Will new investigational drugs change the way we treat Charcot-Marie-Tooth disease?

KEYWORDS:

• Charcot-Marie-Tooth (CMT)
• clinical trials
• gene silencing
• gene therapy
• inherited neuropathies
• novel treatments
• surrogate biomarkers

1. Introduction

Charcot-Marie-Tooth disease (CMT), the most common hereditary neuropathy, currently lacks an FDA/EMA-approved drug, and its management still relies on rehabilitation therapy, surgery for skeletal deformities, and symptomatic treatment [1–3]. However, investigations into numerous drugs and approaches, holding the potential to reshape the CMT therapy landscape in the next decade or two, are ongoing. Indeed, several molecules acting on diverse pathways are under study for various CMT subtypes (Figure 1), although enthusiasm is presently tempered by practical challenges such as proper targeting of neurons (in axonal/neuronal CMT) or Schwann cells (in demyelinating CMT) and clinical trials design [1]. The latter is hindered by the broad genetic heterogeneity, variable disease expression, and difficulties in developing responsive outcome measures and biomarkers due to the disease’s slow progression.

Figure 1. Schematic representation of a motor neuron with its myelinated axon, neuromuscular junction and innervated muscle fiber. A transversal section of an internode with the myelinating Schwann cell is also shown. A list of potential treatments already tested, under evaluation, or under development is provided for each CMT type or related protein. [with permission from ‘Pisciotta C, Pareyson D. Gene therapy and other novel treatment approaches for Charcot-Marie-tooth disease. Neuromuscul Disord. 2023;33(8):627–635. doi:https://10.1016/j.Nmd.2023.07.001E28099’].

Gene therapy currently stands out as one of the most promising approaches, supported by increasingly positive pre-clinical data on gene replacement, insertion, and editing. Nonetheless, substantial limitations persist, particularly in the reliance on viral vectors, the effective targeting of neurons/Schwann cells, and the risk of off-target effects. Careful consideration is crucial when determining the delivery route in order to maximize efficacy and minimize toxicity. Adeno-associated viral (AAV) vectors are presently regarded as the best choice for gene delivery. However, the presence or development of antibodies against AAVs may hinder multiple administrations, and immune reactions could potentially lead to serious side effects, such as thrombocytopenia or hepatotoxicity [1,3].

2. Overview on clinical trials in CMT disease

Despite widespread efforts, only a few clinical trials are currently underway or have just been completed, with the main being: 1) a second phase III trial for CMT1A using PXT3003 – a mixture of baclofen, sorbitol, and naltrexone – to downregulate PMP22 expression (the PREMIER trial, NCT04762758, completed); 2) a phase II/III trial for CMT-SORD, caused by biallelic mutations in the sorbitol dehydrogenase (SORD) gene, using an aldose reductase inhibitor (AT-007) to reduce sorbitol levels (the INSPIRE trial, NCT05397665, ongoing); 3) a phase I/IIa trial for CMT2S involving the AAV9-mediated replacement of the defective IGHMBP2 gene through a single intrathecal injection (NCT05152823, ongoing); 4) a phase I/IIa study for CMT1A employing multiple injections in lower limb muscles of VM202, a plasmid DNA containing the gene encoding human hepatocyte growth factor, thought to act as a neurotrophic factor (NCT05361031, completed – results soon awaited); and 5) a phase II trial to test whether L-serine is an effective drug to slow or halt the disease progression in Hereditary Sensory Neuropathy 1 (HSN1) associated with mutations in the SPTLC1/2 genes (the SENSE trial, NCT06113055, ongoing).

Since CMT1A, the most prevalent CMT subtype, is thought to result from the overexpression of the PMP22 protein, the prospect of its downregulation has always appeared intriguing. PMP22 gene silencing achieved through various methods (ASO, siRNA, shRNA, miRNA, and CRISPR/Cas9 editing) has indeed yielded highly encouraging results in animal models [1–3] (see Table 1). In humans, the administration of PXT3003, deemed to be a PMP22 downregulator, resulted in a significant improvement in the Overall Neuropathy Limitations Scale (ONLS) score in a randomized double-blinded placebo-controlled phase III study involving 323 CMT1A patients [12]. Because of these promising results and its favorable safety profile, PXT3003 was initially regarded as the first potential treatment for CMT. Nonetheless, the FDA requested a confirmatory trial (PREMIER) due to concerns about potential biases arising from drug crystallization and the subsequent dropout of many patients from the high-dose arm. Unfortunately, a recent announcement disclosed that the just-completed novel PREMIER trial did not meet its primary endpoints, mainly due to improvement in both treatment and placebo arms [13]. Criticisms to the trials include the use of the ONLS score, known for its low sensitivity in detecting changes, and of sorbitol, carrying potential neurotoxicity. High levels of sorbitol are, in fact, known to be toxic in diabetic neuropathy and in another CMT subtype, namely CMT-SORD.

