Plexiform Schwannoma of the Oral Cavity: Report of Eight Cases and a Review of the Literature

Abstract

Plexiform schwannoma represents an unusual schwannoma variant, characterized by multinodular growth grossly and/or microscopically. A review of the English-language literature reveals only 28 previously reported cases involving the oral cavity, and herein we present 8 additional cases. Among these 36 patients, the average age at diagnosis was 28 years (range 5 to 62 years), with a female-to-male ratio of 1.4:1. The most frequently involved sites were the tongue (n = 13) and lip (n = 11). Lesion duration prior to presentation averaged 5.3 years (range, 6 weeks to 26 years). The average lesion size was 2.1 cm (range, 0.3 to 16 cm). The typical clinical presentation was a painless mass, although infrequent findings included pain/discomfort, paresthesia, difficulty chewing, and limited buccal mobility. All cases clinically appeared as a solitary mass or localized cluster of tumor nodules, with the exception of one patient who had neurofibromatosis 2 (NF2) and exhibited two distinct nodules on the tongue and buccal mucosa. In addition, extraoral neural neoplasms were evident in four patients, including three with NF2. Typical microscopic findings included multiple well-circumscribed tumor nodules, each surrounded by a perineurium-derived capsule with immunoreactivity for epithelial membrane antigen. The nodules contained characteristically bland and diffusely S-100-positive spindle cells arranged in Antoni A and B patterns; however, modest nuclear pleomorphism was evident in three cases. Most patients (n = 23) were treated by excision or enucleation and curettage, and three patients experienced recurrence. Unlike plexiform neurofibromas, plexiform schwannomas exhibit only a weak association with neurofibromatosis and have no known malignant potential.

Introduction

Plexiform schwannoma represents an unusual schwannoma variant, characterized by multinodular growth grossly and/or microscopically. Alternative terms for this entity include multinodular schwannoma (neurilemoma) and multilobular schwannoma (neurilemoma). In the largest series reported to date, Berg et al. [1] found that this variant represented only 4.3% of all schwannomas in their pathology case files.

The earliest description of plexiform schwannoma appeared in an abstract by Harkin et al. [2] from the 1978 International Congress of the American Association of Neuropathologists. These authors reported six benign peripheral nerve sheath tumors with gross features resembling plexiform neurofibroma but histopathologic features resembling schwannoma. None of these patients had neurofibromatosis; all tumors exhibited benign behavior, and no recurrences were noted following excision. The investigators emphasized the importance of distinguishing these tumors from plexiform neurofibromas, which are pathognomonic for neurofibromatosis 1 (NF1, von Recklinghausen disease) and may undergo malignant transformation. Although plexiform schwannoma is not associated with NF1, it may occur at times within the setting of neurofibromatosis 2 (NF2) or schwannomatosis (neurilemomatosis) [3].

Thus far, more than 100 cases of plexiform schwannoma have been reported in the English-language literature [1, 4]. Similar to conventional schwannoma, the plexiform variant exhibits a predilection for the extremities, head, and neck [1]. Within the head and neck region, plexiform schwannomas most often arise in the skin, subcutis, or other superficial locations and present as innocuous, slowly enlarging, and asymptomatic nodules that are cured readily by conservative excision. However, unusual examples may involve the spinal cord/nerve roots, cranial nerves, major peripheral nerves, orbit, or parotid glands; sometimes deep-seated tumors become especially large, produce significant symptoms, and are difficult to excise in their entirety [4–7].

Although plexiform schwannomas are known to exhibit a predilection for the head and neck, documentation of lesions specifically involving the oral cavity has been limited to small case series or single case reports. In this study, we present a retrospective analysis of oral plexiform schwannomas in our pathology laboratory archives and review the literature regarding these unusual tumors.

