Endoprosthetic reconstruction versus osteosynthesis for metastatic disease of the proximal femur: a population-based cohort study.

Introduction

Metastatic bone disease of the proximal femur (MBDf) is a common complication of advanced malignancy, affecting approximately 10% of cancer patients [1] , [2] , [3] . Owing to the high mechanical load borne by the proximal femur, metastatic involvement frequently results in pain, loss of function, and pathological fracture requiring surgical intervention [4] . Compared with other orthopaedic fractures, MBDf is associated with substantial morbidity and mortality, with reported two-year survival rates of approximately 30% [5] , [6] . Consequently, surgical management in this population must balance limited life expectancy with the need for durable fixation, immediate weight-bearing, pain relief, and preservation of functional independence, while minimizing the risk of extensive re-operation that may delay or interrupt oncologic treatment [3] , [6] .

The two principal surgical approaches for treating MBDf are osteosynthesis- using plates, screws, or intramedullary nails- and endoprosthetic reconstruction, typically involving total hip arthroplasty with or without proximal femoral replacement. Metastatic bone is characterized by impaired healing capacity due to tumour infiltration, osteolysis, and the effects of systemic anticancer therapy. Consequently, fixation strategies that rely on biological fracture healing may be particularly vulnerable to mechanical failure in this setting.

Osteosynthesis nevertheless remains widely used in clinical practice, often based on its perceived lower invasiveness, shorter operative time, and reduced intraoperative blood loss [3] , [6] . Endoprosthetic reconstruction, by contrast, replaces the affected bone segment entirely, providing immediate mechanical stability and allowing unrestricted weight-bearing from the outset. Although accumulating evidence suggests lower implant failure rates, improved functional recovery, and better quality of life following endoprosthetic reconstruction [7] , [8] , most existing studies are retrospective and limited by small, heterogeneous cohorts and inconsistent outcome reporting [3] , [6] . Randomized controlled trials in this field are challenging due to ethical considerations, prognostic heterogeneity, and difficulties in achieving equipoise. As such, robust population-based studies with complete follow-up remain essential to inform surgical decision-making in patients with MBDf.

The aim of this study was to compare outcomes following osteosynthesis and endoprosthetic reconstruction in patients undergoing surgery for MBDf. Specifically, we evaluated implant failure, revision surgery, and overall survival in a large, consecutive, population-based cohort.

Material and methods

Study design and population

In this population-based study, we included all patients undergoing surgery for MBDf in the Greater Copenhagen region of Denmark between January 1, 2013, and December 31, 2019. Data was collected prospectively from 2013 to 2016 and retrospectively thereafter.

All procedures were performed at six public hospitals: five secondary orthopaedic centres and one tertiary musculoskeletal tumour centre. Due to the organization of the Danish healthcare system, all patients with MBDf in the region are treated within these institutions, ensuring a complete, consecutive cohort with no loss to follow-up or emigration. Patients were followed until death or the end of follow-up (September 30, 2024).

Lesions were classified according to the AO/OTA proximal femur classification system. Due to the retrospective design and the fact that many lesions involved both the femoral neck and trochanteric region, further anatomical subdivision was considered unreliable and was therefore not performed [9] . Both impending and manifest pathological fractures were included. Patients treated with resection without reconstruction (Girdlestone procedure) were excluded. Implant failure was defined as complete implant removal or removal of bone-anchored components due to mechanical failure or other implant-related complications.

Revision surgery was defined as any reoperation involving the implant, including implant exchange, removal of bone-anchored components, surgical treatment of infection, or open reduction procedures involving endoprosthetic reconstruction.

The study was conducted in accordance with the STROBE guidelines.

Surgical treatment and follow-up

In the absence of national guidelines for MBDf during the study period, surgical decisions were made at the discretion of the attending surgeon. However, institutional practice favoured endoprosthetic reconstruction in patients with an estimated life expectancy exceeding three months, consistent with British Orthopaedic Association recommendations guidelines [10] . Postoperative rehabilitation was individualized, with equal access to physiotherapy and recovery programs provided through Denmark's universal healthcare system. Clinical, demographic, and imaging data were extracted from comprehensive electronic medical records. Due to the unique social medical system in Denmark, no differentiation was made between patients with respect to indication, treatment, rehabilitation, or oncological care, and the system ensured complete follow-up with respect to implant status and patient survival.

