Aims. The aim of this study was to assess the feasibility of conducting a full-scale, appropriately powered, randomized controlled trial (RCT) comparing internal fracture fixation and distal femoral replacement (DFR) for distal femoral fractures in older patients. Patients and Methods. Seven centres recruited patients into the study. Patients were eligible if they were greater than 65 years of age with a distal femoral fracture, and if the surgeon felt that they were suitable for either form of treatment. Outcome measures included the patients’ willingness to participate, clinicians’ willingness to recruit, rates of loss to follow-up, the ability to capture data, estimates of standard deviation to inform the sample size calculation, and the main determinants of cost. The primary clinical outcome measure was the EuroQol five-dimensional index (EQ-5D) at six months following injury. Results. Of 36 patients who met the inclusion criteria, five declined to participate and eight were not recruited, leaving 23 patients to be randomized. One patient withdrew before surgery. Of the remaining patients, five (23%) withdrew during the follow-up period and six (26%) died. A 100% response rate was achieved for the EQ-5D at each follow-up point, excluding one missing datapoint at baseline. In the DFR group, the mean cost of the implant outweighed the mean cost of many other items, including theatre time, length of stay, and readmissions. For a powered RCT, a total sample size of 1400 would be required with 234 centres recruiting over three years. At six months, the EQ-5D utility index was lower in the DFR group. Conclusion. This study found that running a full-scale trial in this country would not be feasible. However, it may be feasible to undertake an international multicentre trial, and our findings provide some guidance about the power of such a study, the numbers required, and some challenges that should be anticipated and addressed. Cite this article: Bone Joint J 2019;101-B:1408–1415

Introduction. The current investigation includes a retrospective review of the experience of five Institutions with distal femur megaprostheses for tumor over a twenty year period, to analyze the incidence and etiology of failure, using a new classification system based upon the failure modes. Methods. Between 1974 and 2008, 2174 patients underwent primary limb preservation for a benign or malignant extremity tumor using a metallic megaprosthesis at five Institutions, 951 (43.7%) were distal femur replacements. Retrospective analysis of complications according to the Letson and Ruggieri Classification was performed and Kaplan-Meier curves of implant survival were defined. Segmental megaprosthetic reconstruction failures were categorized as mechanical and non-mechanical failures. Results. A total of 951 skeletally mature patients received a segmental endoprosthesis for the treatment of an oncologic condition. Overall 261 (27%) of the primary procedures were considered failures. There were 137 mechanical failures (14.4%): 12 (1.3%) Type 1 (soft tissue failure), 65 (6.8%) Type 2 (aseptic loosening), and 60 (6.3%) Type 3 (structural failure). Non-mechanical causes accounted for 124 failures (13%): 45 (4.7%)Type 4 (tumor progression) and 79 (8.3%) Type 5 (infection). The overall implant survival to all modes of failure was 77% at 10 years and 73% at 20 years. The implant survival to aseptic loosening was over 90% at 10 years. Conclusion. Most frequent cause of failure was infection followed by aseptic loosening and structural failure. The implant survival at long term was quite satisfactory

Background/Aims. The development of extendable prostheses has permitted limb salvage surgery in paediatric patients with bone tumours in proximity to the physis. Prostheses are extended to offset limb length discrepancy as the child grows. Aseptic loosening (AL) is a recognised complication. The implant stem must fit the narrow paediatric medullary canal and remain fixed while withstanding growth and increasing physical demands. Novel designs incorporate a hydroxyapatite (HA) coated collar that manufacturers claim improves bony ongrowth and stability, providing even stress distribution in stem and shoulder regions and providing a bone-implant seal, resulting in decreased AL and prolonged survival. This study aims to assess whether there is a relationship between bony ongrowth onto a HA collar and AL. Hypothesis: Bone ongrowth onto the HA collar of extendable prostheses is associated with more stable fixation and less AL despite patient growth. Methods. Retrospective review of 51 primary partial femoral extendable prostheses implanted over 12 years from 1994–2006 (followed up to death at a mean of 2.5±2.2 years or last clinical encounter at a mean of 8.6 years) and 24 subsequent revisions, to ascertain failure rate and mode, together with a cohort study reviewing bony ongrowth onto the HA coated collar in 10 loose and 13 well fixed partial femoral, humeral and tibial implants. Patient growth was measured as a change in bone:implant-width ratio. Results. 21 (41.2%) primary femoral implants failed at a mean 42.8 months, 5 through AL. 1 secondary implant was revised for AL. 2 implants displayed evidence of progressive AL but had not failed at last follow-up. 5 of 11 tibial component revisions in distal femoral replacement were due to AL. 1 major complication occurred after revision surgery for AL in a primary implant: deep infection requiring 2 stage revision. Bony collar ongrowth was significantly higher in all 4 quadrants (anterior, posterior, medial and lateral) in the well-fixed as opposed to loose group, demonstrating a strong negative relationship in each quadrant between bony ongrowth and AL (p0.001) in the presence of patient growth as shown by increased bone:implant width ratio. In both groups, collar ongrowth was greatest in the posterior quadrant. Summary and Conclusions: AL has been confirmed as a common cause of failure in massive extendable endoprostheses. Revision surgery is difficult and may cause serious complications. For the first time, a significant relationship between a well fixed implant stem and bony ongrowth onto a HA coated collar in the context of massive implants used in tumour surgery has been demonstrated. This newly-proven relationship may result in longer-term implant survival and thus a reduced need for revision surgery. It is hoped that this study will provide the basis for further study of this relationship