Parotid Neurogenic Tumors: MPNST Sarcoma to Schwannoma—Review of Literature and Guidelines in Management

Abstract

Objective

To delineate a distinct clinical, pathological, cytoimmunohistochemical distinguishing features of extracranial parotid neurogenic tumors. Pleomorphic adenoma, the most common parotid neoplasm by itself can mimic a neurogenic tumor, i.e., schwannoma. We have reported such rare entity in earlier publications. This is a systematic review with a case discussion of the third only documented malignant peripheral nerve sheath tumor (MPNST) arising in the parotid in a noncontiguous fashion. We review the management, diagnostic immunohistochemistry of this low-grade sarcoma, which has not been documented thus far.

Purpose

Significant diagnostic and management pitfalls may occur even with fine-needle aspiration or biopsy, as they will not be ideal in diagnosis of parotid neurogenic neoplasms preoperatively. Ultrasound as well as magnetic resonance imaging may not offer a specific facial nerve course in association to the neoplastic lesion. Majority of the neurogenic tumors can be misdiagnosed and hence, improperly managed leading to facial nerve injury and tumor morbidity. Due to the lack of ideal diagnostic modality and morbidity of facial nerve injury, a thorough review of differential diagnosis inclusive of neurogenic tumors of the parotid is to be considered in indications of surgical approaches. We retrospectively review successful resection with preservation of facial nerve function in our parotid neurogenic tumors.

Conclusion

The objective of this paper is to review the guidelines and treatment planning of parotid neurogenic tumors.

Introduction

Malignant peripheral nerve sheath tumor (MPNST) is a malignant soft tissue sarcomas with origins from peripheral nerves. These sarcomas demonstrate histopathological and immunohistochemical characteristic features of the nerve sheath cells, including Schwann cells, perineural cells and fibroblasts [1]. MPNSTs account for 5–10% of all soft tissue sarcomas [2]. The proximal extremities, trunk, retroperitoneum are the predominant anatomical sites, head, and neck, including the scalp, accounts for only 2–9% of cases [3].

Approximately, half of MPNSTs arise from neurofibromas in patients with neurofibromatosis type 1(NF1). However, we know from the literature that MPNSTs can develop sporadically in otherwise normal tissues or some rare cases associated with exposure to or previous treatment with radiation. In NF1, the average age of onset is 10–15 years younger than in non-NF1. Central anatomical locations have an increased incidence of MPNST [4–6]. Alternatively, in non-NF1 cases, MPNSTs develop from large and medium-sized nerve trunks in deep anatomic sites on the proximal extremities or the trunk. There appears to be an existing association with radiation exposure in these areas that also corresponds to the development of a particularly aggressive, rapidly growing and treatment-resistant tumor [7, 8]. The etiology is unknown, but they have a known association with NF‐1, an autosomal‐dominant disorder that involves the NF‐1 tumor suppressor gene, with a known location on chromosome 17. Up to 50% of MPNSTs occur in patients with NF‐1. The incidence is higher in patients with a history of radiation exposure [9–11].

Only a few primary, superficial MPNSTs documented in the literature that arises in the cutaneous or subcutaneous tissues with a unique characteristic of more than half developing on the head and neck with the scalp have the most significant incidence [12]. According to one study, tumors have presented primarily in the peripheral and autonomic nerves of the head and neck (46.0%), connective/soft tissue of the head and neck (31.5%), and nasal cavity and paranasal sinuses (9.6%), skin of head and neck (2.2%), major salivary glands (2.2%) and other head and neck sites (8.6%) [13].

The superficial MPNSTs are thought to derive from cutaneous neurofibromas or small peripheral nerves. They are noted for being markedly aggressive with rapid invasive growth, recurrence after excision, and early hematogenous dissemination [1]. In general, MPNSTs are considered highly aggressive tumors with metastatic potential and poor prognosis despite aggressive therapy. Like other sarcomas, MPNSTs tend to recur and metastasize hematogenously [15].

We performed a systematic review of the literature regarding the clinical presentation, diagnosis, treatment modalities, and outcome for MPNSTs in this anatomic location. In this case review and systematic analysis, we present the clinicopathologic features of a primary MPNST arising in a cutaneous neurofibroma on the scalp with metastasis to the parotid gland in a man without neurofibromatosis type 1.

University of Nebraska institutional review board ethical approval was applied with a review of the presentation and granted the exemption.