Outcomes

Primary outcomes were: • Implant failure, defined as complete implant removal or removal of bone anchored components. • Revision surgery, defined as any reoperation involving the implant, including infection but also open reduction of endoprostheses. • Overall survival, measured from the date of surgery to death from any cause.

Sample size and power considerations

All eligible and consecutively treated patients from all public hospitals in the Capital Region of Denmark were included over a seven-year period. Patients were identified using the surgical planning software system (EPIC).

Statistical analyses

Baseline characteristics were summarized using means (ranges) for continuous variables and frequencies (percentages) for categorical variables. Group comparisons were performed using chi-square or Fisher's exact tests for categorical variables and t -tests or Wilcoxon rank-sum tests for continuous variables, as appropriate. Implant failure and revision surgery were analysed using competing risk methods, with cumulative incidence estimated using the Aalen-Johansen estimator. For implant failure, death was treated as a competing event; for revision surgery, both implant failure and death were considered competing events. Differences between groups were assessed using Gray's test. Overall survival was analysed using Kaplan-Meier methods, with group differences assessed using the log-rank test.

All analyses were performed using R version 4.2.2 (R Development Core Team, Vienna, Austria). In accordance with GDPR regulations, results with five or fewer observations are not reported in detail.

Results

Demographic and clinical characteristics

A total of 287 patients with AO type 31 MBDf underwent surgery during the study period. Six patients were excluded (four Girdlestone procedures and two total femur replacements), leaving 281 patients for analysis (osteosynthesis: 81 [29%]; endoprosthesis: 200 [71%]). Sex distribution was similar between groups. Patients treated with osteosynthesis were slightly older than those treated with endoprosthetic reconstruction (mean age 74 vs. 70 years, p = 0.015).

Manifest fractures were more common in the osteosynthesis group (94% vs. 75%), whereas endoprosthetic reconstruction was more frequently used for impending fractures (26% vs. 6%, p < 0.001). The most common primary malignancies were breast (24%), lung (21%), and prostate cancer (20%). Unknown primary tumours were more frequent in the osteosynthesis group (11% vs. 3%). Among osteosynthesis patients, 69 (85%) had intramedullary nailing, 8 (10%) plating, and 4 (5%) screws. In the endoprosthesis group, 115 (57%) received total hip arthroplasty, 81 hemiarthroplasty (40%) and 4 (2%) had total hip arthroplasty with proximal femoral replacement. The osteosynthesis group had a shorter time from admission to surgery (1 day vs. 4 days, p < 0.001), as well as shorter operative times (115 vs. 130 in min, p = 0.001) and lower intraoperative blood loss (409 mL vs. 696 mL, p < 0.001). Clinical and demographic characteristics are summarized in Table 1 . Table 1 Clinical and demographic characteristics of metastatic bone disease of the femur patients treated with osteosynthesis or endoprosthesis. Osteosynthesis ( n = 81) Endoprosthesis ( n = 202) p -value Age, mean (standard deviation) 74 (11) 70 (12) 0.01 Sex, female, n (%) 41 (51%) 103 (52%) 0.9 Primary cancer, n(%) ▪ Bladder 1 (1%) 3 (2%) ▪ Breast 17 (21%) 50 (25%) ▪ Gastrointestinal 4 (5%) 7 (4%) ▪ Head and neck 0 (0%) 4 (2%) ▪ Lung 16 (20%) 42 (21%) ▪ Lymphoma 2 (3%) 4 (2%) ▪ Melanoma 0 (0%) 4 (2%) ▪ Multiple myeloma 8 (10%) 11 (6%) ▪ Pancreas 0 (0%) 3 (2%) ▪ Prostate 13 (16%) 42 (21%) ▪ Renal 7 (9%) 11 (6%) ▪ Sarcoma 0 (0%) 6 (3%) ▪ Other 4 (5%) 7 (4%) Unknown primary tumour 9 (11%) 6 (3%) Visceral metastases, n(%) a 39 (48%) 94 (47%) 0.2 Bone metastases, n(%) 0.11 ▪ Axial with neurologic deficits 2 (2.5%) 12 (6%) ▪ Axial without neurologic deficits 53 (65%) 129 (60%) ▪ Multiple appendicular 14 (17%) 22 (11%) ▪ Solitary 9 (11%) 42 (21%) ▪ Unknown 3 (3.7%) 4 (2%) Oncologic treatment 0.5 ▪ Chemotherapy 20 (25%) 18 (36%) ▪ Hormone therapy 13 (16%) 39 (20%) ▪ Radiation therapy 12 (15%) 30 (15%) ▪ Targeted therapy 4 (5%) 12 (6%) ▪ Combination 3 (4%) 16 (8%) Karnofsky score, median (interquartile range) 70 (50–80) 70 (60–90) 0.14 Fracture type <0.001 ▪ Manifest 76 (94%) 149 (75%) ▪ Impending 5 (6%) 51 (26%) Admission to surgery (fracture), days, median (interquartile range) 1 (1–2) 4 (1–9) <0.001 Tumour resection 0 (0%) 102 (51%) <0.001 Surgery time, min, median (interquartile range) 96 (61–152) 131 (93–153) 0.001 Blood loss, mL, median (interquartile range) 300 (150–550) 520 (300–900) <0.001 Bolded p -values indicate statistical significance ( p < 0.05).

