Schwannomatosis 1

A number sign (#) is used with this entry because susceptibility to the development of schwannomatosis-1 (SWNTS1) is conferred by germline heterozygous mutation in the tumor suppressor gene SMARCB1 (601607) on chromosome 22q11.

Description

Schwannomatosis, also known as neurilemmomatosis, first reported by Niimura (1973) as neurofibromatosis type 3, is characterized by multiple cutaneous neurilemmomas and spinal schwannomas, without acoustic tumors or other signs of neurofibromatosis I (NF1; 162200) or neurofibromatosis II (NF2; 101000). In neurilemmomas, the tumor consists of Schwann cells. Some patients may develop meningiomas (van den Munckhof et al., 2012).

Genetic Heterogeneity of Schwannomatosis

See also schwannomatosis-2 (615670), conferred by germline heterozygous mutation in the LZTR1 gene (600574) on chromosome 22q11.

Individual schwannoma tumors from patients with schwannomatosis have been found to harbor somatic mutations in SMARCB1 or the neurofibromin-2 gene (NF2; 607379).

Clinical Features

Swensen et al. (2009) reported a family with hereditary schwannomatosis spanning 4 generations associated with a germline duplication in the SMARCB1 gene (601607.0009). Affected individuals developed painful skin lumps in their teenage years. Two family members with mutations had malignant rhabdoid tumors (609322), and a third was believed to have a rhabdoid tumor. These 3 patients all died before age 2 year. Two rhabdoid tumors and several schwannomas showed somatic loss of the SMARCB1 gene.

Bacci et al. (2010) reported a family in which 4 individuals had multiple schwannomas and meningiomas. The proband was a 33-year-old man who had multiple peripheral schwannomas in his legs, first noted at about age 30. Brain imaging showed an extraaxial mass lesion around the left hypoglossal foramen, consistent with either a small meningioma or schwannoma, and spinal imaging showed multiple extraaxial and intradural mass lesions throughout the spinal cord. He also had multiple cafe-au-lait spots and painful lumps on the trunk and legs. At age 55, his father was found to have multiple spinal schwannomas and 2 meningiomas. The proband's paternal aunt had several fibrous meningiomas, and his cousin had multiple spinal schwannomas. Genetic analysis identified a heterozygous germline mutation in the SMARCB1 gene (E31V; 601607.0010) in all patients. Studies of tumor tissue from the proband showed loss of heterozygosity (LOH) for markers on chromosome 22 including both the SMARCB1 and NF2 genes. Bacci et al. (2010) noted that meningiomas are not frequently found in patients with schwannomatosis, but should be considered part of the phenotype.

Christiaans et al. (2011) reported a family in which 5 individuals developed meningiomas (607174), 2 of whom also developed schwannomas. All patients carried a heterozygous mutation in the SMARCB1 gene (P48L; 601607.0011), and meningioma tumors showed loss of the wildtype allele, consistent with the 2-hit hypothesis of tumorigenesis. Meningiomas developed between ages 34 and 56 years, both in the cranium as extra-axial lesions and in the spinal cord as extramedullary lesions. In addition, 1 patient developed multiple chest wall and spinal schwannomas and another developed a single vestibular schwannoma. Two different meningioma tumors from the same patient also carried 2 different heterozygous somatic mutations in the NF2 gene (607379) as well as loss of heterozygosity at the NF2 locus. Christiaans et al. (2011) concluded that the SMARCB1 P48L mutation predisposed the carriers to the development of meningiomas. The mutation may also have predisposed carriers to schwannomas, implying that meningiomas may be part of the schwannomatosis tumor spectrum, as suggested by Bacci et al. (2010), but the schwannomas may also be coincidental findings. The role of the NF2 mutations was uncertain, but may contribute to a 4-hit hypothesis involving 2 genes. Van den Munckhof et al. (2012) provided further studies of the family reported by Christiaans et al. (2011). Reexamination of tumor tissue from 4 meningiomas and 2 schwannomas showed that all tumors had LOH for both SMARCB1 and NF2, consistent with a deletion of a segment of chromosome 22 containing these 2 genes. Three meningiomas and 2 schwannomas were each found to carry somatic mutations in the NF2 gene. Thus, the genetic changes found in the 2 tumor types were the same and characteristic for SMARCB1-mutation positive tumors: retention of the exon 2 mutation, acquisition of an NF2 mutation, and LOH of the wildtype allele of both genes. In addition, van den Munckhof et al. (2012) identified 11 more carriers of the P48L mutation in this family. Eight of these 11 mutation carriers were found to carry 11 lesions suggestive of cranial meningioma and 6 spinal lesions consistent with meningiomas or schwannomas. Nine (82%) of the 11 cranial meningiomas were found in the falx cerebri. Van den Munckhof et al. (2012) concluded that meningiomas should be included in the schwannomatosis tumor spectrum.

