A limb-salvage reconstruction for near-global femoral loss when segmental reconstruction cannot provide a useful femur
- Total femur replacement is an extreme salvage operation, not a routine extension of distal or proximal femoral replacement. Half are revised by five years in the non-oncologic setting, and 34 per cent of all patients across 35 pooled studies reach a defined failure.
- The hip and knee are reconstructed together, but THE HIP IS THE END THAT FAILS - dislocation was the commonest complication at 37 per cent in the largest UK series, and instability plus infection accounted for nine of 12 revisions at Mayo.
- Reconstruct the abductors, capsule, extensor mechanism and collateral or muscle stabilisers at the index operation. Bickels reported 1.7 per cent dislocation with acetabular preservation, a bipolar articulation, capsulorrhaphy over the neck and formal abductor reattachment - but in oncologic patients with native acetabula, so do not expect that number in a scarred revision hip.
- Active infection and poor soft tissue must be treated before definitive replacement; large implants magnify dead-space and biofilm risk. Infection failure runs at 18 per cent in non-oncologic against 8 per cent in oncologic indications.
- Counselling must compare total femur replacement with fusion, hip disarticulation or transfemoral amputation - and must state that DISARTICULATION IS ALSO AN OUTCOME OF THE OPERATION, with a cumulative incidence of 10 per cent by five years.
- “The key indication is near-global femoral loss with a viable limb and a patient for whom a stable reconstructed hip and knee may be more useful than amputation.
- “A total femur prosthesis does not recreate native muscle attachments; use a mesh, flap or muscle sling. Absence of a mesh graft was associated with a higher risk of hip disarticulation in the EMSOS multicentre cohort, though the two hazard models in that paper disagree on the strength of the effect.
- “The common clinical endpoint is a brace-dependent, walking-aid-dependent limb, not a normal knee - about 70 per cent use a walking aid and the mean MSTS score is 66 per cent.
- “Do not sacrifice a functioning distal limb to preserve a reconstruction that cannot be covered or stabilised.
Definition and Indications
Total femur replacement removes and replaces most or all of the femur with a modular endoprosthesis, usually incorporating a proximal femoral or total hip module and a distal femoral or rotating-hinge knee module. It is used most often for extensive tumour resection, but can be considered in non-oncologic salvage after failed infected arthroplasty, massive bone loss, multiple periprosthetic fractures, failed allograft or catastrophic nonunion.
Indications. The operation is considered for:
- Near-global femoral loss with no credible segmental fixation or biological reconstruction.
- Repeated failure of proximal and distal femoral salvage in a patient with a viable distal limb.
- Extensive non-oncologic destruction after infection, only after a staged eradication and coverage plan.
- Selected tumour reconstructions planned by a specialist sarcoma team with oncological margins and systemic treatment integrated.
When to amputate instead. The contraindications, or reasons for amputation, are:
- Active uncontrolled infection or a sinus
- A dysvascular or non-viable limb
- No functional ankle and foot, or a severe neurologic deficit
- A non-reconstructable hip abductor and knee extensor mechanism with no credible soft-tissue substitute
- Inability to cover the implant with viable muscle or flap
- Medical or social inability to tolerate prolonged rehabilitation and repeated revision
- A patient whose functional goals are better served by transfemoral amputation or hip disarticulation
Remove infected implants and non-viable tissue, obtain cultures, manage dead space and obtain durable coverage. A single-stage total femur in uncontrolled infection is usually a limb-threatening gamble.
Amputation is a real alternative. Compare the expected walking, sitting, pain, energy cost, prosthetic potential, wound burden and future operations of each route. Limb salvage alone is not the endpoint.
Assessment and Team Planning
The patient. Document prior procedures, infection organisms, flap history, radiation, vascular operations, fractures, limb length, pain, walking aids, transfers and the patient's expectations. Ask whether the patient can sit, use a brace, protect weight bearing and attend frequent follow-up.