Implant failure after osteosynthesis compared to endoprosthetic reconstruction

At three, six, and twelve months, cumulative incidences of implant failure were 5% (CI 0.2 to 10), 7% (CI 2 to 13), and 7% (CI 2 to 13) after osteosynthesis, compared with 0.5% (CI 0 to 1.5), 0.5% (CI 0 to 1.5), and 1% (CI 0 to 2.4) after endoprosthetic reconstruction ( p = 0.03). The cumulative incidence of implant failure over time is shown in Fig. 1 . Fig. 1 Lower implant failure rates in patients with metastatic bone disease of the proximal femur treated with endoprosthesis compared to osteosynthesis at 3, 6, and 12 months postoperatively. Fig. 1

In the osteosynthesis group, six implants failed, all due to intramedullary nail breakage consistent with fatigue fracture. Owing to the small number of events, further case-level details are not reported to preserve patient confidentiality. Implant failure occurred at a mean of 73 days postoperatively (range 32–129 days), and the mean follow-up duration among patients with failure was 492 days (range 71–1842 days). Among the 44 patients treated with osteosynthesis who survived at least 90 days after surgery, six experienced implant failure, corresponding to a 14% failure rate among 90-day survivors. In the endoprosthesis group, four patients experienced implant failure, with a mean time to failure of 349 days (range 11–738 days). Among patients surviving at least 90 days, this corresponded to an implant failure rate of 3%.

Revision surgery after osteosynthesis compared to endoprosthetic reconstruction

At three months, the cumulative incidence of revision surgery was 5% (95% CI 0.2–10) following osteosynthesis and 3% (95% CI 0.6–5) following endoprosthetic reconstruction. At six months, cumulative incidence increased to 7% (95% CI 2–13) and 4% (95% CI 1–6), respectively, and at twelve months to 9% (95% CI 3–15) and 4% (95% CI 2–7), with no statistically significant difference between groups ( p = 0.10). The cumulative incidence of revision surgery over time is shown in Fig. 2 . Fig. 2 Comparable revision rates in patients with metastatic bone disease of the proximal femur treated with endoprosthesis compared to osteosynthesis at 3, 6, and 12 months postoperatively. Fig. 2

All revisions in the osteosynthesis group were related to implant failure. In contrast, revisions following endoprosthetic reconstruction were primarily due to joint instability and were managed with less extensive procedures, including cup implantation in cases of hemiarthroplasty or by the addition of a 360° anti-dislocation ring in total hip arthroplasty.

