Treatment and Management of Hereditary Neuropathies

Time Course

While HMSN and CMT are characteristically symmetric and slowly progressive, it is important to recognize that not all inherited neuropathies fit this pattern. Patients with hereditary neuropathy with liability to pressure palsies (HNPPs) may present with acute onset of recurrent painless, focal neuropathy at usual sites of entrapment (median at wrist, ulnar at elbow, and peroneal at knee). Similarly, patients with hereditary neuralgic amyotrophy (HNA) (caused by mutations in SEPTIN9 ) may present with acute onset of pain and weakness localizing to the brachial plexus of one arm. Patients with SCN9A-related inherited erythromelalgia may have pain, redness, warmth, and swelling in the feet occurring in an episodic manner. Acute intermittent porphyria (AIP) (caused by mutations in HMBS ) may present with acute onset of weakness in the arms and legs, along with abdominal pain and mental status changes. There can be episodic worsening in the distribution of affected nerves during attacks. Genetic variants can cause neuropathy through many divergent mechanisms, and it is fitting that there is a wide range of presentations.

Distribution

Most inherited neuropathies present with length-dependent symmetric sensorimotor symptoms and signs. However, onset in hands, vocal cord paresis, deafness, optic atrophy, and non-length-dependent sensory ataxia may be the presenting symptoms. These atypical features can suggest specific genetic causes ( Table 14.2 ).

Fiber Types

Most inherited neuropathies involve sensory and motor fibers in a length-dependent manner (HMSN). However, pure motor neuropathies (dHMN), small fiber neuropathies (hereditary sensory neuropathy [HSN]), pure sensory large fiber neuropathies (ataxic neuropathies), small fiber sensory neuropathies (Fabry, amyloidosis, inherited erythromelalgia), autonomic neuropathies (HAN), and sensory and autonomic neuropathies (HSAN) need to be recognized.

Electrophysiology

The original subdivision into CMT1 and CMT2 relied on the presence or absence of slowing of nerve conduction velocity below 38 m/second in an upper extremity motor nerve. Intermediate nerve conduction velocities (between 25 m/second and 45 m/second) occur with CMTX and HNPP (and other intermediate CMTs). Most inherited neuropathies cause uniform and diffuse involvement. This is in contrast to acquired neuropathies, where the findings are nonhomogeneous, can vary between adjacent nerves, and can vary within different segments of the same nerve. Conduction block, a reflection of nonhomogenous involvement within a nerve, is generally a feature of acquired demyelinating neuropathies but may occur with CMTX and CMT1B. In addition to defining an axonal or demyelinating pattern, electrophysiology is helpful in detecting relative involvement of sensory and motor nerves, whether the neuropathy is length dependent or length independent, and in characterizing the severity.

Genetic Studies

Genetic testing is usually best done along with a genetic counselor who can discuss the risks and benefits of genetic testing with multiple family members. Patients need to be well educated regarding the implications of a molecular diagnosis, both for themselves and for family members. Genetic testing is of utmost relevance to family members of childbearing age, because for most well-defined monogenic disorders, preimplantation or in utero genetic testing is available. It is worth noting that the first genetic nondiscrimination law (Genetic Information Nondiscrimination Act [GINA]) was signed into law in 2008, giving patients protection from discrimination by employers and health insurance companies. In most cases, unaffected minors should not be offered genetic testing to determine future health or reproductive risks, unless screening provides a clear and timely medical benefit with minimal psychosocial risks.

Although testing can be expensive and sometimes is not covered by insurance, genetic diagnosis is extremely valuable to many patients and clinicians. The balance and landscape of cost and gene coverage are constantly changing, and helping navigate this is one of the important roles genetic counselors play. The exponential decreases in cost of next generation testing have also made it so that unbiased exome sequencing can compete on price with more focused panel testing. This adds complexity to the decision of how and when to test, which has been covered in a new chapter in this edition ( Chapter 11 ). Carrier screening panels are also starting to include recessive and X-linked neuropathy genes, which may help decrease the incidence of some of the most severe forms of the disease.

