Malignant Peripheral Nerve Sheath Tumor (MPNST)

Malignant peripheral nerve sheath tumour (MPNST) is an aggressive soft-tissue sarcoma arising from the cells of the peripheral nerve sheath. It makes up 5-10% of soft-tissue sarcomas and carries one of the worst prognoses among them, with 5-year survival of 40-60% overall. About 50% arise in neurofibromatosis type 1 (NF1), typically by malignant transformation of a plexiform neurofibroma.

How common. The annual incidence in the general population is about 0.001% (1 per 100,000). In NF1 it is 0.16% overall and 2-5% in adults with a plexiform neurofibroma, and the lifetime risk in NF1 is 8-13% (Evans, population study), concentrated in those with a plexiform neurofibroma, the at-risk precursor lesion. There is a slight male predominance (1.2:1).

Who. NF1 patients present young, at a median of 26-30 years; sporadic disease presents at a median of 40-50 years. In Evans's population series the gap was 26 against 62 years.

Where. The site distribution:

• Proximal extremities: 40-50% (thigh most common, and upper arm)
• Trunk: 25-30% (paraspinal, retroperitoneum)
• Head and neck: 15-20%
• Distal extremities: 5-10%

The brachial and lumbosacral plexuses are commonly involved in NF1.

Three aetiologies. Half of MPNST is NF1-associated, most of the rest sporadic, and a tenth follows radiotherapy. The origin shapes the age of the patient, the genetics and the reported survival.

Is NF1-associated MPNST prognostically worse? The honest answer is that it is widely taught, plausible, and not established by the evidence on this page. Evans reported five-year survival of 21% in NF1 against 42% in sporadic disease, but at p=0.09, in 21 versus 37 patients. Ducatman (1986) did find neurofibromatosis independently adverse at p less than 0.05. Stucky's 175-patient Mayo series found NF1 associated with disease-specific survival on univariate analysis only; it did not appear among the independent multivariate predictors, where size, grade, local recurrence and truncal location did.

What is not in doubt is that NF1 patients present younger and often with larger, deeper, plexiform-derived tumours that are harder to resect, so any survival difference may be mediated by stage and resectability rather than by NF1 biology itself. Counsel on the size, grade and margin in front of you rather than on the NF1 label.

View larger

The NF1 pathway. Transformation is a sequence of genetic hits. A germline NF1 mutation plus a somatic second hit (loss of heterozygosity) produces the benign neurofibroma: slow-growing, soft, painless, stable for years, of low cellularity with wavy spindle cells. Additional genetic changes produce the atypical neurofibroma, hypercellular with mild atypia, still benign but at higher risk of progression and difficult to diagnose histologically; its consensus name, ANNUBP, has its own section below. TP53 inactivation (75% of MPNST) plus PRC2 complex mutations (SUZ12 or EED loss in 70-90%), with CDKN2A deletion in 50-60%, complete the transformation to MPNST.

What each hit does. Each loss contributes differently:

• NF1 loss: neurofibromin deficiency, so Ras-MAPK is hyperactivated and proliferation runs uncontrolled
• TP53 loss: cell-cycle checkpoints and apoptosis are lost, so mutations accumulate
• PRC2 loss (SUZ12/EED): epigenetic dysregulation, with loss of the H3K27 trimethylation mark
• CDKN2A deletion: loss of the p16 tumour suppressor and cell-cycle dysregulation

Sporadic MPNST. Arises with no prior neurofibroma or NF1 diagnosis, through somatic biallelic NF1 inactivation (not inherited), with TP53 mutations common, PRC2 mutations (SUZ12, EED) and CDKN2A deletions.

Radiation-induced MPNST. Median latency is 10-20 years after radiotherapy, at doses typically over 40 Gy, with the field including a peripheral nerve; examples are treatment for breast cancer, lymphoma and childhood malignancies. These tumours are often higher grade with a worse prognosis.

Which neurofibroma transforms. In NF1 the risk depends on the type and site of the neurofibroma; internal plexiform lesions carry the highest malignant potential and cutaneous neurofibromas rarely transform.

Macroscopic. The tumour is usually over 5 cm at diagnosis (mean 8-10 cm), with poorly defined, infiltrative margins, unlike the encapsulated schwannoma. The cut surface is tan-grey, fleshy and heterogeneous; necrosis is common in high-grade tumours (50-70%) and haemorrhage is frequent. The nerve of origin may be visible in a smaller tumour and is obscured in a large mass.

