Dermatofibrosarcoma protuberans

Meet the patient

A 38-year-old man has had a firm, slowly enlarging plaque on his shoulder for six years that his GP called a keloid. Over the last year it has thickened into a multinodular blue-brown mass with a depressed centre, and a new satellite nodule has appeared at the edge. He has never had it biopsied.^[1][3]

Two questions decide this case and frame every DFSP stem: is this the rare but recognisable truncal sarcoma, or one of its benign mimics? — answered by CD34 and the fusion — and can it be completely excised the first time? — answered by Mohs, because the subclinical honeycomb extension defeats every standard wide excision.^[1][6]

One fusion, one stain, one surgery — the three things that earn marks

The molecular engine is the t(17;22) COL1A1-PDGFB fusion. The COL1A1 promoter on chromosome 17 — a highly expressed collagen gene — is juxtaposed to PDGFB on chromosome 22, driving constitutive overexpression of PDGFB, which activates PDGFR-beta on the same cell in an autocrine loop. Downstream RAS-MAPK and PI3K-AKT signalling drives the slow, locally infiltrative, scar-like tumour. That fusion — and its ring-chromosome form r(17;22) — is the target of imatinib, and it is found in giant cell fibroblastoma of childhood, which is now regarded as the paediatric form of the same entity.^[1][6]

The immunohistochemical profile closes the diagnosis. DFSP is CD34 strongly and diffusely positive, Factor XIIIa negative, S100 negative, SMA negative, desmin negative, STAT6 negative, SOX10 negative. That single inversion — DFSP is CD34 positive where dermatofibroma is CD34 negative — is the working diagnostic matrix, and it is what excludes melanoma (S100 positive), leiomyosarcoma (desmin positive), and solitary fibrous tumour (CD34 positive but STAT6 positive with a NAB2-STAT6 fusion).^[1][3]

Mohs micrographic surgery is the standard of care because the tumour extends invisibly. Spindle cells crawl along connective-tissue septa into subcutaneous fat in a honeycomb (lace-like) pattern that runs 1 to 3 cm beyond the visible margin in 80 to 90 percent of cases. Bread-loaf sectioning of a standard wide excision examines under 1 percent of the margin; the rest is unchecked. Mohs examines 100 percent of the margin — with CD34 immunostaining to highlight tumour cells — which is why it drops recurrence from 10 to 20 percent to 0 to 2 percent.^[1][5]

Etymology for viva gold: protuberans is Latin for 'to push forward' — the name describes the natural history, the slow evolution from a flat plaque to a multinodular protuberant mass over years. The older name 'progressive and recurrent dermatofibroma' was retired because it understated the malignant biology, but it captures the lesson examiners test: a 'dermatofibroma' that recurs on the trunk is DFSP.^[3]

The classic lesion, read at the bedside

The presentation is, for years, insidious — a small, firm, skin-coloured, pink, red, or blue-brown plaque or nodule that the patient and their family physician dismiss as a cyst, a lipoma, a keloid, a scar, or a dermatofibroma. Over months to years it enlarges slowly, thickens, and develops multinodular or protuberant surface change; satellite nodules may appear at the periphery; late lesions ulcerate. The lesion is firm, indurated, tethered to the overlying skin (the skin does not glide over it as it would over a lipoma) but slightly mobile over the deep fascia, and the centre is often atrophic or depressed — a key clue.^[1][3]

^[1]

The site distribution is itself a discriminator from dermatofibroma: roughly 50 percent on the trunk (chest, abdomen, back, flank, shoulder girdle, groin), 20 to 35 percent on the proximal extremities (shoulder, upper arm, thigh, buttock), and 10 to 15 percent on the head and neck — where recurrence is higher because of the anatomic constraints on margin. Acral and genital sites are uncommon but recognised.^[3][5][6]

Risk factors are largely host-related, not environmental. Prior trauma at the site — a surgical scar, burn, vaccination site, insect bite, blunt trauma — is reported in 10 to 20 percent, biologically plausible given the scar-like behaviour of the tumour. Skin of colour is over-represented in several large series, the Bednar (pigmented) variant is more common in patients of African descent, and the t(17;22) fusion is acquired (somatic), not germline, so family history is not a risk factor.^[1][2][6]

The subtypes — the fibrosarcomatous variant is the one that changes everything

The fibrosarcomatous variant (FS-DFSP) is the one that upgrades the disease. In roughly 10 to 15 percent of cases, areas within a DFSP show increased cellularity, a herringbone pattern, more than 5 mitoses per 10 high-power fields, and loss of CD34 staining in the high-grade areas. The fusion persists — imatinib is still worth trying — but the metastatic risk climbs from under 5 percent to 15 to 30 percent. FS-DFSP mandates wider margins, adjuvant radiotherapy, CT chest staging, and sarcoma-MDT discussion.^[1][3][6]