The limb. Map scars, sinuses, skin mobility, muscle viability and anticipated coverage, then test and record baseline function:
- Hip abductors, flexors and extensors
- Quadriceps, patellar tendon and collateral stability
- Ankle and foot function
- Sciatic, femoral and common peroneal nerve function, and pulses
Examine the opposite limb and the spine as well, because these determine rehabilitation.
The team. A total femur case should include adult reconstruction or tumour surgery, infection, plastic surgery, vascular surgery when indicated, anaesthesia, rehabilitation, physiotherapy and orthotics. Before incision the team agrees the resection levels, implant, bearing, mechanism repair, flap, antimicrobial plan, thrombosis prevention, blood strategy and the fallback to amputation.
Anatomy and Biomechanics
The total femur has no native periosteal healing interface. Stability and function come from the prosthesis, its fixation and the soft-tissue reconstruction, so infection and soft-tissue failure are especially consequential.
The hip. An abductor mechanism, capsule or synthetic sling is needed to centre the proximal module and limit dislocation. Abductor and capsular reconstruction is what controls the proximal joint, and it needs planned attachment points.
The knee. A rotating hinge or another constrained bearing is required when the native cruciates and collateral structures are absent. The quadriceps and patellar tendon determine active extension: their failure produces an extensor lag even when the hinge is mechanically intact. Quadriceps, patellar tendon and collateral reconstruction controls the knee, and like the hip it needs planned attachment points.
Nerves and vessels. Previous infection, scars and repeated lengthening bring the sciatic, femoral and peroneal nerves and the major vessels close to the operative field. The sciatic and femoral nerves can be shortened, scarred or exposed after previous surgery, so length restoration must be gradual and planned, and excessive lengthening avoided. The femoral and popliteal vessels may be adherent to scar, cement or tumour bed; obtain vascular imaging when their course is uncertain.
Cover and the distal link. The thigh compartments and skin must provide vascularised coverage over a long implant and its dead space. The distal tibia, ankle and foot become the only distal load-bearing link, so a plantigrade foot and adequate ankle function are essential.
Imaging and Planning
Obtain AP and lateral radiographs of the entire femur and the adjacent hip and knee, long-leg alignment views when possible, and images of the opposite side for length and joint-level planning. CT defines bone loss, old hardware, and canal and vessel relationships. CT angiography is indicated for prior vascular surgery, absent pulses, scarred vessels or a flap plan.
The plan. The implant should be templated to a useful, stable limb, not simply to the length of the removed femur, and a small planned shortening can be safer than excessive nerve tension. Before theatre the plan should settle:
- Hip centre, offset, version and leg length
- Knee joint line, patellar height and the tibial tray relationship
- Proximal and distal fixation or cement mantle
- Extensor, abductor, capsule and collateral attachments
- Hinge or dual-mobility bearing
- Limb rotation and foot progression
- Soft-tissue flap and dead-space management
- Extraction and staged infection pathways
Operative Technique
Total femur replacement sequence
Use lateral or supine positioning according to the approach, secure the pelvis and leave the foot visible for rotation, length and plantigrade checks. Prep from above the iliac crest to the foot, including potential flap and vascular fields.
Coordinate blood conservation, cell salvage, neuromonitoring, urinary catheterisation and postoperative intensive care if required.
Confirm AP and lateral imaging of the hip, total femur and knee, and have modular proximal, shaft, distal and hinge components available with backup sizes. Verify implant side, length, hip centre, knee joint line and tibial component compatibility against the plan.
Keep a temporary spacer or amputation pathway available in case the resection reveals active infection or non-viable tissue.
Give antibiotics after multiple deep cultures when the patient is stable; treat sepsis immediately and follow the infection plan. Excise sinus tracts and contaminated scar en bloc, and prepare the flap, blood and vascular-control strategy.
Confirm whether the case is definitive or a staged reconstruction, and define the endpoint for proceeding.
Use a planned extensile incision that preserves viable skin bridges and permits muscle or flap coverage. Identify the sciatic, femoral and peroneal nerve pathways and the major vessels before resection.
