Overview & Rationale
Expandable (growing) prostheses are specialised endoprosthetic devices designed for skeletally immature children undergoing limb salvage surgery after resection of primary malignant bone tumours — most commonly osteosarcoma and Ewing sarcoma. The fundamental challenge in paediatric orthopaedic oncology is that radical resection of a tumour involving or adjacent to a growth plate will cause progressive limb length discrepancy (LLD) as the contralateral limb continues to grow. A fixed-length prosthesis implanted in a young child will result in a leg length discrepancy of several centimetres by skeletal maturity — causing gait disturbance, scoliosis, and functional impairment. Expandable prostheses address this by incorporating a mechanism that allows controlled lengthening of the implant over time, compensating for the growth of the contralateral limb.
- Historical development: the first expandable prosthesis for paediatric bone tumours was introduced in 1975 by Professor John T Scales at the Birmingham Bone Tumour Service; early designs required repeated open surgical procedures under general anaesthesia to lengthen the prosthesis — each `conventional` lengthening required a surgical incision over the prosthesis, engagement of the lengthening mechanism, and closure; this meant children required multiple anaesthetics over their growing years; the shift to non-invasive (electromagnetic) lengthening has been one of the most significant advances in paediatric oncological orthopaedics
- Tumour epidemiology in children: osteosarcoma is the most common primary malignant bone tumour in children and adolescents, with a peak incidence in the second decade; the most frequent sites are the distal femur (~40%) and proximal tibia (~20%) — precisely the periarticular regions that include the most active growth plates; Ewing sarcoma has a broader diaphyseal distribution but also commonly affects the femur and tibia; both tumours respond to neoadjuvant chemotherapy, enabling limb salvage in approximately 80–90% of cases
- Indications for expandable prosthesis: skeletal immaturity (open physes; significant predicted growth remaining — minimum 3–4 cm predicted LLD as a threshold in most centres); tumour requiring resection of the physis (growth plate) with adequate bone stock for stem fixation; non-metastatic disease (two-thirds of surgeons decline expandable prostheses in the presence of metastatic disease due to expected shorter survival); age typically >6 years (minimum median age in published series approximately 6.5 years — below this, bone stock is insufficient for prosthetic fixation and alternative reconstruction is preferred); primary tumour at the distal femur or proximal tibia (most common; soft tissue coverage is adequate; proximal femur and humerus prostheses also exist but with more limited data)
Types of Expandable Prostheses
| Type | Mechanism | Examples | Advantages / Disadvantages |
|---|---|---|---|
| Conventional (surgical) expandable | Surgical incision required to engage the lengthening mechanism; a key or Allen wrench unlocks a ratchet and the prosthesis is manually distracted | Historical Kotz/Repiphysis designs; early Stanmore prostheses | Proven long-term data (Kotz design — 27 years experience at Vienna); requires repeated surgical procedures under GA; infection risk at each lengthening; largely superseded by non-invasive designs |
| Non-invasive electromagnetic (NIE) | An internal electric motor (powered by electromagnetic induction from an external coil placed over the prosthesis) rotates a power screw within the prosthesis, telescopically extending the implant; no skin incision required; performed in outpatient clinic; typical session adds 2–5 mm per visit | Stanmore JTS (Juvenile Tumour System — Stanmore Implants Ltd); MUTARS (Implantcast); Phenix (Lépine); Repiphysis (Wright Medical — electromagnetic version) | Eliminates repeated GA; outpatient clinic procedure; reduced infection risk per lengthening; allows more frequent, smaller incremental lengthenings; titanium alloy body with NdFeB rare-earth magnetic disc; gearbox in stainless steel; HA collar at bone-prosthesis junction |
| Bioexpandable (distraction osteogenesis) | A modular distraction nail is integrated into the prosthesis; sequential electromagnetic activation lengthens a remaining bone segment by callotasis (distraction osteogenesis), rather than telescoping the implant body itself; as the bone segment grows, the proportion of native bone increases | Fitbone (Wittenstein); Baumgart bioexpandable system | As bone proportion increases over time, lever arm forces on the prosthesis decrease — potentially improving long-term durability; technically demanding; less widely used than NIE designs |
Calculating Required Lengthening
- Growth prediction — the Moseley straight-line method and Anderson-Green-Messner charts: the expected limb length discrepancy at skeletal maturity is calculated based on the child`s age, gender, current LLD, and the growth contribution of the affected physis; the distal femoral physis contributes approximately 9–10 mm/year (37% of total lower limb growth) and the proximal tibial physis approximately 6–7 mm/year (28%); for a child of 8 years with a distal femoral resection (removing the distal femoral physis), the expected LLD at maturity is approximately 7–8 years × 9 mm/year = ~63–72 mm = 6.3–7.2 cm; this is the total lengthening target for the expandable prosthesis
- Teleoradiograph / scanogram: a full-length standing lower limb scanogram (or EOS imaging system) is performed at regular intervals post-operatively; it measures current LLD precisely; the lengthening schedule is adjusted to keep pace with contralateral limb growth; most non-invasive prostheses can be lengthened by 2–5 mm per session in the outpatient clinic; the frequency of lengthening is tailored to the child`s growth velocity (more frequent sessions during growth spurts — puberty)