Table 1. PMP22 gene silencing approaches for treating CMT1A in animal models.

Approach	Method and route of administration	Animal model	Comment	Reference
ASO	Repeated subcutaneous injections	C22 mice, CMT1A rats	Need for repeated administration	[4]
Squalenoyl siRNA nanoparticles	Repeated intravenous injections	JP18 and JP18/JY13 mice	Need for repeated administration	[5]
Fatty-acid-siRNA conjugated	Repeated intravenous injections	C3 mice	Need for repeated administration	[6]
AAV2/9 shRNA	Single intraneural injection	CMT1A rats	Multiple sites of injection required; risk of nerve damage	[7]
LV-miRNA (miR-381)	Single intraneural injection	C22 mice	Risks associated with lentiviral vectors; risk of nerve damage	[8]
scAAV9 miRNA (miR-871)	Single lumbar intrathecal injection	C61 mice	Greater diffusion ability than by intraneural administration	[9]
CRISPR/Cas9 liposomal PMP22 TATA box RNP-mediated editing	Single intraneural injection	C22 mice	Risk of nerve damage	[10]
CRISPR/Cas9 PMP22 TATA box AAV-mediated editing	Single intravenous or intratechal injection	C22 and C3 mice, CMT1A rats	Greater efficacy in intravenous compared to intratechal administration	[11]

Efficacy demonstrated on behavioral, electrophysiological, and histological level for the majority of the approaches (only for the CRISPR/Cas9 approach no data on behavioral effects are reported). General remarks: need to target a sufficient number of Schwann cells avoiding at the same time potential unwanted off-target effects; risk of HNPP if excessive downregulation.

Abbreviations: AAV=adeno-associated viral (vector); ASO=antisense oligonucleotides; CRISPR/Cas9=clustered regularly interspaced short palindromic repeats/associated protein 9; HNPP=hereditary neuropathy with liability to pressure palsies; LV=lentiviral (vector); miRNA=microRNA; RNP=ribonucleoprotein; shRNA=short hairpin RNA; siRNA=short interfering RNA.

In the latter, biallelic mutations in SORD cause a loss of function in the enzyme converting sorbitol to fructose. This malfunction results in increased sorbitol levels, which ultimately lead to a predominantly motor neuropathy. A promising approach is to use drugs inhibiting aldose reductase, which converts glucose into sorbitol and acts before the SORD-mediated step, to reduce the production of sorbitol. The biochemical rationale for the employment of newer generation aldose reductase inhibitors to prevent sorbitol toxicity appears robust. Preclinical studies have demonstrated that aldose reductase inhibitors reverse the disease phenotype in a Drosophila model of SORD deficiency and are able to reduce the intracellular sorbitol levels in patient-derived cultured fibroblasts [14,15]. Moreover, an interim analysis of the ongoing phase II/III INSPIRE trial using AT-007, a novel aldose reductase inhibitor, revealed a 52% mean reduction in serum sorbitol levels in treated patients, confirming the correct target engagement of the drug [15,16]. The challenge in this case involves confirming the effective translation of preclinical and biochemical findings into clinical efficacy.

Another interesting example of a biochemical correction strategy in metabolic neuropathies, with a strong rationale, is the oral supplementation of L-serine in SPTLC1/2-related HSN1. SPTLC1/2 mutations cause serine palmitoyltransferase to develop a higher affinity for alanine and glycine than for serine in the sphingolipids’ synthesis pathway, leading to the production of neurotoxic deoxysphingolipids. Unfortunately, despite preclinical data suggesting that L-serine supplementation reduces deoxysphingolipid levels, a pilot trial on 18 patients did not meet its primary endpoint, probably due to the small sample size [17]. Nevertheless, the analysis revealed a clear, though non-significant, trend toward drug efficacy (p = 0.09), prompting a new, larger trial (NCT06113055) which is under way in the UK. It is also interesting to note that decrease of deoxysphingolipids is emerging as a target also in diabetic neuropathy in mice [18].