Overall patient survival after osteosynthesis compared to endoprosthetic reconstruction

At three months, survival was 65% (95% CI 58–72) following endoprosthetic reconstruction compared with 54% (95% CI 44–65) after osteosynthesis. Corresponding survival at six months was 51% (95% CI 44–57) versus 37% (95% CI 27–48), and at twelve months 39% (95% CI 32–45) versus 26% (95% CI 16–36) ( p = 0.01). Overall survival probabilities over time are presented in Fig. 3 . Fig. 3 Increased survivalrates in patients with metastatic bone disease of the proximal femur treated with endoprosthesis compared to osteosynthesis at 3, 6, and 12 month spostoperatively. Fig. 3

Discussion

In this population-based, multicentre study, we found that surgical treatment with osteosynthesis for MBDf was associated with a 7–14-fold increased risk of implant failure requiring revision to a mega prosthesis. We found no difference in the risk of overall revision. These findings extend prior retrospective studies and registry data suggesting favourable outcomes following endoprosthetic reconstruction in patients with MBDf [3] .

Implant failure following osteosynthesis occurred early and plateaued between six and twelve months. In studies with longer follow-up, failure rates after osteosynthesis have been shown to increase progressively over time, whereas endoprosthetic reconstructions demonstrate more stable long-term outcomes [3] . In our cohort, endoprostheses were associated with a low risk of failure, with only 1% experiencing implant failure at twelve months. This observation is consistent with previous reports and supports the view that endoprosthetic reconstruction provides a durable treatment option that is largely independent of the compromised bone healing environment characteristic of metastatic disease [3] . The ongoing PERFORM trial (NCT0686319) is expected to provide prospective randomized evidence comparing osteosynthesis and endoprosthetic reconstruction in patients with MBDf and will help to clarify whether the associations observed in population-based and observational studies are confirmed in a randomized setting. The question is, however, to what extent the results of a randomized control trial will be externally valid, given the substantial selection bias inherent in determining eligibility for studies of this kind.

Revision surgery was infrequent in both treatment groups and did not differ significantly between surgical strategies. However, the clinical implications of revision differed between groups. Revisions following osteosynthesis were exclusively related to implant failure and typically required conversion to endoprosthetic reconstruction, representing a major secondary procedure in a frail oncologic population. In contrast, revisions after endoprosthetic reconstruction were primarily related to joint instability and were generally managed without revision of bone-anchored components. Although a systematic review has reported revision rates after osteosynthesis ranging from 5% to more than 25% [3] , the revision rates observed in our cohort were lower, potentially reflecting differences in patient selection, surgical technique, or perioperative management. While definitions of implant failure and thresholds for revision vary across studies, the overall pattern observed between osteosynthesis and endoprosthetic reconstruction in our cohort aligns with previously published trends [11] .

Several mechanisms may contribute to the observed survival difference. Prior studies have suggested that the immediate mechanical stability provided by endoprosthetic reconstruction facilitates early mobilization, reduces immobilization-related complications, and enables uninterrupted oncologic therapy, factors that may influence survival independently of baseline prognosis [2] , [12] . Although institutional practice generally favoured endoprosthetic reconstruction in patients with longer expected survival, introducing the possibility of prognostic selection bias, we found no significant differences between groups in established prognostic indicators, aside from a higher prevalence of complete fractures in the osteosynthesis group [13] . This may indicate that routinely recorded clinical variables do not fully capture selection bias, or that the two cohorts were more comparable than initially assumed.

Time to surgery was substantially longer in the endoprosthesis group, approximately four times that observed in the osteosynthesis group and exceeding the 48-h threshold commonly recommended in hip fracture guidelines [14] . While the applicability of this benchmark to patients with MBDf remains uncertain, prolonged immobilization is unlikely to be benign in a frail oncologic population. In our cohort, surgical delays were most likely attributable to referral pathways for complex reconstruction, particularly in cases requiring megaprosthetic replacement due to extensive bone loss. These findings support the potential value of institutional and regional guidelines aimed at expediting the diagnostic work-up and referral processes for patients with metastatic fractures of the proximal femur.

