S100-Positive Schwann-Cell Tumour
- The GRANULAR CELL TUMOUR is an UNCOMMON, usually BENIGN soft-tissue tumour of SCHWANN-CELL (neural) origin - which is why it is S100-POSITIVE (and SOX10-positive) - composed of cells with characteristic abundant GRANULAR EOSINOPHILIC (lysosome-rich) cytoplasm.
- It typically presents as a SMALL, FIRM, often solitary DERMAL or SUBCUTANEOUS NODULE; the classic site is the TONGUE (and head/neck), but it also arises in the SKIN and SOFT TISSUE of the trunk and limbs, the breast, and (uncommonly) the gastrointestinal tract and other viscera. THE DEMOGRAPHICS ARE WORTH KNOWING: across 42 cases with a systematic review of 1,499, the mean age at diagnosis was 45.8 YEARS (range 6 to 69) and 60 PERCENT were FEMALE.
- PUT A NUMBER ON 'RARE MALIGNANT VARIANT' - AND NOTE THAT THE HISTOLOGICAL CRITERIA OVER-CALL IT. Applying the Fanburg-Smith criteria to that pooled literature classified 72 percent benign, 17 PERCENT ATYPICAL and 11 PERCENT MALIGNANT. But TRUE malignancy, defined by actual METASTASIS, occurred in only about 2.5 PERCENT - so the histological grading labels roughly FOUR TIMES as many tumours malignant as ever behave that way. Of the six criteria, NECROSIS and MITOTIC ACTIVITY were the most effective at identifying genuinely aggressive tumours.
- HISTOLOGY shows nests or sheets of polygonal cells with abundant granular eosinophilic cytoplasm; immunohistochemistry confirms the neural (Schwann-cell) origin with S100/SOX10 positivity, and the overlying epithelium may show PSEUDOEPITHELIOMATOUS HYPERPLASIA, which can be mistaken for squamous cell carcinoma on a superficial biopsy.
- The vast majority are BENIGN, but a RARE MALIGNANT variant exists and is the dangerous one: features that raise concern include LARGE size (often over about 4-5 cm), RAPID growth, necrosis, increased mitotic activity/pleomorphism and infiltrative margins - malignant granular cell tumours carry a HIGH risk of local recurrence and METASTASIS and significant mortality.
- The DIFFERENTIAL includes other granular/eosinophilic lesions and, because of the pseudoepitheliomatous hyperplasia, squamous cell carcinoma on a small/superficial biopsy - so adequate sampling and S100/SOX10 staining are important to confirm the diagnosis and avoid over-diagnosis of carcinoma.
- A POSITIVE MARGIN IS NOT AUTOMATICALLY A RE-EXCISION. In the pooled follow-up, about 20 PERCENT of excisions had positive margins, and of those only about 20 PERCENT recurred locally - and in the authors' own 42 cases, 32 were INCOMPLETELY excised yet none recurred or progressed, with the single exception of the one malignant tumour. So an incompletely excised benign granular cell tumour can reasonably be observed rather than re-operated, and the decision should turn on the histological grade rather than the margin alone.
- MANAGEMENT is COMPLETE surgical EXCISION with CLEAR MARGINS, which is usually curative for benign lesions; SMALL lesions (under about 1 cm) may be monitored in selected cases, but LARGER lesions and any with worrying features warrant WIDE excision, and a MALIGNANT granular cell tumour requires wide resection and surveillance (with consideration of adjuvant therapy) given its high recurrence/mortality.
- “Granular cell tumour = uncommon, usually BENIGN soft-tissue tumour of SCHWANN-CELL (neural) origin → S100/SOX10 POSITIVE; cells with abundant GRANULAR eosinophilic cytoplasm. Classic site = TONGUE (also skin/soft tissue/limbs, rarely GI).
- “Overlying pseudoepitheliomatous hyperplasia can MIMIC squamous cell carcinoma on a superficial biopsy - confirm with S100/SOX10 and adequate sampling.
- “RARE MALIGNANT variant = the dangerous one: large (often over ~4-5 cm), rapid growth, necrosis/mitoses → high recurrence/metastasis/mortality. Treatment = complete excision with clear margins (wide resection + surveillance for large/malignant).