Distinction from Neurofibromatosis type II

Schwannomas are benign tumors of the peripheral nerve sheath that usually occur singly in otherwise normal individuals. Multiple schwannomas in the same individual suggest an underlying tumor predisposition syndrome. The most common such syndrome is neurofibromatosis II. The hallmark of NF2 is the development of bilateral vestibular nerve schwannomas, but two-thirds or more of all NF2-affected individuals develop schwannomas in other locations, and dermal schwannomas (or neurilemmomas) may precede vestibular tumors in NF2-affected children (Evans et al., 1992; Mautner et al., 1993; Parry et al., 1994). MacCollin et al. (1996) reviewed reports of individuals with multiple schwannomas who do not show evidence of vestibular schwannomas, and suggested that schwannomatosis is a clinical entity distinct from other forms of neurofibromatosis.

Sasaki and Nakajima (1992) described multiple cutaneous neurilemmomas in an 8-year-old Japanese girl and a 5-year-old Japanese boy. The boy's father had bilateral acoustic neuromas, suggestive of NF2, and also had multiple skin tumors that were diagnosed as neurilemmomas histopathologically. There were no pigmented macules. The multiple skin tumors had been present in both children since birth. However, Jacoby et al. (1997) commented that NF2 could not be excluded in these children, since vestibular schwannoma may not be apparent until adolescence. In addition, Jacoby et al. (1997) noted that there had been no previous reports of schwannomatosis meeting their clinical criteria (see DIAGNOSIS) who had a positive family history of NF2.

Evans et al. (1997) reported 5 families with schwannomatosis inherited in an autosomal dominant pattern. The phenotype was consistent, with multiple skin and spinal tumors and relative sparing of the cranium. Members of a sixth family, who initially appeared to have schwannomatosis, developed bilateral acoustic neuromas and were later classified as having NF2. Genetic linkage in 2 large families with schwannomatosis showed linkage to chromosome 22q12.2 in the region of the NF2 gene. Evans et al. (1997) noted the difficulty in distinguishing the 2 disorders and suggested that young patients thought to have schwannomatosis may have a variant form of NF2.

Diagnosis

The criteria used by Jacoby et al. (1997) for the diagnosis of schwannomatosis were as follows: 2 or more pathologically proved schwannomas and lack of radiographic evidence of vestibular nerve tumor at age more than 18 years was taken as evidence of definite schwannomatosis. For presumptive or probable schwannomatosis the criteria were 2 or more pathologically proved schwannomas, without symptoms of eighth-nerve dysfunction at an age of more than 30 years; or 2 or more pathologically proved schwannomas in an anatomically limited distribution (single limb or segment of the spine), without symptoms of eighth-nerve dysfunction at any age.

Mapping

Because molecular analysis of tumor specimens from affected individuals in kindreds with familial schwannomatosis revealed a pattern of somatic NF2 inactivation incompatible with germline NF2 alteration, MacCollin et al. (2003) performed linkage analysis in 6 families with schwannomatosis. They obtained a maximum lod score of 6.60 near marker D22S1174 in the proximal portion of chromosome 22 centromeric to the NF2 gene. MacCollin et al. (2003) concluded that schwannomatosis is a distinct entity from neurofibromatosis type II, and that the NF2 region is not the inherited genetic locus responsible for familial schwannomatosis.

Molecular Genetics

Mutations in the SMARCB1 Gene

Since the NF2 locus had been excluded as the germline event underlying familial schwannomatosis, and the gene placed centromeric to NF2 on chromosome 22, Hulsebos et al. (2007) investigated the SMARCB1 gene in a father and daughter with the disorder. Both were found to be heterozygous for a inactivating germline mutation of this gene (601607.0005). In 2 of 4 investigated schwannomas from these patients, inactivation of the wildtype INI1 allele by a second mutation in exon 5 of the gene (601607.0006) or by loss of the gene was found, consistent with the Knudson 2-hit hypothesis.