Case Review

A 76-year Caucasian male reported a history of lumps on the anterior skull after minor trauma in his fourth decade of life. A rapid escalation in size was noted in 2008, which led to a soft tissue biopsy from the scalp's vertex. Pathology was consistent with low-grade malignant peripheral nerve sheath tumor arising within a neurofibroma, representing a sporadic case. Wide resection was compatible with NF with tumor present at the margins. An extensive resection with skin grafting a year later of the scalp had cellular NF present with Meissner-like bodies at the margins. Once again, a component of low-grade malignant peripheral nerve sheath tumor arose in an NF present to 0.1 cm at the deep margin. An indeterminate NF at the margin led to resections for margin clearance. This had more diffuse NF with atypia and persistent positive margins for NF, but not the malignancy. The oncologist did not recommend radiation therapy then.

Multiple local follow ups of small nodule excisions were consistent with a malignant peripheral nerve sheath tumor. Immunohistochemistry staining was diffusely positive for S100 and no significant staining for neurofilament, Melan-A (Fig. 1). Increasing nodules led to extensive resections yearly from 2013 to 2018 (largest 8 × 15 cm). Pathology was consistent with MPNST, epithelioid type with the focal presence of tumor at the deep margin. Staining was positive for both S100 and SOX10 with no significant staining for EMA, melan-A, or desmin. Posterior scalp biopsy also revealed MPNST with negative margins on the section examined. The largest nodule had a diameter of about 5.3 cm and was confluent with smaller nodules with diameters of 0.5–2 cm. There was a relatively normal expansive skin separating the 2 sides of the multinodular mass. However, this relatively normal 5.5 cm area had multiple nodules present when it was cross-sectioned.

Fig. 1 .

Left to right: a H&E, b S100, c MelanA, d Neurofilament. a H&E staining demonstrates a poorly defined spindled to epitheloid appearing lesion arranged in irregular fascicles. b S100 staining highlights spindle cells that are diffusely positive, demonstrating origin within neural tissue. c Stains for neurofilament (neuronal cells) and d Melan-A (melanoma) were negative

The patient referred to our institution in 2018 had an MRI of the brain performed and revealed a large lobulated right scalp heterogeneous mass measuring 9.8 × 8.9 × 6.3 cm adjacent to the outer calvarial table. The anterior left forehead had a lobulated frontal subcutaneous bulk measuring 2.8 × 6.7 × 2 cm. We found no intracranial masses, but a left parotid mass of 2 cm was seen on the MRI. A PET CT had avidity of the right anterolateral apex of the skull, the paracentral frontal scalp, and along the posterior margin of the temporalis in the right scalp, and the large left intra-parotid mass regions.

An extensive resection as illustrated was performed of an 18.1 cm × 12.7 cm × 5.0 cm multinodular scalp mass from the supraorbital ridge bilaterally eroding the skull in the midline with extensive cortical destruction extending posteriorly to the vertex with outer cortical stripping (Fig. 2). The parotidectomy revealed a left parotid mass that was not attached to any peripheral nerve. The scalp mass was diagnosed as MPNST measuring 18.1 cm in the most significant dimension with negative peripheral margins but sarcoma focally extending to the deep margin (Fig. 3). The parotid gland specimen demonstrates a metastatic malignant peripheral nerve sheath tumor measuring 2.6 cm in the greatest dimension, invading a lymph node and focally extending to the peripheral margin (Fig. 4).

Fig. 2.

Microscopic appearance of MPNST stained with H&E a low power, b medium power, c high power. Areas of high cellularity, myxoid stroma, and nuclear palisading are shown. Hyperchromatic and wavy features can be appreciated at mid and high power. Pleomorphism is variable

Fig. 3.

a Primary MPNST arising in a cutaneous neurofibroma on the scalp. b MRI coronal section of scan showing the scalp tumor with a metastasis to the left parotid gland in a man without neurofibromatosis type 1

Fig. 4.

a High power picture of tumor-lymph node interface. Metastasis or direct invasion is impossible to tell. b High power view of tumor metastatic to parotid gland. A cell undergoing mitosis is seen near the centre of the picture. Inside the thickened blood vessel at the left of the picture, the red blood cells can be used to compare the variability of nuclear size in the invasive tumor (1–4 times larger than a red blood cell)

Materials and Methods

The focused research question was: "Malignant peripheral nerve tumors with parotid metastasis?". A systematic review to evaluate the current studies on head and neck malignant peripheral nerve tumors and their surgical management is the objective. Knowledge of this subject will allow clinicians to make better treatment planning decisions and provide more predictable outcomes for the patient, including the unreported adjuvant therapy.