- “Quote the proportions: mean age 45.8, 60% female. Fanburg-Smith grades 72% benign / 17% atypical / 11% malignant, but only ~2.5% METASTASISE - a fourfold over-call, and necrosis + mitoses are the most useful features. Nasser criteria are more conservative (93% benign). ~20% of excisions have positive margins and only ~20% of those recur, so a positive margin alone is not an indication to re-operate.
What It Is and Where It Arises
The tumour. The granular cell tumour is an uncommon, usually benign soft-tissue tumour of Schwann-cell (neural) origin, which is why it is S100- and SOX10-positive. Its cells have abundant granular, eosinophilic, lysosome-rich cytoplasm.
Who. Across 42 institutional cases and a systematic review of 1,499 more, the mean age at diagnosis was 45.8 years (range 6 to 69) and 60 percent were female.
Where. It typically presents as a small, firm, often solitary dermal or subcutaneous nodule. The classic site is the tongue (and the head and neck), but it also arises in the skin and soft tissue of the trunk and limbs, in the breast, and uncommonly in the gastrointestinal tract and other viscera.
Histology and the Immunoprofile
The pattern. Nests or sheets of polygonal cells with abundant granular eosinophilic cytoplasm. In the benign tumour the cells are bland.
The confirming immunoprofile. Conventional granular cell tumour is S100-, SOX10- and CD68-positive (the last reflecting its lysosomes), inhibin- and calretinin-positive, with PAS-positive, diastase-resistant granules. It is negative for keratins, desmin, SMA and HMB45/Melan-A, and that panel is what separates it from the epithelial and muscle mimics.
The skin can fool the pathologist. The epithelium overlying a cutaneous or mucosal granular cell tumour frequently shows pseudoepitheliomatous hyperplasia, which can be mistaken for squamous cell carcinoma on a small or superficial biopsy and lead to over-treatment. Adequate sampling and S100/SOX10 confirmation of the underlying granular cell tumour prevent the error.
Not every granular lesion is neural. A distinct primitive (non-neural) granular cell tumour is S100-negative, often ALK-rearranged/ALK-positive, and not of Schwann-cell origin. Do not assume that every granular-cell lesion is the S100-positive neural tumour.
The Granular and Eosinophilic Differential
The mimics are the other granular and eosinophilic lesions.
- Alveolar soft part sarcoma - crystalline, PAS-positive (diastase-resistant) cytoplasmic crystals and TFE3-positive. Granular cell tumour can also be TFE3-positive, so rely on morphology and the wider panel.
- Rhabdomyoma and rhabdomyosarcoma - desmin- and myogenin-positive, S100-negative.
- Granular-cell variants of leiomyoma and leiomyosarcoma - SMA- and desmin-positive.
- Melanoma - S100- and SOX10-positive too, so distinguish it with HMB45/Melan-A and the clinical picture. In the hand and foot it presents as subungual melanoma.
- Histiocytic lesions (e.g. reticulohistiocytoma) - CD68-positive but S100-negative.
The neural tumours it sits alongside are schwannoma and neurofibroma.
The Malignant Variant and Its Grading
The dangerous one. The rare malignant granular cell tumour carries a high risk of local recurrence, metastasis and significant mortality, and it is the one to identify. Concern is raised by large size (often over about 4-5 cm), rapid growth, necrosis, increased mitotic activity and pleomorphism, and infiltrative margins. Size and growth rate are the practical red flags that change the plan from a simple local excision to a wide resection with surveillance.
- Benign (the majority)
- Usually small (often under 2 cm)
- Malignant (rare)
- Often large (over ~4-5 cm)
- Benign (the majority)
- Slow
- Malignant (rare)
- Rapid
- Benign (the majority)
- Bland granular cells
- Malignant (rare)
- Necrosis, increased mitoses, pleomorphism, infiltration
- Benign (the majority)
- Cured by complete excision
- Malignant (rare)
- High recurrence/metastasis; significant mortality
- Benign (the majority)
- Complete excision (clear margins)
- Malignant (rare)
- Wide resection + surveillance (+/- adjuvant)
The Fanburg-Smith criteria. Malignancy is graded histologically on six features:
- Necrosis
- Spindling of the tumour cells
- Vesicular nuclei with large nucleoli
- Increased mitotic activity, more than 2 mitoses per 10 high-power fields
- A high nuclear-to-cytoplasmic ratio
- Pleomorphism
Meeting none of the six is benign, one or two is atypical, and three or more is histologically malignant.