High-grade MPNST (90% of cases). Dense fascicles of spindle cells in a whorled pattern, a "marbled" alternation of densely cellular and myxoid zones, perivascular accentuation (cells clustered around vessels) and infiltrative margins into the surrounding tissue. The cells have hyperchromatic, wavy, buckled nuclei like those of benign Schwann cells, with moderate to severe pleomorphism and a high nuclear-to-cytoplasmic ratio. Mitoses typically exceed 10 per 10 HPF with atypical forms common, and geographic necrosis is present in 50-70% and predicts aggressive behaviour.

View larger

Low-grade MPNST (10% of cases). A rare variant that is difficult to separate from atypical neurofibroma: fascicles of spindle cells with mild to moderate nuclear atypia, a low mitotic count (under 5 per 10 HPF), minimal necrosis, hypercellularity relative to neurofibroma and, the key feature, an infiltrative growth pattern. Molecular markers (H3K27me3 loss) help, the tumour remains aggressive despite its grade, and the diagnosis requires expert sarcoma pathology review.

View larger

Immunohistochemistry. The neural markers are present but patchy, which is the recurring trap: S100 is positive in only 50-70% and focally, so a negative stain does not exclude MPNST, whereas a schwannoma stains strongly and diffusely.

View larger

Reading H3K27me3. Loss of the mark reflects PRC2 inactivation and supports the diagnosis, and it is retained in neurofibroma and the main mimics, so complete loss in a spindle-cell lesion is meaningful. It is a weak negative: 31% of MPNSTs retained it in Prieto-Granada's series, only about 60% of NF1-related tumours lose it, and all the epithelioid MPNSTs in that series retained it. Retained staining therefore does not exclude MPNST, and least of all in an NF1 patient or an epithelioid tumour.

View larger

Negative markers. These exclude the look-alikes:

• Desmin negative (excludes rhabdomyosarcoma)
• Cytokeratin negative (excludes synovial sarcoma and epithelioid sarcoma)
• CD34 typically negative
• EMA negative (unlike perineurioma)

Variants. MPNST with rhabdomyoblastic differentiation, the malignant Triton tumour, is 5-10% of MPNST, contains a malignant skeletal-muscle component that is desmin, myogenin and MyoD1 positive in the rhabdomyoblastic areas, and has a worse prognosis than conventional MPNST. Epithelioid MPNST has epithelioid rather than spindle cytology, is more common in superficial locations, is often S100 positive and must be distinguished from melanoma and epithelioid sarcoma.

Differential diagnosis. MPNST is a spindle-cell sarcoma that overlaps morphologically with several other neoplasms, and immunohistochemistry and clinical context are what separate them.

The discriminators to carry: diffuse strong S100 favours schwannoma or melanoma over MPNST; H3K27me3 loss favours MPNST and is retained in melanoma, neurofibroma and most mimics; epithelial markers (cytokeratin, EMA, TLE1) point to synovial sarcoma; and smooth-muscle markers point to leiomyosarcoma.

Grading. MPNST is graded by the FNCLCC system, summing differentiation, mitotic count and necrosis.

The total score gives the grade, and the distribution explains why most MPNST is called high grade at diagnosis:

• Grade 1 (low): 2-3 points, rare in MPNST (under 10%)
• Grade 2 (intermediate): 4-5 points (20%)
• Grade 3 (high): 6-8 points (70%)

The intermediate lesion on the road from neurofibroma to MPNST deserves a name and a definition, because it is the diagnostic grey zone that determines whether a lesion is watched, marginally excised, or treated as a sarcoma. The benign neurofibroma itself is covered in the dedicated neurofibroma topic; here the focus is the precursor.

The consensus term. A 2017 international consensus replaced the vague "atypical neurofibroma" with atypical neurofibromatous neoplasm of uncertain biologic potential (ANNUBP), the histological and biological bridge between a benign neurofibroma and MPNST. It is a recognised premalignant lesion, especially within large internal or plexiform neurofibromas in NF1.

Diagnostic criteria. ANNUBP is diagnosed when at least two of four features are present:

• Cytologic atypia
• Loss of the normal neurofibroma architecture
• Hypercellularity
• Low-level mitotic activity, present but below the MPNST threshold (fewer than 3 mitoses per 10 high-power fields), with no necrosis

The distinction from low-grade MPNST is essentially quantitative: a higher mitotic count (3 or more per 10 HPF) and/or necrosis push a lesion into MPNST. This is a genuinely difficult call that requires expert sarcoma pathology.

