Unraveling schwannomas

In this issue, Williams et al. provide new and very interesting information on the molecular tumorigenesis of a substantial subset of sporadic schwannomas.¹ Before discussing the results of this study in somewhat more detail, it may be good to briefly look back at the genesis of the concept of schwannomas. It was Antoni van Leeuwenhoek, a multitalented Dutch microbiologist, who discovered the myelination of nerve fibers in 1717. More than a century later, the German anatomist and physiologist Theodor Schwann suggested the association between myelin and the “lemmocyte,” a cell type that later on became known as Schwann cell.^2,3 The term “schwannoma” was coined by the French Canadian histopathologist Pierre Masson in 1923, and Jose Juan Verocay, a Uruguayan neuropathologist, played an important role in the more precise description of these tumors.⁴ The architectural pattern of alternating cellular areas with nuclear palisading including Verocay bodies (Antoni A) and loosely organized areas with myxomatous and cystic changes (Antoni B) was documented in 1920 by the Swedish neurologist Nils Antoni (Figure 1).⁵

Figure 1.

Histology of and underlying genetic alterations in schwannomas. The H&E-stained section shows alternating compact (Antoni A, #) and more loosely organized areas (Antoni B, *). Typically, palisading of tumor cell nuclei is seen in the Antoni A component (arrowhead), Verocay bodies representing an “exaggerated” form of such palisading. In the upper part of this figure genes and their protein products are indicated that are already known for some time to play a causal role in the tumorigenesis of schwannomas, all or not in the context of a genetic tumor syndrome. The bottom lists more recently identified (fusion) genes, including the SOX10 indel mutations as reported by Williams et al.¹ The inset shows positivity of the nuclei of schwannoma cells in an immunohistochemical SOX10 staining. Of note, this latter staining does not discriminate between the different genetic drivers underlying the tumorigenesis of schwannomas.

Schwannomas are usually solitary, benign, encapsulated tumors, attached to a nerve, and almost entirely composed of neoplastic Schwann cells. The pathological diagnosis of these neoplasms can generally easily be made on H&E-stained sections. However, a variety of morphological patterns have been described, including ancient, cellular, plexiform, epithelioid, microcystic/reticular, and neuroblastoma-like. Also, hybrid tumors with both a schwannoma and a neurofibroma or perineurioma component occur. Malignant transformation is exceedingly rare. Using immunohistochemistry, schwannomas typically show strong and extensive S100 and SOX10 staining.^6,7

Different molecular alterations can underlie the tumorigenesis of schwannomas with the drivers NF2 (encoding for merlin), SMARCB1 (INI1/BAF47), and LZTR1 (LZTR1) located on the long arm of chromosome 22. More recently identified drivers include mutations involving ARID1A, ARID1B, and DDR, as well as SH3PXD2A::HTRA1 and NONO::TFE3 fusion genes (Figure 1).^6–8 Sporadic schwannomas most commonly carry NF2 mutations (in about 50% of the cases).^6,9 Inactivation of INI1 is known to occur in a few sporadic cases, morphologically described as epithelioid variant of schwannoma.¹⁰ Importantly, schwannomas may occur in the context of a genetic tumor syndrome, especially neurofibromatosis type 2 (NF2) or, much less frequently, schwannomatosis (SMARCB1 or LZTR1). Those patients typically have multiple schwannomas (± other neoplasms).^6,7,9

Williams et al. now report that 30% of the sporadic schwannomas in their series harbored a recurrent insertion–deletion (indel) mutation of the SOX10 gene promoting impaired transactivation of tissue-specific gene programs responsible for myelination.¹ The Sox10 transcription factor is well known for its role in the development and maintenance of the peripheral nervous system and its strong, nuclear expression in (non-neoplastic and neoplastic) Schwann cells. In a way, it may therefore not come as a surprise that the corresponding gene, which is also located on chromosome 22 (band 22q13.1), can be involved in the tumorigenesis of schwannomas. Interestingly, however, this alteration has not been identified before.

In the study of Williams et al., the SOX10 indel mutations were only identified in sporadic, nonvestibular schwannomas and were mutually exclusive with other known drivers of these tumors. The authors, therefore, speculate that these mutations drive a particular subtype of schwannomas, which may be related to divergent subtypes of specialized Schwann cells in different anatomic regions or a differential developmental cell state. Indeed, in different studies, DNA methylation signatures of schwannomas show epigenomic subgroups, often associated with the anatomic site.^1,7

While NF2-mutant schwannomas typically have a “double-hit” alteration (often monosomy or copy-neutral loss of heterozygosity of chromosome 22q eliminating the wild-type allele), most SOX10-mutant schwannomas appeared to have an intact SOX10 wild-type allele. Also, the authors found that schwannomas harboring SOX10 indel mutations were different from NF2-mutant schwannomas by hypomethylated CpG islands at the 5’ end of MEIS1, a gene involved in nervous system development.

Up till now, schwannomas and other nerve sheath tumors were primarily thought to arise due to hyperactivation of the RAS growth factor signaling pathway. In contrast, the tumorigenesis of SOX10 indel-mutant schwannomas seems to be based on blocked differentiation resulting from mutations in the transcription factor responsible for Schwann cell fate determination. Williams et al. hypothesize that because of this dominant oncogenic effect, alternative therapeutic strategies (eg, by exploiting antisense oligonucleotide technology) may be considered, especially so when complete surgical resection is not an option.¹

Last but not least, Williams et al. speculate that these SOX10 mutations may not have been detected before because they were overlooked due to the main focus in prior studies and common sequencing pipelines on single nucleotide variants or small indels (ie, indels with a length of less than 3 base pairs). Most SOX10 indel mutations in schwannomas in this study, however, had a length of 6–69 base pairs. As the authors suggest, it may thus be worthwhile to have a closer look at such larger indels (whether or not in genes encoding for transcription factors) in tumors of which the driver has not yet been identified.

In conclusion, the study of Williams et al. provides a new paradigm in schwannoma tumorigenesis, with SOX10 indel mutations leading to impaired transactivation of tissue-specific gene programs responsible for myelination. This study not only further elucidates the variety of molecular mechanisms responsible for schwannoma development, but also suggests a potential avenue for improved therapy of SOX10-mutant schwannomas and a clue for optimization of molecular diagnostic pipelines.

Conflict of interest statement

None declared.

Declaration

The text is the sole product of the authors and no third party had input or gave support to its writing.