Having a molecular diagnosis continues to have many benefits. First, it can potentially give a definitive diagnosis not available by any other means and can obviate invasive testing (e.g., nerve biopsy) and unnecessary treatment (e.g., intravenous gamma globulin). More recently, the advent of FDA-approved therapies for hereditary transthyreitin amyloidosis (hATTR) has also introduced the possibility that genetic diagnosis could identify a potentially treatable cause of neuropathy. This has been another driver toward affordable panel and exome testing, spurring collaborations between pharmaceutical companies and genetic diagnostic companies to defray the cost to patients and make it easier to test.

Knowing the cause of their symptoms can often empower patients to follow research progress, take part in natural history studies, and enroll in clinical trials. This chapter will discuss genetic tests that are commercially available. The National Center for Biotechnology Information’s Genetic Testing Registry Web site ( https://www.ncbi.nlm.nih.gov/gtr/ ) is a valuable resource for identifying gene tests, both those available clinically and those performed on a research basis, and includes online reviews of most diseases (GeneReviews).

Imaging Studies

Magnetic resonance imaging (MRI) studies are able to show increased size of the nerves in CMT patients and normal magnetic resonance signal without enhancement. This is opposed to acquired neuropathies such as chronic inflammatory demyelinating polyneuropathies (CIDPs) in which enlarged nerves are associated with enhancement or hyperintensity of the T2 signal. Nerve root hypertrophy may cause compressive myelopathy or radiculopathy in CMT patients. Ultrasound is being used more frequently in clinical practice to show that the cross-sectional areas of the nerves are increased.

Treatment

Unfortunately, most syndromic and nonsyndromic inherited neuropathies do not yet have specific pharmacologic therapies available. Given that the first genes causing CMT were first discovered in the early 1990s, and given the magnitude of discoveries achieved since then, there is certainly reason to be optimistic that therapies will continue to enter clinical trials and the clinic in the coming years.

In addition to targeted potentially disease-modifying therapies, symptomatic therapies are available, mostly to treat the orthopedic complications that arise. These psychiatric and surgical therapies are covered in detail in Chapter 8, Chapter 9 , respectively, and will only briefly be discussed here as they relate to specific disorders. Neuropathic pain is uncommon in most inherited neuropathies and is treated in the same way as in acquired neuropathies; however, in a few inherited neuropathies, pain may be the presenting symptom, particularly in Fabry disease and FAP. Chapter 6 discusses treatment of painful neuropathy.

Importantly, there are several patient support groups that can offer valuable services to patients and their families. The Muscular Dystrophy Association ( www.mdausa.org/ ) supports patients with CMT and FRDA, and this organization can be very helpful to patients. Other patient resources include the Charcot-Marie-Tooth Association ( www.charcot-marie-tooth.org/ ), the Hereditary Neuropathy Foundation ( www.hnf-cure.org/ ), and the Neuropathy Association ( www.neuropathy.org/ ).

Gene Replacement Strategies

Loss-of-function mutations can be treated with gene or protein replacement strategies, whereas gain-of-function mutations require a strategy to reduce the toxicity of the mutated or duplicated gene product (e.g., antisense oligonucleotides [ASOs] in CMT1A). There are viral and nonviral gene replacement strategies ( Fig. 14.3 ). Adeno-associated viral vectors have been the first to reach the clinic, with FDA-approved AAV9 vectors for SMA and AAV2 vectors for RPE65 -linked retinal dystrophy. The adeno-associated virus (AAV) serotype reflects different capsid proteins and influences viral tropism, infectivity, and immunogenicity. The mode of administration also influences the therapeutic dose, distribution, and potential toxicity. AAV2- RPE65 is delivered by subretinal injection ( ), for example, while AAV9-SMN1 is delivered intravenously ( ). Many groups are working to improve the tropism of AAV serotypes to the central and peripheral nervous system. Several AAV subtypes have also been shown to effectively infect motor and sensory neurons with local injection, intrathecal injection, and systemic administration, although delivery to Schwann cells has proven more difficult.