Molecular correlate. Progression from neurofibroma to ANNUBP is marked by biallelic CDKN2A (p16) inactivation; loss of p16 by immunohistochemistry supports the diagnosis and dovetails with the CDKN2A deletions seen in established MPNST. H3K27me3 is usually retained at the ANNUBP stage, its global loss being a later event that comes with PRC2 inactivation as the lesion becomes MPNST, so an atypical lesion that has lost H3K27me3 should raise concern for transformation.

Clinical relevance and management. ANNUBP often corresponds to a focally concerning area within a plexiform neurofibroma, for example a region of intermediate PET avidity (SUVmax 2.5-3.5) or new growth, and may be what a targeted biopsy of that area yields. Unlike MPNST, ANNUBP is managed by complete (marginal) excision with clear margins alone: it does not require the wide 2 cm margins, nerve sacrifice, radiotherapy or chemotherapy used for MPNST, but it does mandate negative margins and ongoing surveillance because of its malignant potential.

MRI is the gold standard for local staging. The protocol is T1-weighted for anatomy, T2-weighted with fat suppression for tumour extent and oedema, and T1 post-gadolinium with fat suppression for the enhancement pattern, covering the entire compartment plus the joint above and below. It guides resection planning and biopsy location.

What MRI shows. On T1 the tumour is isointense to muscle with heterogeneous signal from necrosis and haemorrhage, irregular infiltrative margins unlike the well-defined benign tumours, and loss of the surrounding fat planes. On T2 the signal is heterogeneous, the uniform hyperintensity of a benign neurofibroma is lost, central necrosis gives fluid signal and aggressive tumours carry peritumoral oedema. After gadolinium the enhancement is heterogeneous, with non-enhancing necrotic areas, rim enhancement around them and enhancement of the involved nerve proximally and distally.

View larger

View larger

Benign neurofibroma or MPNST. The features that separate the two on MRI:

FDG PET-CT. This is the critical tool for diagnosing malignant transformation in NF1, where the question is which of many plexiform neurofibromas has changed. Its indications:

• Known plexiform neurofibroma with concerning clinical features
• Surveillance of high-risk plexiform neurofibromas
• Preoperative staging
• Detection of metastases

Reading the SUV. Ferner showed that FDG-PET and PET-CT separate benign from malignant well (sensitivity 0.89, specificity 0.95) but also that SUVmax did not predict tumour grade, and the commonly quoted cut-offs (about 2.5 and 3.5) come from pooled and subsequent analyses rather than that primary report. Treat them as convention: a high SUV raises concern and a low SUV lowers it, but neither is a decision rule on its own, and a clinically worrying lesion is biopsied whatever the number.

What PET adds and where it fails. It guides the biopsy to the highest-SUV area of a heterogeneous tumour, identifies occult metastases on whole-body staging, serves as a surveillance tool in high-risk NF1 patients and monitors treatment response. Inflammation and infection give false positives, low-grade MPNST may have a lower SUV and give false negatives, and a PPV of 77% means a 23% false-positive rate, so it is not a replacement for biopsy. Cost and radiation exposure limit its use for routine surveillance. An SUV over 3.5 should trigger urgent MDT discussion and biopsy.

Staging CT. The lung is the most common metastatic site (50% at presentation or during follow-up), the liver less common, bone rare, and retroperitoneal or mediastinal disease is looked for on the same study. The protocol is CT chest with intravenous contrast, and CT abdomen and pelvis for a trunk primary or concerning symptoms; a dedicated chest CT rather than a chest radiograph, because the lung metastasis risk is high. Staging is completed before biopsy so that treatment can be planned in full.

View larger

View larger

View larger

Biopsy. The principles are inviolable because the tract becomes part of the specimen.

What to ask of the pathologist. The report must give the FNCLCC grade, which drives prognosis and adjuvant decisions, together with S100, Ki67 and H3K27me3 status, and it must exclude the other spindle-cell sarcomas, synovial sarcoma and fibrosarcoma in particular.

AJCC staging. Most MPNST presents as stage IIIB (large, deep, high grade) or stage IV (metastatic).

History. The presentation is a rapidly enlarging mass over weeks to months, the key feature separating it from a benign lesion, with severe pain in 60-70% (neurogenic, in the distribution of the nerve) and a progressive neurological deficit. Symptoms have usually been present for less than a year. In NF1 the story is a neurofibroma that was stable for years and has suddenly grown.