Materials and Methods

Our study consisted of the following: (1) a retrospective review of all oral plexiform schwannomas diagnosed by the Medical University of South Carolina (MUSC) College of Dental Medicine’s Oral Pathology Laboratory and the Texas A&M University College of Dentistry’s Oral Pathology Laboratory during the period from January 1, 2002 through April 1, 2018, (2) a review of the English-language literature for previously reported cases of oral plexiform schwannoma. The protocol was approved by the Institutional Review Boards of MUSC and Texas A&M University. For the purpose of this study, we defined plexiform schwannoma as a schwannoma exhibiting multinodular growth grossly and/or microscopically, and we defined schwannoma as per the diagnostic criteria of the World Health Organization (WHO) [8]. Also, our review was restricted to lesions involving the oral cavity, jaws, or lip vermilion. The oropharynx (including soft palate and base of tongue), skin, subcutis, and areas overlying the oral cavity (e.g., “cheek,” masseter muscle) were excluded from our definition of oral cases. In addition, we excluded tumors for which the oral subsite could not be determined from the information provided.

In order to identify acceptable cases in our archives, we queried our laboratory’s electronic database for the diagnostic terms plexiform schwannoma (neurilemoma), multinodular schwannoma (neurilemoma), or multilobular schwannoma (neurilemoma) among cases accessioned during the study period. The pathology reports, specimen submission forms (completed by the clinician), and microscopic slides were reviewed in order to confirm the diagnosis and examine the clinical and pathologic findings. Cases for which records and/or microscopic slides were not available for review were excluded from the study.

If not done previously, immunohistochemical studies for S-100 protein, epithelial membrane antigen (EMA), and neurofilament protein (NFP) were performed on available material, using 4-µm sections of formalin-fixed, paraffin-embedded tissue and the Discovery Ultra System (Ventana Roche Diagnostics, Indianapolis, IN) as per the manufacturer’s recommendations. The following antibodies were used: S-100 (clone EP32, 1:1,000, Leica Biosystems, Buffalo Grove, IL), EMA (clone E29, prediluted, Biocare Medical, Pacheco, ca.), and NFP (clone 2F11, prediluted, Biocare Medical, Pacheco, ca.). Standard positive and negative controls were employed throughout.

For our review of the English-language literature, the keywords plexiform schwannoma OR plexiform neurilemoma OR multinodular schwannoma OR multinodular neurilemoma OR multilobular schwannoma OR multilobular neurilemoma AND oral OR jaws were searched in PubMed, Embase, Scopus, and Google Scholar. The references cited by the retrieved articles were reviewed as well in order to identify additional cases.

For the pooled set of oral plexiform schwannomas identified from our laboratory files and literature review, the following parameters were recorded: patient age at diagnosis, gender, anatomic subsite, clinical findings, lesion size, clinical diagnostic impression, radiographic findings, pathologic findings, treatment, and follow-up information. We defined the “lip” to include lip vermilion, labial mucosa, and commissure. Descriptive statistics were performed for each of these parameters. In order to calculate the average lesion size, we used the greatest maximum diameter noted by clinical, radiographic, and/or gross examination for a given tumor overall (rather than the maximum diameter of an individual nodule within a multinodular tumor).

Results

In our pathology laboratory archives, 8 cases of oral plexiform schwannoma met the study inclusion criteria. These cases represented 9% of all oral schwannomas diagnosed and < 0.01% of all accessions during the study period. In addition, our review of the English-language literature revealed 28 acceptable cases of oral plexiform schwannoma [1, 9–25]. Note that we excluded the oral plexiform schwannoma reported by Ide et al. [26] because the microscopic description and photomicrographs suggested that the lesion had a multi-fascicular rather than multinodular growth pattern. Also, the cases documented by Nisa et al. [27] and Dhua [28] were excluded because multinodularity was not evident grossly or microscopically.