Nonviral gene delivery systems are also under preclinical development but have not yet been taken to the clinic for neuromuscular diseases. These approaches often require liposomal vehicles to facilitate transfer of mRNA or linear or plasmid DNA. These approaches would require repeat dosing but do not have the risks associated with immune response to the virus, and these approaches do allow dose titration and re-dosing.

Antisense Oligonucleotide And Rnai-Based Approaches (This And Gene Therapy Are Discussed In Detail In Chapter 11 )

The success of nusinersen for the treatment of SMA has further increased enthusiasm for ASO therapies for neuromuscular diseases and neuropathies. Nusinersen binds to an intronic sequence in the SMN2 gene to prevent splicing of exon 7 and encourage production of full-length protein ( ). Eteplirsen binds to exon 51 in transcripts from the Duchenne muscular dystrophy-associated gene (DMD) and facilitates skipping of mutant exon 51 containing a premature termination codon or frameshift and results in restoration of the reading frame ( ). This demonstrates two ways ASOs can be used to restore protein function by modulating splicing in opposite ways, exon inclusion and exon skipping. ASO therapies can also be used to facilitate RNaseH-mediated degradation of target mRNA.

Small interfering RNA (siRNA) and short hairpin RNA (shRNA) therapies are less versatile because they facilitate RNA-induced silencing complex (RISC)-mediated mRNA degradation and cannot be used to modulate splicing or expression. shRNAs are generally delivered in vectors and converted into siRNAs by Dicer. siRNA, shRNA, and ASOs can be used to drive degradation of mRNAs before they are translated into toxic proteins. Both the first FDA-approved siRNA therapy, patisiran, and one of the first FDA-approved ASO therapeutics, inotersen, target mutant and wild-type hTTR. Allele-specific knockdown of only the mutant allele using ASO, siRNA, or shRNA therapies is possible, but achieving efficient knockdown with good specificity has proven challenging for missense variants and only a few such examples have been published ( ). Allele-specific knockdown also requires more tailored therapeutics development to match missense changes or associated single nucleotide polymorphisms (SNPs) linked to disease. Nevertheless, it is an appealing approach in diseases caused by mutations in genes that are required for survival and function of the targeted cell type (e.g. MFN2, AARS).

Small Molecule Approaches

While rational and platform-based biologic therapeutic approaches have stolen much of the spotlight in recent years, there continue to be advantages to small molecule approaches. Most notably, the parameters for bioavailability, tissue access, and penetrance are well known to drug developers. In addition, massively parallel screening is possible with small molecules. There are generally two approaches to small molecule drug discovery: target-based approaches, and phenotype or expression-screening approaches. With confidence in the target and mechanism, several complementary approaches are often used to identify the best medicines. The technologies for drug discovery are constantly improving and multiple approaches are often used simultaneously to find hits, to verify hits and characterize toxicity to identify lead compounds, and to perform lead optimization to select compounds ready for clinical testing.

Once a mechanism of pathogenesis is proposed and animal models are developed, the drug development process can begin using either a target-based approach or a phenotype-driven approach (see Fig. 14.1 ). Target-based approaches utilize our understanding of the basic biology and disease pathogenesis to screen molecules for a desired effect on a protein target. Transthyreitin protein in the liver functions as a tetramer, and disease pathogenesis requires dissociation of monomeric protein from the tetramer and release from the liver. Several groups performed screens for small molecules that increase the stability of the transthyreitin tetramer. Increasing its stability prevents dissociation of the monomer from its tetrameric structure, release of the mutant monomer from the liver, and accumulation of the amyloidogenic monomer in heart and peripheral nerve.