Red flags in NF1. A neurofibroma that enlarges rapidly, changes from soft to firm, becomes newly and severely painful having been painless, or produces weakness or numbness is a red flag for transformation. Constitutional symptoms (weight loss, fatigue, in 10-20%) are rare in localised disease, suggest metastatic disease when present, and warrant full staging.

Examination. The mass is large, typically over 5 cm, firm and fixed to underlying structures; the overlying skin is normal or, in advanced disease, tethered or ulcerated, and large tumours may show venous engorgement. On palpation the firm to hard consistency against the soft neurofibroma is the critical distinguishing feature; the mass is non-mobile, fixed to deep tissues, poorly defined at its margins, tender in 60-70%, and without pulsation, which separates it from a vascular lesion.

Neurological assessment. A motor deficit in the distribution of the involved nerve is present in 30-40% and a sensory deficit in 40-50%, with muscle atrophy if chronic and reduced or absent deep tendon reflexes. Document the preoperative neurological status, because it is the baseline for informed consent. Brachial or lumbosacral plexus involvement gives weakness and sensory loss across multiple nerve distributions. Lymphadenopathy is rare, since MPNST rarely spreads to lymph nodes (under 5%).

Red flags requiring urgent investigation. Any of these calls for imaging now:

• Size greater than 5 cm
• Deep (subfascial) location
• Rapid growth over weeks to months
• Fixation to underlying structures
• Neurological deficit

Radiotherapy. It is indicated for high-grade MPNST (most cases, 90%), positive or close margins (less than 1 mm), large tumours (over 5 cm) and deep (subfascial) location. It improves local control from 60-70% to 80-85% but does not improve overall survival, and it is recommended for most MPNST given the high local recurrence risk.

Before or after surgery. Preoperative radiotherapy is 50 Gy in 25 fractions: a smaller treatment volume and potentially better local control, against wound complications in 30-40% and a delay to surgery of 6-8 weeks after radiotherapy. Postoperative radiotherapy is 60-66 Gy in 30-33 fractions: immediate surgery and fewer wound complications (15-20%), against a larger treatment volume that includes the entire surgical bed. If a positive margin cannot be re-resected, a boost of a further 10-16 Gy is given to the positive margin area.

Chemotherapy. The settings and what they achieve:

What the response rate means. Response rates of 20-30% were long read as evidence that MPNST is intrinsically chemoresistant. The EORTC pooled analysis (Kroep) refutes that: response was 21% in MPNST against 22% in other soft-tissue sarcomas, so the benefit is modest in the way it is modest for every soft-tissue sarcoma. Adjuvant chemotherapy remains unproven, and Kroep's data are from advanced disease only.

Novel therapies. MEK inhibitors (selumetinib, trametinib) target the Ras-MAPK pathway but have modest activity in established MPNST (response under 10%); EZH2 inhibitors (tazemetostat) target the PRC2 pathway with limited single-agent activity; combination strategies are under investigation. Checkpoint inhibitors (anti-PD1, anti-CTLA4) have limited data and low response rates (under 10%), and tumour mutational burden is typically low. No targeted therapy or immunotherapy is standard of care, and clinical trial enrolment is encouraged.

MEK inhibition treats the precursor, not the sarcoma. The real, practice-changing role of MEK inhibition is in the benign plexiform neurofibroma, and the distinction is a high-yield one. Selumetinib, an oral MEK1/2 inhibitor that blocks the Ras-MAPK pathway driven by NF1 loss, is approved for paediatric NF1 patients with symptomatic, inoperable plexiform neurofibromas. In the pivotal SPRINT phase II trial, 35 of 50 children (70%) had a confirmed partial response (a reduction of at least 20% in tumour volume), most responses were durable, and there were meaningful improvements in pain, function and quality of life (Gross et al., N Engl J Med 2020, PMID 32187457).

The logic is a two-compartment distinction. In the benign plexiform neurofibroma, MEK inhibition can shrink the tumour, relieve symptoms and reduce the at-risk substrate; in established MPNST the same drugs, and the other targeted and immune agents, have minimal activity, and complete surgical resection remains the only curative modality. The danger is mistaking a transforming or frankly malignant mass for a benign plexiform neurofibroma and managing it medically: a plexiform neurofibroma that grows, becomes painful, develops a focally intermediate-to-high SUV, or shows worrying MRI change must be biopsied to exclude ANNUBP or MPNST before it is attributed to benign disease and offered a MEK inhibitor. Detailed medical management of plexiform neurofibroma belongs to the neurofibromatosis topic; the point here is the boundary with sarcoma.