Although osteosynthesis is often perceived as less invasive due to shorter operative time and reduced intraoperative blood loss [15] , these advantages were modest in our cohort, amounting to approximately 15 min and 300 mL, respectively. Such differences must be weighed against potential clinical trade-offs. Internal fixation frequently necessitates restricted weight-bearing, which may delay mobilization and increase the risk of complications such as thromboembolism, pulmonary morbidity, and prolonged hospitalization [1] , [12] . In contrast, endoprosthetic reconstruction allows immediate full weight-bearing and may support more rapid functional recovery.

Taken together, the modest procedural advantages of osteosynthesis may be outweighed by its limitations in implant durability and functional recovery, considerations of particular importance in a frail and comorbid patient population with limited life expectancy.

Strengths and limitations

The strengths of this study include its multicentre design and population-based sampling, which reduced referral bias, enhanced external validity, and ensured complete inclusion of eligible patients within a defined geographic region. The inclusion of both secondary surgical centres and a tertiary referral hospital reflects the diversity of real-world treatment environments in which patients with MDBf are managed. In addition, the use of competing risk analysis allowed more accurate estimation of implant-related events in a population characterized by high early mortality.

Several limitations should be considered when interpreting the findings of this study.

First, retrospective design introduces an inherent risk of selection bias. Treatment allocation was based on clinical decision-making rather than randomization, and differences in patient characteristics may therefore have influenced outcomes.

Second, detailed information regarding systemic oncological treatment, including chemotherapy, targeted therapy, immunotherapy, radiotherapy, and treatment sequencing, was not consistently available and could not be included in the analysis. These factors may influence both survival and surgical outcomes.

Third, osteosynthesis procedures were analysed as a single group despite including different fixation constructs, such as intramedullary nails, plates, and screw fixation. These implants possess different biomechanical characteristics and failure mechanisms, which may have influenced the observed results.

Fourth, patient-reported outcome measures were not included in the final analysis due to insufficient response rates and substantial follow-up completeness between treatment groups, creating a high risk of bias. Consequently, the study cannot provide direct insight into patient-perceived quality of life or functional recovery.

Finally, the study was not designed to evaluate cost-effectiveness. While implant costs differ between reconstructive strategies, broader economic consequences related to rehabilitation, hospital stay, complications, and revision surgery were not assessed.

Conclusion

In this population-based cohort study, EPR reconstruction resulted in a lower risk of implant failure compared with osteosynthesis for metastatic disease of the proximal femur. Although revision surgery rates were similar between groups, the findings suggest that EPR reconstruction may provide a more durable reconstructive option in appropriately selected patients. Given the challenges of prognostication in metastatic bone disease, treatment decisions should remain individualized and made within a multidisciplinary framework. Prospective studies incorporating patient-reported outcomes, oncological treatment variables, and cost-effectiveness analyses are warranted. Importantly, revisions following osteosynthesis were related to implant failure and typically required more extensive secondary procedures, whereas revisions after EPR reconstruction were primarily related to instability and could often be managed without revision of bone-anchored components. In a clinically heterogeneous population with compromised bone integrity, these population-based data provide relevant insights into surgical outcomes and may support clinical decision-making.

CRediT authorship contribution statement

Pardis Zarifkar: Writing – original draft, Formal analysis, Data curation. Thea Hovgaard Ladegaard: Writing – review & editing, Validation, Methodology, Investigation, Funding acquisition, Data curation. Afrim Iljazi: Writing – review & editing, Methodology, Formal analysis, Data curation. Michael Mørk Petersen: Writing – review & editing, Supervision, Project administration, Funding acquisition. Michala Skovlund Sørensen: Writing – review & editing, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Ethical approval

This study was conducted in accordance with the Declaration of Helsinki and approved by The Danish Patient Safety Authority (3-3013-2820/1), the Data Protection Agency of the Capital Region of Denmark (VD-2019-132).