The salient clinical features of the 36 current and previously reported cases of oral plexiform schwannoma are summarized in Table 1. Patient age and sex were provided in 30 cases, with an average age at presentation of 28 years (range, 5 to 62 years) and a female-to-male ratio of 1.4:1. The most frequently involved locations were the tongue (n = 13) and lip (n = 11). Additional sites included the buccal mucosa/vestibule (n = 4), hard palate (n = 4), floor of mouth/submandibular region (n = 2), mandible (n = 2), maxilla (n = 1), and gingiva (n = 1). Four tumors exhibited involvement of major nerves (i.e., mental, inferior alveolar, hypoglossal, and greater palatine) [1, 18, 21, 25]. The tumor size was stated for 23 lesions, with an average of 2.1 cm (range, 0.3 to 16 cm). Vera-Sempere and Vera-Sirera [18] described a “large” tumor involving the tongue, although they did not provide exact tumor dimensions. Eight patients reported a history of slow or gradual enlargement. The approximate lesion duration prior to presentation was noted in 17 cases, with an average of 5.3 years (range, 6 weeks to 26 years); one case apparently was congenital [10]. The most common clinical presentation was an asymptomatic, painless, or nontender swelling (n = 17). Infrequent findings included pain when talking, eating, and/or chewing (n = 2); paresthesia (of the lower lip and angle of mouth in a patient with an intraosseous lesion of the mandible) (n = 1); difficulty chewing (n = 1); and limited mobility of the buccal mucosa (n = 1) [12, 14, 20, 29]. The color of the overlying mucosa was described as pink/“normal” (n = 7), red (n = 2), white (n = 1), or yellow-gray (n = 1). One lesion exhibited an ulcerated surface [12].

Table 1.

Summary of clinical features for current and previously reported cases of oral plexiform schwannoma

Year of Publication and Author(s)	Age (years)	Sex	Site	Treatment	Follow-up	Comments
Barbosa and Hansen (1984)	12	M	Tongue	Excision	N.E.D. > 13 yrs
	36	M	Upper lip	Excision	Lost to follow-up
	20	F	Buccal mucosa	Excision	N.E.D. 3 yrs 4 mos
Fletcher and Davies (1986)	26	M	Tongue	N.S.	N.E.D. 2.5 yrs
Heifetz et al. (1991)	6	F	Buccal mucosa	Excision	N.E.D. 2 yrs
Krolls et al. (1994)	21	F	Hard palate	Excision	Recurred after 3 yrs.
Ishida et al. (1998)	24	F	Upper lip, forearm, abdominal skin (4), foot skin	N.S.	N.S.	Patient with NF2, also had bilateral acoustic nerve schwannomas and an “ordinary” schwannoma of the hand
	52	M	Tongue	N.S.	N.S.	Also had an “ordinary” schwannoma of the mediastinum and a meningioma; findings did not fulfill criteria for NF2
Di Giovanni et al. (2006 )	16	M	Oral vestibule	Excision	6 mos later developed an additional nodule in the vestibule and 3 minor lesions at the lip commissure; N.E.D. 7 yrs after re-excision
Hashiba et al. (2007)	12	F	Upper lip	Extracapsular dissection	Recurred twice (9 mos, then 7 mos after 2nd operation; 2nd recurrence was not resected until 10 years after 2nd operation); 27 mos after third resection N.E.D.
Ferreti Bonan et al. (2008)	46	F	Tongue	Excision	N.E.D. 12 mos
Berg et al. (2008)	N.S.	N.S.	Lip	N.S.	N.S.
	5	F	Lip and thigh	N.S.	N.E.D. 1 yr	Patient with NF2
	N.S.	N.S.	Lip	N.S.	N.S.
	N.S.	N.S.	Lip	N.S.	N.S.
	N.S.	N.S.	Tongue	N.S.	N.S.
	N.S.	N.S.	Tongue	N.S.	N.S.
	N.S.	N.S.	Right mental nerve	N.S.	N.S.
Lobo et al. (2009)	9	M	Lip mucosa	Excision	N.E.D. 9 mos
Vera-Sempere and Vera-Sirera (2010)	46	F	Mandible (intraosseous, arose from IAN)	Enucleation	N.E.D. 2 yrs, mild lower lip paresthesia
Bavle et al. (2011)	54	M	Mandible (intraosseous)	Enucleation and curettage	N.E.D.1 yr
dos Santos et al. (2011)	3	F	Hard palate	Excision	N.E.D. 14 mos
Al-Mahdi et al. (2012)	27	M	Tongue, FOM, and submandibular region; arose from hypoglossal nerve	Excision with sacrifice of hypoglossal nerve	N.S.
Lambade et al. (2013)	33	M	Maxilla	Excision	N.S.
Dezfuli et al. (2017)	34	F	FOM	Excision	N.S.
Amer et al. (2018)	13	F	Tongue	Excision	N.E.D. “up to the present time” (~3 yrs)
Alotaiby et al. (2019)	25	M	Palatal gingiva	N.S.	N.S.
Sergheraert et al. (2019)	28	F	Hard palate (greater palatine nerve)	Excision	N.E.D. 5 yrs
Chi et al. (2020)	27	F	Hard palate	Excision	Not available
	27	F	Lateral border of tongue	Excision	Not available
	58	M	Dorsal tongue	Excision	Not available
	41	M	Upper lip mucosa	Excision	Not available
	25	F	Lateral border of tongue	Excision	Not available
	62	F	Tongue	Excision	Not available
	18	M	Lower lip	Excision	Not available
	32	F	Dorsal tongue and buccal mucosa	Excision	Not available	Patient with NF2