Another approach is phenotype or expression-based screening. For example,

PMP22 duplication is the most common cause of CMT and results in excess expression levels of PMP22. Researchers performed an expression-based screen for FDA-approved molecules that reduce PMP22 expression. The result was the identification of three FDA-approved compounds (fenretinide, olvanil, and bortezomib) that reduced PMP22 mRNA and protein ( ). The criticism of phenotype and expression-based systems is that it is difficult to determine the molecular target of the effect, and therefore, it is difficult to improve upon the molecule in the ways that traditional drug development is performed. To our knowledge, these compounds have not been tested clinically. Bortezomib, one of the compounds that were found to decrease PMP22 expression in this screen, is known to cause a severe acquired neuropathy and would therefore be a poor candidate for clinical testing.

Drug development in other therapeutic areas may identify compounds that will slow disease progression of hereditary neuropathies. Therapeutics targeting generally neuroprotective pathways will be promising for neuropathy patients. Since the publication of the last edition, our understanding of the pathways mediating axon degeneration has improved significantly. Axodegenerative programs mediated by the SARM1 protein that trigger Wallerian degeneration are attractive therapeutic targets and, if found, may serve as effective therapies in many hereditary neuropathies. Conversely, candidate compounds with general neuroprotective or neuroregenerative properties identified in research efforts for hereditary neuropathies may also be effective in acquired neuropathies.

Gene Editing

Gene editing has great therapeutic potential, but many challenges remain before it will be a viable therapeutic approach in hereditary neuropathy. Even with gene editing, delivery of the editing proteins to relevant tissues will continue to be a challenge and will rely on the successes of gene therapies that precede them to ensure delivery of Cas9 and related proteins and short guide RNAs. Another caveat is that gene editing targets that require correction rather than simply elimination of an exon or repeat will not be early targets because of the challenges of recombination-based repair and correction. Second- and third-generation editing tools may skirt the requirement for repair or recombination by performing specific substitutions in DNA or RNA, but transformation of the cells with exogenous editing proteins will still be early hurdles. As a result, hereditary neuropathies will not be early targets for CRISPR-based approaches.

Mechanism Agnostic Approaches

Some drug development paradigms for dominant diseases have skirted the need to definitively parse the contributions of gain and loss of function. These include gene editing, and a “knockdown and replace” strategy ( Fig. 14.3 ). In addition to the hurdles discussed in the section on gene editing earlier, replacing portions of genomic DNA with wild-type sequence will require either homology-directed repair with an oligo template or homologous recombination with a double-stranded template. The efficiencies of these approaches will need to be improved before they can be used therapeutically.

The “knockdown and replace” approaches use viruses to deliver oligonucleotides that knock down endogenous mRNA and simultaneously deliver gene replacement with transcripts that harbor synonymous mutations that make them impervious to knockdown. Several CMT2 genes are essential genes and lend themselves to this approach.

Clinical Development

After candidate drugs are identified and undergo rigorous toxicity, efficacy, and dosage testing in animal models, they can then be tested in clinical trials. An often-underappreciated part of drug development to complement bench science and preclinical development is concomitant definition and characterization of natural history, biomarkers, and quantitative disease metrics ( Fig. 14.1 ). This is critical so that investigators can identify subpopulations whose disease progresses rapidly enough to show an effect in a feasible study duration.

Most drugs fail in phase 2 clinical studies, where efficacy signal is needed for advancement. The reasons are multifold, but without a good understanding of a disease’s natural history, selection of the correct patient population, ability to predict effect size, and adequate power, even the most effective preclinical therapies will fail and result in disappointment for patients and clinicians. These are particularly challenging obstacles for treatments for inherited neuropathies. These diseases individually are rare, and adequately powering a study can be challenging. The development of large consortia that can enroll many patients from sites around the world is now allowing much larger studies to be performed with more uniform tracking of outcomes in many modalities. The slowly progressive nature of many of these disorders with varied and clinical parameters that can be difficult to measure makes it difficult to quantitate the treatment effects. Another obstacle to the advancement of biologic therapies into the clinic has been the absence of tools to facilitate large molecule delivery to motor and sensory neurons and Schwann cells.

