What is the Cumulative Incidence of Revision Surgery and What Are the Complications Associated With Stemmed Cementless Nonextendable Endoprostheses in Patients 18 Years or Younger With Primary Bone Sarcomas About the Knee
Ahmed Mohamed El Ghoneimy,Amin Mohamed Shehab,Nesma Farid
DOI: https://doi.org/10.1097/CORR.0000000000002150
2022-07-01
Abstract:Background: Osteosarcoma and Ewing sarcoma are more common diagnoses in preadolescent and adolescent children compared with the adult population. A greater percentage of patients are treated with limb salvage and reconstruction using modular tumor endoprostheses. Implant-to-bone fixation can be cemented or cementless. Cementless tumor endoprostheses rely on biologic osteointegration for implant stability, and chemotherapy during childhood and adolescence can disturb the bone turnover rate and reduce bone mineral density, which in turn may predispose patients with uncemented endoprostheses to a high rate of revision surgeries. Questions/purposes: (1) What is the cumulative incidence of revision operations for any cause (wound dehiscence, periprosthetic fracture, hinge breakage, aseptic loosening, infection, local recurrence, implant removal, and amputation) of cementless tumor endoprostheses around the knee? (2) What is the cumulative incidence of aseptic loosening, periprosthetic fracture, hinge breakage, and infection, and what proportion of patients had other complications? (3) What was the mean limb length discrepancy (LLD) at the time of skeletal maturity? (4) What was the median Musculoskeletal Tumor Society (MSTS) score at most recent follow-up or just before implant removal/amputation if implant removal/amputation were performed? Methods: Between 2008 and 2019, we treated 328 patients younger than 18 years for a primary bone sarcoma around the knee at our institution. Of those, 138 were treated with resection and reconstruction using two different types of modular tumor endoprostheses. During this period, our general indications for an endoprosthesis were patients who were candidates for an intraarticular resection of the distal femur or proximal tibia and who were at least 10 years of age. Uncemented fixation was always preferred. Cemented fixation was only done when intraoperative press-fitting of a cementless stem was not possible. Among uncemented implants, 26 patients died before completing 2 years of follow-up with intact implants and without further surgery, three were lost to follow-up before 2 years, and four patients received implants as a secondary salvage surgery after a failed primary biologic reconstruction with a vascularized fibular bone graft, leaving 94 patients for evaluation in this retrospective study. The mean age was 15 ± 2 years and the median (interquartile range) follow-up duration was 51 months (39 to 74). We did a competing risks analysis to tally cumulative incidence of all-cause revision procedures and cumulative incidence of aseptic loosening, periprosthetic fracture, hinge breakage, and infection. Other complications, including wound dehiscence, local recurrence, and stem breakage, were characterized descriptively and ascertained by review of electronic records of a longitudinally maintained institutional database by the treating surgeons. LLD was measured by serial clinical assessments and CT scans, starting since primary salvage surgery and until the latest follow-up of every patient. For the analysis of remaining LLD, we included only patients who were skeletally immature at time of primary resection and who had reached skeletal maturity by their latest follow-up (73% [69 of 94]). Functional outcome was assessed using MSTS scores obtained from a review of electronic records of a longitudinally maintained institutional database. Results: The 8-year cumulative incidence of revision surgery for any cause was 32% (95% confidence interval 23% to 42%). The 8-year cumulative incidence of aseptic loosening was 5% (95% CI 2% to 11%), periprosthetic fracture was 9% (95% CI 4% to 15%), hinge breakage was 19% (95% CI 12% to 28%), and infection was 7% (95% CI 3% to 14%). Other complications included wound dehiscence in 2% (2 of 94), stem breakage in 2% (2 of 94), and local recurrence in 2% (2 of 94) of patients. Stress shielding of the cortical bone around implanted stems was observed in 26% (24 of 94). The mean LLD for those who were skeletally immature at the time of primary resection and who reached skeletal maturity was 3.5 ± 2.6 cm. At latest follow-up, the median (IQR) MSTS score for all patients, excluding those who had complete implant removal or amputation, was 26 (24 to 27) of a maximum score of 30. Conclusion: We observed a high rate of early revision and relatively frequent complications associated with the use of cementless fixation, and although this was not a comparative study, the findings were not superior to those reported by others who have studied cemented fixation for this indication. Furthermore, there may be some disadvantages with cementless fixation, such as stress shielding. Comparative studies about fixation methods are needed. The prevalence of bushing breakage in the current study highlights the importance of future modifications in the hinge design of both types of prostheses used in this study. Patients who were skeletally immature at the time of primary surgery had a LLD no more than 5 cm at skeletal maturity; consequently, nonexpandable endoprostheses may be appropriate for some adolescent patients who have limited remaining growth, although which patients are best suited for this approach would require specific study. Level of evidence: Level IV, therapeutic study.