Charcot-Marie-Tooth disease: CMT2, CMT4, and others

Francisco de Assis Aquino Gondim MD MSc PhD (

Dr. Gondim of Universidade Federal Ceará & Unichristus, Fortaleza, Brazil, received consulting fees from Alnylan.

Florian P Thomas MD MA PhD MS (

Dr. Thomas of Hackensack University Medical Center, Hackensack Meridian School of Medicine at Seton Hall University, received consulting fees from Pharnext and Acceleron.

Louis H Weimer MD, editor. (

Dr. Weimer of Columbia University has received consulting fees from Roche.

Originally released January 5, 2004; last updated December 4, 2018; expires December 4, 2021

This article includes discussion of multiple subtypes of CMT2, CMT3, CMT4, congenital hypomyelination neuropathy, Dejerine-Sottas syndrome, Dejerine-Sottas syndrome, DI-CMT, DI-CMT, HMSN3, and HMSN-Lom. The foregoing terms may include synonyms, similar disorders, variations in usage, and abbreviations.

Historical note and terminology

Hereditary peripheral neuropathies were first described independently by Charcot and Marie in France (Charcot 1886; Marie 1906) and by Tooth in England (Tooth 1886) and have become known as Charcot-Marie-Tooth diseases. Several earlier descriptions had been published, including a 6-generation pedigree (Eichhorst 1873) and a clinicopathological study (Friedreich 1873).

The heterogeneous nature and different forms of inheritance of the condition were soon appreciated. In 1889, Herringham recognized a family with X-linked Charcot-Marie-Tooth disease (Herringham 1889). Dejerine and Sottas described a more severe infancy-onset disease, which now bears their name (Déjerine and Sottas 1893), and Roussy and Levy described cases associated with tremor (Roussy 1926) that were defined genetically (Auer-Grumbach et al 1998; Plante-Bordeneuve et al 1999). Different forms of inheritance were later categorized (Allan 1939).

With the advent of modern neurophysiologic testing in the late 1960s, Charcot-Marie-Tooth disease was divided into 2 groups, 1 with slow nerve conduction velocities and histologic features of a hypertrophic demyelinating neuropathy (HMSN1 or CMT1) and another with relatively normal velocities and axonal and neuronal degeneration (HMSN2 or CMT2) (Dyck 1968; Thomas and Calne 1974; Buchthal and Behse 1977). The features of CMT1 and CMT2 patients were outlined in 2 landmark publications detailing the genetic and clinical characteristics of over 200 patients (Harding and Thomas 1980a; Harding and Thomas 1980b). CMT1 patients had median motor nerve conduction velocities below 38 m/sec, and CMT2 patients had velocities above 38 m/sec. As a dividing value between both forms, nerve conduction velocities of 38 m/sec are used by some and nerve conduction velocities of 42 m/sec by others (Harding and Thomas 1980b; Kaku et al 1993). Nerve conduction velocities in CMT1 patients typically are uniformly slowed along individual nerves and between different nerves of an individual patient, distinguishing CMT1 patients from those with acquired demyelinating neuropathies such as Guillain Barré syndrome or chronic inflammatory demyelinating polyneuropathy (Lewis and Sumner 1982; Kaku et al 1993).

Although the separation of neuronal and non-neuronal forms is an important etiologic and pathogenic distinction, even in CMT1, the clinical deficits correlate better with progressive axonal degeneration than with slowed nerve conduction. This is not surprising considering that demyelination secondarily disturbs axonal structure and transport. In some CMT families, patients have median motor nerve conduction velocities of 25 to 45 m/sec, and thus, are difficult to classify by electrophysiological criteria. This type of Charcot-Marie-Tooth disease was designated “intermediate CMT” (Davis et al 1978). The distinction between demyelinating and nondemyelinating Charcot-Marie-Tooth disease was further challenged by reports of relatively normal conduction velocities suggestive of CMT2 in younger members of a family with a myelin protein zero mutation, whereas older relatives had severely slowed conduction consistent with CMT1 (De Jonghe et al 1997).

