Expandable Prostheses in Pediatric Oncology

Expandable endoprostheses allow limb salvage in skeletally immature patients with malignant bone tumours while accommodating the growth still to come. Without them, a child with significant growth remaining would face a limb-length discrepancy (LLD) greater than 4cm after a tumour resection that includes the growth plate.

Who and where. Primary bone sarcomas are rare, approximately 8-9 per million children and adolescents per year worldwide, and osteosarcoma and Ewing sarcoma account for the majority of those requiring a growing reconstruction. Osteosarcoma peaks at 15-19 years and Ewing sarcoma at 10-15 years. The common sites are:

• Distal femur - 40%
• Proximal tibia - 20%
• Proximal humerus - 15%

History. Scales and Sneath developed the first expandable prosthesis at the Royal National Orthopaedic Hospital (Stanmore) in 1976, and the early designs needed open surgery to exchange modular segments. Non-invasive magnetic technology (REPIPHYSIS) arrived in the early 2000s and significantly reduced the operative burden.

Indications. The indications for an expandable prosthesis are:

• Predicted LLD greater than 4cm at skeletal maturity
• A malignant bone tumour whose resection takes the physis
• A patient and family who accept multiple procedures and prolonged follow-up
• An adequate soft-tissue envelope for reconstruction

A pathological fracture through the tumour is a relative contraindication.

Physeal contribution. Accurate prediction of the discrepancy is essential for planning, because resecting a physis arrests the growth it would have supplied, and that contribution varies by site.

The multiplier method. Paley's multiplier method gives the most accurate prediction of LLD, using skeletal age:

1. Bone-age radiograph of the left hand and wrist
2. Read the multiplier from the published tables for skeletal age and sex
3. Remaining growth = current limb length x (multiplier - 1)
4. Predicted LLD = the share of that remaining growth the resected physis would have supplied, since it supplies none once resected

A worked example. A 10-year-old boy (skeletal age 10) with osteosarcoma of the distal femur needs a resection that takes the physis. His femur is 35cm long and the multiplier for a 10-year-old boy is 1.28, so the femur has 35 x (1.28 - 1) = 9.8cm still to grow. The distal femur would have supplied 70% of that, a predicted LLD of 6.9cm, and he needs an expandable prosthesis.

Timing of expansion. Lengthening is typically done every 3-4 months to match physiological growth, more often in younger children who grow faster, aiming for 5-10mm per episode.

The soft tissues set the limits. Nerve stretch, muscle contracture and skin tension limit a single lengthening to 10-15mm at most, and the soft-tissue envelope limits the total a prosthesis can deliver, typically 80-100mm. If growth outruns that capacity, the prosthesis may need to be exchanged.

The five modes. Henderson and colleagues reviewed 2,174 tumour endoprostheses across five institutions (534 failures) and defined five primary modes of failure, each stratified by amputation risk and treatment urgency. It is the shared vocabulary examiners expect for endoprosthetic complications, and the one registries use to standardise reporting.

Mapping the complications. The expansion-mechanism breakage shown under Complications is a Type 3 structural failure, and the elbow dislocation after distal humeral reconstruction is a Type 1 soft-tissue failure. The aseptic loosening (20-30%) and deep infection (10-15%) figures are Type 2 and Type 4.

What is commonest depends on where you look. Infection was the most common mode in Henderson's own series, whereas aseptic loosening dominated in the pooled literature. Failure mode depends on anatomic location and on time to failure, so cumulative reporting that lumps all sites together masks site-specific trends: report failures per anatomic location, not in aggregate.

The trial. PARITY was a blinded, multicentre randomised superiority trial across 48 sites in 12 countries. It enrolled patients with a primary bone tumour, a sarcoma invading the femur or tibia, or oligometastatic femoral or tibial disease who required excision and endoprosthetic reconstruction, the population that includes paediatric and adolescent limb-salvage patients. They were randomised to a 1-day or a 5-day postoperative regimen of intravenous cephalosporin (cefazolin or cefuroxime).

The result. Prolonging prophylaxis to 5 days did not reduce surgical site infection: SSI within 1 year occurred in 15.0% of the 5-day group and 16.7% of the 1-day group (hazard ratio 0.93, 95% CI 0.62 to 1.40, P equals 0.73). Antibiotic-related complications were significantly more common with the longer course, 5.1% against 1.6% (hazard ratio 3.24, P equals 0.02).

What to take from it. A short, approximately 24-hour, prophylactic regimen is appropriate; extending antibiotics does not lower infection and causes more harm. An SSI rate of roughly 15-17% in a controlled trial also shows how infection-prone tumour endoprostheses are, consistent with the 10-15% deep-infection figure. PARITY settles the five-days-versus-one question for oncologic endoprostheses; general perioperative infection-prevention bundles (skin preparation, normothermia, timing of the first dose) are covered in the surgical-site-infection topic.

Oncological assessment. The staging imaging is set out under Investigations. Biopsy is by core needle, with the tract planned so that it can be excised with the tumour. The response to neoadjuvant chemotherapy is assessed clinically and radiologically, and the resection is planned for wide margins, a minimum of 2cm from the tumour.