- Maximum extension capacity: non-invasive expandable prostheses have a maximum extension range of approximately 60–80 mm total; if the predicted LLD exceeds the prosthesis extension capacity, a larger initial prosthesis body or a `double-length` extension module must be planned; in practice, children implanted at 6–8 years with distal femoral resections require the maximum extension capacity of the device
Complications — Henderson Classification
| Henderson Type | Failure Mode | Management |
|---|---|---|
| Type I — Soft tissue failure | Wound dehiscence, extensor mechanism failure (proximal tibial resections), poor soft tissue coverage, aseptic wound problems | Soft tissue reconstruction; gastrocnemius flap for proximal tibial coverage; extensor mechanism repair |
| Type II — Aseptic loosening | Mechanical failure of the bone-stem interface; progressive radiolucency on X-ray; stem loosening | Stem revision with longer/larger stem; cement augmentation |
| Type III — Structural failure | Mechanical failure of the prosthesis itself — stem fracture, hinge failure, bearing wear; gearbox jamming in non-invasive prostheses | Component replacement or revision prosthesis |
| Type IV — Infection | Deep periprosthetic infection; the most common reason for failure in the Stanmore JTS series (deep infection in 10 of 51 prostheses at 3–5 year follow-up — Gilg et al.); more common in proximal tibial resections (poor soft tissue coverage) | Two-stage revision (antibiotic cement spacer + reimplantation) or amputation in severe/refractory cases |
| Type V — Tumour progression | Local recurrence or distant metastasis necessitating change in surgical management | Extended resection; amputation if local recurrence cannot be controlled |
Outcomes & Alternatives
- Survival data: 3-year revision-free survival approximately 81.7% and 5-year approximately 61.6% in the largest published series (Gilg et al., 51 JTS prostheses); MSTS (Musculoskeletal Tumour Society) functional scores range from 52–96% (mean approximately 80–84%) in published series; functional outcomes are broadly comparable to other limb salvage methods (biological reconstruction, allograft-prosthesis composite) with the advantage of immediate stable fixation and preserved joint function
- Rotationplasty: the `Van Nes rotationplasty` — the distal limb is rotated 180° and reattached, so that the ankle joint functions as a knee; provides an excellent functional outcome and a biological reconstruction with no implant-related complications; traditionally rejected by patients for cosmetic reasons, but functional scores (including for sports participation) are often superior to prosthetic reconstruction; increasingly used in very young children (<6 years) and for cases with failed implant-related infection
- Joint-sparing surgery: where the tumour does not involve the physis and does not reach the joint, joint-sparing intercalary resection with a custom expandable intercalary prosthesis (bridging the diaphyseal defect, retaining both ends of the bone and both joints) preserves the growth plate and provides the best functional outcome with lowest risk of joint complications; this approach requires precise preoperative planning with MRI to confirm adequate tumour-to-physis distance (>2 cm safe margin in most protocols) and intraoperative navigation for bone cuts
- Biological reconstruction alternatives: vascularised fibula free flap graft (VFF) — autologous vascularised fibula to bridge the diaphyseal defect; undergoes hypertrophy with loading over time; no implant-related infection risk; technically demanding; allograft-prosthesis composite (APC) — structural allograft provides bone stock while a prosthesis provides the joint; less commonly used in paediatric oncology due to infection and fracture risk in immunocompromised chemotherapy patients
Exam Pearls
- Expandable prosthesis rationale: skeletal immaturity + physis resection → progressive LLD; expandable prosthesis compensates for contralateral growth; minimum 3–4 cm predicted LLD as threshold; minimum age ~6–7 years
- Non-invasive NIE prosthesis (Stanmore JTS): electromagnetic coil activates internal motor → power screw extends telescopic body; outpatient clinic; no GA; 2–5 mm per session; titanium alloy body; NdFeB magnet; HA collar at bone junction
- Growth calculation: distal femoral physis 9–10 mm/year (37%); proximal tibial physis 6–7 mm/year (28%); use Moseley straight-line method or Anderson-Green charts; serial scanograms to monitor LLD and schedule lengthening; maximum extension ~60–80 mm total
- Henderson classification: Type I (soft tissue); Type II (aseptic loosening); Type III (structural failure — stem fracture, gearbox jam); Type IV (infection — most common in JTS series); Type V (tumour progression)
- Proximal tibial resection highest infection risk: poor soft tissue coverage; gastrocnemius flap recommended for coverage at index surgery; non-invasive prosthesis reduces per-lengthening infection risk vs conventional surgical expandable
- 5-year revision-free survival ~61.6% (Gilg et al.); MSTS scores 80–84% mean; infection most common failure (Type IV); outcomes broadly comparable to other limb salvage methods
- Rotationplasty: Van Nes; 180° rotation of distal limb; ankle acts as knee; biological, no implant; superior functional scores for sports; cosmetically challenging; preferred in very young (<6 years) and failed infected prostheses
- Joint-sparing intercalary resection: where tumour does not involve physis; preserves growth plate; custom intercalary expandable prosthesis; requires >2 cm tumour-to-physis safe margin on MRI; lowest LLD risk; best functional outcome when feasible