The grading over-calls malignancy. Applied to a pooled literature of about 1,500 cases, the Fanburg-Smith criteria classified 72 percent benign, 17 percent atypical and 11 percent malignant. True malignancy, defined by metastasis, occurred in only about 2.5 percent, so histology labels roughly four times as many tumours malignant as ever behave that way. The alternative Nasser criteria are more conservative, calling 93 percent benign and 7 percent of uncertain malignant potential.
What to trust. Of the six features, necrosis and mitotic activity were the most effective at identifying genuinely aggressive tumours. The definitive proof of malignancy remains metastasis, to regional nodes, lung or bone.
Management
Diagnosis first. Management starts from an adequate biopsy with S100/SOX10 staining, which confirms the neural origin and guards against the carcinoma mimic described above.
Benign lesions. Complete surgical excision with clear margins is usually curative. Small lesions, under about 1 cm, may be monitored in selected cases.
Large, worrying or malignant lesions. Larger lesions and any with worrying features warrant wide excision. A malignant granular cell tumour requires wide resection with clear margins and surveillance, with consideration of adjuvant therapy, given its high recurrence, metastasis and mortality. These patients are managed at a specialist sarcoma centre, and any large or deep lesion should travel the pathway in soft-tissue masses and sarcoma referral.
The positive margin. A positive margin on a benign lesion is not an automatic re-excision. In the pooled follow-up about 20 percent of excisions had positive margins, and of those only about 20 percent recurred locally. In the authors' own 42 cases, 32 were incompletely excised yet none recurred or progressed, with the single exception of the one malignant tumour.
So let the grade decide. An incompletely excised benign granular cell tumour can reasonably be observed rather than re-operated, and the decision should turn on the histological grade rather than the margin alone.
Mnemonics & Memory Aids
GRANULAR
Hook:GRANULAR: Granular cytoplasm, S100 (neuRal), Anywhere (tongue), Nodule, Usually benign (PEH mimics SCC), Large=maLignant, Adequate margins (but grade decides re-excision), Really malignant = metastasis (~2.5%).
NECROSThe Six Fanburg-Smith Criteria - NECROS
Hook:NECROS grades the granular cell tumour: none benign, one or two atypical, three or more histologically malignant - but metastasis is the only proof.
MARGINThe Positive Margin - MARGIN
Hook:MARGIN: the margin matters less than the grade - 20% positive, 20% of those recur, and grade (not the margin) drives the re-excision.
Clinical Decision Scenarios
Practise clinical reasoning and management decisions out loud
“A firm subcutaneous nodule is reported as a granular cell tumour. What is its origin, and how do you decide management?”
What it is
- Uncommon, usually benign soft-tissue tumour of Schwann-cell (neural) origin
- S100/SOX10-positive; granular eosinophilic cytoplasm
- Classic site: tongue (also skin/soft tissue/limbs; rarely GI tract)
Diagnostic pitfall
- Overlying pseudoepitheliomatous hyperplasia mimics squamous cell carcinoma
- Confirm with S100/SOX10 and adequate sampling
- Avoid over-treatment as carcinoma
Malignant variant (rare)
- Large (often over ~4-5 cm), rapid growth
- Necrosis and increased mitoses are the most useful features; also pleomorphism, infiltration
- Fanburg-Smith calls 11 percent malignant but only ~2.5 percent metastasise - a fourfold over-call
- Metastasis is the definitive criterion; malignant tumours carry high recurrence and mortality
Management
- Complete surgical excision with clear margins (curative for benign)
- Very small lesions may be monitored; a positive margin on a benign lesion is not an automatic re-excision (~20 percent of those recur)
- Large/malignant: wide resection + surveillance; specialist centre
Evidence & Key Studies
Malignant granular cell tumour - a rare case and review of the literature
- Granular cell tumours are rare, mostly benign soft-tissue neoplasms; only a small proportion (about 8%) occur in the gastrointestinal tract, and the diagnosis is confirmed by S100-positive staining (reflecting the Schwann-cell/neural origin).