With these in mind, it is not surprising that hTTR amyloidosis has been an attractive target. The liver, the source of amyloid protein production, is an easier tissue to target than anterior horn cells, dorsal root ganglion cells, or Schwann cells, for both small molecules and biologics, and the disease course is rapidly progressive based on existing natural history data.

The development of quantitative scales to measure disease severity and the ability to assay biomarkers that are relevant to disease pathogenesis are improving the quality of clinical trials in these diseases. A validated primary clinical outcome used in current studies is the CMT Neuropathy Score (CMTNS) ( ), which was improved with an updated second version (CMTNS2) ( ). There is also the overall neuropathy limitation score (ONLS) ( ). Oftentimes, both the CMTNS and ONLS are used concomitantly in clinical studies. There is also the modified Neuropathy Impairment Score +7 (mNIS+7), which is tailored to trials of FAP ( ).

Biomarkers are another important part of clinical development that can be used to measure target engagement or to more sensitively measure the desired therapeutic effect. There has been significant work to improve quantitative assays for neurofilament light (NfL) chain and to characterize serum and CSF levels in control and patient populations. In CMT, plasma NfL has been established as a reliable biomarker of disease activity that stratifies with disease severity and genetic subtypes ( ). There is also great interest in its use in many neurodegenerative disease of the central (e.g., multiple sclerosis and Alzheimer disease) and peripheral nervous systems.

Thus, through continued advancements in our understanding of inherited neuropathies, and building on the success of the FDA-approved therapy for hTTR amyloidosis and improvements in the design and implementation of clinical trials, effective therapies for inherited neuropathies can be expected to advance through the development pipeline.

Clinical Presentation

Most patients with CMT develop slowly progressive weakness and atrophy in their feet beginning in childhood or early adulthood. Pain or sensory loss is variable but is usually not a chief complaint. Foot deformities (pes cavus, or high arches, and hammertoes) are common and may lead to disability (see picture in Chapter 9 ). Mildly affected patients may be limited only in their ability to walk on their heels and play sports, whereas more severely affected patients will have foot drop and steppage gait (hence the initial description of this disease as peroneal muscular atrophy). However, due to the insidious nature of this disease, many patients do not complain of motor or sensory symptoms until late in the disease course, and most remain ambulatory. Thus, it is not uncommon to find asymptomatic family members with classical examination and electrophysiologic features of CMT.

On examination, patients typically have distal weakness and atrophy in the feet, areflexia, and length-dependent sensory loss of both large and small-fiber sensory modalities. More severely affected patients will develop sensory ataxia or tremor (Roussy-Levy syndrome), palpably enlarged nerves (demyelinating CMT), and weakness, atrophy, and sensory loss of the hands. Patients with motor findings only and without evidence of sensory involvement on examination or electrodiagnostic testing are classified as having HMN, whereas patients with sensory findings only and without evidence of motor involvement are classified as HSN or HSAN (see Table 14.2 and later). Less commonly, patients can have a severe form of CMT (CMT3) that is congenital (congenital hypomyelinating neuropathy [CHN]) or with onset before age 3 (Dejerine-Sottas syndrome [DSS]). Some forms of CMT have unusual features, such as central nervous system (CNS) involvement in CMTX, vocal cord paralysis with CMT2C, and specific thenar involvement in CMT2D (see Table 14.2 ).

Whereas CMT1A is caused by a duplication of the PMP22 locus, HNPP is caused by a deletion of the same locus ( ). As the name suggests, these patients often present with painless, recurrent entrapment neuropathies, often precipitated by minor trauma or compression. HNPP patients usually also have a mild length-dependent demyelinating neuropathy. Another disorder occasionally confused with HNPP is HNA. These patients present essentially the same as those with idiopathic brachial neuritis (Parsonage-Turner syndrome) with episodic, painful weakness and numbness of one upper extremity, often triggered by an infection, exercise, or stress.