The MDT. The discussion that precedes planning must include:

• A sarcoma surgeon (orthopaedic oncology or surgical oncology)
• A medical oncologist
• A radiation oncologist
• A musculoskeletal radiologist
• A pathologist with sarcoma expertise
• An NF1 specialist, if applicable

Planning. The MRI is reviewed for tumour extent and neurovascular involvement, the relationship to major structures (vessels, bone, joints), and the feasibility of limb salvage against amputation. The functional cost of resecting the nerve is assessed and the reconstruction options discussed (nerve, soft tissue, bone if needed), with vascular surgery on standby if major vessel involvement is anticipated and an anaesthetic evaluation completed.

Consent. Realistic expectations are essential given the prognosis, and the discussion covers:

• The nature of MPNST and its poor prognosis (5-year survival 40-60%)
• Nerve sacrifice is mandatory, so a permanent neurological deficit is expected
• The functional consequence for the nerve concerned, for example foot drop with the sciatic nerve or hand weakness with the median or ulnar nerve
• The alternative of amputation, with its better local control and the same survival
• The need for adjuvant radiotherapy
• The risk of recurrence and of metastasis

Wide excision. The goal is an R0 resection, microscopically negative margins with a 2 cm margin in all directions. Under general anaesthesia, positioned for the tumour's location, with a tourniquet for an extremity (released before closure to assess perfusion) and vascular surgery available if a major vessel is involved, the operation runs in five steps:

1. Incision. A longitudinal incision incorporating the biopsy site, with a wide ellipse around the tract so it is excised en bloc, and exposure at least 5 cm beyond the tumour margins proximally and distally.
2. Dissection and margins. Develop the fascial planes 2 cm from the tumour edge, measured with a ruler, and resect the tumour en bloc with its cuff of normal tissue. The involved nerve is transected 2 cm proximal and distal to the tumour margins, involved muscle compartments are sacrificed, and major vessels are preserved if possible by dissecting them free with an adventitial margin.
3. Vessel and bone. An encased vessel is resected and reconstructed with a vein graft or synthetic conduit; a vessel that is adherent but not invaded is dissected free with its adventitia, with frozen section of the adventitia if uncertain. Bone that is abutted is resected extraperiosteally; bone that is invaded is resected en bloc.
4. Margin assessment. Orient the specimen with sutures or clips (superior, lateral, deep), send frozen section of any concerning margin, and re-resect if a margin is positive.
5. Reconstruction and closure. Primary closure if tension-free, otherwise local flaps or free tissue transfer for a large defect. Nerve reconstruction is rarely performed, because this is a cancer operation and a nerve graft would lie within the radiation field; vascular and bone reconstruction follow if those structures were resected. Haemostasis, drains and a layered closure.

Limb salvage is achieved in over 95% of extremity MPNST, but with a significant functional deficit.

View larger

View larger

View larger

Amputation. It is indicated when the neurovascular bundle is completely encased with no possibility of limb salvage, when extensive soft-tissue involvement prevents a functional reconstruction, for recurrent MPNST after failed limb salvage, and by patient preference after informed discussion. The level is proximal to the tumour with an adequate margin, for example an above-knee amputation for a distal femoral MPNST, and the proximal margin is confirmed negative by frozen section of bone and soft tissue.

Why it is offered. Local control after amputation is superior to limb salvage (95% versus 80%), while overall survival is no different from limb salvage that achieves negative margins. Prosthetic rehabilitation is required and patient acceptance is often poor, so extensive counselling is needed: quality of life with a prosthesis may exceed that with a severely impaired limb, the decision is the patient's, and amputation is presented as an alternative option, not a failure.

Treatment-related. The neurological deficit is expected with nerve resection, which is why it sits in the consent discussion; the rest are managed as they arise.

Local recurrence. It occurs in 30-50% at 5 years despite wide resection, the risk factors being positive margins, size over 10 cm, trunk location and high grade. The median time to recurrence is 12-24 months, with 80% within 2 years. Management is re-resection if feasible, radiotherapy if not previously given, amputation, or palliative care.

Distant metastasis. It occurs in 40-65% at 5 years, at a median of 18 months. The lung takes 80-90% of metastases; bone, liver and brain are rare sites. Management is palliative chemotherapy, metastasectomy in selected cases, or best supportive care.

