Desmoid Tumor (Aggressive Fibromatosis)

Desmoid tumour is a rare benign fibroblastic neoplasm: a monoclonal proliferation of myofibroblasts that grows by local infiltration and has no metastatic potential. Its difficulty lies elsewhere, in an unpredictable natural history, real local aggressiveness, and a high recurrence rate after resection.

Two names, one tumour. "Desmoid" comes from the Greek desmos, a band or tendon, for the collagenous, band-like consistency of the lesion. "Aggressive fibromatosis" emphasises the locally aggressive behaviour despite benign histology. The terms are synonymous, and both appear in examination questions.

The paradigm shift. Over the past decade the historical aggressive surgical approach has given way to surveillance first, because most desmoids need no treatment. In the prospective GRAFITI trial 28% regressed and 32% stabilised on active surveillance, so roughly 60% required no intervention; the earlier retrospective Fiore series reported a 5-year progression-free survival of about 50%. The two figures are not in tension: they measure different endpoints, and the prospective one is the more favourable. Treatment-related morbidity may exceed disease-related morbidity in many patients, and that recognition is what changed practice.

Who. The peak is at 20-40 years, median 30. Women predominate 2-3:1 overall, with a strong female bias in abdominal wall tumours, attributed to hormonal influence, and a more equal distribution in the extremities.

Where. Location shapes both presentation and management:

• Extra-abdominal, 60% - shoulder and thigh most common; head and neck account for 5-10% of extra-abdominal cases
• Abdominal wall, 25% - the rectus, with a postpartum association
• Intra-abdominal, 15% - the mesentery, where FAP-associated tumours arise

The pathway. Desmoid tumorigenesis is driven by aberrant activation of Wnt-beta-catenin signalling, reached by two distinct genetic routes. In the normal cell beta-catenin is captured by the destruction complex (APC-Axin-GSK3beta), phosphorylated, and so marked for ubiquitin-mediated degradation, which keeps cytoplasmic levels low. When beta-catenin cannot be phosphorylated, or the complex that would phosphorylate it has lost APC, the protein accumulates, translocates to the nucleus, binds the TCF/LEF transcription factors and activates target genes that promote fibroblast proliferation. That nuclear accumulation is what beta-catenin immunostaining detects.

View larger

Sporadic tumours: CTNNB1, 85-90%. Mutation of the CTNNB1 gene on chromosome 3p21, which encodes beta-catenin, is the hallmark of the sporadic desmoid. The mutations sit in exon 3, predominantly at codons 41 (T41A) or 45 (S45F), and create a beta-catenin that cannot be phosphorylated and therefore cannot be degraded.

Genotype and behaviour. Get the direction right; it is regularly examined.

• Three hotspots account for almost all sporadic mutations: T41A (the commonest), S45F and S45P
• S45F is the adverse genotype. It predicts progression on surveillance and higher recurrence after surgery; in the prospective GRAFITI cohort it carried a hazard ratio of 6.24 (95% CI 1.92-20.30) for needing to start active treatment, alongside tumour size of 5 cm or more (HR 2.38)
• T41A behaves comparatively well: the most frequent mutation, and associated with a more indolent course
• Wild-type CTNNB1 in a convincing desmoid should prompt a search for an APC germline mutation

Practically, an S45F tumour warrants closer surveillance intervals and a lower threshold to treat; it does not by itself mandate surgery.

FAP-associated tumours: APC, 5-10%. A germline mutation of APC on chromosome 5q21-22, inherited as an autosomal dominant trait, causes familial adenomatous polyposis and its desmoids. APC is the tumour-suppressor component of the destruction complex, so its loss prevents beta-catenin degradation and produces the same nuclear accumulation as a CTNNB1 mutation. The tumours that result behave differently from sporadic disease:

• Occur in 10-20% of FAP patients
• Typically intra-abdominal (mesenteric), and often multifocal
• Associated with prior abdominal surgery, classically colectomy
• Generally more aggressive than sporadic desmoids
• The leading cause of death in FAP after prophylactic colectomy

APC genotype. Mutations beyond codon 1400 carry a higher desmoid risk; codons 1310-1580 carry the greatest predisposition; 3' APC mutations reach a desmoid incidence of up to 25%. FAP patients need counselling about desmoid risk before any abdominal surgery.