Preserve or mark the abductors, capsule, quadriceps, patellar tendon, collateral and muscle attachments for later repair.
Remove implants, cement and non-viable bone while keeping control of the proximal hip and the distal knee or tibial components. Obtain multiple deep samples and histology, debride to bleeding viable tissue and manage dead space.
Control the femoral vessels, or involve vascular surgery, before dividing scarred structures near a previous graft or flap.
Prepare the acetabulum, or retain a stable acetabular component only when infection and compatibility permit, and prepare the tibial canal or distal fixation for the hinge component. Choose cemented or uncemented fixation according to remaining bone, implant system and infection status, and bypass stress risers and old holes.
Prepare mesh, allograft, tendon graft or muscle sling attachment points for the mechanisms and capsule.
Assemble trial proximal, shaft and distal modules, and check hip centre, length, offset, rotation, knee joint line, patellar tracking and foot progression. Test hip stability and hinge motion through flexion and extension, and confirm that the sciatic and femoral nerves are not under excessive tension.
Compare sitting, limb clearance and the expected gait, and adjust modules before fixation.
Avoid forceful traction, release retractors frequently and check distal perfusion. Keep the hinge and stems away from the neurovascular bundle and maintain a vascularised layer over the metal.
If coverage is not possible, stop and revise the plan: a bare implant is not a closed operation.
Fix the proximal and distal components with the planned cement or press-fit technique, and lock the modular junctions once the mechanics are confirmed. Reconstruct the abductors and capsule around the hip, and the quadriceps, patellar tendon, collateral or extensor mechanism around the knee, using mesh, graft or strong sutures.
Place the bearing or rotating hinge, confirm stable motion and close dead space with vascularised muscle or flap.
Obtain AP and lateral images of the complete implant and adjacent joints. Check hip centre, length, rotation, knee joint line, stem fixation, hinge position, patellar tracking and fracture.
Reassess pulses, nerve function and soft-tissue tension before closure.
Close muscle, fascia and skin over durable coverage without tension, and apply a brace or immobiliser that protects the repaired mechanisms. Continue antimicrobials, thrombosis prevention, flap monitoring and serial imaging.
Complications and Failure Management
Report failures in Henderson types. Never report them as one pooled number. The five modes are:
- Type 1, soft tissue
- Type 2, aseptic loosening
- Type 3, structural
- Type 4, infection
- Type 5, tumour progression
The 2014 ISOLS revision adds subclassification, a biological-reconstruction system and a paediatric category.
Why one number misleads. Henderson's own analysis of 2,174 patients found mode of failure statistically dependent on anatomical location and on time, which is precisely why a single "total femur failure rate" is misleading. The two ends of this implant fail differently: the hip early and by instability, the knee later and by loosening or structural failure.
The rates to expect. The pooled failure rate is 34 per cent across 1,002 patients, but it splits by indication: infection failure runs at 18 per cent in non-oncologic against 8 per cent in oncologic cases. In the largest UK non-oncologic series, dislocation occurred in 37 per cent, 66 per cent suffered at least one complication and 55 per cent needed at least one further operation. Cumulative implant revision is 24 per cent at two years and 35 per cent at five; cumulative hip disarticulation is 4 per cent at two years and 10 per cent at five.
