Intraosseous Lipoma

Intraosseous lipoma is a rare benign tumour of mature adipose tissue within the medullary cavity of bone, accounting for under 0.1% of all bone tumours. It is worth knowing well because the imaging is characteristic: central calcification on the radiograph and fat signal on MRI allow a confident diagnosis without biopsy in most cases.

Who. A middle-aged adult, typically 40-60 years, with no sex predilection. Under 5% of cases are children. The lesion is usually solitary, and multiple lipomas are exceptionally rare.

Where. About a third arise in the calcaneus, the commonest single site, and the proximal femur comes next.

• Calcaneus, 32% of the pooled Campbell series. Usually asymptomatic, found incidentally or with heel pain.
• Proximal femur, 20%, in the intertrochanteric region, where a large lesion carries a risk of pathological fracture.
• Proximal tibia, 5%, metaphyseal and usually asymptomatic.
• Other sites, 15%: fibula, ribs, skull and ilium, all rare; fibula, ribs and skull under 5% each.
• Never in the spine, unlike soft-tissue lipomas.

Why the calcaneus. The predilection is unexplained. It may relate to the high proportion of fatty marrow in the calcaneus, to repetitive microtrauma from weight-bearing, or to a vascular anatomy that predisposes to focal ischaemia. Whatever the mechanism, the calcaneus is the first site to consider when you see an intraosseous lipoma.

View larger

View larger

Origin. The origin is unknown. The most accepted theory is metaplasia: bone marrow mesenchymal cells transform into adipocytes in response to trauma or microtrauma, ischaemia or infarction, or an unknown stimulus. The alternative, a congenital nidus of ectopic fat cells that slowly expands, is less favoured because most cases present in adulthood rather than childhood.

Why the calcification is central. Fat at the centre of the lesion is furthest from its blood supply, and it necroses there, whether from ischaemia or because the lesion outgrows its supply. The necrotic fat undergoes dystrophic calcification, which spreads from the centre outward as the lesion involutes. A bone infarct is the opposite: its calcification is peripheral and serpentine, delineating the margin of the dead segment like a geographic map.

Gross appearance. Yellow, greasy, soft tissue indistinguishable from normal adipose tissue, sometimes with areas of white chalky calcification where fat has necrosed. Lesions are typically 2-5 cm across, rarely larger.

Microscopy. Histologically the lesion is indistinguishable from a soft-tissue lipoma: both are mature adipose tissue. What identifies it is the intramedullary location and the central calcification pattern.

• Mature adipocytes: large cells with a single lipid vacuole displacing the nucleus to the periphery, with minimal atypia
• Fat necrosis (stages 2-3): ghost cells and loss of cell membranes
• Dystrophic calcification: calcium deposited in the necrotic fat
• Reactive bone: woven bone at the periphery in some cases

The histological differential. Normal marrow fat is excluded because the lipoma is an expansile mass, not simply fatty marrow. Liposarcoma contains lipoblasts, cells with scalloped hyperchromatic nuclei, which are not present in a benign lipoma. A bone infarct shows geographic necrosis of bone and marrow with calcification at the periphery.

View larger

Milgram's histological classification (1988) is the gold standard. It describes the natural involution of the lesion from viable fat through necrosis to calcification, and does not indicate biological behaviour or malignant potential: all three stages are equally benign.

How common each stage is. No reliable proportion exists. Milgram described the three stages but did not publish their relative frequencies, and the two surgical series cited on this page disagree sharply.

• Stage 1: 10 of 29 (34%) in Radl, 9 of 14 (64%) in Aycan
• Stage 2: 14 of 29 (48%) and 4 of 14 (29%), not a clear majority
• Stage 3: the least frequent in both, 3 of 29 (10%) and 1 of 14 (7%)

Both are operated cohorts, so they over-represent symptomatic lesions and cannot give a population distribution. Stage 2 is called the classic appearance, and it is the one an examiner will show, but its prominence reflects how recognisable it is rather than a measured predominance.

Why stages 1 and 3 are harder. A stage 1 lesion has no calcification to see. A stage 3 lesion is the hardest of all, because heavy calcification loses the fat signal that otherwise makes the diagnosis straightforward.