F female, FOM  floor of mouth, M  male, mos  months, IAN  inferior alveolar nerve, N.E.D. no evidence of disease, NF2 neurofibromatosis 2, N.S. not stated, WNL  within normal limits, yrs  years

All cases of oral plexiform schwannoma clinically appeared as a solitary mass or localized cluster of tumor nodules, with the exception of one of our patients who had two distinct nodules (on the tongue and buccal mucosa) and a history of NF2. In addition, extraoral neural neoplasms were evident in four patients, including three with NF2.[1, 13, 29] Extraoral tumor types in the NF2 patients included conventional schwannoma, bilateral acoustic schwannomas, plexiform schwannoma, meningioma, optic nerve glioma, and neurofibroma. The third patient had a meningioma and mediastinal “nonplexiform type” schwannoma but lacked additional features necessary for a diagnosis of NF2.

Imaging studies were performed in 8 cases [11, 12, 18–22, 25]. The three intraosseous jaw lesions appeared radiolucent, with two of these lesions described as multilocular, well-defined (or “distinct”), and expansile [18, 20, 22]. Other radiographic findings included root resorption (n = 2), “diffuse” borders (n = 1), and cortical thinning (n = 1).

The pre-operative diagnostic impression was stated for 19 cases. The most frequent considerations were fibroma (n = 8) and salivary gland tumor (of unspecified type or pleomorphic adenoma) (n = 6). Additional impressions included benign mesenchymal (or soft tissue) neoplasm (n = 2), lipoma (n = 2), neurofibroma (n = 2), “neural tumor” (n = 1), lymphoepithelial cyst (n = 1), and peripheral odontogenic tumor (n = 1). Among the intraosseous cases, various benign odontogenic lesions (e.g., odontogenic keratocyst, residual cyst, ameloblastoma, central odontogenic fibroma) and central giant cell granuloma were considered.

Initial treatment consisted of excision or enucleation (n = 22), enucleation and curettage (n = 1), or extracapsular dissection (n = 1). In two cases, excision was accompanied by removal of an associated nerve. The average follow-up period (among the 14 cases for which this information was provided or available) was approximately 3.9 years (range, 9 months to > 13 years). Recurrence of an oral plexiform schwannoma was noted in three patients, including one with a lesion that recurred twice [12, 14, 15]. The recurrent tumors typically were resected, although in the case reported by Di Giovanni et al. [14] a small satellite tumor nodule deep to the muscle fascia was not removed.

Grossly, a multinodular or multilobular mass was evident in 19 cases, and the tumors were described as well-circumscribed and/or encapsulated in 7 cases. There was variability in tumor consistency ( firm = 6, soft n = 2, rubbery n = 1); external color (tan n = 1, brown n = 1, gray-yellow n = 1, yellow-white n = 1, creamy white n = 1); and color of the cut surface (gray-tan n = 1, yellow-white n = 1, gray-white to yellow n = 1, solid cream n = 1, white = 1).

