Endoprosthetic Reconstruction (Tumour Megaprosthesis)

Indication. Segmental reconstruction of a long bone after wide resection for a primary bone sarcoma, for aggressive benign or metastatic disease with massive bone loss, or for a failed prior reconstruction. The reconstruction is always secondary to an oncologically adequate wide (Enneking) margin — the operation is planned from the tumour outwards. Primary bone sarcoma — the classic indications in children and young adults: osteosarcoma and Ewing sarcoma (reconstruction follows wide resection after neoadjuvant chemotherapy), chondrosarcoma (surgical resection is the mainstay as it is chemo- and radio-resistant), and high-grade spindle-cell sarcoma of bone or dedifferentiated chondrosarcoma. Aggressive benign or metastatic disease — metastatic bone disease with extensive bone loss or an impending or pathological fracture not amenable to internal fixation (especially a solitary renal or thyroid metastasis with a good prognosis), aggressive giant cell tumour with joint destruction unsuitable for curettage, and failed prior reconstruction with massive bone loss (revision megaprosthesis). Common anatomical sites — distal femur (the commonest site, rotating-hinge knee), proximal femur (bipolar or total-hip articulation; abductor reattachment critical), proximal tibia (highest infection risk; extensor mechanism and gastrocnemius flap mandatory), proximal humerus (commonest upper-limb site; shoulder stability and abduction limited), and total femur for massive disease or a failed prior segmental reconstruction. The one decision that matters — what do you rebuild with? The choice balances immediate function (favours an endoprosthesis) against long-term biological durability (favours allograft or biological options), modified by patient age, oncological prognosis, defect location and surgeon expertise.

Pre-operative planning is the operation on paper. Staging and imaging: a whole-bone MRI defines intramedullary tumour extent, the soft-tissue mass, neurovascular involvement and skip lesions, and the resection length is planned from this; CT chest (pulmonary metastases) and bone scan or PET-CT complete systemic staging; plain radiographs of the whole bone are used for templating and contralateral length comparison; the biopsy is image-guided (core or open) with the tract in line with the definitive incision and planned for en-bloc excision. Templating and implant selection: measure the resection length from the planned osteotomy (a wide margin from the tumour edge on MRI), restore limb length and joint offset templated against the contralateral limb, order a modular system so segment length and articulation can be adjusted intra-operatively, and plan fixation (cemented versus uncemented stem, compress-type fixation, extracortical bridging). Consent and setup. Positioning and approach follow the resection: an extensile utility approach incorporating the biopsy tract, allowing wide en-bloc resection and protection of the major neurovascular bundle. The limb-salvage decision is made at the multidisciplinary tumour board, and the honest discussion of limb salvage versus amputation — or, in a child with a distal femoral tumour, rotationplasty — is mandatory.

The goal is an oncologically adequate wide resection with en-bloc excision of the biopsy tract, protection of the major neurovascular bundle, and a modular reconstruction that restores limb length and offset with stable fixation — then a soft-tissue reconstruction that dictates the functional result. The exposure and resection are laid out in the timeline below; the site-specific detail and the dangers at each site follow it.

View larger

The general sequence is common to every site, but the articulation, the key soft-tissue problem and the danger structures differ at each. The four common sites are laid out below. Distal femur — the commonest site. A rotating-hinge knee tolerates the loss of the collateral ligaments after resection. The extensor mechanism is usually preservable, so protect the quadriceps and patellar tendon. Cemented or uncemented femoral and tibial stems are used, with extracortical bone bridging or an HA collar at the bone-implant junction to reduce aseptic loosening; reattach the gastrocnemius and soft tissues and close over a drain.

View larger

Proximal femur. A bipolar or total hip articulation is used; a dual-mobility or constrained liner reduces dislocation given the loss of the abductors and capsule. Reattach the abductors to the prosthesis via a trochanteric attachment device or synthetic mesh, and consider capsular reconstruction. Restore offset and limb length to tension the soft tissues and stabilise the hip. A cemented or uncemented stem runs to the femoral diaphysis.

View larger

Proximal tibia — the highest-risk site. It has the highest infection rate of all sites because of the thin anteromedial soft-tissue envelope and frequent prior radiotherapy. Extensor mechanism reconstruction is mandatory: reattach the patellar tendon to the prosthesis or a tendon anchor, then rotate a medial gastrocnemius flap to reinforce the repair and provide durable soft-tissue coverage. A rotating-hinge knee is used; protect the common peroneal nerve laterally.