Symptoms. Between 50 and 70% are asymptomatic, found on imaging done for other reasons. The other 30-50% have a dull, aching, activity-related pain, lasting months to years; its mechanism is unclear, and may be microfracture or pressure. There are never systemic symptoms such as fever, weight loss or malaise. Pathological fracture is rare (under 5%) but possible when a large lesion has thinned the cortex.

Natural history. The lesion stays the same size, without the progressive enlargement of a malignant lesion. It involutes slowly, and may calcify and become more apparent over years.

Examination. There is usually little to find. The lesion is intramedullary, so there is no visible swelling or external mass, and no erythema, warmth or overlying soft-tissue abnormality. There may be mild focal tenderness over the lesion. Neurovascular examination is normal and the adjacent joints move fully. A painful calcaneal lesion may produce an antalgic gait. Weight-bearing is tolerated unless there is a fracture or severe pain, and there is no deformity unless a pathological fracture has occurred.

The principle. Most intraosseous lipomas can be observed. Treatment is guided by symptoms and fracture risk, not by Milgram stage: an asymptomatic lesion is observed, and a symptomatic or structurally threatened one is curetted and grafted. Site shapes the threshold. A calcaneal or proximal tibial lesion is observed unless symptomatic, a proximal femoral lesion with cortical thinning is considered for prophylactic curettage, and lesions elsewhere are managed case by case.

Observation. Observation is appropriate when:

• The lesion is asymptomatic and discovered incidentally
• The imaging is classic, leaving no diagnostic uncertainty
• It is small, with no risk of pathological fracture
• The patient prefers to avoid surgery

Repeat the radiograph at 6 months to confirm stability and again at 1 year; if it is still stable, discharge. Explain that the lesion is benign and ask the patient to return if new pain or symptoms develop. Most lesions never become symptomatic, and mild pain from a small lesion with an intact cortex and no fracture risk is managed conservatively with analgesia and activity modification.

Surgery. Curettage is indicated for:

• Persistent pain unresponsive to conservative measures
• Pathological fracture or imminent fracture risk
• A large lesion with cortical thinning (prophylactic treatment)
• Diagnostic uncertainty (biopsy and curettage)

Technique. Curettage proceeds in four steps.

1. Approach: direct, through a small cortical window made with a curette or drill, preserving the structural integrity of the bone and avoiding neurovascular structures
2. Curettage: remove all fatty tissue and calcification, curette the walls to ensure complete excision, and send the specimen for histology to confirm the diagnosis
3. Grafting: fill a large defect with iliac crest autograft for structural defects, cancellous allograft chips for small defects, or a calcium phosphate or sulfate substitute
4. Closure: irrigate and close the soft tissues in layers; a drain is usually not needed for small defects

Completeness. Recurrence occurs when curettage is incomplete, so curette every wall thoroughly. Wide margins are not needed for a benign lesion, and nor are adjuvants such as phenol or burr (see Controversies).

Prophylactic fixation. Consider internal fixation when:

• The lesion is large, greater than 50% of the bone diameter
• Cortical thinning is significant after curettage
• The bone is weight-bearing (proximal femur, tibia)
• The location carries a high fracture risk (femoral neck, subtrochanteric)

The options, depending on location, are an intramedullary nail, a plate and screws, or percutaneous screws.

Liposclerosing myxofibrous tumour is the key differential at the intertrochanteric femur. That matters because the proximal femur is the second most common site of intraosseous lipoma, and the two lesions behave differently.

A different, mixed lesion. LSMFT is a benign fibro-osseous lesion with a mixed histology: variable proportions of lipoma, fibrous tissue, myxoid change, fat necrosis, cyst formation, ischaemic ossification and ground-glass, fibrous-dysplasia-like bone, rather than the pure mature fat of an intraosseous lipoma. It is characteristically located in the intertrochanteric region of the proximal femur, in the great majority of cases.

The MRI discriminator. Because of its mixed matrix, LSMFT usually does not follow fat signal uniformly. It shows heterogeneous T1 and T2 signal, often with a fibrous-dysplasia-like or myxoid appearance and sometimes a sclerotic rim, so the reassuring high T1 that fully suppresses with fat saturation is typically absent.