Growth and function. The bone age, limb-length measurement and multiplier calculation predict the discrepancy at skeletal maturity. Document the baseline joint range of motion for later comparison.

The patient and family. Five domains are assessed:

Expandable or not. Five factors decide it.

Radiographs. Full-length views of the affected limb, and a chest radiograph as a baseline.

MRI of the entire bone, so that skip lesions are not missed. It also shows joint involvement, soft-tissue extension and the proximity of the neurovascular structures.

CT of the chest stages pulmonary metastases, and CT of the limb provides 3D planning for a custom prosthesis. Systemic staging is completed with a bone scan or PET-CT per protocol.

Growth assessment. A bone-age radiograph of the left hand and wrist, a scanogram or CT for accurate limb-length measurement, and then the multiplier calculation.

View larger

Four families of mechanism lengthen the implant, and they trade the number of operations against mechanical reliability.

Non-Invasive (Magnetic)Modular (Invasive)Self-Expanding

Why they are preferred. Non-invasive prostheses are the current preferred technology where available, because lengthening becomes an outpatient procedure with no incision. That removes the repeated operations, reduces infection risk and carries a psychological benefit for the child.

What they cost. Magnetic systems have higher mechanical failure rates than modular designs, and traditional designs are MRI incompatible. They are expensive and not available in all centres.

REPIPHYSIS (Wright Medical/Stryker). An external rotating magnetic field (ERC) activates an internal gearbox. Expansion is done as an outpatient without anaesthesia, takes 15-30 minutes and typically gains 4-8mm per session, with up to 100mm of total expansion.

Juvenile Tumour System (JTS, Stanmore Implants). An external electromagnetic field drives an internal motor. The design is modular for customisation, and MRI-conditional designs are now available.

Mechanical Failure

Non-invasive mechanisms have 15-20% mechanical failure rates. Revision to modular prosthesis or exchange may be required. Regular radiographic monitoring for expansion function is essential.

View larger

View larger

View larger

View larger

Stanmore modular system. The traditional approach lengthens the implant by surgically exchanging spacer segments, as day surgery or a short stay, gaining 10-20mm per procedure. It has high mechanical reliability and a proven long-term track record, it is available at most sarcoma centres, and it is MRI compatible. The steps are under Surgical Technique.

The price is repetition. Lengthening every 3-6 months means repeated anaesthetics and a psychological burden for the child, and the infection risk accumulates with each operation.

Cumulative Infection Risk

Each modular expansion procedure carries 2-3% infection risk per operation. With 10+ procedures over growth, cumulative infection risk reaches 20-30%. This is the major advantage of non-invasive systems.

Phenix. Stored mechanical energy in a compressed spring extends the prosthesis by a predetermined amount at each stage, with several stages built in. Neel's early multicentre experience (Evidence) describes it as a non-invasive, spring-loaded electromagnetic device lengthened almost always as an outpatient. The mechanism is simpler than the magnetic systems and each stage is predictable, but total expansion capacity is limited and a spring is less precise than active expansion.

MUTARS (Modular Universal Tumour And Revision System). Available in expandable configurations, it combines modular exchange with anchor mechanisms and is versatile for complex reconstructions.

View larger

View larger

The expandable endoprosthesis is one of several ways to reconstruct a skeletally immature limb, and examiners expect you to compare them and justify your choice.

1. Confirm the oncology first. Staging complete, biopsy tract excisable, wide (R0) margins achievable: the reconstruction is chosen only after oncological clearance is planned (Morrison 2026).
2. Calculate the predicted LLD from the bone age, the multiplier method and the expected growth of the resected physis.
3. Branch on the discrepancy. A predicted LLD of 4cm or less: consider a non-expandable or short modular reconstruction, with or without contralateral epiphysiodesis. Greater than 4cm: an expandable endoprosthesis is indicated.
4. Choose the mechanism. Non-invasive magnetic where available, for fewer procedures and lower infection exposure; modular where reliability and availability dominate, remembering the Ruggieri survival caveat under Controversies.
5. Weigh the alternatives honestly. In the very young child, with poor soft tissues, an irradiated field (Jeys 2007 - postoperative radiotherapy raised infection from 9.8% to 35.3%) or limited follow-up resources, rotationplasty, biological reconstruction or amputation may serve better.
6. Plan the whole pathway at the outset - the expansion schedule to maturity, surveillance and conversion to an adult prosthesis - and document it with the family before the first incision.

Failure of the lengthening mechanism is specific to these implants. The Henderson classification in the next section gives all of them a common vocabulary.

Infection. Management follows the timing:

Single-stage revision is rarely successful.

View larger

Mechanical failure. The failure modes differ by design:

• Non-invasive systems - motor burnout (most common), gear mechanism slippage, telescoping section jamming, electromagnetic coil malfunction
• Modular systems - locking mechanism failure, taper corrosion, component dissociation

Management is revision to an alternative expansion mechanism, conversion to a non-expandable prosthesis if near maturity, or a custom prosthesis for complex failures.

