|Home | About | Journals | Submit | Contact Us | Français|
We describe a patient with both Neurofibromatosis type 1 and Charcot-Marie-Tooth disease type 1B. While one might expect an overwhelming tumor burden due to the combination of these two disorders, the two mutations did not appear to interact.
Neurofibromatosis type 1 (NF1; OMIM 162220) is an autosomal dominant disorder characterized by multiple tumor types, particularly neurofibromas. It is caused by loss-of-function mutations in the NF1 gene on chromosome 17.1 The large size of the NF1 gene makes detection of some mutations problematic, but the most comprehensive genetic testing available can detect NF1 mutations in 95% of patients with NF1.2 NF1 is diagnosed based upon the presence of 2 or more of the following criteria: multiple café-au-lait macules, 2 or more neurofibromas (or 1 plexiform neurofibroma), a family history of NF1, axillary/inguinal freckling, optic nerve glioma, 2 or more Lisch nodules, or bone dysplasia.3 Charcot-Marie-Tooth disease type 1B (CMT1B; OMIM 11820) is an autosomal dominant demyelinating neuropathy caused by mutations in the Myelin Protein Zero gene (MPZ) on chromosome 1. Here we describe a patient with both NF1 and CMT1B.
A 50-year-old man sought evaluation for 3 months of progressive foot numbness. He had been diagnosed with NF1 in childhood on the basis of multiple neurofibromas, including a plexiform neurofibroma that had been resected from his arm in young adulthood. He has never been tested for mutations in NF1. He was adopted and did not know the health issues of his biological family. He had no children. He had multiple café-au-lait macules and multiple subcutaneous neurofibromas. There was mild weakness of toe extension (4+/5), and he could not walk on his heels or tandem walk. Cold perception was decreased distal to the knees, and vibration perception was absent on the toes and decreased at the ankles. Reflexes were absent at the ankles and 1+ elsewhere. The Romberg sign was present. His CMT neuropathy score4 was 12. Motor nerve conduction studies showed uniform and severe slowing and proportional prolongation of distal latencies, with decreased amplitudes in the distal lower extremity muscles (Table 1). Sensory nerve action potentials (SNAPs) were absent except for a radial SNAP that was severely decreased in amplitude and moderately slowed. Needle EMG showed length-dependent, chronic, active denervation. The following tests were normal: vitamin B12 levels, Lyme antibody, serum protein electrophoresis, and glucose tolerance test. Genetic testing for causes of CMT1 and CMT4 (PMP22 duplication/deletion, and sequencing of PMP22, LITAF, EGR2, GJB1/CX32, PRX, GDAP1, and SH3TC2) was negative except for a 449-1G>A in MPZ. This is a novel mutation that affects the splice site acceptor for exon IV (Figure 1). It would be predicted to cause skipping of this exon, which encodes the transmembrane domain of P0 (Figure 2), the protein encoded by MPZ.5
Peripheral nervous system symptoms, particularly mononeuropathies and radiculopathies related to nerve tumors, are well-recognized in NF1.6 The association between NF1 and polyneuropathy is less clear. Approximately 1% of patients with NF1 have a mild, predominantly sensory, indolently progressive polyneuropathy that is not otherwise explained.7,8 Given the prevalence of idiopathic polyneuropathy in the general population,9 some of these cases may be coincidental, so it has been suggested that patients with NF1 who develop polyneuropathy should be evaluated for an underlying cause; neurofibromatous neuropathy is a diagnosis of exclusion.1
P0 is the most abundant myelin protein and is the molecular glue of the myelin sheath. More than 100 different MPZ mutations have been identified,10 and their associated phenotypes vary considerably.11 The mildest phenotypes result from mutations that likely cause simple loss-of-function (e.g., E71X, D75frameshift, and V102frameshift) and thereby produce a demyelinating neuropathy owing to haplotype insufficiency. That haplotype insufficiency of Mpz also causes a late-onset demyelinating neuropathy in mice12,13 strongly supports the idea that the proper amount of P0 in compact myelin is crucial. The mutation found in our patient (449-1G>A) affects the splice site acceptor of exon 4 and is associated with a mild phenotype. Two other mutations that affect the same base have been reported previously - 449-1G>C 14 and 449-1G>T,15,16 but the clinical phenotypes of these patients are not described in sufficient detail to allow comparison to our patient. All three splice site mutations would be expected to result in the loss of exon 4 and hence the transmembrane domain of P0. If these P0 transcripts splice to exon 5 or 6, then it is possible that some mutant P0 protein is made and produces more than a simple loss-of-function. The severity of neuropathy in a 55 year old patient from a family with a mutation at the +2 site of intron 4 (c.614 +2T>G) with demonstrated skipping of exon 4 was similar to our patient (CMT neuropathy score 13). The motor conduction velocities in this family (39–42 m/s in the upper extremities) are faster than those in our patient (17–27 m/s); this discrepancy could indicate that the 449-1G>A mutation results in more than a simple loss-of-function (i.e. a dominant-negative effect) or, alternately, that the coexisting NF1 modifies our patient’s neuropathy.
NF1 and CMT each affect about 1 in 2500 persons, so that both diseases should occur together in about 1in 6 million persons. The co-occurrence of NF1 and CMT1 has been reported,17 including a prior report of two unrelated patients with both CMT1A and NF1.18 In the latter cases, the patients had the typical PMP22 duplication (17p11.2p12); this should not affect the NF1 gene, which is on the opposite side of the centromere (17q11.2). Another family had both NF1 and co-existing CMT, likely CMT1X based upon the electrophysiological and clinical characteristics, but genetic testing was not done.19 The case we describe represents the first known co-occurrence of CMT1B and NF1. Since MPZ mutations are found in about 5% of individuals with CMT,20 the specific coincidence of NF1 and CMT1B should occur in about 1 in 125 million.
Neurofibromas are complex, with several cell types, but several lines of evidence show Schwann cells are the primary cell type affected by NF1 mutations.21 First, the Schwann cells in the tumors, but not other cell types, show loss of heterozygosity for NF1.22 Second, Schwann cells from neurofibromas invade through basement membranes and trigger angiogenesis.23,24 Third, the conditional deletion of Nf1 only in Schwann cells results in neurofibromas initiated by non-myelinating Schwann cells (those associated with unmyelinated axons).25,26 While Schwann cell proliferation has not been specifically measured in CMT1B, demyelination and remyelination are always, to the best of our knowledge, accompanied by Schwann cell proliferation;27 this has been demonstrated for insults as varied as systemic tellurium intoxication,28 intraneural lysolecithin injection, or a genetic mutation in PMP22.29–32 Although one might suspect that conditions associated with increased Schwann cell proliferation, such as CMT1,33 would result in an overwhelming disease burden, this did not appear to be the case in our patient or in the other published cases of patients who have both CMT1 and NF1.17–19 Conversely, even if Schwann cells that are haplo-insufficient for Nf1 have increased proliferation,34 this consequence has not yet been detected in patients with CMT1.
This work was supported the NIH (RO1 NS43174 and NS55284 to S.S.S. and T32 DA022605-01 to E.L.). We thank Dr. Dev Batish (Athena Diagnostics) for a helpful discussion and Figure 1.