Funding

This research was funded by the Research Fund (Rigshospitalets Forskningsfond) at Copenhagen University Hospital, Rigshospitalet . The funders had no role in the study design, data collection, analysis, interpretation, or manuscript preparation. The authors have no conflicts of interest.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Thea Hovgaard Ladegaard reports financial support was provided by Musculoskeletal tumour section, Orthopaedic department Rigshospitalet. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

1 Jacofsky D.J. Haidukewych G.J. Management of pathologic fractures of the proximal femur: state of the art J. Orthop. Trauma 18 2004 459 469 15289695 10.1097/00005131-200408000-00013

2 Guzik G. Oncological and functional results after surgical treatment of bone metastases at the proximal femur BMC Surg. 18 2018 1 8 29370790 10.1186/s12893-018-0336-0 PMC5784608

3 Iljazi A. Andersen M.S. Brorson S. Petersen M.M. Sørensen M.S. Surgical management of metastatic lesions in the proximal femur: a systematic review EFORT Open Rev. 10 2025 104 114 40071964 10.1530/EOR-24-0138 PMC11825154

4 Hovgaard T.B. Horstmann P.F. Petersen M.M. Sørensen M.S. Patient survival following joint replacement due to metastatic bone disease - comparison of overall patient and prostheses survival between cohorts treated in two different time-periods Acta Oncol. 57 2018 839 848 29293034 10.1080/0284186X.2017.1420910

5 Skovlund Sørensen M. Incidence of surgical interventions for metastatic bone disease in the extremities: a population-based cohort study Acta Oncol. 58 2019 456 462 30632859 10.1080/0284186X.2018.1549368

6 Putnam D.S. Philipp T.C. Lam P.W. Gundle K.R. Treatment modalities for pathologic fractures of the proximal femur pertrochanteric region: a systematic review and meta-analysis of reoperation rates J. Arthroplast. 33 2018 3354 3361 10.1016/j.arth.2018.06.012 30232017

7 Iljazi A. Andersen M.S. Brorson S. Petersen M.M. Sørensen M.S. Surgical management of metastatic lesions in the proximal femur: a systematic review EFORT Open Rev. 10 2025 104 114 40071964 10.1530/EOR-24-0138 PMC11825154

8 Sørensen M.S. Horstmann P.F. Hindsø K. Petersen M.M. Use of endoprostheses for proximal femur metastases results in a rapid rehabilitation and low risk of implant failure. A prospective population-based study J. Bone Oncol. 19 2019 100264 10.1016/j.jbo.2019.100264 PMC6911858 31871883

9 Müller M.E. Koch P. Nazarian S. Schatzker J. The Comprehensive Classification of Fractures of Long Bones 1990 10.1007/978-3-642-61261-9

10 British Orthopaedic Oncology Society and British Orthopaedic Association Metastatic Bone Disease: A Guide to Good Practice 2015 Revision

11 Sørensen M.S. Horstmann P.F. Hindsø K. Petersen M.M. Use of endoprostheses for proximal femur metastases results in a rapid rehabilitation and low risk of implant failure. A prospective population-based study J. Bone Oncol. 19 2019 100264 10.1016/j.jbo.2019.100264 PMC6911858 31871883

12 Ito K. Recovery of ambulation in older adults undergoing fracture surgery for femoral metastasis: a multicenter retrospective cohort study Sci. Rep. 15 2025 1 7 40360634 10.1038/s41598-025-01126-8 PMC12075596

13 Sørensen M.S. Gerds T.A. Hindsø K. Petersen M.M. Prediction of survival after surgery due to skeletal metastases in the extremities Bone Joint J. 98B 2016 271 277 10.1302/0301-620X.98B2.36107 26850435

14 Klestil T. Impact of timing of surgery in elderly hip fracture patients: a systematic review and meta-analysis Sci. Rep. 8 2018 1 15 30224765 10.1038/s41598-018-32098-7 PMC6141544

15 Gusho C.A. Internal fixation versus endoprosthetic replacement of the proximal femur for metastatic bone disease: single institutional outcomes: proximal femur replacement versus internal fixation J. Orthop. 28 2021 86 90 34992334 10.1016/j.jor.2021.11.003 PMC8695262

Data availability

Anonymized data will be made available upon reasonable request.