View larger

View larger

Loosening. Aseptic loosening at the stem is why surveillance is lifelong and extends beyond the completion of growth.

View larger

View larger

Soft-tissue failure includes instability and dislocation (Henderson Type 1).

View larger

Growth-related problems. Asymmetric growth produces angular deformity, joint contracture limits function, and the lengthening can over- or under-correct the limb.

The implant is not permanent. Expect conversion to an adult prosthesis at skeletal maturity and multiple revisions over a lifetime.

View larger

View larger

View larger

View larger

The first days. An inpatient stay of days 0-3 for wound monitoring, with:

• DVT prophylaxis, mechanical and pharmacological
• Drain management
• Pain control, PCA then oral
• Non-weight bearing initially

Rehabilitation. Physiotherapy referral starts a four-phase programme that ends in lifelong surveillance.

Conversion to an adult prosthesis. The expandable implant is exchanged at or near skeletal maturity, or earlier when its expansion capacity is exhausted or mechanical failure precludes further expansion. The conversion is an elective revision to a non-expandable prosthesis; it often needs a longer stem for fixation, may need bone grafting of expansion gaps, and is the moment to reassess joint constraint.

View larger

Oncological outcomes. Local recurrence is 5-10% with wide margins and higher with inadequate ones, and surveillance MRI is recommended. Five-year survival is 65-70% for osteosarcoma and 70-75% for Ewing sarcoma, and the response to chemotherapy is the key prognostic factor.

Implant survival at 5, 10 and 15 years:

Function. Mean Musculoskeletal Tumor Society scores in the paediatric systematic review ranged from 71.0% to 86.8%, depending on reconstruction type and anatomical site. MSTS is surgeon-rated, not a patient-reported outcome. Scores are lower than for non-expandable prostheses in adults, and are affected by the multiple procedures. Limb salvage succeeds in 85-90%, with some activity restriction recommended and high-impact sports avoided.

Quality of life. Generally good when limb salvage succeeds, though repeated procedures have a psychological impact and support services are important.

Global Epidemiology

Primary bone sarcomas of childhood and adolescence are rare worldwide (approximately 8-9 per million children/adolescents annually). Osteosarcoma and Ewing sarcoma dominate, with peak incidence during the pubertal growth spurt. The proportion needing a growing reconstruction is driven by age at presentation and tumour site - distal femoral and proximal tibial lesions most often require growth-accommodating implants because they straddle the most active lower-limb physes.

Side-by-Side Guidance

Across all major bodies the principles converge: referral to a specialist sarcoma centre, MDT decision-making, R0 resection before reconstruction, and shared decision-making with family. Genuine differences are organisational (network structure) rather than oncological.

Registry and Outcome Notes

There is no dedicated international registry for paediatric expandable prostheses; evidence is dominated by single-centre and multi-institutional case series (Level IV). Adult tumour-endoprosthesis registries and the Henderson failure-mode classification (soft tissue, aseptic loosening, structural, infection, tumour progression) are widely applied to standardise reporting. Collaborative groups (e.g. PARITY trial network for infection prophylaxis) are improving the evidence base.

High- vs Limited-Resource Practice Variation

• Well-resourced centres: Non-invasive magnetic/electromagnetic systems (REPIPHYSIS, JTS, Stanmore non-invasive) with outpatient lengthening; 3D-printed custom implants; routine MDT and long-term surveillance.
• Limited-resource settings: Higher reliance on modular/minimally invasive expansion or non-expandable reconstruction; rotationplasty and amputation remain important, durable, cost-effective options where growing implants or repeated lengthening logistics are not feasible.
• Universal principle: Implant choice should match the local capacity for repeated follow-up, expansion, and revision - a sophisticated implant without reliable surveillance can underperform a simpler, durable reconstruction.

Non-invasive versus modular survival. Non-invasive systems reduce surgical episodes and infection exposure, yet some series (e.g. Ruggieri, where the open Kotz outperformed the Repiphysis) report lower implant survival for early magnetic devices. The optimal mechanism remains unsettled and is centre- and era-dependent.

Growth prediction. The multiplier method assumes normal growth, which treatment can blunt (Growth Prediction Uncertainty, above). Over- or under-estimation of the final LLD is common, and how aggressively to oversize expansion capacity is debated.

Cemented or cementless. Cementless or compress fixation may improve long-term survival in young active patients, but the evidence is limited and stem fixation in the immature skeleton is not standardised.

When to stop expanding. There is no consensus on the LLD to accept against the risks of repeated lengthening (nerve palsy, contracture, mechanical failure). Some advocate accepting a minor LLD with a shoe raise over pushing for every millimetre.

What the literature cannot yet answer.

• No randomised data compare expandable endoprosthesis with rotationplasty or biological reconstruction; all comparisons are Level IV.
• Patient-reported and long-term quality-of-life outcomes are inconsistently reported, limiting true device comparison.
• PARITY answers five days against one, but it was a largely adult trial that tested no shorter regimen, so the optimal prophylaxis duration for these high-risk implants in children is still being defined.