The risk of change. A true intraosseous lipoma has essentially zero malignant potential. LSMFT has a described, though contested and probably referral-biased, risk of malignant transformation to malignant fibrous histiocytoma or osteosarcoma. A symptomatic, enlarging or aggressive-appearing intertrochanteric LSMFT is therefore treated with a lower threshold for biopsy and closer surveillance than a classic intraosseous lipoma.

Why it is examined. It is the classic intertrochanteric fibro-osseous lesion that overlaps radiologically with intraosseous lipoma and fibrous dysplasia. The point being tested is that a heterogeneous, not purely fatty, intertrochanteric lesion is not automatically a benign lipoma to be ignored.

A fatty mass on the bone surface. Parosteal (juxtacortical) lipoma is a benign lipoma arising on the outer surface of the cortex or periosteum, most often around the radius, femur, humerus and pelvis. Like the intraosseous type it has fat signal on CT and MRI, but it is juxtacortical, not intramedullary. It frequently provokes a reactive bony excrescence, cortical hyperostosis or an osteochondroma-like spur at its base, a useful radiographic clue.

The clinical catch: nerve compression. Its surface position means it can compress an adjacent peripheral nerve, a functional consequence the intramedullary lipoma never causes. The classic and most examined example is a parosteal lipoma of the proximal radius compressing the posterior interosseous nerve, producing a PIN palsy or radial-tunnel-type presentation; similar compressive syndromes occur elsewhere.

Diagnosis and treatment. MRI shows a fat-signal mass abutting the cortex with the reactive bony change, confirming the diagnosis without biopsy in typical cases. Because the lesion is benign, marginal excision is curative. It is indicated when the lipoma is symptomatic or compressing a nerve, with neurolysis of the involved nerve, and asymptomatic incidental lesions can be observed.

Radiographs are taken at 6 weeks and 3 months after curettage. Annual follow-up is not required once the lesion is confirmed as a benign lipoma.

After curettage. Curettage with or without bone graft is curative in over 95% of cases, and 95% of patients have complete pain relief. Grafted defects heal within 6-12 weeks, and patients return to full activity by 3-6 months.

Long-term prognosis. Excellent. The lesion does not grow, does not metastasise, does not affect other organs and does not cause death. It may calcify and involute over time, from stage 2 to stage 3. For practical purposes it has no meaningful malignant potential, and patients can be reassured on that point; the presumed cases of malignant transformation are discussed under Controversies.

Global Epidemiology

Intraosseous lipoma is reported worldwide with a consistent profile: peak presentation in middle age (mean ~43-48 years across pooled series), near-equal sex distribution, and a lower-limb predominance (~71% of cases), with the os calcis the single most common site. It is genuinely uncommon (classically quoted at under 0.1% of primary bone tumours) but is widely regarded as underdiagnosed because involuted (Milgram III) lesions are mistaken for cysts or bone infarcts. There is no dedicated tumour registry for this benign entity; epidemiology derives from radiology and orthopaedic-oncology case series rather than national arthroplasty/implant registries.

Guideline Framework (Side-by-Side)

No major society mandates routine biopsy or specialist referral for a radiologically classic intraosseous lipoma.

True neoplasm or involutional process? Campbell and colleagues questioned whether intraosseous lipoma is a genuine benign tumour of fat or the end-result of involution of pre-existing lesions and biomechanical stress, pointing to the constant os calcis location at the critical angle. The histogenesis remains unsettled.

Does malignant transformation exist? Milgram reported four cases of presumed malignant transformation (liposarcoma or malignant fibrous histiocytoma) within bone lipomas. These are exceptionally rare and contested. For practical purposes the lesion is treated as having no meaningful malignant potential, but rapid bone destruction in a previously lucent lesion warrants reassessment.

Observe or operate? There are no randomised data comparing observation with curettage. Practice is guided by symptoms and fracture risk, as set out under Management.

Adjuvants. Phenolisation and extended (burr or argon) curettage are sometimes used by analogy with other bone lesions. The Radl cohort found no added benefit from phenol, consistent with the benign biology, and aggressive adjuvants are not justified.