Metastasectomy. In selected patients it improves survival, to 5-year survival of 30-40% against 10% with chemotherapy alone. The selection criteria:

• Limited lung metastases (1-3 nodules)
• Disease-free interval greater than 12 months
• Complete resection possible
• Good performance status

The first weeks. The inpatient phase is about the wound and the deficit; the outpatient phase is about the histology and what it changes.

Surveillance. The schedule is front-loaded because most recurrence is early, and early detection may allow metastasectomy in selected cases.

Survival. MPNST has the worst prognosis among soft-tissue sarcomas. Five-year survival by aetiology is given in the table in the Overview: 21% NF1-associated, 42% sporadic and 20-30% radiation-induced, with the caveat there about how secure the NF1 comparison is.

Margin status. It is the single most important modifiable prognostic factor: 5-year survival is 65% after R0 (negative margins), 40% after R1 (microscopic positive) and 30% after R2 (gross residual).

Favourable factors. The features that predict a better outcome:

• Complete surgical resection (R0 margins)
• Small size (less than 5 cm)
• Superficial location (above the fascia)
• Low grade (rare, under 10% of MPNST)
• Distal extremity location
• Sporadic (non-NF1) origin, with the caveat in the Overview

Unfavourable tumour factors. Large size (over 10 cm), deep (subfascial) location, high grade (grade 3, 70% of MPNST), trunk or proximal extremity location, necrosis over 50% and a mitotic count over 20 per 10 HPF.

Unfavourable patient factors. NF1 association (taught as adverse, but not established, as set out in the Overview), radiation-induced disease, the malignant Triton variant and age over 50 years.

Unfavourable treatment factors. Inadequate initial surgery (marginal excision), delay in diagnosis and treatment, inability to deliver adjuvant radiotherapy and poor response to chemotherapy.

1. Ducatman BS, Scheithauer BW, Piepgras DG, et al. Malignant peripheral nerve sheath tumors. A clinicopathologic study of 120 cases. Cancer. 1986;57(10):2006-2021. PMID:3082508. doi:10.1002/1097-0142(19860515)57:10%3C2006::aid-cncr2820571022%3E3.0.co;2-6

2. Evans DGR, Baser ME, McGaughran J, et al. Malignant peripheral nerve sheath tumours in neurofibromatosis 1. J Med Genet. 2002;39(5):311-314. doi:10.1136/jmg.39.5.311

3. Lee W, Teckie S, Wiesner T, et al. PRC2 is recurrently inactivated through EED or SUZ12 loss in malignant peripheral nerve sheath tumors. Nat Genet. 2014;46(11):1227-1232. doi:10.1038/ng.3095

4. Ferner RE, Golding JF, Smith M, et al. [18F]2-fluoro-2-deoxy-D-glucose positron emission tomography (FDG PET) as a diagnostic tool for neurofibromatosis 1 (NF1) associated malignant peripheral nerve sheath tumours (MPNSTs): a long-term clinical study. Ann Oncol. 2008;19(2):390-394. doi:10.1093/annonc/mdm450

5. Prieto-Granada CN, Wiesner T, Messina JL, et al. Loss of H3K27me3 expression is a highly sensitive marker for sporadic and radiation-induced MPNST. Am J Surg Pathol. 2016;40(4):479-489. doi:10.1097/PAS.0000000000000564

6. Stucky CC, Johnson KN, Gray RJ, et al. Malignant peripheral nerve sheath tumors (MPNST): the Mayo Clinic experience. Ann Surg Oncol. 2012;19(3):878-885. doi:10.1245/s10434-011-1978-7

7. Kroep JR, Ouali M, Gelderblom H, et al. First-line chemotherapy for malignant peripheral nerve sheath tumor (MPNST) versus other histological soft tissue sarcoma subtypes and as a prognostic factor for MPNST: an EORTC soft tissue and bone sarcoma group study. Ann Oncol. 2011;22(1):207-214. doi:10.1093/annonc/mdq338

8. Anghileri M, Miceli R, Fiore M, et al. Malignant peripheral nerve sheath tumors: prognostic factors and survival in a series of patients treated at a single institution. Cancer. 2006;107(5):1065-1074. doi:10.1002/cncr.22098

9. Wasa J, Nishida Y, Tsukushi S, et al. MRI features in the differentiation of malignant peripheral nerve sheath tumors and neurofibromas. AJR Am J Roentgenol. 2010;194(6):1568-1574.

10. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Soft Tissue Sarcoma Version 2.2024.