- Clues
- Drainage, sinus, fever, abscess or painful loosening
- Response
- Cultures, imaging, debridement and specialist review
- Salvage
- Staged implant removal/reconstruction or amputation when eradication fails
- Clues
- Early instability or recurrent reduction
- Response
- Reduce, image and assess abductors/capsule/hip centre
- Salvage
- Mechanism reconstruction, bearing or component revision; consider salvage if repeated
- Clues
- Extensor lag, patellar instability or inability to straight-leg raise
- Response
- Brace and assess repair and hinge
- Salvage
- Mesh or tendon reconstruction, revision and prolonged protection
- Clues
- Acute pain, deformity, stem or hinge failure
- Response
- Protect limb and image full implant
- Salvage
- Fix fracture or revise modules and bypass stress risers
- Clues
- Foot drop, sensory loss, femoral weakness
- Response
- Document, release compression, image and assess length
- Salvage
- Remove causative hardware or revise excessive length; rehabilitate
- Clues
- Necrosis, dehiscence, exposed metal
- Response
- Urgent plastic-surgical and infection assessment
- Salvage
- Debride and obtain vascularised coverage or remove the implant
Rehabilitation and Functional Expectations
A staged recovery. Rehabilitation is prolonged and staged. Early goals are wound and flap protection, pain control, early safe transfers, ankle and foot motion, respiratory and thrombosis prevention, and preservation of the opposite limb.
Weight bearing is protected and physiotherapy follows a staged plan; avoid aggressive hip abduction or knee flexion until the repairs heal. A brace may protect the hip or knee mechanisms for weeks, followed by graded range of motion and gait training.
What success looks like. The patient may require a walking aid, brace, shoe modification and long-term surveillance. Functional success is a pain-controlled, covered, stable limb that supports transfers and ambulation; normal quadriceps power, hip abductor strength and proprioception should not be promised.
Guidelines, Registries & Global Practice
Global evidence. Total femur replacement is rare, and most evidence comes from oncological series, small non-oncologic cohorts and case reports. Outcomes are highly selected and depend on infection, soft tissue, mechanism reconstruction and rehabilitation. There is no universal functional advantage over amputation.
Practice principles.
- Use total femur replacement only when segmental alternatives are not credible and the distal limb is useful.
- Stage active infection and obtain durable coverage before definitive implantation.
- Plan both hip and knee stability, plus abductor and extensor reconstruction.
- Protect nerves and vessels during length restoration and monitor flap/soft tissue closely.
- Compare with fusion, hip disarticulation and transfemoral amputation in a shared functional decision.
Registries. Arthroplasty registries rarely separate total femur replacement from other megaprostheses. Specialist tumour and limb-reconstruction databases should record indication, infection, mechanism reconstruction, bearing, flap, revision and amputation.
Global practice. Custom and modular total femur implants require specialist manufacturing and surgical support. When unavailable, staged fusion or amputation may be more reliable. The safe endpoint is a viable, covered and useful limb, not a particular implant.
MCQ Practice Points
Q: When is a total femur replacement considered, and what changes the risk most?
A: When near-global femoral loss cannot be reconstructed segmentally, the hip, distal limb, neurovascular supply and soft tissues can support a useful limb, and the patient accepts a high complication burden. The indication itself is the strongest risk stratifier. In the Lari meta-analysis of 1,002 patients, infection (Henderson type 4) failure was 18 per cent in non-oncologic against 8 per cent in oncologic patients - roughly double - while functional scores were no different. So the multiply revised arthroplasty patient and the sarcoma patient are being offered the same operation with materially different risk.
Q: What survivorship should a non-oncologic patient be given?
A: Half of these reconstructions are revised by five years. The Mayo series of 24 gives survival free of any revision of 66 per cent at two years, 51 per cent at five and 34 per cent at ten. The Sheffield series of 38 gives a 66 per cent complication rate and 55 per cent requiring at least one further operation. And the reconstruction can itself end in amputation: the EMSOS multicentre study of 143 patients found a cumulative incidence of hip disarticulation of 4 per cent at two years and 10 per cent at five. Disarticulation is not only the alternative discussed beforehand - it is an outcome of the operation.
Q: Which end of a total femur replacement causes the trouble, and by what mechanism?
A: The hip, and by instability. Dislocation was the commonest complication in the Sheffield series at 37 per cent, and five of the 12 revisions in the Mayo series were for hip dislocation with four more for infection. The knee end fails later and differently - by aseptic loosening and structural failure. Henderson's warning follows directly: do not report or think about a pooled "total femur failure rate", because the two ends fail by different mechanisms at different times and a single number hides both.