The other variants are viva currency that do not change prognosis but do change the differential. Bednar tumour (pigmented DFSP) — melanin-laden dendritic cells scattered among the spindle cells, S100 positive in the pigmented cells not the tumour cells, more common in patients of African descent, identical prognosis to classic DFSP.^[2] Myxoid DFSP — prominent myxoid stroma that mimics myxoid liposarcoma and myxofibrosarcoma until CD34 and the fusion rescue the diagnosis. Atrophic DFSP — a depressed sclerotic plaque rather than a protuberant mass, easily mistaken for morphea or a burn scar, where diagnostic delay is the rule.^[1]

Giant cell fibroblastoma (GCF) of childhood is the paediatric form of the same disease — it shares the COL1A1-PDGFB fusion, the CD34-positive immunoprofile, and the locally infiltrative behaviour. The WHO 4th edition lists GCF and DFSP as a combined entity. A dermal or subcutaneous nodule or plaque in a child should raise GCF as well as DFSP.^[6]

Face-off — DFSP versus dermatofibroma, and the rest of the spindle-cell list

The central axis is DFSP versus dermatofibroma, settled by CD34 and Factor XIIIa. Beyond that, the lesion must be distinguished from the other dermal and subcutaneous spindle-cell tumours, from scar and keloid, and from the mimics of protuberant nodular disease.^[1][3][4]

^[1]

The discriminator line: CD34 positive on the trunk is DFSP; CD34 negative on the lower leg with a dimple sign is a dermatofibroma; CD34 positive plus STAT6 positive is a solitary fibrous tumour; S100 positive is melanoma. Order the panel, do not guess.^[1]

Investigations — diagnosis is histological, and the biopsy technique matters

The clinical picture raises suspicion, the biopsy confirms it, and the IHC plus molecular work-up closes it. Take a 4 mm punch biopsy of the most indurated, most infiltrative area; an incisional biopsy is acceptable for very large lesions. A shave biopsy is inadequate — it does not sample the deep margin and may miss the fibrosarcomatous component — and it is the commonest reason a DFSP is initially under-diagnosed.^[1][3][4]

^[1]

The histology reads in a specific order: dermal spindle-cell tumour in a monotonous storiform (cartwheel) pattern, infiltrating subcutaneous fat in a honeycomb (lace-like) pattern, with uniform cells, minimal atypia and fewer than 5 mitoses per 10 high-power fields in classic DFSP. The epidermis is usually spared (a grenz zone is common). FISH for the COL1A1-PDGFB fusion (break-apart probes on 17q22 and 22q13) is the standard molecular confirmation; RT-PCR and next-generation sequencing panels identify the rare variant fusions (COL1A2-PDGFB, EMILIN2-PDGFB, and others) that still predict imatinib sensitivity when PDGFR is the 3-prime partner.^[1][8]

^[1]

The classic trap: CD34 loss in the fibrosarcomatous component does not mean the fusion is gone — the fusion persists, and imatinib remains effective in many FS-DFSPs. Do not dismiss imatinib in FS-DFSP on the basis of CD34 loss alone.^[1]

Management — Mohs first, imatinib for the rest

^[1]

The 2024 European interdisciplinary guideline (Saiag, EJC 2025), the NCCN guideline, and the EADO/EORTC consensus agree on the same stepwise framework.^[1][3][6][7]

Mohs micrographic surgery is the standard of care for localised disease. Complete 100 percent margin examination — increasingly with CD34 immunostaining to highlight tumour cells at the margin — detects the subclinical infiltration that defeats standard wide local excision. Reported 5-year local control is 95 to 100 percent, with recurrence of 0 to 2 percent in pooled series.^[1][3][5]

Wide local excision with 2 to 3 cm peripheral margins to the deep fascia (or the next anatomical barrier — periosteum, perichondrium) is the alternative when Mohs is not available. The traditional margin is 2 to 3 cm, not the 4 to 5 cm cited in older texts. Recurrence after WLE is 10 to 20 percent at 5 years, reflecting the subclinical infiltration missed by bread-loaf sectioning; CD34 immunostaining of the WLE margin reduces false-negative margins.^[1][3][4]

Imatinib is the targeted therapy for unresectable, recurrent, or metastatic disease — and increasingly for neoadjuvant use in large or anatomically constrained lesions. It is a small-molecule tyrosine kinase inhibitor of PDGFR-alpha, PDGFR-beta, BCR-ABL and KIT; in DFSP the relevant target is the PDGFR-beta encoded by the PDGFB portion of the fusion.^[1][4]

