Adamantinoma

What it is. Adamantinoma is a rare low-grade malignant bone tumour, under 0.5% of all primary bone tumours and under 1% of primary malignant bone tumours, with a distinctive biphasic histology of epithelial cells set within an osteofibrous stroma. Its predilection for the tibial diaphysis is remarkable: 80-85% arise there, characteristically involving the anterior cortex.

Who. Peak age is 20-50 years, with a range of 2-86 years, and there is a slight male predominance (5:4).

Risk factors. None has been established. There is no identified genetic predisposition and no association with radiation or chemical exposure. A history of prior trauma is reported in 60% of patients, of uncertain significance, and there is a possible association with osteofibrous dysplasia (OFD), which is taken up under Pathophysiology.

Biphasic histology. Adamantinoma is defined by two components, epithelial and osteofibrous, intermingled in various proportions. The epithelial cells form nests, cords or tubules, sometimes with peripheral palisading, and take one of four patterns (basaloid, tubular, spindle-cell or squamous) set out under Classification. They are positive for cytokeratins 14 and 19 and negative for CK8 and CK18.

The osteofibrous component. A fibrous stroma containing woven bone trabeculae and scattered osteoclast-like giant cells, positive for vimentin. On routine histology it resembles osteofibrous dysplasia, so it is the epithelial component, highlighted on cytokeratin immunostaining, that separates the two.

View larger

View larger

View larger

Relationship to osteofibrous dysplasia. The relationship is controversial but clinically important, and the three lesions are separated by age, by how much epithelium is present and by how it stains:

The controversy. The Leiden group demonstrated keratin-positive epithelial cells emerging individually from the osteofibrous stroma, with progressively organised basement membrane around cohesive epithelial islands, and read this as evidence that OFD-like adamantinoma may be a precursor to classic adamantinoma via fibrous-to-epithelial transformation (Hazelbag, Hum Pathol 1997). The Mayo Clinic counterargument is follow-up: in 80 OFD cases followed for up to 31 years, no lesion progressed to adamantinoma and two matured into fibrous dysplasia, suggesting OFD is a variant of fibrous dysplasia rather than an obligate precursor (Park & Unni, Hum Pathol 1993).

The reconciled view. OFD, differentiated (OFD-like) adamantinoma and classic adamantinoma form a histological continuum, but pure OFD does not inevitably become malignant. Hence cytokeratin immunohistochemistry is essential: it detects the epithelial component on which the distinction rests.

View larger

View larger

Molecular features. The malignant clone is epithelial and the osteofibrous tissue is reactive, and three findings show it:

• Aneuploidy is confined to the epithelial cells. DNA cytometry shows aneuploid nuclei only in cells of epithelial phenotype, while the fibrous cells remain diploid (Hazelbag, Am J Pathol 1995).
• p53 abnormalities (immunoreactivity and loss of heterozygosity at the p53 locus) are detectable in roughly half of tumours and again localise to the epithelial component (Hazelbag, Am J Pathol 1995).
• Lung metastases contain only keratin-positive epithelial cells, with no osteofibrous element, confirming the epithelial cells as the metastasising population (Hazelbag, Am J Pathol 1995).

Shared chromosomal gains. Both OFD and adamantinoma carry the same extra copies of chromosomes 7, 8, 12, 19 and 21. This shared pattern of numerical gains is the strongest genetic argument that the two sit on one biological spectrum rather than being unrelated lesions that happen to occupy the same piece of tibia, and it is the finding to quote when asked why the continuum hypothesis is taken seriously despite the Mayo follow-up data.

What is absent. No specific recurrent fusion gene has been identified, unlike the EWSR1 rearrangement of Ewing sarcoma, which is a useful diagnostic discriminator. The GNAS mutations characteristic of fibrous dysplasia are not a feature either, which helps separate adamantinoma and OFD from fibrous dysplasia.

History. The classic story is slowly progressive swelling over the anterior shin with a dull, aching pain that has often been present for months to years; approximately 30% have had symptoms for more than 5 years before diagnosis. Duration matters prognostically: a shorter symptom duration (less than 5 years) is associated with a higher recurrence risk.

Red flags. Features that should raise concern:

• Rapid increase in swelling
• New or worsening pain
• Constitutional symptoms (weight loss, fever), rare but concerning
• Pathological fracture

Examination. Expect a palpable, firm mass arising from and fixed to the anterior tibial shaft, tender on palpation, with overlying skin that is normal or tense and possibly anterior tibial bowing. Warmth and erythema are usually absent unless there has been a fracture. Knee and ankle movement is usually preserved and the neurovascular examination is usually normal; check the dorsalis pedis, and palpate the popliteal and inguinal nodes.

Differential diagnosis. The lesions that share the tibial diaphysis or the lytic appearance, and what separates each from adamantinoma:

Plain radiographs. The first-line investigation for tibial swelling. The characteristic appearance is a "soap-bubble", multilocular lytic lesion, eccentric in the anterior cortex and well defined, with sclerotic margins and cortical expansion. Look along the rest of the bone, because multifocality (skip lesions) may be shown.

View larger

View larger

CT. Better cortical definition than the radiograph: CT defines the extent of bone destruction and identifies a pathological fracture, which makes it the planning study. CT of the chest stages for pulmonary metastases.