Microscopic descriptions for the previously reported cases typically mentioned well-circumscribed and/or encapsulated tumor nodules (n = 26). Likewise, all 8 cases in the current series exhibited multiple well-circumscribed and partially to completely encapsulated tumor nodules (Figs. 1a, b and 2a, d). In one of our cases, the individual nodules were comprised of small, round “micronodules” (Fig. 3a, b). Antoni A and B patterns were evident in variable proportions (Antoni A predominant in 20 of the current and previously reported cases) (Figs. 1b, 3b), with a proliferation of bland spindle cells. However, occasional “benign pleomorphism” was noted in the two cases reported by Ishida et al. [13], and one case in our archives exhibited modest nuclear enlargement and pleomorphism (Fig. 4). Nevertheless, no additional features characteristic of degenerative or “ancient” change (i.e., perivascular hyalinization, hemorrhage, cyst formation, calcification, histiocytic infiltrate) and no features of malignancy (i.e., overt atypia, tumor necrosis, atypical mitoses, increased mitotic activity) were noted in any of the current or previously reported cases.

Fig. 1.

Representative photomicrographs for a case in our series. a Low-power (hematoxylin and eosin, magnification 40x) and b medium-power (hematoxylin and eosin, magnification × 200) views showing multiple well-circumscribed and encapsulated tumor nodules with a predominantly Antoni A growth pattern. c There was strong, diffuse immunoreactivity for S-100 protein (magnification × 200). d Immunohistochemistry for EMA highlights a thin perineurium-derived capsule (magnification × 200)

Fig. 2.

Representative photomicrographs for another case in our series. Low-power views showing a multiple well-circumscribed and encapsulated tumor nodules (hematoxylin and eosin, magnification × 40), b strong immunoreactivity for S-100 protein (magnification × 40), and c immunoreactivity for EMA among the perineurium-derived capsules (magnification × 40). Corresponding medium-power views for one of the tumor nodules in Fig. 2a–c are shown in d (hematoxylin and eosin, × 200), e (S-100 immunohistochemistry, × 200), and f (EMA immunohistochemistry, × 200)

Fig. 3.

A micronodular growth pattern was evident in one case: a hematoxylin and eosin (× 200), b hematoxylin and eosin (× 400)

Fig. 4.

Photomicrograph showing modest nuclear enlargement and pleomorphism in one case in our series (magnification × 400)

The results of immunohistochemistry are summarized in Table 2. Major findings included expression of S-100 protein among the tumor cells (21 out of 21 cases, with most exhibiting strong and/or diffuse immunoreactivity) (Figs. 1c, 2b, e) and expression of EMA among the perineurium-derived capsular cells surrounding the tumor nodules (13 out of 15 cases) (Figs. 1d and 2c, f). In addition, focal NFP immunoreactivity was noted at the periphery (among capsular or subcapsular cells) and/or more centrally within the tumor nodules in 10 out of 14 cases (Fig. 5).

Table 2.

Results of immunohistochemistry for current and previously reported cases

Antigen	Results for current cases	Results for previously reported cases
S-100	8/8 positive (strong, diffuse, nuclear and cytoplasmic)	13/13 positive, with 11 described as strong and/or diffuse
EMA	7/8 positive in perineurium-derived capsular cells surrounding tumor nodules	6/7 positive in perineurium-derived capsular cells surrounding tumor nodules; 1/7 also positive in Antoni A areas
NFP	5/8 positive among isolated peripheral and central cells	4/6 focally, mainly, or rarely positive at periphery and/or within septa; 1/6 positive among residual axons within tumor nodules
CD34	NP	3/4 positive at periphery of some tumor nodules, 1/4 focally positive
GFAP	NP	1/2 positive in perineurium and some tumor cells
vimentin	NP	2/3 positive
CD56	NP	1/1 positive in tumor and perineurium
CD57	NP	2/2 positive
CD68	NP	1/1 “variably” positive
CD117	NP	2/2 positive (focally, among mast cells)
SOX10	NP	1/1 positive (strong and diffuse)
laminin	NP	1/1 positive
AE1/AE3	NP	0/2 positive
SMA	NP	0/2 positive
desmin	NP	0/1 positive
calponin	NP	0/1 positive
Factor XIIIa	NP	0/1 positive
Bcl-2	NP	0/1 positive
SMARCB1	NP	1/1 positive (strong)
Ki-67 proliferation index	NP	3/3 low (“<4–5%,” “rare,” or “mild”)

NP  not performed

Fig. 5.