As for gene replacement therapy, the only ongoing clinical trial involves replacing the defective IGHMBP2 gene through a single intrathecal AAV9-mediated administration in CMT2S, a rare and severe early-onset form of CMT allelic to SMARD1 (Spinal Muscle Atrophy with Respiratory Distress type I). Encouraging results have just been published from a similarly designed trial involving intrathecal delivery of the AAV9-embedded gigaxonin gene in 14 patients with giant axonal neuropathy (GAN); the study demonstrates a potential benefit in motor function scores and other outcomes [19]. As mentioned earlier, gene replacement therapy using AAV-9 is not without risks and yields optimal results when antibodies against AAVs are absent, which is often the case in children. Its feasibility in treating a severe pediatric-onset CMT form thus offers a favorable balance between risks and benefits. However, its application appears still limited in other CMT forms, particularly those with adult onset and milder severity, where the risk/benefit ratio may not be as advantageous until safer and more effective techniques are developed.

All the discussed approaches share a common, notable, limitation: they are specific to certain CMT subtypes and, therefore, not transversally applicable. Additionally, another major challenge in CMT therapy is to address axonal degeneration, which is the final common pathway of both primary axonopathies and myelinopathies. To date, no therapy has proven effective in reversing this degeneration. However, SARM1 is a known mediator of axonal degeneration and might play a role at least in some CMT types [20]; therefore, SARM1 inhibitors are being actively developed. Another recently proposed approach involving histone deacetylase-6 (HDAC6) inhibitors has shown promise, primarily due to their impact on microtubules and heat shock protein 90 acetylation resulting in improved axonal transport and reduced cellular stress. HDAC6 inhibitors seem able to reverse motor and sensory deficits in different axonal and demyelinating CMT mouse models by improving axonal transport, mitochondrial function, protein folding and even myelination [21–25], thus holding the potential to be applied to all CMT types, irrespective of the specific underlying mutation and regardless of whether the primary damage occurs in the axon or the myelin. Phase I studies are underway (see NCT03713892).

3. Expert opinion

In summary, therapy availability seems possible within the next two-three years for metabolic neuropathies like CMT-SORD or SPTLC1/2-HSN1, where correction of the biochemical defect may be achieved. For other forms of CMT, this goal seems to be more distant, with the greatest advancement expected in CMT1A by PMP22 silencing within the next five years, as phase I/IIa trials are planned. Recent success in the treatment of transthyretin-related amyloidosis offers promising possibilities for progress in this field, mainly driven by technological advancements. However, PMP22 downregulation needs to be strictly controlled, as an overly excessive reduction poses the risk of developing hereditary neuropathy with liability to pressure palsy (HNPP). In the medium term, gene therapy emerges as a great hope for treating selected CMT subtypes, thanks to the continuous advancements that improve safety, vectors, and efficient delivery.

It remains crucial to refine clinical trial design, as certain treatments might not meet the clinical endpoints, but they could prove effective when employing more responsive surrogate measures, such as blood biomarkers and quantitative muscle MRI to accurately assess denervation-related fat fraction. The CMT research community is working on developing responsive surrogate biomarkers that correlate with clinical disability and predict clinical outcomes [26]. If accepted by regulatory authorities, these biomarkers could play a pivotal role in the registration of novel treatments.

In conclusion, we are confident that the first treatments for at least some CMT types will be available within the next five years, ushering in a revolution in CMT management and prognosis during the 2030s. Accelerating genetic diagnosis and deeply characterizing the neuropathy will be crucial to achieve personalized treatment and to initiate it ideally in the earliest phases of the disease or even before its onset.

Declaration of interests

Amedeo De Grado, Chiara Pisciotta and Paola Saveri have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. Davide Pareyson acknowledges payments or reimbursement for consultancy and participation in Advisory Boards of Inflectis, Alnylam, Akcea, Arvinas, Augustine Tx, DTx, speaker honorarium from Alnylam, participation as local PI in clinical trials sponsored by Alnylam, Ionis, AT Therapeutics.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.