MRI. The gold standard for local staging. It defines intramedullary extent and soft-tissue extension and detects skip lesions; the tumour is low signal on T1, high signal on T2 and enhances with contrast.

View larger

Bone scan and PET-CT. Bone scan shows increased uptake and identifies multifocal disease, and is used to screen for metastases. The tumour is FDG avid, so PET-CT is useful for staging and for recurrence, and has a place in surveillance.

Biopsy. Core needle biopsy, CT- or ultrasound-guided, is preferred, with open biopsy if the core is inconclusive. Place the tract in line with the planned surgical incision, because it must be excised with the definitive resection.

Laboratory studies. FBC, ESR and CRP are usually normal, and LDH and ALP are usually normal or mildly elevated. Hypercalcaemia is rare and paraneoplastic when it occurs.

Pathological fracture occurs in 16-23% of cases. Stabilise the fracture, then proceed to definitive resection.

Local recurrence is the most significant complication, at the rates set out under Management. Assessment before any decision:

1. Imaging: MRI of the local site, CT chest, bone scan or PET
2. Biopsy to confirm the recurrence histologically
3. Surgical planning: review the margins from the primary surgery

Treatment is wide re-excision if margins are achievable, amputation if limb salvage is not possible, and palliative care for unresectable or metastatic disease.

Metastases. Pulmonary metastases occur in 15-30% and lymph node metastases in 5-10%. Pulmonary metastases are treated by surgical metastasectomy if feasible, lymph node metastases by regional lymphadenectomy, and multiple metastases by systemic therapy, for which the options are limited.

View larger

View larger

Surgical complications.

Days 0-3. DVT prophylaxis, mechanical and pharmacological. Drains come out when output falls under 50 ml per 24 hours; analgesia is PCA then oral; the affected limb is non-weight bearing.

Weeks 1-6. Watch the wound for infection and monitor for DVT and PE. Weight bearing is protected, to a degree that depends on the reconstruction, and physiotherapy works on range of motion and muscle strengthening.

Reconstruction-specific loading. An allograft is kept non-weight bearing for 6-12 weeks, and the construct is protected until the allograft-host junction unites, which may take 12-18 months. A vascularised fibula is loaded progressively as it hypertrophies. An endoprosthesis allows earlier weight bearing, at 2-4 weeks.

Return to activity.

• Desk work at 4-6 weeks, depending on pain control
• Manual work at 3-6 months
• Sports: low-impact activities are generally recommended
• Driving once off opioids and able to weight bear for an emergency stop

Survival. Five-year survival is 95-98% and 10-year survival 85-92% across series; single-institution cohorts differ, and the companion imaging page quotes 85-87%. Metastatic disease significantly worsens the prognosis. Recurrence-free survival at 10 years is approximately 72%, and late recurrence is a significant concern.

Prognostic factors. The factors associated with a worse outcome, and their better-prognosis counterparts:

Function and quality of life. Limb salvage is achievable in 80-90% of cases at specialised centres, with an MSTS functional score typically 70-85% of normal. Quality of life is generally good after successful limb salvage; patients may need walking aids initially, and long-term surveillance can cause anxiety.

Global Epidemiology

Adamantinoma accounts for well under 1% of all primary bone tumours, making it one of the rarest malignant bone neoplasms encountered worldwide. The disease is remarkably consistent across populations: the tibia is the dominant site (around 70-85% of cases), the peak incidence is in the second to fifth decades, and there is no described geographic, ethnic or environmental predisposition. The tumour has not been linked to radiation or to a hereditary cancer syndrome. Because the absolute numbers are so small, no national cancer registry generates adamantinoma-specific incidence figures of high precision, and the evidence base is built almost entirely from single-institution and tertiary-referral case series rather than trials.

Side-by-Side Guideline and Society Guidance

There is no disease-specific randomised evidence; all major bodies converge on the same principles, derived from retrospective series (evidence level IV) and expert consensus. Differences are largely organisational rather than oncological.

Registry and Centralisation Evidence

Because adamantinoma is so uncommon, the strongest registry-level lesson is centralisation: outcomes for primary bone tumours improve when biopsy and surgery are performed at high-volume sarcoma units, and a poorly planned biopsy outside such a unit can compromise limb salvage. National and supranational sarcoma networks (for example the European EURACAN reference network and equivalent national bone tumour services) exist precisely to pool these rare cases. There is no arthroplasty-style implant registry relevant to adamantinoma, but limb-salvage reconstruction outcomes (allograft, vascularised fibula, endoprosthesis) are tracked within institutional and bone tumour databases.

Practice Variation

The principal real-world variation is in reconstruction technique and resource availability rather than oncological strategy. Wide resection is universal, but the choice between intercalary allograft, vascularised fibular autograft, allograft-prosthesis composite and endoprosthesis depends on resection length, patient age and - importantly in lower-resource settings - access to bone banks and custom implants. Where advanced prostheses and allografts are unavailable, free vascularised fibular grafting is a well-described and effective alternative (Farooque, J Orthop Case Rep 2024). Across all systems, lifelong surveillance for late local recurrence and pulmonary metastasis is the shared standard, reflecting documented recurrences decades after index surgery.