^{[1] [4] [7]}

Imatinib monitoring: FBC and LFTs every 2 weeks for the first 2 months, then monthly; weight, oedema and symptom check at every visit. Toxicity is oedema (often periorbital), nausea, diarrhoea, myelosuppression, hepatotoxicity, fatigue, muscle cramps; less commonly cardiac failure, hypothyroidism, and hepatitis B reactivation. Continue imatinib until progression or unacceptable toxicity; consider maintenance in responders, with periodic re-imaging every 3 months.^[1][7]

Fibrosarcomatous transformation changes the plan: wider surgical margins, adjuvant radiotherapy (50 to 60 Gy) for positive margins or unresectable disease, consideration of imatinib (response less predictable than in classic DFSP), and staging with CT chest and MRI — mandatory. Metastatic disease is treated with imatinib 400 mg daily (escalable to 800 mg split dose for progression); second-line options are limited (sunitinib, sorafenib, pazopanib), and clinical-trial enrolment is preferred.^[1]

Follow-up — the late recurrence that defines surveillance

Clinical review every 6 months for 5 years, then annually for a further 5 years — 10 years of follow-up in total. Local recurrence can be very late, and the 5 to 10 year window is when the late recurrences present. MRI or ultrasound of the primary site at each review, with comparison to baseline post-operative imaging; CT chest annually in fibrosarcomatous, large, or recurrent disease. The patient returns promptly for any new nodule at the surgical site, any change in the scar, or any new truncal plaque.^[1][3]

How patients with DFSP come to harm — the preventable list

• A DFSP misdiagnosed as a dermatofibroma and under-excised, recurring locally for years before the sarcoma is recognised — the classic preventable error.^[1]
• A shave biopsy that misses the deep infiltration and the fibrosarcomatous component, sending the patient down the wrong pathway.^[1]
• A standard WLE with a 1 cm margin that leaves subclinical honeycomb infiltration behind — the commonest surgical error.^[1]
• Failure to stage the lungs in fibrosarcomatous, recurrent, or large primary tumours.^[1]
• CD34 loss in the FS component read as 'no longer DFSP', when the fusion persists and imatinib is still worth trying.^[1]
• Diagnostic delay in skin of colour, where plaque-stage and Bednar variants are more common and the diagnosis is missed for years.^[2]

Special populations

Skin of colour — DFSP more often presents as plaque-stage (less protuberant), the Bednar (pigmented) variant is more common, and diagnostic delay is the rule, not the exception. Maintain a low threshold to biopsy any chronic indurated truncal plaque.^[2]

Pregnancy — DFSP may enlarge rapidly due to hormonal and stromal influences; imatinib is teratogenic and generally avoided. Mohs surgery is feasible in the second trimester with appropriate positioning; defer definitive surgery to the second trimester or postpartum for non-urgent cases.^[1]

Paediatric and congenital DFSP — typically the plaque-stage or giant cell fibroblastoma (GCF) phenotype, sharing the COL1A1-PDGFB fusion and the CD34-positive immunoprofile. Mohs or WLE with 1 to 2 cm margins with paediatric surgery input; imatinib is generally avoided in young children because of growth-plate effects.^[6]

Anticoagulated patients — imatinib has antiplatelet effects via PDGFR inhibition on megakaryocytes; bleeding risk is modest but real. Do not stop anticoagulation for a skin biopsy of a suspicious lesion.^[1]

Prognosis and regional differences

Disease-specific survival for localised DFSP is over 99 percent at 5 years and remains over 95 percent at 10 years. For metastatic DFSP, 5-year survival is 10 to 30 percent in older series, improving with imatinib. The prognostic factors, in order of importance: fibrosarcomatous transformation, positive margins, size over 5 cm, head and neck location, deep invasion to fascia or muscle, and the fusion variant (classic COL1A1-PDGFB predicts imatinib response).^[1][3][6]

^[1]

Landmark evidence. The original description of the t(17;22) translocation and the COL1A1-PDGFB fusion (Simon 1997; O'Brien 1998) established the molecular basis and the rationale for PDGFR-targeted therapy. The phase II imatinib trials in DFSP (Rutkowski 2010; Stacchiotti 2016) defined the 400 mg daily dose, the 50 percent objective response rate, and the 2 to 3 month median time to response. The 2024 European interdisciplinary guideline (Saiag, EJC 2025) is the most contemporary multidisciplinary consensus. The 2025 review of rare fusion variants (Beyond COL1A1::PDGFB) summarises the imatinib-relevant molecular biology of the rare fusions.^[1][7][8]