Immunohistochemistry for neurofilament protein (NFP) highlights an isolated intratumoral axon (magnification 400x)

Discussion

The clinical presentation of plexiform schwannomas overall –whether involving extraoral or intraoral sites—is variable. The tumor has been reported over a broad age range (infancy through the eighth decade) [1, 30], and a given patient may have solitary or, at times, multiple lesions. Similar to conventional schwannomas, most examples arise in the skin or superficial soft tissue, with a predilection for the extremities as well as the head and neck region [29]. In superficial locations (including the oral mucosa), the tumors tend to be innocuous, painless, and slowly growing nodules, with a low risk for recurrence after conservative excision. However, as noted in our review of oral lesions, pain/discomfort, paresthesia, functional disturbance, and recurrence are possible. Furthermore, plexiform schwannomas at times may involve the spinal cord/major nerves, deep soft tissue, or viscera [6, 31]. In such locations, the tumors may cause significant morbidity and may be difficult to excise with negative margins.

Plexiform schwannomas arising within the setting of NF2 or schwannomatosis have been reported in a minority of patients, although the exact frequency is unclear. Berg et al. [1] noted that 5% of individuals in their large series of plexiform schwannomas had NF2 or schwannomatosis, and Evans [32] commented that only 4 of > 450 NF2 patients whom he examined exhibited plexiform schwannomas. In our review of oral plexiform schwannomas, a history of NF2 or schwannomatosis was evident in 4 of 36 (11%) of cases. In one example from our archives, a patient with NF2 presented with two distinct plexiform schwannomas, involving the tongue and buccal mucosa. In two previously reported cases, NF2 patients exhibited oral as well as extraoral plexiform schwannomas (involving the skin, forearm, and thigh); additional extraoral findings included conventional schwannoma and bilateral acoustic schwannomas [1, 13]. Furthermore, Ishida et al. [13] described a schwannomatosis patient with a plexiform schwannoma on the tongue, a conventional schwannoma in the mediastinum, a meningioma, and no personal or family history of NF2. Overall, although plexiform schwannomas appear to exhibit a weak association with NF2 or schwannomatosis, thorough examination of any patient diagnosed with a plexiform schwannoma in order to rule out such systemic conditions would appear to be prudent.

The neurofibromatoses represent a group of heterogeneous neurogenetic disorders, characterized by a predisposition for developing multiple nerve sheath tumors [33]. They exhibit an autosomal dominant inheritance pattern, but sporadic cases are possible. Major disorders within this group include NF1, NF2, and schwannomatosis. NF1 is most common, occurring in approximately 1 of 2500 to 3000 births [34]. It is caused by mutations in the NF1 tumor suppressor gene located on chromosome 17q11.2, which encodes the protein neurofibromin. Major clinical findings include neurofibromas (including conventional and plexiform types); pigmentary abnormalities (café au lait macules, freckling, Lisch nodules); optic glioma; brainstem glioma; learning, behavioral, and attention deficits; and long bone dysplasia. In addition, affected individuals exhibit an increased risk for developing malignant peripheral nerve sheath tumors and other malignancies. NF2 is caused by mutations in the NF2 tumor suppressor gene, which is located on chromosome 22 and encodes the protein merlin (schwannomin). Individuals with this condition exhibit a predisposition for developing acoustic (vestibular) schwannomas; schwannomas of other cranial, spinal, and peripheral nerves; intracranial and intraspinal meningiomas; certain low-grade central nervous system malignancies (e.g., ependymomas, low-grade gliomas); cerebral calcifications; juvenile posterior subcapsular lenticular opacities and peripheral cortical cataracts; retinal hamartomas; and large café au lait macules (typically fewer in number compared to NF1) [34]. Neurofibromas are possible as well [32]. Schwannomatosis is a rare condition associated with mutations in the LZTR1 and SMARCB1 (INI1) genes on chromosome 22 [34]. More than 95% of cases are sporadic [35]. Affected individuals tend to develop peripheral and spinal schwannomas, meningiomas, and chronic pain; however, they do not meet the diagnostic criteria for NF2 (e.g., they exhibit no evidence of vestibular schwannoma or NF2 mutation) [36].