Q: What must be reconstructed beyond the bone, and is there evidence it matters?
A: Hip abductors, capsule, quadriceps, patellar tendon and collateral or muscle attachments - bone replacement alone does not restore function. There is quantitative support for the hip half. Bickels reported 1.7 per cent dislocation in 57 patients using acetabular preservation, a bipolar articulation, Dacron capsulorrhaphy over the prosthetic neck and formal abductor reattachment - against 37 per cent in a modern non-oncologic series. The EMSOS study found absence of a mesh graft associated with higher risk of hip disarticulation. Read the difference cautiously: Bickels' patients had native acetabula and sharply divided planes, not scarred revision tissue.
Q: What are the absolute contraindications?
A: Uncontrolled infection, dysvascularity, non-viable soft tissue or an unusable distal limb. Definitive total femur metal should not be used to postpone a safer staged pathway or an amputation. Note the direction of the evidence: infection is the commonest reason these reconstructions fail, and in the Mayo series four patients reached hip disarticulation for recalcitrant infection despite the reconstruction - with 12 of 24 held on chronic antibiotic suppression rather than cured.
Q: What do patients most often misunderstand?
A: That restored length equals a normal leg. About 70 per cent require a walking aid afterwards, and the mean MSTS score across 1,002 pooled patients is 66 per cent. The realistic gain is captured better by the Sheffield mobility data - the proportion of mobile patients rose from 52 to 65 per cent and every patient could at least transfer from bed to chair. That transfer ability, not a normal gait, is what the operation is buying against disarticulation.
Exam Viva Scenarios
Practise clinical reasoning and management decisions out loud
“A patient has a failed proximal femoral replacement, a destructive distal femoral nonunion and a draining sinus. The ankle and foot are sensate and plantigrade, but the hip abductors and quadriceps are severely deficient. What do you offer?”
“Two weeks after total femur replacement, the patient dislocates the hip while transferring. The radiograph shows the stem and knee hinge are intact. How do you assess and manage?”
“A 70-year-old with a painful non-union and multiple failed reconstructions asks whether total femur replacement is better than amputation. How do you counsel?”
Indication
- Near-global femoral loss with no credible segmental or biological reconstruction
- Viable useful distal limb, controllable infection, coverage and rehabilitation potential
- Compare proximal/distal replacement, allograft composite, fusion and amputation
- Oncologic cases require a sarcoma pathway; non-oncologic cases still have high risk
Plan
- Whole-limb radiographs, CT and CT angiography when vascular/flap risk
- Hip centre, length, offset, rotation, knee joint line and hinge/bearing
- Abductor, capsule, quadriceps, patellar tendon and collateral reconstruction
- Staged infection, flap, fracture, nerve and amputation contingencies
Operative technique
- Position, full-limb imaging, cultures, blood and coverage plan
- Approach through viable tissue, protect nerves/vessels and resect non-viable femur
- Prepare acetabulum/tibia, trial whole-limb mechanics and avoid excessive lengthening
- Fix modules, rebuild both mechanisms, cover metal, image and protect rehabilitation
Failure
- Infection, hip dislocation, extensor failure, fracture, nerve palsy and flap loss
- Treat infection and soft tissue as part of the implant failure
- A stable stem does not imply a stable hip or knee
- Amputation remains appropriate when salvage is non-viable or non-functional
Evidence Base
Comparative Outcomes and Failure Rates of Total Femur Replacement in Oncologic and Nononcologic Indications
- Systematic review and meta-analysis of 35 studies and 1,002 patients, of whom 63.7 per cent had a total femur replacement for an oncologic indication