View larger

Proximal humerus — the commonest upper-limb site. Resection sacrifices the rotator cuff and deltoid attachments and the capsule. Articulation options are hemiarthroplasty, a reverse total shoulder (if the deltoid and axillary nerve are intact), or a constrained spacer, with a suspension reconstruction using mesh or tendon. Active abduction is usually limited, but hand, wrist and elbow function are preserved, which is the main functional goal of upper-limb salvage. Protect the axillary and radial nerves and the brachial plexus and axillary vessels.

Fixation and growing prostheses. The fixation strategy is chosen for the patient and the site: cemented stems give immediate fixation and early weight-bearing and suit older or metastatic patients and the post-chemotherapy setting; uncemented stems give biological fixation, with extracortical bone bridging and hydroxyapatite collars at the transcortical junction promoting bone ongrowth and reducing aseptic loosening; compress-type fixation applies controlled compression across the bone-implant interface to stimulate cortical hypertrophy and osseointegration and is useful in young patients. In skeletally immature children a growing (expandable) prosthesis maintains limb-length equality (lengthened non-invasively by electromagnetics or with minor procedures) at the cost of higher mechanical-failure and reoperation rates. Silver-coated megaprostheses provide an antimicrobial surface that reduces periprosthetic infection, especially at high-risk sites.

Rehabilitation. In the lower limb (distal femur, proximal femur, proximal tibia), mobilise early when fixation allows — immediately with cemented stems, and protected weight-bearing with biological fixation. At the proximal tibia, protect the extensor reconstruction in an extension brace and delay active knee flexion until the patellar tendon and gastrocnemius repair is secure. At the proximal femur, observe hip precautions and abductor protection and expect a period of abductor lurch. Co-ordinate rehabilitation with the chemotherapy schedule, because wound healing and infection risk are higher during myelosuppression. In the upper limb (proximal humerus), use sling immobilisation initially to protect the suspension reconstruction, focus rehabilitation on the preserved hand, wrist and elbow function, accept limited shoulder abduction, and begin pendulum and assisted-range exercises as the soft-tissue reconstruction allows. Complications — the Henderson failure modes. The Henderson classification is the universal language for reporting megaprosthesis failure and frames the differential when a reconstruction fails.

Modes 1 to 3 are mechanical; modes 4 and 5 are non-mechanical. Infection (Type 4) and tumour progression (Type 5) are the modes most likely to end in amputation. The proximal tibia has the highest combined failure rate, driven by infection and extensor mechanism (Type 1) failure.

These are the critical decisions, danger structures and exam traps that recur in the limb-salvage viva.

Setting the reconstruction in context. Endoprosthetic reconstruction is one limb of the reconstructive spectrum after segmental bone resection. The choice between it and the biological options (osteoarticular allograft, allograft-prosthetic composite, vascularised fibula) turns on the balance between immediate, reproducible function and long-term biological durability, weighed against patient age, oncological prognosis, the defect location and surgeon expertise. Functional outcomes (MSTS). The Musculoskeletal Tumour Society (MSTS) score is the standard functional outcome measure; its domains include pain, function, emotional acceptance, supports, walking and gait (lower limb), and hand positioning, dexterity and lifting (upper limb). Limb-salvage survival at 5 to 10 years is acceptable but punctuated by reoperations, and many young patients undergo multiple revisions over a lifetime. By site, the distal and proximal femur tend to achieve the best functional scores, the proximal tibia is limited by extensor lag, and the proximal humerus preserves hand and elbow function but has limited shoulder abduction. Endoprosthesis versus amputation and rotationplasty. Limb salvage generally gives equal or better function than amputation in selected patients, but at the cost of more reoperations and complications. Rotationplasty (in young children with distal femoral tumours) converts the ankle into a knee, giving a durable, high-function, low-maintenance reconstruction that tolerates high activity and growth, and is chosen where a growing prosthesis would face years of reoperations. Amputation remains appropriate when a wide margin cannot preserve a functional, sensate limb, after failed salvage, or for uncontrolled infection, and gives reliable, durable function. Critically, oncological survival is equivalent between limb salvage and amputation when margins are adequate — the choice is functional, not oncological.

Key evidence. The Henderson classification (2011) defined the five universal failure modes and was extended (2014) to biological and expandable reconstructions; it is the standard reporting framework. Jeys (2005) quantified periprosthetic infection at 11% in 1240 tumour prostheses with amputation in 37% of infected cases; Schmolders (2016) reported a lower infection rate with silver-coated lower-limb megaprostheses. Zhang (2018) showed the distal femur outperforms the proximal tibia for implant survival (86.1% versus 66.9% at five years) and validated patellar-tendon reattachment with a medial gastrocnemius flap, with a mean MSTS of 22.9 of 30. Enneking (1993) described the MSTS scoring system.