View larger

Risk factors. Beyond FAP, the recurring themes are injury and hormones.

Macroscopic. A firm to rubbery, poorly circumscribed mass that resembles scar tissue or keloid, with a whorled, glistening white-grey cut surface reflecting its collagen. There is no true capsule: the borders are irregular and infiltrative, the tumour follows fascial planes and invades between muscle fibres. Size at diagnosis is typically 5-15 cm, and can be larger.

View larger

Low power. Long, sweeping, infiltrative fascicles of fibroblasts in collagen, with irregular margins that invade adjacent tissue and dissect between muscle fibres. Vascularity is variable and often low.

High power. Uniform spindle cells, myofibroblasts, with bland cytology and minimal nuclear atypia; the nuclei are vesicular with small nucleoli. Mitoses are rare to absent and necrosis is absent, which is the key to excluding sarcoma. Collagen is abundant and becomes keloid-like in older lesions.

View larger

Immunohistochemistry. The panel confirms the myofibroblastic line and excludes the mimics.

Reading the beta-catenin stain. Cytoplasmic beta-catenin is a normal, non-specific finding; it is nuclear accumulation that reflects Wnt pathway activation, and nuclear staining is the pattern to look for. It is only about 80% sensitive and it is not specific, so on its own it confirms nothing: it must sit beside the morphology and the clinical context. Negative staining should prompt reconsideration of the diagnosis, and FAP-associated tumours show the same nuclear pattern through APC loss.

View larger

Extra-abdominal, 60%. A painless, slowly enlarging mass over months to years, prior trauma to the area (sports injury, surgery) in 30%, found incidentally or noticed by the patient: firm, non-mobile, fixed to the underlying structures. Growth is slow but may run in periods of rapid enlargement followed by stabilisation, and prior trauma to the area (sports injury, surgery) is reported in 30%. Pain appears in 30%, when nerve compression develops, and a large tumour or one involving the shoulder or thigh limits function. The usual sites:

• Shoulder girdle - deltoid, pectoralis, scapular muscles
• Thigh - quadriceps, adductors
• Chest wall, back, gluteal region
• Head and neck, rarely

Abdominal wall, 25%. A firm mass in the anterior abdominal wall, usually in the rectus, and usually painless, so that cosmetic concern predominates. The patient is often a postpartum woman, and a history of caesarean section or prior abdominal surgery is present in 30%. The risk factors are the postpartum period (hormonal change plus rectus trauma), prior caesarean or abdominal surgery, multiple pregnancies and female sex.

Intra-abdominal, 15%. Abdominal pain, bloating and early satiety, with obstructive symptoms when the tumour is mesenteric and a palpable mass when it is large. The tumour may present through its complications: small bowel obstruction, ureteric obstruction with hydronephrosis, and rarely portal vein compression or gastrointestinal bleeding. A history of FAP is present in 50%.

Examination. The mass is firm to hard and non-tender unless a nerve is compressed, with poorly defined margins and an infiltrative feel, fixed to the underlying fascia or muscle rather than freely mobile. There are no overlying skin changes, unlike malignancy, and no lymphadenopathy, because the tumour is benign. In the abdominal wall, Carnett's sign (pain increasing with contraction of the abdominal muscles) separates an abdominal-wall mass from intra-abdominal pathology. Complete the assessment with the range of motion for an extremity lesion, a neurological examination where nerve compression is suspected, and a vascular examination where the tumour lies near major vessels.

Algorithm

View larger

Why surveillance comes first. The natural history set out in the overview is the argument: most tumours stay still or shrink, surgery carries functional loss and a high recurrence rate, and treatment complications may exceed the morbidity of the disease. A benign tumour that never metastasises gives you time to watch, and progression after a period of stability remains manageable.