- Combined failure rate 34 per cent; limb salvage rate 89 per cent; mean Musculoskeletal Tumor Society score 66 per cent
- Infection (Henderson type 4) was significantly commoner in non-oncologic than oncologic patients - 18 per cent (95 per cent CI 12 to 26) against 8 per cent (95 per cent CI 6 to 12)
- Functional scores did not differ between the two groups, and failure modes 1, 2 and 3 did not differ when compared independently
- Approximately 70 per cent of patients required a walking aid after surgery
Contemporary Results of Total Femoral Arthroplasties for Non-Oncological Conditions
- 24 total femoral arthroplasties from an institutional registry, mean age 66 years, mean BMI 33, median follow-up four years - 17 of the 24 had a previously documented periprosthetic joint infection
- Survival free of ANY revision was 66 per cent at two years, 51 per cent at five years and 34 per cent at ten years
- Survival free of ANY REOPERATION is the figure to quote at the two-year mark because it is WORSE than the revision figure: 53 per cent at two years, 47 per cent at five and 34 per cent at ten - so nearly half have been back to theatre for something within two years
- Of 12 revisions, five were for hip dislocation and four for periprosthetic joint infection - instability and infection together accounted for three quarters
- Harris Hip Score improved from 48 to 69 (p = 0.002), but the Knee Society Score change from 56 to 71 did NOT reach significance (p = 0.110)
- Four patients ended in hip disarticulation for recalcitrant infection despite the reconstruction
Total Femoral Arthroplasty for Non-Oncological Indications
- 38 total femoral arthroplasties from a tertiary revision unit, mean age 73 years, mean follow-up ten years - the largest UK non-oncological series
- Indications were infection in 53 per cent and periprosthetic fracture in 26 per cent
- 66 per cent suffered at least one complication and DISLOCATION WAS THE COMMONEST AT 37 PER CENT; 55 per cent required at least one further operation
- 63 per cent were classed a success on the Musculoskeletal Infection Society outcome tool, and 70 per cent of those done for infection were infection-free at final review
- The proportion of mobile patients rose from 52 per cent to 65 per cent, and every patient could at least transfer from bed to chair
Which Factors Associate With Implant Revision and Hip Disarticulation After Total Femur Replacement? A Retrospective Multicentric EMSOS Study
- 143 patients from 14 international centres operated between 1990 and 2024, mean age 44 years, 76 per cent for oncologic indications - analysed in a competing-risk framework with death as the competing event
- Cumulative incidence of implant revision was 24 per cent at two years (95 per cent CI 16 to 31) and 35 per cent at five years (95 per cent CI 26 to 44)
- Cumulative incidence of HIP DISARTICULATION was 4 per cent at two years (95 per cent CI 1 to 7) and 10 per cent at five years (95 per cent CI 4 to 16) - the reconstruction itself carries a one-in-ten chance of ending in amputation by five years
- Advanced age was associated with hip disarticulation in both models (subdistribution HR 1.03 per year, 95 per cent CI 1.01 to 1.05, p less than 0.001; cause-specific HR 1.05, 95 per cent CI 1.02 to 1.07, p less than 0.001) - this is the one finding on which the two analyses agree
- For MAJOR COMPLICATIONS the two models diverge exactly as they do for mesh: age reached significance only in the cause-specific model (HR 1.01, 95 per cent CI 1.00 to 1.03, p = 0.023) and not in the subdistribution model (HR 1.01, 95 per cent CI 0.98 to 1.02, p = 0.17)
- Absence of a mesh graft was associated with a higher risk of hip disarticulation (subdistribution HR 0.30 for mesh, 95 per cent CI 0.10 to 0.92, p = 0.035)
Reconstruction of Hip Stability After Proximal and Total Femur Resections
- 57 patients undergoing proximal or total femur resection between 1980 and 1996, average follow-up 6.5 years
- The acetabulum was PRESERVED and not resurfaced in every patient; 49 had a bipolar hemiarthroplasty and eight a fixed unipolar