Despite clinical similarities among CMT1 patients, it was soon discovered that the group was genetically heterogeneous, as linkage studies demonstrated CMT1 loci on both chromosome 1 (Bird et al 1982) and chromosome 17 (Raeymaekers et al 1989; Vance et al 1989; Middleton-Price et al 1990). In 1991, 2 groups showed that CMT1A, the most common form of CMT1, was associated with a 1.5 mB duplication within chromosome 17p11.2 (Lupski et al 1991; Raeymaekers et al 1991). Some 70% of CMT1 and 90% of CMT1A cases result from this duplication (Brice et al 1992; Hallam et al 1992; Wise et al 1993; Patel and Lupski 1994). Mutations in the PMP22 gene, contained within the 1.5 kB duplication on chromosome 17, have been demonstrated to cause demyelinating neuropathies in Trembler and Trembler-J mice (Suter et al 1992a; Suter et al 1992b) as well as in some families with a CMT1 phenotype (Valentijn et al 1992; Roa et al 1993; Nelis et al 1994). Moreover, transgenic mice and rats that over-express PMP22 develop neuropathies resembling CMT1 (Huxley et al 1996; Magyar et al 1996; Sereda et al 1996); therefore, it is now believed that the extra PMP22 gene copy within the 1.5 mB duplication on chromosome 17 causes the majority of cases of CMT1A. CMT1A also occurs with partial or complete trisomy for the short arm of chromosome 17, as part of a multiorgan phenotype with developmental and growth delay, craniofacial and skeletal anomalies, and heart defects (Fernandez-Torre et al 2001; Stankiewicz et al 2001).

The second most common subtype, X-linked recessive CMTX1, was found to result from mutations in the gap junction protein beta 1/connexin-32 on chromosome Xq13.1 (Bergoffen et al 1993). Two other loci for X-linked recessive Charcot-Marie-Tooth disease have been proposed: Xp22.2 for CMTX2 and Xq26 for CMTX3 (Ionasescu et al 1992).

The 1990s also saw the identification of other Charcot-Marie-Tooth genes. CMT1B and some cases of Déjerine Sottas syndrome, known to be linked to chromosome 1q22-q23 (Lebo 1998), were found to be associated with mutations in the myelin protein zero gene (Hayasaka et al 1993; Kulkens et al 1993; Su et al 1993). Mutations in the zinc-finger domain containing transcription factor early growth response 2 gene (EGR2 or Krox20) on chromosome 10q21.1-q22.1 were linked to CMT1D, Déjerine Sottas syndrome, and congenital hypomyelinating neuropathy (Warner et al 1998). Homozygous EGR2 knockout mice show peripheral hypomyelination and block of Schwann cells (Warner et al 1998). Deletion of the PMP22 gene locus was associated with hereditary neuropathy with liability to pressure palsies and several other phenotypes (Chance et al 1993). A similar condition, hereditary brachial plexus neuropathy (or hereditary neuralgic amyotrophy with predilection for the brachial plexus) is not linked to the PMP22 locus but was mapped to chromosome 17q25 (Chance et al 1994; Pellegrino et al 1997). Mutations of all of these genes have been associated with several overlapping clinical phenotypes. For instance, Déjerine Sottas syndrome is associated with PMP22, Cx32, or myelin protein zero mutations or deletions (Nelis et al 1996; Warner et al 1996; De Jonghe et al 1997; Reilly 1997; Warner et al 1998).

Peripheral myelin protein 22 relative to the myelin membrane with characteristic mutations indicated Image: Peripheral myelin protein 22 structure

After the initial description of the most common autosomal dominant CMT1 and CMT2 phenotypes, the quest for the most severe and rare phenotypes led to the diagnosis of autosomal recessive variants. Demyelinating autosomal recessive phenotypes are called CMT4 (or alternatively ARCMT1), whereas autosomal recessive axonal variants are called ARCMT2. In regions where consanguinity is important, up to 30% of the CMT cases are autosomal recessive (Tazir et al 2013). These forms overlap with DSS (CMT3) and CHN phenotypes. Several new disease linkages and genes have been identified. As of this review, at least 50 different CMT genes have been recognized, including 14 autosomal dominant CMT2 variants and 11 CMT4 subtypes (Table 1).

