The family of benign peripheral nerve sheath tumors includes schwannoma, neurofibroma, solitary circumscribed neuroma, perineurioma and dermal nerve sheath myxoma. The first three are regularly diagnosed in routine practice, whereas the latter two are uncommon and have been well characterized only in the past decade,1, 3 despite being first described over 30 years ago.4, 5 This review will cover the most recent developments in this area: the description of hybrid benign peripheral nerve sheath tumors6-12 and the introduction of a new schwannoma variant, the microcystic/reticular schwannoma.13 We will also briefly discuss the results of a recent immunohistochemical study of the specificity of GLUT-1, a marker used in the diagnosis of perineurioma.14 The concept of hybrid peripheral nerve sheath tumors was introduced with the description of hybrid neurofibroma–schwannoma,7 a rare neoplasm in which an otherwise typical neurofibroma shows focal areas of well-developed schwannian differentiation. The underlying neurofibroma in these cases is often plexiform, and as expected, many patients therefore have type 1 neurofibromatosis (NF1). The tumors arise over a wide age range and may involve superficial or deep soft tissue. Their clinical behavior is benign, and the importance of recognizing this morphologic variant lies primarily in distinguishing it from malignant transformation within a neurofibroma, particularly given the increased incidence of the latter phenomenon in NF1 patients. More recently, several groups have described hybrid benign peripheral nerve sheath tumors with perineurial and schwannian components.6, 8-12 Reports of schwannoma–perineurioma hybrids have increased as perineuriomas have become more widely recognized,8 and this type of hybrid now appears to be more common than the neurofibroma–schwannoma hybrids. Most schwannoma–perineurioma hybrids present in adulthood, but children may be affected.6 Only one patient was known to have had NF1.11 Any anatomic site may be involved and although the majority is subcutaneous, visceral lesions have been reported.12 They behave in a clinically benign fashion; no recurrences or metastases have been reported to date. The tumors range from less than 1 to several centimeters in size, and most are well circumscribed but unencapsulated. Two histologic forms have been described: biphasic tumors (known as schwannoma–perineuriomas), in which distinct schwannian and perineurioma-like areas are identifiable, and ‘monophasic’ tumors, in which the schwannian and perineurial components are intimately admixed.6, 8, 12, 11 The nosology of the latter form is variable, depending on whether the schwannian component is interpreted as representing a schwannoma or a neurofibroma. Some use the term ‘schwannoma–perineurioma’ (as for the biphasic hybrid tumors), which may be appropriate, given the absence of axons,6 while other authors prefer ‘neurofibroma–perineurioma’ for these lesions, although they acknowledge that the distinction between schwannian and neurofibromatous differentiation is somewhat arbitrary.8, 9 The monophasic hybrid tumors often have a storiform pattern, but short fascicular, lamellar and whorled arrangements may also be noted. The stroma is typically finely collagenous and palely eosinophilic. On close examination, two cell types are generally identifiable: short spindle cells with dark tapered or wavy nuclei (the schwannian component) and elongated spindle cells with ovoid nuclei and variably prominent eosinophilic cytoplasm (the perineurial component) (Fig. 1). The bipolar cytoplasmic processes of the perineurial component may be visible on routine sections, particularly when the stroma is myxoid or hyalinized. Occasionally, mild degenerative-appearing nuclear atypia (‘ancient change’) similar to that seen in other benign peripheral nerve sheath tumors is present.10 Hybrid schwannoma–perineurioma (‘monophasic’ pattern). The low-power view may suggest a perineurioma because of the eosinophilic background and loose fascicular architecture (A), but closer examination reveals both short spindle cells with wavy nuclei (B, schwannian component) and elongated spindle cells with long cytoplasmic processes in whorled and lamellar arrangements (C, D; perineurial component). Although discrete foci in which one cell type predominates may be identified, the Schwann cells and perineurial cells are admixed throughout the tumor. Immunohistochemistry is generally required to confirm the presence of dual differentiation in the monophasic type of hybrid schwannoma–perineurioma. Immunostains reveal S100-positive and EMA-positive cells admixed throughout the tumor (Fig. 2). Some cases also show a smaller population of CD34-positive cells, possibly representing a fibroblastic component similar to that seen in pure neurofibromas10, 11 (although perineurial cells may be CD34 positive, the cytomorphology and comparison with an EMA immunostain help distinguish these cells from fibroblasts). Characteristic features of classical schwannomas, including Antoni A/B areas, perivascular hyalinization, hemorrhage and cystic degeneration, are not present. From a practical point of view, the monophasic hybrid tumors most often resemble a perineurioma at first glance because of the loose storiform or whorled architecture and eosinophilic background, but on further examination, a population of cells with short, wavy nuclei is identified, admixed with elongated spindle cells. Immunohistochemistry revealing the two cell populations is confirmatory. Hybrid schwannoma–perineurioma (‘monophasic’ pattern). Immunostains demonstrate significant numbers of both S100-positive cells (A) and EMA-positive cells (B) intermingled within the lesion (same case as in Fig. 1). In biphasic hybrid tumors (known as schwannoma–perineuriomas), discrete perineurial and schwannian foci are clearly discernible on routine sections.8, 9, 11 The perineurial component is made up of elongated spindle cells with ovoid nuclei and eosinophilic cytoplasm forming lamellar, whorled and/or retiform patterns. As expected, EMA and claudin-1 are positive in these areas. The schwannian areas form discrete foci composed of wavy spindle