- Although rare in some sites, malignant granular cell tumours carry a high risk of mortality and recurrence.
- Small lesions (under about 1 cm) can be monitored, but larger or malignant ones require surgical resection - the case underscores early detection and intervention to prevent severe outcomes.
Granular cell tumor: a study of 42 cases and systematic review of the literature
- 42 institutional cases plus 1,499 cases reviewed to PRISMA guidelines. Mean age at diagnosis 45.8 years (range 6 to 69), 60 percent female; sites in descending frequency were skin and subcutaneous tissue, bronchus, oesophagus, breast, tongue, larynx, pharynx, gingiva, trachea, colon, vulva and hypopharynx.
- By Fanburg-Smith criteria the pooled cases were 72 percent benign, 17 percent atypical and 11 percent malignant; by the more conservative Nasser criteria, 93 percent benign and 7 percent of uncertain malignant potential. True malignancy defined by metastasis occurred in only about 2.5 percent, so the histological criteria over-call malignancy roughly fourfold. Necrosis and mitotic activity were the most effective individual features.
- About 20 percent of excisions had positive margins and about 20 percent of those recurred locally. In the authors' own series, 32 of 42 were incompletely excised yet none recurred or progressed apart from the single clinically malignant tumour, whose patient died 2 years after diagnosis. The follow-up is retrospective and of unstated duration for many pooled cases.
Malignant granular cell tumor of soft tissue: diagnostic criteria and clinicopathologic correlation (the Fanburg-Smith criteria)
- 73 malignant, atypical and multicentric granular cell tumours assessed on six histological criteria: necrosis, spindling, vesicular nuclei with large nucleoli, increased mitotic activity (over 2 mitoses per 10 HPF at 200x), high nuclear-to-cytoplasmic ratio, and pleomorphism - three or more = histologically malignant, one or two = atypical, focal pleomorphism alone = benign.
- 46 cases were histologically malignant, 21 atypical (3 multicentric), 6 benign (all multicentric); no benign or atypical case metastasised or caused a tumour death.
- Among the malignant cases with follow-up: 11 of 28 (39%) died of disease at a median of 3 years, local recurrence in 9 of 28 (32%) and metastases in 14 of 28 (50%), each at a median of 2 years.
- Immunohistochemistry: 100% vimentin, 98% S-100 protein - establishing the neural immunoprofile.
Malignant granular cell tumor: a look into the diagnostic criteria (the Nasser refinement)
- 48 cases classified first by Fanburg-Smith (37 benign, 7 atypical, 4 malignant), then reclassified using ONLY necrosis and/or mitoses into 44 benign and 4 of uncertain malignant potential (GCT-UMP).
- Under Fanburg-Smith the malignant cases were significantly younger, predominantly male, and significantly larger, with higher Ki-67 expression.
- Supports a more conservative two-tier scheme that reserves concern for necrosis and mitotic activity rather than the full six-feature panel.
The nature of granular cell tumours (rare, mostly benign soft-tissue neoplasms confirmed by S100 positivity, with only a small proportion in the gastrointestinal tract), the high recurrence/mortality risk of the malignant variant, and the management principle (monitor small lesions, resect larger/malignant ones) come from the cited Usman report. The Schwann-cell (neural) origin with S100/SOX10 positivity, the granular eosinophilic cytology, the classic tongue/skin/soft-tissue sites, the pseudoepitheliomatous-hyperplasia mimic of squamous cell carcinoma, and the malignancy criteria (large size, rapid growth, necrosis, mitoses) with complete/wide excision are standard, well-established teaching. The six criteria themselves and the malignant-variant outcome data (39% dying of disease, 50% metastasising in the follow-up subset) come from the cited Fanburg-Smith study, and the necrosis-and-mitoses refinement from the cited Nasser study. The demographics, the Fanburg-Smith and Nasser classification proportions, the 2.5 percent metastasis figure, the margin and recurrence data and the finding that necrosis and mitoses are the most useful individual criteria come from the Mobarki study, whose follow-up is retrospective and of unstated duration for many of the pooled cases. No randomised trial addresses margin width, re-excision for a positive margin, surveillance interval or adjuvant therapy in this tumour, and no cohort establishes a size threshold at which excision becomes mandatory - so none is quoted here as though settled.