PubMed, Medline, and Embase databases for studies in English using the terms "MPNST" or "MPNST parotid," "Parotid neurogenic sarcoma," etc. 428 initial reports were identified for possible study inclusion. The vast majority (n = 402) excluded after reviewing abstracts or titles. After reviewing the relevant studies' full texts, 18 studies were excluded due to one or more of the following reasons: not addressing the primary study question or analysis was not metastatic to the parotid. The majority of studies reviewed was level 3 cohort studies. No Randomized clinical trials were identified. A modified Delphi's criteria list entailing of seventeen criteria was used to evaluate the quality of the studies.

Results

Of the cases reviewed with parotid gland involvement by MPNST, only two were involved by metastasis. The most extensive database review of 160 cases reported by the French sarcoma study group in 2016, led by Le Guellec et al. in American journal of the pathology of both NF-1 and sporadic ones, have established an excellent immunohistochemical differential of this rare diagnosis. Most studies are case reports with the literature review, and there is no recorded parotid metastasis, especially in the NF-1 type. Studies varied in immunohistochemistry confirmation of the metastasis, and the S100 was positive and nonreactive to HMB-45 in most cases. There was a large variation of the treatment of the metastatic MPNST and the primary lesion in rendering adjuvant chemoradiation. There are only 2 previous cases of an MPNST metastasizing to the parotid gland have been reported (Table 1) [10, 11, 30].

Table 1.

Literature review of studies on patients with MPNST metastasized to the parotid gland

Study	Sex	Age	Neurofibromatosis type 1	Tumor location	Tumor size (cm)	Immunohisto chemistry	Bone infiltration	Local recurrence	Metastasis (location)	Outcome	Treatment
Colville 2003 [11]	Male	31	NA	Temple	1.9	NA	NA	NA	Parotid, Lung, Bone	Dead of disease (6 years later)	Excision, palliative chemoradiation
Colville 2003 [11]	Male	19	NA	Preauricular	NA	NA	NA	No	Parotid	Alive, free of disease (16 mo later)	Excision, parotidectomy, modified radical neck dissection, radiation
Present case	Male	78	No	Scalp	20	S100 and SOX-10	Yes	Yes	Parotid	Alive with disease	Excision, parotidectomy

The mechanism of parotid infiltration and the metastatic behaviour in not documented in literature review. And that is our effort in this systematic analysis. The scans and imaging have no continuity of tumor between the scalp and parotid. And we have provided the most representative images. And as avidity on PET can be as sensitive to represent any contiguous tumor, the parotid is isolated from the scalp lesion in the work up and post resection review.

Discussion

There are five tissue layers that compose the scalp: pericranium, loose areolar tissue, aponeurosis, connective tissue, and skin. Like tissues elsewhere in the body, benign and malignant tumors may be associated with any of the scalp's layers [12]. Globally MPNST has a prevalence of 0.001% in the general population. Only a few cases of MPNSTs have been reported to have originated in the skin accounting for approximately 2–9% of all MPNSTs, of which 50% are in the head and neck with the scalp as the main location [9].

Surgical ablation by resection is the first choice of treatment for MPNST and the only way to control the tumor completely; we could find no patients who survived without radical surgery. In the head and neck area, not only tumor size but also an invasion of the non resectable, critical regions, such as the skull base, internal cervical artery, or cervical vertebrae, determine surgical extent, surgical margin, and, finally, prognosis. As reported, the histologic grade was predictive of the development of distant metastasis and survival [27, 28].

However, a complex anatomy of the head and neck, its vital structures constituents, and the potential for microscopic local spread, obtaining negative margins can be problematic. The margin clearance challenges have led to higher local recurrence percentages (22–52%) and distant metastasis (18–33%) outcomes. So a multimodality treatment approach is most favorable [17–19].

