Periosteal Osteosarcoma

Periosteal osteosarcoma is a malignant bone-forming tumour that arises from the deep layer of the periosteum and grows outward on the cortical surface. Its phenotype is predominantly chondroblastic and its histological grade intermediate (grade 2), which makes it the middle child of the surface osteosarcoma family.

Origin. The cell of origin lies in the cambium layer of the periosteum, so the tumour sits on, and erodes into, the outer cortex rather than arising within the medulla. Its local behaviour is dominated by the soft-tissue component.

Site. The diaphysis or metadiaphysis of the femur (approximately half) and the tibia (approximately one third); the humerus, ulna, ribs and jaw are described. The lesion typically occupies one aspect of the shaft, and circumferential involvement is uncommon.

Age. The peak is in the second and third decades, essentially overlapping conventional osteosarcoma and a decade younger than parosteal.

Spread. It metastasises to the lung in a minority of patients. Regional nodal spread is not a feature.

Macroscopic. A lobulated, glistening, blue-grey cartilaginous mass sits on the cortex, with gritty bone spicules extending into it from the cortical surface and a firm reactive base. The medulla, when sectioned, is normal.

Microscopic. There are three things to look for.

1. Lobules of atypical cartilage form most of the tumour volume: hypercellular hyaline cartilage with binucleate cells and nuclear enlargement, corresponding to the low-density areas on CT.
2. Peripheral spindling. At the edge of each cartilage lobule the cells become spindled and more atypical, blending into a sarcomatous stroma. This is where the diagnosis is made.
3. Fine, lace-like tumour osteoid deposited directly by those malignant spindled cells between the lobules and along the spicules. Osteoid produced by tumour cells, not entrapped reactive bone, is what makes this an osteosarcoma rather than a chondrosarcoma.

Grade. Intermediate, grade 2 of 3. Mitoses are present but not frenzied, and necrosis is limited.

Molecular. There is no specific immunohistochemical marker. Unlike parosteal osteosarcoma, periosteal lesions do not show consistent MDM2/CDK4 amplification. A negative MDM2 FISH is therefore useful against parosteal osteosarcoma and against dedifferentiated parosteal lesions, and supports periosteal over parosteal when the imaging is ambiguous, but it does not exclude conventional osteosarcoma. IDH1/IDH2 mutation is a feature of conventional cartilage tumours, and its absence supports osteosarcoma in an equivocal chondroid lesion.

Parosteal, periosteal and high-grade surface osteosarcoma are the three surface (juxtacortical) osteosarcomas, and they differ in site, density, grade and molecular profile.

Stage before the needle. Whole-bone MRI with contrast and a staging chest CT come before any tissue is taken, so that local extent and metastatic status are defined ahead of biopsy. The systemic survey is:

• CT chest, mandatory as the metastatic target organ: thin-slice, non-contrast, at diagnosis and at intervals thereafter
• Whole-body bone scintigraphy or FDG-PET/CT for the skeleton; PET also provides a baseline SUV if chemotherapy is used
• Bloods: alkaline phosphatase and LDH are usually normal or mildly raised, and markedly elevated values raise suspicion of a higher-grade lesion

Enneking. On the Enneking (MSTS) surgical staging system, a periosteal osteosarcoma with an extraosseous soft-tissue mass and no metastasis is typically Stage IIB: high grade on Enneking's two-tier system, extracompartmental, M0. The binary grading forces an intermediate lesion into the high-grade box, and a viva answer should say so explicitly rather than pretend the system is finer than it is.

AJCC. T is assigned by size about an 8 cm threshold, G by grade, and M by site of metastasis, lung versus other.

What the meeting settles before theatre. The aim is a single agreed plan covering margin, reconstruction and systemic therapy, and it needs these decisions:

• Grade confirmation on adequate tissue
• True marrow status
• Osteotomy levels
• Reconstruction strategy
• Whether chemotherapy is offered and, if so, neoadjuvantly (which allows assessment of necrosis in the resected specimen) or adjuvantly

Fertility and growth. If chemotherapy is planned in an adolescent, arrange fertility counselling and gamete or ovarian tissue preservation before the first cycle. In a skeletally immature patient, account for remaining growth when choosing the reconstruction.

After resection. Confirm the margins on the specimen and, if neoadjuvant chemotherapy was given, the percentage necrosis. Clear margins give an excellent prognosis; marrow-positive or intralesional disease shifts management towards conventional high-grade osteosarcoma protocols and closer surveillance.

Where it goes wrong. A badly placed diaphyseal biopsy contaminates compartments and can convert an intercalary resection into an amputation or a joint sacrifice. Biopsy performed outside the centre that will resect contaminates planes that were never going to be excised, and it is the commonest avoidable cause of a compromised reconstruction.