The term “plexiform” generally is used by pathologists to describe multinodular growth at the macroscopic and/or microscopic level. Gross examination of plexiform lesions may show multiple entwined tumor nodules likened to a “bag of worms,” and histopathologic examination may reveal multiple discrete tumor lobules simulating a network or “plexus” [37]. In formulating the inclusion criteria for the present study, we adhered to this generally accepted definition of multinodular growth grossly and/or microscopically. However, we recognize that some authors have proposed more stringent application of such terminology. For example, within the context of palisaded encapsulated neuromas, Argenyi et al. [38] distinguished between the terms “plexiform” (repetitive appearance of cross, oblique, and longitudinal sections of markedly expanded, sometimes nodular nerve fascicles arranged in close juxtaposition) and “multinodular” (nodular fascicles arranged in a perpendicular fashion). Also, some investigators have used serial or step sections in order to demonstrate a lack of continuity or branching among individual tumor nodules within plexiform lesions [9]; in the current study, no continuity among tumor nodules was evident upon examination of multiple levels of sections. With regard to plexiform neurofibroma, Goldblum et al. [39] commented that this terminology should be restricted to large lesions distorting or contorting large segments of a nerve into a “bag of worms,” as opposed to smaller lesions with a plexiform pattern evident microscopically but not macroscopically.

The microscopic differential diagnosis for plexiform schwannoma includes other neural tumors, including plexiform neurofibroma, palisaded encapsulated neuroma, and traumatic neuroma. Microscopically, both plexiform neurofibromas and plexiform schwannomas exhibit well-delineated tumor nodules encapsulated by perineurium, although plexiform neurofibromas may “spill out” into the surrounding tissues to form areas with a diffuse growth pattern as well. Characteristically, neurofibromas exhibit interlacing bundles of spindle cells with wavy nuclei, ropey strands of collagen (likened to “shredded carrots”), and a variably myxoid background. Also, neurofibromas lack the alternating Antoni A (cellular proliferation of spindle cells arranged in ordered fascicles, often with nuclear palisading and formation of Verocay bodies) and Antoni B (more loosely arranged and myxoid) patterns characteristic of schwannomas. Immunohistochemistry typically exhibits more uniformly intense reactivity for S-100 protein in schwannomas than neurofibromas; this difference in extent of S-100 immunoreactivity is explained by the more uniform tumor population derived mainly from Schwann cells in schwannomas, as opposed to the more varied admixture of Schwann cells, fibroblasts, perineurial-like cells, and axons in neurofibromas. Importantly, in contrast to plexiform schwannoma, which is weakly associated with NF2 or schwannomatosis and has no malignant potential, plexiform neurofibroma is strongly associated with NF1 and does have the potential for malignant transformation. Indeed, when applying the stringent definition for plexiform neurofibroma proposed by Goldblum et al., the lesion is nearly pathognomonic for NF1 [39]. Plexiform growth also can be seen in tumors with hybrid features of neurofibroma and schwannoma (so-called “hybrid neurofibroma/schwannoma”) [35]. Many of these hybrid tumors have arisen in patients with schwannomatosis, NF2, or NF1. Interestingly, investigators have identified monosomy 22 in nearly half of hybrid neurofibromas/schwannomas subjected to molecular analysis [40].