Who is suitable. Surveillance is the default for the patient who is asymptomatic or minimally symptomatic, without functional impairment or an acute complication such as obstruction or compression, who prefers it after an informed discussion and understands what monitoring involves. On the tumour side it suits:

• Any size
• Stable disease on serial imaging
• A location where surgery would cause major morbidity
• An extra-abdominal site (better prognosis)
• A first presentation rather than a recurrence

The schedule. Intervals lengthen as the tumour declares itself, but surveillance never ends, because late behaviour is unpredictable.

Triggers for intervention. Treatment is considered for:

• Progressive growth on serial MRI, greater than 20% volume increase
• Increasing symptoms - pain, functional impairment
• A developing complication - nerve compression, vascular compromise
• Bowel or ureteric obstruction in intra-abdominal disease
• Patient anxiety despite stable disease, through shared decision-making

Counselling. Tell the patient what the evidence says: the tumour is benign and never spreads to other organs, the odds favour it staying the same or shrinking without treatment, monitoring is a safe and active approach, treatment can begin at any point if the tumour progresses, and surveillance is a long-term commitment. Surgery, when it is chosen, carries recurrence and functional loss, and that is part of the same conversation.

Algorithm

View larger

Who needs it. Systemic therapy is reserved for:

• Unresectable tumours with documented progression
• Recurrent tumours not amenable to further surgery
• Symptomatic intra-abdominal desmoids
• Multifocal FAP-associated desmoids
• Patient preference to avoid surgery

Rank the options by their evidence. Two agents have a positive, placebo-controlled phase 3 randomised trial: nirogacestat (DeFi), which is also the only agent approved specifically for desmoid tumours, and sorafenib. Pazopanib has a randomised phase 2 (DESMOPAZ). Everything else - NSAIDs, anti-oestrogens, methotrexate-vinblastine, imatinib, interferon - rests on uncontrolled series.

Radiotherapy is rarely used in modern desmoid management, and when it is used it is the last resort:

• An unresectable tumour failing systemic therapy
• A recurrent tumour not amenable to surgery or systemic therapy
• Symptomatic progression despite all other modalities
• Older patients, in whom late effects matter less

Technique. 50-56 Gy in 25-28 fractions, delivered daily over 5-6 weeks with conformal planning (IMRT or proton therapy) to minimise normal-tissue exposure.

Outcomes. Local control is 70-80% at 5 years, with a response (shrinkage or stabilisation) in 50-60%, and a median time to response of 6-12 months.

Toxicity is significant:

• Acute: skin reaction, fatigue
• Subacute: fibrosis at 3-12 months
• Chronic: fibrosis and joint stiffness in 40-50%, lymphoedema in 10-20%, muscle atrophy
• Radiation-induced sarcoma: rare, 0.5-1% at 10 or more years

Surgery is now reserved for selected cases after surveillance has failed.

Function over margins. This is the critical departure from sarcoma surgery, where wide margins are mandatory. Because the tumour is benign, margin status has no survival impact, a recurrence is manageable with surveillance or systemic therapy, and a positive margin is therefore acceptable wherever clearance would cost nerve, vessel or critical muscle. Function is the priority, and the surgical decision-making that follows serves long-term function rather than the pathology report.

A major nerve (median, ulnar, sciatic, femoral). Attempt preservation if a clearance of greater than 1 mm is possible; if the tumour encases the nerve, preserve the nerve and accept the positive margin. A major nerve is never sacrificed for a negative margin, because the functional loss is permanent and devastating.

A major vessel (femoral, popliteal, brachial). Dissect the tumour off the vessel if a plane exists; if it is adherent, preserve the vessel and accept the positive margin. Vascular reconstruction is possible but adds morbidity, so a positive margin is the better trade.

Muscle and bone. Non-critical muscle can be resected; functionally critical muscle is preserved where possible, so the deltoid, which the shoulder depends on, is kept while the sartorius can be sacrificed. A tumour abutting bone is resected extraperiosteally, and bone is not resected for margin in a benign tumour.

Frozen section rarely changes intraoperative management, because a positive margin is often anticipated and accepted; its value is in documenting margin status for adjuvant planning.