Why a core can mislead. Periosteal osteosarcoma has an abundant chondroid matrix. A core taken from the cartilaginous lobules alone reads as a low-grade cartilage tumour, and the ossified periphery yields only reactive bone, so the target is the non-mineralised soft-tissue component.

When the tract is already wrong. Excise the whole tract with the specimen, and accept a larger soft-tissue defect, flap cover or, in the worst case, the loss of a limb-salvage option. A transverse incision, or a tract crossing an intermuscular plane or neurovascular bundle, forces a marginal excision; re-resect to negative margins where anatomy allows, and only once radiology, pathology and surgery agree on further treatment.

Tempo. Symptoms have typically been present for several months, commonly 3 to 12. That is slower than conventional osteosarcoma and faster than parosteal, which may smoulder for years.

Pain and mass. The pain is dull, activity-related and felt in the mid-shaft of the thigh or shin. Night pain is variable, and its absence does not reassure. The mass is firm and fixed to bone, often first noticed by the patient or after minor trauma, and the trauma is a red herring that delays referral.

What is missing. There are no systemic features; weight loss, fever or malaise should prompt thought of Ewing sarcoma or osteomyelitis. Function is near normal until the mass is large, and pathological fracture is rare because the cortex is thickened, not destroyed.

Inspection. Examine the patient standing and then supine, with the limb fully exposed and compared with the other side. Look for asymmetric fullness of the anteromedial thigh or the subcutaneous tibial border, overlying venous prominence, and skin tethering, which is a late sign.

Palpation. Sweep the flat of the hand longitudinally along the shaft, then define the lesion with the fingertips. A periosteal osteosarcoma is firm, deep and immobile relative to bone, with mobile overlying muscle, and tender on direct pressure. Measure its craniocaudal and transverse dimensions with a tape; this is the baseline for the referral letter.

Joints. Compare hip, knee and ankle range side to side. Diaphyseal tumours characteristically spare the joint, so an effusion or fixed flexion suggests a very large mass, an alternative diagnosis, or joint contamination.

Nerves, vessels and nodes. Palpate the distal pulses and assess the sciatic and femoral, or peroneal and tibial, nerve distributions. In the proximal tibia and posterior femur, document peroneal and sciatic function before any biopsy, which is medicolegally important. Examine the inguinal and popliteal basins; nodal disease is essentially never osteosarcoma, so pathological nodes mean rethinking the diagnosis.

Differential diagnosis. The lesions to separate it from, and what separates each one, are these.

Wide en-bloc resection is the treatment. Because the medulla is spared, most tumours are amenable to diaphyseal (intercalary) segmental resection preserving both adjacent joints, a major functional advantage over conventional osteosarcoma.

The soft-tissue margin. Take an uninvolved layer of muscle or fascia across the entire surface of the mass, and never dissect on tumour. The biopsy tract comes out en bloc with the specimen.

The bone cuts. Plan the resection on MRI with the periosteal reaction and reactive rim included. On the coronal or sagittal T1, mark the proximal and distal extent of tumour and reactive rim, then add the planned margin, commonly 2 to 3 cm of normal marrow, to derive the osteotomy levels. Check each level against a fixed landmark (joint line, tubercle) for intraoperative reproducibility, and confirm it with frozen section of marrow from each cut end.

Hemicortical resection. If a portion of cortex is uninvolved circumferentially and the tumour is truly limited to one aspect, a hemicortical resection with hemicortical allograft is described and preserves continuity. It is only for cases where MRI proves circumferential sparing and margins are achievable; most surgeons reserve it for parosteal lesions and are more conservative with periosteal disease.

Amputation is reserved for tumours encasing the neurovascular bundle beyond reconstruction, extensive contamination from unplanned surgery, or fungation and infection.

Reconstruction. Once the segment is out, the defect is filled in one of these ways.

The problem. Conventional osteosarcoma chemotherapy protocols (doxorubicin, cisplatin, high-dose methotrexate, with or without ifosfamide) were validated in high-grade disease. Periosteal osteosarcoma is intermediate grade and rare, and it has never been the subject of a randomised trial.

What the literature shows. Large retrospective and cooperative-group analyses of periosteal osteosarcoma have not demonstrated a clear survival advantage for chemotherapy, and pathological necrosis response rates are generally poorer than in conventional osteosarcoma. Series consistently identify surgical margin adequacy, not chemotherapy, as the dominant determinant of outcome.

Weighing it for one patient. These are the inputs that push the decision one way or the other.

The viva answer. "There is no randomised evidence that chemotherapy improves survival in periosteal osteosarcoma, and retrospective series have not shown a consistent benefit. Surgery with a wide margin is the mainstay. Many centres nonetheless offer conventional osteosarcoma chemotherapy to young patients, particularly where the tumour is large, the grade is heterogeneous, or the margin is compromised. I would take the case to the sarcoma multidisciplinary meeting and consent the patient with an honest account of the uncertainty."