cells in a pale background. Verocay bodies or hemorrhage may be present in these regions, and the cells show strong, diffuse S100 staining. In both forms of schwannian–perineurial hybrids, mitoses are rare and necrosis is absent. Axons have not been detected with special stains or immunohistochemistry in either form.6, 8, 9 Mutations in the NF2 gene on chromosome 22 were detected in one case,8 which is not surprising as schwannomas, neurofibromas and perineuriomas in their pure forms all frequently have deletions or mutations in this region.15-18 Finally, very rare tumors with a perineurial component and a granular cell component of presumably schwannian origin have been described and may be considered a distinct form of hybrid benign peripheral nerve sheath tumor.19, 20 In these lesions, elongated spindle cells in a storiform or whorled pattern typical of perineurioma are admixed with a second population of large eosinophilic cells with granular cytoplasm. The granular cells are S100 positive, which distinguishes these hybrid tumors from the recently reported granular cell perineurioma, in which both the typical perineurial cells and the granular cells are EMA positive.21 Microcystic/reticular schwannoma shows a predilection for the gastrointestinal tract but may arise at any site, including the dermis,22 subcutaneous/deep soft tissue, bronchus and adrenal gland.13 The age range is wide, and no association with NF1 or NF2 has been reported. No tumors have recurred or metastasized. When these tumors arise in the skin or soft tissue, they tend to be well circumscribed and often encapsulated, whereas those in the gastrointestinal tract are unencapsulated, similar to classical schwannomas at this site. The tumors are composed of spindle cells with ovoid or tapered nuclei, delicate eosinophilic cytoplasm and ill-defined cell borders, and most cases show both reticular and microcystic patterns. Some cases have focally epithelioid cytomorphology. The matrix is typically predominantly myxoid with focal hyalinized collagenous areas, but infrequently, the entire tumor may be hyalinized. Foci reminiscent of classic schwannoma are present in a minority of cases. Microcystic/reticular schwannomas are diffusely S100 positive and show focal to diffuse expression of GFAP.13, 22 EMA is generally negative in tumor cells but highlights the tumor capsule (if present). The differential diagnosis of microcystic/reticular schwannoma in the skin or soft tissue includes nerve sheath myxoma, reticular perineurioma, soft tissue myoepithelioma and extraskeletal myxoid chondrosarcoma. Nerve sheath myxoma has a diffusely myxoid stroma, often shows a reticular architecture and is S100 negative and GFAP positive, similar to microcystic/reticular schwannoma, but it lacks a microcystic pattern, is characteristically multinodular and has frequent multinucleated cells and ring-like vacuolated cells.3 However, there is undoubtedly histologic overlap with predominantly reticular schwannomas, and Fetsch et al3 acknowledge that nerve sheath myxomas may ultimately prove to be a subtype of schwannoma. Reticular perineurioma shares morphologic similarities with predominantly reticular schwannomas but is easily distinguished by the presence of EMA staining and the lack of S100 reactivity. Soft tissue myoepithelioma can have spindle cells forming a reticular pattern in a myxoid or hyalinized background, but similar to pleomorphic adenoma (its salivary gland counterpart), it typically shows marked architectural and cytologic heterogeneity, in contrast to the relatively uniform features of microcystic/reticular schwannoma. Furthermore, all myoepithelial tumors stain for at least one epithelial marker (ie, cytokeratin or EMA), in addition to reactivity for S100 and, in some cases, GFAP. Extraskeletal myxoid chondrosarcoma displays an intricate reticular pattern in a myxoid matrix, but microcystic areas are not seen and S100 is negative or shows only weak, focal reactivity. EMA is the gold standard immunostain for perineuriomas, as nearly all tumors are positive for this marker. However, other tumors in the differential diagnosis of perineurioma – particularly low-grade fibromyxoid sarcoma – may express EMA,23, 24 and additional immunostains are sometimes required. Second-line markers used for the diagnosis of perineurioma include claudin-1 and GLUT-1. Claudin-1, a tight junction protein expressed in normal perineurial cells, has a lower sensitivity than EMA (ranging from 29% to 92%)1, 25 but may be useful if the EMA reactivity is equivocal or if the differential diagnosis includes low-grade fibromyxoid sarcoma, which is claudin-1 negative. GLUT-1 is also expressed by normal perineurial cells and has been considered relatively specific for perineurial differentiation.26-28 In fact, the detection of GLUT-1 reactivity in epithelioid sarcomas was used to support the hypothesis that these neoplasms show perineurial differentiation.29 However, it is well recognized that many carcinomas are GLUT-1 positive,30 and the specificity of this antibody among mesenchymal tumors had not been fully investigated until recently. Ahrens et al14 evaluated GLUTl-1 expression in 247 benign and malignant soft tissue tumors and reported positivity in a majority of perineuriomas, chordomas, epithelioid sarcomas and undifferentiated pleomorphic sarcomas as well as in a minority of several other tumor types. In the high-grade pleomorphic sarcomas, GLUT-1 reactivity was most often noted near zones of necrosis, consistent with the known upregulation of this gene in hypoxic states, and the authors speculated that its expression in epithelioid sarcomas may have a similar etiology.14 Although these data indicate that GLUT-1 lacks specificity among mesenchymal tumors, it may still be useful in certain contexts, such as the distinction of perineurioma from low-grade fibromyxoid sarcoma or dermatofibrosarcoma protuberans, both of which appear to be consistently GLUT-1 negative.
No takes yet. Share an insight, caveat, or question.
Phan et al. (2008) studied this question.
Synapse has enriched 4 closely related papers on similar clinical questions. Consider them for comparative context: