The classification of Charcot-Marie-Tooth diseases, a never-ending story: CMT4?

doi:10.1093/brain/awy207 BRAIN 2018: 141; 1–3 | e70 LETTER TO THE EDITOR The classification of Charcot-Marie-Tooth diseases, a never-ending story: CMT4? Jean-Michel Vallat,1,2 Meriem Tazir,3 Laurent Magy,1,2 Gwendal Le Masson4,5 and Stéphane Mathis4,5 Correspondence to: Jean-Michel Vallat Department of Neurology, University Hospital Dupuytren, 2 avenue Martin Luther King, 87042 Limoges, France E-mail: Sir, We read with interest the recent findings of Rebelo et al. detailing the observation of two unrelated patients with earlyonset axonal predominantly motor polyneuropathy, carrying compound heterozygous variants in the SCO2 (cytochrome c oxidase assembly) gene (Rebelo et al., 2018). Homozygous mutations in SCO2 are known to produce fatal infantile cardiomyopathy (OMIM 604377) (Papadopoulou et al., 1999), whereas heterozygous SCO2 mutations are a cause of highgrade myopia (OMIM 608908) (Tran-Viet et al., 2013). The two unrelated, sporadic patients described by Rebelo et al. harbour missense mutations (recessive transmission) near the conserved copper-binding motif (a common pathogenic variant E140K and a novel change D135G), which could explain the distinct phenotype observed. Moreover, the authors clearly show that SCO2 mutations did not appear to affect mitochondrial respiration, but caused severe copper deficiency in fibroblasts (Rebelo et al., 2018). As a consequence, they make a parallel between SCO2 and ATP7A [ATPase Cu (2 + )-transporting alpha polypeptide], another gene linked to copper metabolism, which gives rise to distal motor neuropathy (OMIM 300489). This paper, underlining the importance of copper homeostasis in the peripheral nervous system, increases knowledge and furthers understanding of the pathophysiology of Charcot-Marie-Tooth diseases (CMT). It is another example of the expanding number of genes implicated in the genesis of CMT. As we have noted, this large number of genes, probably 4100 in a just a few years, represents a problem for the classification of CMT (Mathis et al., 2015). At the present time, the classification of CMT is based on electrophysiological findings (motor nerve conduction velocities, MNCV) and mode of inheritance; therefore, CMT patients are classified as autosomaldominant (AD) demyelinating (CMT1; MNCV 538 m/s) or axonal (CMT2; MNCV 438 m/s) neuropathy. However, autosomal-recessive (AR) forms of CMT2 (‘ARCMT2’) have also been reported (AR and AD forms of CMT2 were finally grouped under the same name ‘CMT2’ in the current classification), as well as intermediate forms of CMT (MNCV between 30 and 40 m/s) (Mathis et al., 2015). Indeed, in 1993, several Tunisian cases of the AR form of demyelinating CMT were reported, called ‘CMT4’ by Ben Othmane et al. (1993) because, at that time, only three forms of CMT (CMT1, CMT2 and CMT3) were well identified: unfortunately, this new ‘CMT4’ designation was somewhat confusing. These CMT4 cases were characterized by slow motor nerve conduction velocities, severe hypomyelination upon nerve biopsy with basal lamina ‘onion bulbs’ and no myelin outfolding; since the transmission Advance Access publication August 1, 2018 ß The Author(s) (2018). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For permissions, please email: 1 Department of Neurology, University Hospital Dupuytren, 2 avenue Martin Luther King, 87042 Limoges, France 2 National reference center ‘neuropathies périphériques rares’, University Hospital Dupuytren, 2 avenue Martin Luther King, 87042 Limoges, France 3 Department of Neurology, University Hospital Mustapha Bacha, Algiers, Algeria 4 Department of Neurology, Nerve-Muscle Unit, CHU Bordeaux (Pellegrin University Hospital), University of Bordeaux, Place Amélie Raba-Léon, 33000 Bordeaux, France 5 National reference center ‘maladies neuromusculaires du grand sud-ouest’, CHU Bordeaux (Pellegrin University Hospital), University of Bordeaux, Place Amélie Raba-Léon, 33000 Bordeaux, France e70 | BRAIN 2018: 141; 1–3 Letter to the Editor Table 1 Current denomination of CMT and new proposals (adapted from Mathis et al., 2015) Demyelinating CMT Axonal CMT Current denomination Our proposal CMT1 (A, B, C, etc) CMT4 (A, B, C, etc) CMT2 (A, B, C, etc) AD-CMTde-gene AR-CMTde-gene AD-CMTax-gene AR-CMTax-gene AD-CMTin-gene AR-CMTin-gene XL-CMT-gene Intermediate CMT CMT-X (1, 2, etc) Figure 1 Sural nerve biopsy; transversal section, electron micrography. Myelin sheaths are too thin and surrounded by numerous myelin foldings, with the presence of ‘onion bulbs’ constituted of fragments of flat Schwann cell cytoplasms and of basal lamina. These lesions are induced by a MTMR2 mutation: in the current classification ‘CMT4B1’, which could be called ‘ARCMTde(or dys)-MTMR2’. was autosomal recessive, such cases should have been better described as ‘AR-CMT1’ (Vallat et al., 2016b), which means recessive transmission and a demyelinating process. Figure 1 shows a nerve biopsy from one of our ‘CMT4’ cases induced by a c.331dupA (p.Arg111LysfsX24) mutation of the MTMR2 gene: it shows a dysmyelinating process with numerous myelin outfoldings; as we will discuss later, we would classify this patient as ‘AR-CMTde-MTMR2’ rather than ‘CMT4B1’. Finally, the increasing knowledge of CMT has led to many changes in the way in which CMT is classified since the first bases of the current classification were established in the 1970s. Rebelo et al. described their two patients as having ‘CMT4’ (Rebelo et al., 2018), but they consider that both patients have an axonal polyneuropathy; in view of the axonal process, they should not be called ‘CMT4’ but ‘CMT2’, and more precisely ‘AR-CMT2’. All so-called ‘CMT4’ patients clearly present a primary demyelinating process, such as CMT4K (linked to mutations of SURF1, Surfeit 1), cited in the paper of Rebelo et al. (EchanizLaguna et al., 2013). In fact, it would be probably more correct to discuss a dysmyelinating process rather than a demyelinating one, as these ‘CMTde’, both ‘AD-CMT1’ (or ‘CMT1’) and ‘AR-CMT1’ (or ‘CMT4’) are induced by gene mutations that probably affect normal myelination. Unfortunately, many often chaotic modifications and discussions in the literature have led to an over complex classification, thus increasing the risk of an error of denomination (Mathis et al., 2016). Finally, considering SCO2-related CMT as AR-CMT2, it becomes difficult to include it in the current classification of CMT2 because CMT2Z is still attributed to MORC2-related CMT (OMIM 616688). As a consequence, it is now impossible to add other forms of CMT2 in the current classification. For these reasons, we have proposed a new way of classifying CMT. Our aim was to define the mode of transmission, the type of disease (‘CMTde’ for ‘demyelinating CMT’, ‘CMTax’ for ‘axonal CMT’, or even another phenotype such as ‘SPG’ for ‘hereditary spastic paraplegia’), then the causative gene (Vallat et al., 2016a). Thus, it becomes possible to name (...truncated)