Radiotherapy has no established role in primary management. It is reserved for unresectable disease, positive margins where re-excision is impossible, or palliation.

Prognosis. Long-term survival is 85 to 90% with a wide, negative margin, intermediate between parosteal (greater than 90%) and high-grade surface or conventional osteosarcoma. Pulmonary metastases develop in 10 to 15%, typically within the first 3 years, but late events occur.

The margin is the prognosis. Local recurrence after an inadequate margin correlates strongly with metastasis and death. The adverse factors are:

• Inadequate margin
• Local recurrence
• Large tumour volume
• High-grade foci
• True medullary invasion
• Metastasis at presentation

Lung metastases. Resectable pulmonary metastases should be considered for metastasectomy; prolonged survival after complete resection of limited lung disease is well documented across osteosarcoma generally.

Surveillance. Local and chest imaging are most frequent in the first two years, then spaced out, and the reconstruction is watched for life.

Complications of treatment come from the resection, the reconstruction and the chemotherapy.

Global epidemiology. Osteosarcoma incidence is broadly similar worldwide (roughly 3 to 4 per million per year across all subtypes). Estimates of the periosteal share vary with the denominator and the era: this page quotes 1 to 2%, while the SEER-based survey cited above puts it at 1 to 6% — quote a range and say which source you are using rather than a single figure. Because absolute numbers are tiny either way — that SEER query found just 54 patients in the whole United States registry — no single centre accumulates a large series and evidence is inevitably retrospective and multi-institutional. Outcome differences between regions are driven far more by time to referral and access to reconstruction than by tumour biology.

Society and network guidance (where it genuinely addresses this topic):

Registry and network evidence. There are no arthroplasty-style implant registries for this disease. The relevant datasets are national and supranational bone sarcoma registries and cooperative group databases (European Musculoskeletal Oncology Society registries, cooperative osteosarcoma study groups, national cancer registries), which pool rare variants to generate the survival and chemotherapy-response figures quoted above. Where segmental endoprostheses are used, national implant and tumour-prosthesis audits contribute long-term revision data.

Practice variation by resource setting.

• High-resource centres: MRI-planned intercalary resection, allograft banking, microvascular reconstruction, PET staging, fertility preservation, protocol chemotherapy where chosen.
• Intermediate-resource settings: intercalary resection with autograft, non-vascularised fibula, or bone transport where allograft banks and microsurgery are unavailable; CT chest available but PET is not.
• Low-resource settings: late presentation with large masses, limited pathology expertise (risk of misdiagnosis as a cartilage tumour), and higher amputation rates. The single highest-yield intervention in these settings is early referral and correct biopsy, not access to chemotherapy.

1. Does chemotherapy help? The central unresolved question. No randomised data; retrospective series conflict and are underpowered. Practice ranges from surgery alone in all cases to routine conventional protocols in all young patients.
2. What counts as medullary invasion? MRI reactive oedema is over-called. Some argue any T1 marrow change reclassifies the tumour; others require confluent replacement plus histological confirmation. The threshold materially changes both grade assignment and chemotherapy decisions.
3. Is periosteal osteosarcoma a distinct entity or a spectrum? Some pathologists regard it as one end of a continuum with high-grade surface osteosarcoma, with the grade determined by sampling. Inadequate biopsy sampling can misgrade either way.
4. Hemicortical versus segmental resection. Hemicortical excision preserves continuity and function but the oncological safety margin in an intermediate-grade tumour with periosteal spread is less secure than in parosteal disease. Most units resect segmentally.
5. Best intercalary reconstruction. Allograft, vascularised fibula, Capanna construct, intercalary endoprosthesis and bone transport all have advocates; comparative data are small-series and centre-dependent. Choice is driven by defect length, host biology, chemotherapy exposure and local microsurgical capability.
6. Duration of surveillance. Late metastasis and late reconstruction failure both occur; whether chest imaging should continue beyond 10 years is not settled.

Related pages. The tumour it is most confused with, and the distinction that matters: Parosteal Osteosarcoma — low grade, metaphyseal, classically the posterior distal femur, MDM2/CDK4 amplified — against this one, which is intermediate grade, diaphyseal, chondroblastic and MDM2 negative. The parent disease: Osteosarcoma, whose chemotherapy protocols are the ones this tumour probably does not need. Reading the film: Periosteal Reaction Patterns for the perpendicular sunburst reaction, and Osteochondroma and Myositis Ossificans for the benign surface lesions on the differential. Getting the tissue and the operation right: Biopsy Principles and Techniques — the two-to-twelve-fold hazard above is the reason — plus Limb Salvage Surgery Principles and Allograft-Prosthetic Composite Reconstruction. Other sarcomas of the same age group: Ewing Sarcoma and Chondrosarcoma.