Sestini et al. (2008) identified a de novo germline deletion/insertion in the SMARCB1 gene (601607.0007) in 1 of 21 unrelated patients with schwannomatosis. Three different tumors derived from this patient showed the deletion/insertion and a somatic NF2 mutation on the same allele, but no other SMARCB1 mutations. In addition, 2 of the tumors had somatic loss of heterozygosity (LOH) encompassing the SMARCB1 and NF2 region. Sestini et al. (2008) also found somatic mutations in tumor tissues from 3 additional patients who did not have germline mutations. Tumor tissue from 1 of these patients contained respective somatic mutations in the SMARCB1 and NF2 genes as well as LOH, indicative of complete monosomy of chromosome 22q in this tumor. Tumor tissue from a second patient had somatic mutation in the NF2 gene as well as LOH for chromosome 22q. Tumor tissue from the third patient had a somatic mutation in the SMARCB1 gene and LOH for chromosome 22q. Based on these results, Sestini et al. (2008) postulated that a 4-hit mechanism involving 2 distinct but linked tumor suppressor genes, SMARCB1 and NF2, may underlie the development of tumors in a subset of patients with schwannomatosis. However, given the low frequency of SMARCB1 germline mutations, there may also be additional loci involved.

In 5 (33.3%) of 15 families with schwannomatosis and 2 (7.1%) of 28 individuals with sporadic schwannomatosis, Hadfield et al. (2008) identified germline mutations in the SMARCB1 gene (see, e.g., 601607.0008). In all of these individuals in whom tumor tissue was available, tumor tissue showed a second hit with loss of SMARCB1. In addition, all of these patients had biallelic somatic inactivation of the NF2 gene. Similar to the report of Sestini et al. (2008), the findings suggested that 4 hits of these 2 genes are usually necessary to develop schwannomas. Germline SMARCB1 mutations were associated with a higher number of spinal tumors in patients with a positive family history (p = 0.004).

Somatic Mutation in the NF2 Gene

Honda et al. (1995) analyzed the peripheral leukocytes and tissue from cutaneous neurilemmomas of 7 patients with neurilemmomatosis, using DNA markers for different regions of chromosome 22. They detected allele losses in 3 of 7 tumors from 7 patients with a probe for the NF2 region and the germline mutations in 2 of 3 tumors from the same 3 patients. They described 2 mutations in the NF2 gene (607379.0017; 607379.0018). They concluded that neurilemmomatosis is a form of NF2.

Jacoby et al. (1997) undertook a molecular genetic investigation of the relationship of schwannomatosis to NF2. They examined the NF2 locus in 20 unrelated schwannomatosis patients and their affected relatives. Tumors from these patients frequently harbored typical truncating mutations of the NF2 gene and loss of heterozygosity of the surrounding region of chromosome 22. Surprisingly, unlike patients with NF2, no heterozygous NF2 gene changes were seen in normal tissues. Examination of multiple tumors from the same patient revealed that some schwannomatosis patients were somatic mosaics for NF2 gene changes. In contrast, other individuals, particularly those with a positive family history of schwannomatosis, appeared to have an inherited predisposition to formation of tumors that carry somatic alterations of the NF2 gene. This tendency was biased toward the occurrence of different mutations in the same, coinherited allele within a given family, combined with loss of the trans allele in any given individual. Although the family data were consistent with linkage of this trait to the NF2 locus, the studies implied that the primary event in the tumors lay outside the coding region of the NF2 gene.

In 28 schwannoma tumor specimens from 17 affected individuals in 8 families with familial schwannomatosis, MacCollin et al. (2003) identified 20 different somatic mutations in the NF2 gene, 18 of which were truncating mutations. None of the mutations detected in tumor specimens were detected in the paired blood specimens and no tumor was found to have 2 mutations in the NF2 gene. In the 6 instances in which multiple tumors were from the same patient, no 2 tumors shared the same mutation. Microsatellite analysis showed that all but 4 of the tumors had loss of heterozygosity for NF2. Ten of 18 tumors had loss of all markers, consistent with monosomy or mitotic nondisjunction.

Other Genes

Buckley et al. (2005) investigated the genetic factors underlying the differences between schwannomatosis and NF2. They comprehensively profiled the DNA copy number in samples from patients with sporadic and familial schwannomatosis, patients with NF2, and a large cohort of normal controls. Using a tiling-path chromosome 22 genomic array, they identified 2 candidate regions of copy number variation, which were further characterized by a higher resolution PCR-based array. In DNA derived from peripheral blood of a schwannomatosis patient and a sporadic schwannoma sample, they detected rearrangements of the immunoglobulin lambda (IGLC1; 147220) locus, which were thought not to be due to a B-cell-specific somatic recombination of IGLC1. Analysis of normal controls indicated that these IGLC1 rearrangements were restricted to the schwannomatosis/schwannoma samples. In a second candidate region spanning the GSTT1 (600436) and CABIN1 (604251) genes, they observed a frequent copy number polymorphism at the GSTT1 locus. They also identified missense mutations in the CABIN1 gene that were specific to samples from schwannomatosis and NF2.

For discussion of a possible association between variation in the COQ6 gene and susceptibility to schwannomatosis, see 614647.0007.