Preoperative planning. Review the MRI with radiology, identify the neurovascular structures at risk, plan an extensile incision for adequate exposure, and consent for the acceptance of a positive margin and the risk of recurrence.

The operation.

1. Exposure - wide exposure of the entire tumour through an extensile incision with adequate working room, identifying and protecting the critical structures early
2. Dissection - aim for complete macroscopic excision along tissue planes where they exist; the tumour is often densely adherent to fascia and muscle, and piecemeal resection is accepted if that is what preserves function
3. Critical structures - dissect the tumour carefully from nerve and vessel; if a structure is encased, preserve it and leave tumour on its surface; neurolysis where a nerve is compressed but separable; never sacrifice a structure for margin
4. Specimen - orient with sutures, document any margin concerns, and send for permanent sections
5. Reconstruction - primary closure if tension-free, local flaps for soft-tissue cover, mesh reconstruction of the abdominal wall if the rectus has been excised, and drains for dead space

After surgery. Early mobilisation, drain removal once output is under 30 ml per 24 hours, wound monitoring, and then the final pathology for margin status, on which surveillance or adjuvant therapy is planned.

Outcomes. Recurrence after resection runs at 20-40% even after complete excision, and it is the commonest complication of surgery.

Recurrence also varies by site:

• Extra-abdominal: 20-40%
• Abdominal wall: 20-30%
• Intra-abdominal: 40-60%, the worst prognosis

Why a positive margin is not re-resected. The crude figures above are higher with a positive margin, yet many patients with positive margins never recur, and in Crago's multivariate analysis the microscopic margin did not independently predict recurrence at all. Re-resection adds morbidity without clear benefit, recurrence is manageable with surveillance or systemic therapy, and function matters more than clearance. Examiners expect this principle stated plainly.

Managing a recurrence. The first line is active surveillance, exactly as for the primary tumour. Re-resection is for the recurrence that is progressive and symptomatic, systemic therapy is considered especially after multiple recurrences, and radiotherapy remains the last resort.

Of the disease. The complications follow the site.

Of surgery. Recurrence, covered above, is the commonest. The others:

• Wound infection: 10-15%
• Nerve injury: 5-10%, higher if the tumour is adherent
• Functional impairment: 20-40% (shoulder, thigh)
• Abdominal wall hernia: 10% where mesh reconstruction was needed

Of systemic therapy and radiotherapy. Each agent's toxicity is set out with the agent above: the GI upset and hot flushes of NSAIDs and tamoxifen, the rash, fatigue, hypertension and diarrhoea of sorafenib together with hand-foot syndrome in 50%, the myelosuppression and nausea of methotrexate-vinblastine, and the fibrosis, lymphoedema and secondary malignancy of radiotherapy.

Global Epidemiology

Desmoid-type fibromatosis is a rare tumour. Population-based and registry data give an estimated incidence of roughly 3 to 5 cases per million per year, with a female predominance and a peak in the third and fourth decades. The prospective Dutch GRAFITI cohort and large referral-centre series confirm that the majority of patients managed with active surveillance never require active treatment, and that genotype (S45F CTNNB1) and tumour size predict the need for intervention (Schut et al., Ann Surg 2023; DOI). The defining molecular feature, CTNNB1 mutation in approximately 85% of sporadic cases, is consistent across international cohorts (Gronchi et al., Ann Oncol 2014; DOI).

Side-by-Side Guideline Guidance

Registry & Prospective Evidence

Practice Variation & Global Access

International practice has converged on a surveillance-first, multidisciplinary model coordinated through specialist sarcoma centres, though access to newer agents varies by jurisdiction. Suspected desmoid tumours should be discussed with a specialist sarcoma multidisciplinary team, with image-guided core needle biopsy arranged locally or at the referral centre. Tamoxifen is widely available and inexpensive, and nuclear beta-catenin immunohistochemistry is available through standard anatomical pathology services. Sorafenib and nirogacestat regulatory and reimbursement status differs between countries, so confirm local availability before counselling patients on systemic options.