Palisaded encapsulated neuromas also can appear very similar to plexiform schwannomas, in that they are typically well-circumscribed and sometimes exhibit a plexiform or multilobular growth pattern [41]. However, a lesser degree of nuclear palisading, fewer stromal changes (such as myxoid change or stromal hyalinization), a higher frequency of intratumoral axons, and a more delicate and discontinuous EMA-positive, perineurium-derived capsule would favor palisaded encapsulated neuroma over plexiform schwannoma.

An additional consideration in the differential diagnosis is the traumatic neuroma, which represents a non-neoplastic proliferation resulting from nerve injury. Identification of the associated damaged nerve should aid in distinguishing a traumatic neuroma from a schwannoma. Also, traumatic neuromas lack the alternating Antoni A and Antoni B patterns characteristic of schwannomas and exhibit participation of all elements of a nerve fascicle rather than a predominance of Schwann cells.

Traditionally, schwannomas have been purported to derive exclusively from Schwann cells. Ultrastructural studies showing a virtually pure population of cells with schwannian differentiation suggest that these tumors develop eccentrically from nerve surfaces and, thus, lack axons [39]. However, more recent studies have challenged this concept. In particular, Nascimento and Fletcher [42] demonstrated NFP-positive axons in 55% of conventional schwannomas and 30% of plexiform schwannomas. Rare to multifocal axons were detected centrally (within the tumor proper) and/or peripherally. Similarly, in 5 out of the 8 oral plexiform schwannomas in the current series, we noted isolated NFP immunoreactivity peripherally (among capsular and subcapsular cells) as well as more centrally within the tumor nodules. Such NFP-positive axons merely could represent entrapped nerve fibers, although it is also intriguing to consider that schwannomas might contain a somewhat more diverse cell population than previously believed.

It is acceptable for schwannomas to demonstrate increased cellularity and/or modest nuclear pleomorphism, and such findings should be not be interpreted as evidence of malignancy [13, 41]. Interestingly, Goldblum et al. [39] commented that plexiform schwannomas in particular tend to exhibit greater cellularity than conventional schwannomas. Although we did not encounter this feature among the oral plexiform schwannomas in our study, there was one case with somewhat enlarged, pleomorphic nuclei (Fig. 4). Findings that generally aid in distinguishing plexiform schwannoma from a malignant peripheral nerve sheath tumor include the presence of well-circumscribed tumor nodules surrounded by perineurium-derived capsules, the absence of necrosis, a low Ki-67 proliferation index, and strong and diffuse S-100 protein expression. However, Woodruff et al. [43] cautioned that an unusual subset of childhood plexiform schwannomas may exhibit hypercellularity, nuclei three times larger than those in typical neurofibromas, tissue infiltration by small tumor nodules, Ki-67 proliferation index as high as 30%, and mitotic activity ranging from 4 to 31 per 10 high-power fields. These “troublesome” pediatric cases nevertheless had benign—albeit at times locally aggressive-behavior.

Excision of oral plexiform schwannomas typically is curative, although recurrence has been reported in three (8% of) cases [12, 14, 15]. Because of the inherent multinodularity of these tumors, it is possible that reported recurrences actually may represent persistence of incompletely excised lesions. Complete surgical removal usually can be performed without significant morbidity, although sacrifice of an involved major nerve may be necessary at times. Mild post-surgical paresthesia has been described in one case [18]. Malignant transformation of plexiform schwannoma has not been reported.

In conclusion, we have presented a series of 8 oral plexiform schwannomas and reviewed the literature regarding such lesions. Within the oral cavity, plexiform schwannomas exhibit a predilection for the tongue and lip. Most lesions present as slowly growing, asymptomatic nodules, although pain, discomfort, paresthesia, and other symptoms are possible. Three patients with NF2 were noted among the 36 current and previously reported cases of oral plexiform schwannoma in our study. Unlike plexiform neurofibromas, plexiform schwannomas exhibit only a weak association with neurofibromatosis and have no known malignant potential.

Funding

No funding obtained.

Compliance with Ethical Standrads

Conflict of interest

No conflicts of interest to disclose.