There are only a few reports of MPNSTs arising primarily in the parotid gland [20–23]. Apart from our case, only two previous cases of an MPNST metastasizing to the parotid gland are documented (Table 1) [10, 11, 30]. The current opinion is that these tumors metastasize by the hematogenous route, not through the lymph nodes. However, it is recognized that rarely the path they take is via lymph nodes. But that is rare. In our case, to make it more exceptional, it metastasized to the intra-parotid lymph node. It is impossible to tell in this case if it actually metastasized to the parotid gland (which is vascular rich) and then infiltrated the lymph node by direct extension. The opposite could also be true, as known in large cutaneous epithelial and secretory parotid tumors. Another possibility would be that it arose de novo, as a secondary malignant tumor, in an intra-parotid nerve, and then the cancer took over the typical architecture. The above behavior is not documented in the literature review. And that is our effort in this systematic analysis. The scans and imaging have no continuity of tumor between the scalp and parotid. And we have provided the most representative images. And as avidity on PET can be as sensitive to represent any contiguous tumor, the parotid is isolated from the scalp lesion in workup and post-resection review. Malignant Peripheral Nerve Sheath Tumor (MPNST) terminology is the current identification as a name of neurofibrosarcoma, neurogenic sarcoma and malignant schwannoma as reported entities. Grossly, MPNSTs are poorly defined and tend to have infiltration along adjacent nerves or adjacent soft tissue. The cut surfaces are tan-white, fleshy to gelatinous with focal areas of hemorrhage and necrosis. The microscopic examination consists of regions of cellular spindle cells with fascicular growth pattern combined with alternating hypercellular and hypocellular zones with myxoid stroma (Fig. 1). Nuclear palisading and whorled nodules of spindle cells are frequently noted with spindle cells showing hyperchromatic wavy nuclei that may vary in terms of pleomorphism (from minimal to prominent) [13, 14].

Immunohistochemically, S100 staining is positive in up to 50%. There is a variable staining pattern with CD34 and GFAP. Staining should be negative for keratin, SMA, desmin, HMB-45 [15]. Up until now, genetic testing has demonstrated no difference between NF1-associated and sporadic tumors and is negative for MDM2 amplification [14].

Of the cases reviewed with parotid gland involvement by MPNST, only two were involved by metastasis. Of the cases reported with metastasis to the parotid, none appears to have had neurofibromatosis. Our patient was diagnosed with MPNST in 2011 and his morbidity has been increasing for 7 years after his first excision of tumor. Noteworthy is that his recurrences have been treated by re-excision of tumor and with a recent parotidectomy. One patient who had survived his parotid metastasis was treated with excision of tumor and parotidectomy with modified radical neck dissection and radiation. The other patient with parotid metastasis was treated with excision of the tumor but died after distant metastasis 6 years later to his first excision [11].

In general, MPNST is an aggressive tumor that has a poor prognosis. Because of poor prognosis and insensitivity to radiotherapy and chemotherapy, complete surgical resection of the tumor and the negative margin is the primary treatment of MPNST. Loco-genional rate of recurrence and distant metastasis is very high, and distant metastasis mainly shifts to the lungs. The clinical characters of MPNST indicating poor prognosis include tumor size > 5 cm, high tumor grade, NF1 correlation, distant metastasis, surgical margin positive, S100 negative, and Ki-67 index > 20% [14, 16–19]. Multiple studies have shown that the 5-year overall survival rate of MPNST is 44–57%, the 5-year disease-specific survival rate is 32–50%, and the prognosis of head and neck MPNST is worse [24–28].

A study of 140 cases of MPNST in the head and neck region found that those with advanced AJCC stage and negative S-100 have a low 5-year overall survival rate. Additionally, those with positive margins, a Ki-67 greater than 20%, and those that did not receive chemotherapy had a greater recurrence rate. Overall, MPNST is associated with poorer prognosis when cancer is advanced, lesions with positive margins, and if Ki-67 is greater than 20% [14].

A study of 11 patients with MPNST of the scalp describes increased local recurrences and mortality in cases with histologically high-grade tumors, larger size (> 5 cm), incomplete surgical excision, and intracranial tumor extension. Another study found better survival rates in cutaneous MPNSTs versus deep MPNSTs, with 62% of cases showing no indication of disease after a follow up of 3.3 years. The difference in prognosis between cutaneous and deep MPNSTs is attributed to the early manifestation and treatment of the cutaneous MPNSTs. Given the rapid growth and potential metastasis, surgical treatment with margins greater or equal to 2 cm and adjuvant radiation must be considered standard treatment [14–17].

As an adjunct to surgical excision, many studies have cited the importance of radiotherapy as adjunctive treatment, especially for improving local control [40, 41, 43–45]. Adjuvant radiation dose regimens as reviewed in studies are around 50–60 Gy. Consensus oncological group streamlining treatment policy for MPNSTs, endorses adjuvant radiotherapy for all "intermediate to high-grade lesions and for low-grade tumors a marginal excision [43]. Even though the group consensus states that radiotherapy utility is for local control. Radiotherapy has not noted to be effective for long-term survival, while a small number of studies have also reported survival benefit [40, 41, 43–45].

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflict of interest associated with this publication.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.