Abstract
Background Proximal femur replacements (PFRs) are an effective surgical option to treat primary and metastatic tumors causing large bony defects in the proximal femur. Given the relative rarity of these indications, current studies on PFR for oncologic indications are generally limited by patient volume or relatively short-term follow-up. Because recent advances in systemic therapy have improved the prognosis of patients who undergo limb salvage surgery for musculoskeletal tumors, data on the long-term durability of endoprosthetic reconstructions have become increasingly important.Questions/purposes (1) How does the long-term survival of cemented bipolar PFRs compare with patient survival in patients who underwent PFR for benign, aggressive, and metastatic tumors? (2) What are common reasons for revisions of primary PFRs? (3) Which factors are associated with survival of primary PFRs? (4) What is the survivorship free from conversion of bipolar PFRs to THA?Methods Between January 1, 1980, and December 31, 2020, we treated 812 patients with an endoprosthetic reconstruction for an oncologic indication. All patients who underwent a primary PFR for an oncologic indication were included in this study. The study cohort consisted of 122 patients receiving a primary PFR. Eighteen patients did not reach a censored endpoint such as death, revision, or amputation within 2 years. Thirty-three patients died within 2 years of their surgery. Of the 122 patients with primary PFRs, 39 did not reach a censored endpoint and have not been seen within the past 5 years. However, the mean follow-up time for these patients was longer than 10 years. The Social Security Death Index was queried to identify any patients who may have died but might not have been captured by our database To allow for adequate follow-up, endoprosthetic reconstructions performed after December 31, 2020 were excluded. The mean age at the time of the index surgery was 48 +/- 22 years. The mean follow-up time of surviving patients was 7 +/- 8 years. All PFRs were performed using a bipolar hemiarthroplasty with a cemented stem, and all implants were considered comparable. Demographic, oncologic, procedural, and outcome data including prosthesis survival, patient survival, complication rates, and rates of conversion to THA were analyzed. Patient, prosthesis, and limb salvage survival rates were generated, with implant revision as the endpoint and death as a competing risk. Statistical significance was defined as p < 0.05.Results Generally, patients with benign or low-grade (Stage I) disease outlived their implants (100% patient survival through 30 years; p = 0.02), whereas the opposite was true in patients with high-grade, localized Stage II disease (64% patient survival at 5 years [95% CI 49% to 76%]; p = 0.001) or widespread Stage III metastatic disease (6.2% patient survival at 5 years [95% CI 0.5% to 24%]; p < 0.001). Primary PFR implant survival at 5, 10, 20, and 30 years was 97% (95% CI 90% to 99%), 81% (95% CI 67% to 90%), 69% (95% CI 46% to 84%), and 51% (95% CI 24% to 73%), respectively. Eight percent (10 of 122) of primary PFRs were revised for any reason. The most common causes of revision were aseptic loosening (3% [four of 122]), infection (3% [three of 122]), breakage of the implant (2% [two of 122]), and tumor progression (1% [one of 122]). Follow-up time was the only factor that was associated with revision of primary PFRs. Neither segment length nor stem length were associated with revision of primary.
Six percent (seven of 122) of PFRs were converted to THA at a mean 15 +/- 8 years from the index procedure. Survivorship free from conversion to THA (accounting for death as a competing risk) was 94% (95% CI 85% to 99%), 86% (95% CI 68% to 94%). and 77% (95% CI 51% to 91%) at 10, 20, and 30 years, respectively.Conclusion Cemented bipolar PFRs for an oncologic indication are a relatively durable reconstruction technique. Given the relative longevity and efficacy of PFRs demonstrated in our study, especially in patients with high-grade or metastatic disease where implant survival until all-cause revision was longer than patient survival, surgeons should continue to seriously consider PFRs in appropriate patients. The relative rarity of these reconstructions limits the number of patients in this study as well as in current research; thus, further multi-institutional collaborations are needed to provide the most accurate prognostic data for our patients.Level of Evidence Level III, therapeutic study.