Several loci have been identified in families with dominant intermediate Charcot-Marie-Tooth disease, ie, autosomal dominant Charcot-Marie-Tooth disease with conduction velocities between 24 and 45 m/sec. These include DI-CMTB on chromosome 19p12-p13.2 (Kennerson et al 2001) and DI-CMTA, which is associated with both large fiber loss and regeneration clusters as well as onion bulbs and uncompacted enlarged myelin lamellae on chromosome 10q24.1-q25.1 (Malandrini et al 2001; Verhoeven et al 2001). A recessively inherited severe form of Charcot-Marie-Tooth disease with intermediate conduction velocities has been linked to chromosome 10q23 (Rogers et al 2000). Intermediate conduction velocities also occur with myelin protein zero and neurofilament light subtype gene mutations (De Jonghe et al 1999; De Jonghe et al 2001). We reported that DI-CMTC is caused by disrupted function and axonal distribution of mutant tyrosyl-tRNA synthetase, in 3 unrelated families in the United States and Bulgaria (Jordanova et al 2003b; Jordanova et al 2006).

Since the early 1990s, over 100 genes have been found to be defective in Charcot-Marie-Tooth disease patients. Discovery has been facilitated by diagnostic tools such as target-enrichment next-generation sequencing with copy number assessment (Wang et al 2016). The boundaries between CMT1 and CMT2 are less clear than originally believed and the spectrum of pathological mechanisms involved in these conditions is growing. Even when mutations reside in the nuclear genome, some CMT subtypes represent mitochondrial disorders, ie, the encoded proteins function in the mitochondria (Züchner and Vance 2006). Bienfait and colleagues illustrated the major challenges of diagnosing CMT2 in a large series of CMT2 patients (N=61, 18 families), where they found mutations in only 3 families (Bienfait et al 2007). They stressed the difficulties in clinically distinguishing CMT1 from CMT2. CMT2 had later disease onset, less complete areflexia, foot deformities, and weakness of knee extensors and foot dorsal flexors (Bienfait et al 2006). Rapid discovery of new genes has dramatically changed the field for several reasons. First, different mutations in a single gene can lead to CMT1, CMT2, DI-CMT, or CMT4 phenotypes, as well as milder or more severe phenotypes. Second, there is accumulating evidence for genes that can modify and complicate phenotype-genotype relations of known mutations. Third, the number of mutated genes leading to different variants has surpassed the number of letters in the alphabet; therefore, various new nomenclatures are being considered that rely on gene mutations and inheritance pattern rather than clinical phenotype (Magy et al 2018).

This article focuses on CMT2, intermediate CMT, and several other subtypes. Inherited neuropathies in which autonomic or sensory features predominate, conditions in which the neuropathy is part of a multiple-organ disturbance, and neuropathies with specific metabolic dysfunction are not discussed. For discussion of CMT1A, CMT1B, CMTX1, and HNPP the reader is referred to the summaries devoted to those subtypes (Charcot-Marie-Tooth disease type 1A; Charcot-Marie-Tooth disease type 1B and mutations of the myelin protein zero; Charcot-Marie-Tooth disease type X; Hereditary neuropathy with predisposition to pressure palsy).

The content you are trying to view is available only to logged in, current MedLink Neurology subscribers.

If you are a subscriber, please log in.

If you are a former subscriber or have registered before, please log in first and then click select a Service Plan or contact Subscriber Services. Site license users, click the Site License Acces link on the Homepage at an authorized computer.

If you have never registered before, click Learn More about MedLink Neurology  or view available Service Plans.