrophy. J Clin Invest.

2007;117:3952–3957.

39. 39van Deutekom JC, Janson AA, Ginjaar IB, et al. Local dystrophin

restoration with antisense oligonucleotide PRO051. N Engl J Med.

2007;357:2677–2686. CrossRef

40. 40Hoffman EP. Skipping toward personalized molecular medicine. N

Engl J Med. 2007;357:2719–2722. CrossRef

41. 41Pergolizzi RG, Crystal RG. Genetic medicine at the RNA level:

modifications of the genetic repertoire for therapeutic purposes by

pre-mRNA trans-splicing. C R Biol. 2004;327:695–709. MEDLINE |

CrossRef

42. 42Hai M, Bidichandani SI, Hogan ME, Patel PI. Competitive binding

of triplex-forming oligonucleotides in the two alternate promoters of

the PMP22 gene. Antisense Nucleic Acid Drug Dev. 2001;11:233–246.

MEDLINE

43. 43Sabet A, Li J, Ghandour K, Pu Q, et al. Skin biopsies

demonstrate MPZ splicing abnormalities in Charcot–Marie–Tooth

neuropathy 1B. Neurology. 2006;67:1141–1146. CrossRef

44. 44Estrada LD, Soto C. Inhibition of protein misfolding and

aggregation by small rationally-designed peptides. Curr Pharm Des.

2006;12:2557–2567. MEDLINE | CrossRef

45. 45Frid P, Anisimov SV, Popovic N. Congo red and protein

aggregation in neurodegenerative diseases. Brain Res Rev. 2007;53:135–

160. MEDLINE | CrossRef

46. 46Reilly MM, de Jonghe P, Pareyson D. 136th ENMC International

Workshop: Charcot–Marie–Tooth disease type 1A (CMT1A), 8–10 April

2005, Naarden, The Netherlands. Neuromuscul Disord. 2006;16:396–402.

Full Text | Full-Text PDF (93 KB) | MEDLINE | CrossRef

47. 47Shy ME, Blake J, Krajewski K, et al. Reliability and validity

of the CMT neuropathy score as a measure of disability. Neurology.

2005;64:1209–1214.

48. 48Shy ME, Chen L, Swan E, et al. Neuropathy progression in

Charcot–Marie–Tooth disease type 1A (CMT1A). Neurology. 2008;70:378–

383. CrossRef

49. 49Li J, Bai Y, Ghandour K, et al. Skin biopsies in myelin-related

neuropathies: bringing molecular pathology to the bedside. Brain.

2005;128:1168–1177. CrossRef

50. 50Wegman AC, van der Windt DA, Stalman WA, de Vries TP.

Conducting research in individual patients: lessons learnt from two

series of N-of-1 trials. BMC Fam Pract. 2006;7:54. MEDLINE | CrossRef

51. 51Zucker DR, Schmid CH, McIntosh MW, D'Agostino RB, Selker HP,

Lau J. Combining single patient (N-of-1) trials to estimate

population treatment effects and to evaluate individual patient

responses to treatment. J Clin Epidemiol. 1997;50:401–410. Abstract |

Full-Text PDF (1087 KB) | MEDLINE | CrossRef