The mantra, and the memory devices

^{[1] [1]}

The mantra: trunk plaque, CD34 positive, honeycomb fat, COL1A1-PDGFB — Mohs first, imatinib for the rest.^[1]

Ward-round test — three stems, thirty seconds each

Stem 1 — the 'keloid' that came back (answer)ShowHide

A 40-year-old woman had a 'keloid' excised from her chest wall three years ago. A new firm nodule has appeared in the scar, with two satellite papules at the periphery. What is the diagnosis, and what is the first step? Model: A recurrent nodule at the site of a previous 'keloid', 'scar', or 'dermatofibroma' excision on the trunk is DFSP until proven otherwise — the classic stem. The first step is a punch or incisional biopsy (never a shave) with CD34, Factor XIIIa and S100 immunohistochemistry, and FISH for the COL1A1-PDGFB fusion if the diagnosis is equivocal. DFSP is CD34 strongly and diffusely positive, Factor XIIIa negative; the fusion is the molecular hallmark and the imatinib target. Definitive treatment is Mohs micrographic surgery (standard of care, around 1 percent recurrence) or wide local excision with 2 to 3 cm margins to the deep fascia, with sarcoma-MDT discussion.^[1]

Stem 2 — the changing histology (answer)ShowHide

A 45-year-old man has a biopsy of a truncal plaque reported as DFSP, but the report adds 'areas of fibrosarcomatous transformation with CD34 loss, over 5 mitoses per 10 HPF, herringbone pattern'. How does this change management? Model: Fibrosarcomatous transformation (FS-DFSP) upgrades the disease. The metastatic risk rises from under 5 percent in classic DFSP to 15 to 30 percent. Management escalates: wider surgical margins (2 to 3 cm or more), adjuvant radiotherapy (50 to 60 Gy) for positive margins or unresectable disease, mandatory CT chest and MRI staging, and sarcoma-MDT discussion. Imatinib is still worth trying despite the CD34 loss, because the COL1A1-PDGFB fusion persists in the fibrosarcomatous component — CD34 loss does not mean the fusion is gone.^[1]

Stem 3 — the unresectable truncal mass (answer)ShowHide

A 50-year-old man has a 12 cm DFSP of the shoulder infiltrating the deltoid, deemed unresectable without major functional loss. What is the systemic therapy, its target, its dose, and its monitoring? Model: Imatinib — a small-molecule tyrosine kinase inhibitor of PDGFR-beta — is first-line for unresectable, recurrent, or metastatic DFSP, targeting the constitutive PDGFR activation driven by the COL1A1-PDGFB fusion from t(17;22) or r(17;22). The dose is 400 mg orally once daily (escalable to 400 mg twice daily for resistant disease). Neoadjuvant use to shrink the tumour before a less morbid resection is well described. Monitoring: FBC and LFTs every 2 weeks for the first 2 months then monthly; watch for oedema (often periorbital), myelosuppression, hepatotoxicity, and hepatitis B reactivation (screen HBsAg and anti-HBc first). Objective response is about 50 percent and disease control about 80 percent, with a median time to response of 2 to 3 months.^[1][4]

References8ShowHide

1. [1]Saiag P, Lebbe C, Brochez L, et al. Diagnosis and treatment of dermatofibrosarcoma protuberans. European interdisciplinary guideline - update 2024 Eur J Cancer, 2025.PMID 39904126
2. [2]Alchorne MMA, Conceição KDC, Barraza LL, et al. Dermatology in black skin An Bras Dermatol, 2024.PMID 38310012
3. [3]Allen A, Ahn C, Sangüeza OP Dermatofibrosarcoma Protuberans Dermatol Clin, 2019.PMID 31466588
4. [4]Acosta AE, Vélez CS. Dermatofibrosarcoma Protuberans Curr Treat Options Oncol, 2017.PMID 28795284
5. [5]Tillman BN, Liu JC. Cutaneous Sarcomas Otolaryngol Clin North Am, 2021.PMID 33602520
6. [6]Jozwik M, Bednarczuk K, Osierda Z Dermatofibrosarcoma Protuberans: An Updated Review of the Literature Cancers (Basel), 2024.PMID 39335097
7. [7]Remiszewski P, et al. Dermatofibrosarcoma Protuberans (DFSP): Current Treatments and Clinical Trials Curr Treat Options Oncol, 2025.PMID 41042442
8. [8]Das S, et al. Beyond COL1A1::PDGFB: Rare fusions and their clinical implications in dermatofibrosarcoma protuberans World J Clin Cases, 2025.PMID 41356086