- Soft-tissue reconstruction was systematic - Dacron tape capsulorrhaphy over the prosthetic neck, reattachment of the abductor mechanism to the prosthesis, and extracortical bone fixation
- Dislocation occurred in ONE patient (1.7 per cent), against dislocation being described as the commonest complication of this reconstruction
- Aseptic loosening in three (5.3 per cent), limb salvage rate 98 per cent, and 81 per cent had a good to excellent functional result
Failure Mode Classification for Tumor Endoprostheses: Retrospective Review of Five Institutions and a Literature Review
- 2,174 patients across five institutions yielded 534 endoprosthetic failures, classified into five modes: type 1 soft tissue, type 2 aseptic loosening, type 3 structural, type 4 infection, type 5 tumour progression
- The commonest mode in this series was INFECTION, but in the accompanying literature review it was ASEPTIC LOOSENING - the same question answered differently by two datasets
- Mode of failure was statistically dependent on anatomical location, and on time to failure
- Types 1, 2, 3 and 4 differed significantly in incidence between polyaxial and uniaxial joints
- The authors conclude that CUMULATIVE reporting of segmental failures should be avoided because anatomy-specific trends are lost
Aseptic Loosening in Cemented Custom-Made Prosthetic Replacements for Bone Tumours of the Lower Limb
- 1,001 custom-made prostheses - 493 distal femoral, 263 proximal femoral, 245 proximal tibial - reviewed for aseptic loosening as a mode of failure
- Probability of surviving aseptic loosening at 120 months was 93.8 per cent for proximal femoral, 67.4 per cent for distal femoral and 58 per cent for proximal tibial replacements
- For distal femoral replacements BOTH younger age at operation AND a higher percentage of femur resected predicted aseptic loosening
- For proximal tibial replacements the percentage of bone removed mattered but age did not
- For proximal femoral replacements NEITHER age nor the percentage resected was a factor
References
- Lari A, Esmaeil A, AlSalem Y, et al. Comparative outcomes and failure rates of total femur replacement in oncologic and nononcologic indications: a systematic review and meta-analysis. JBJS Rev. 2024;12(7). PMID: 38968379. DOI: 10.2106/JBJS.RVW.24.00022.
- Owen AR, Bettencourt JW, Wyles CC, et al. Contemporary results of total femoral arthroplasties for non-oncological conditions. Bone Joint J. 2025;107-B(4):449-454. PMID: 40164183. DOI: 10.1302/0301-620X.107B4.BJJ-2024-0959.R1.
- Murray J, Jeyapalan R, Davies M, et al. Total femoral arthroplasty for non-oncological indications. Bone Joint J. 2023;105-B(8):888-894. PMID: 37524348. DOI: 10.1302/0301-620X.105B8.BJJ-2022-1372.R1.
- Valentini M, Svehlik M, Leithner A, et al. Which factors associate with implant revision and hip disarticulation after total femur replacement? A retrospective multicentric EMSOS study. Clin Orthop Relat Res. 2026;484(4):695-705. PMID: 41588606. DOI: 10.1097/CORR.0000000000003819.
- Bickels J, Meller I, Henshaw RM, Malawer MM. Reconstruction of hip stability after proximal and total femur resections. Clin Orthop Relat Res. 2000;(375):218-230. PMID: 10853173. DOI: 10.1097/00003086-200006000-00027.
- Henderson ER, Groundland JS, Pala E, et al. Failure mode classification for tumor endoprostheses: retrospective review of five institutions and a literature review. J Bone Joint Surg Am. 2011;93(5):418-429. PMID: 21368074. DOI: 10.2106/JBJS.J.00834.
- Henderson ER, O'Connor MI, Ruggieri P, et al. Classification of failure of limb salvage after reconstructive surgery for bone tumours: a modified system including biological and expandable reconstructions. Bone Joint J. 2014;96-B(11):1436-1440. PMID: 25371453. DOI: 10.1302/0301-620X.96B11.34747.
- Unwin PS, Cannon SR, Grimer RJ, Kemp HB, Sneath RS, Walker PS. Aseptic loosening in cemented custom-made prosthetic replacements for bone tumours of the lower limb. J Bone Joint Surg Br. 1996;78(1):5-13. PMID: 8898118.