Is Proximal Tibial Hemiarthroplasty Reconstruction Effective in Minimizing Limb Length Discrepancy Among Skeletally Immature Patients With Primary Bone Sarcomas?

BACKGROUND Proximal tibial resection and reconstruction with a hinged knee megaprostheses may result in severe limb length discrepancy (LLD) in very young children because of the removal of the distal femoral and proximal physis. An alternative reconstruction using a proximal tibial hemiarthroplasty reconstruction has been proposed and reduces the degree of LLD because the distal femoral physis is preserved. However, there are very few reports on the results from this reconstruction, and it is not certain that the disadvantages of a more unstable knee are outweighed by the possibility of reducing limb length inequality. QUESTIONS/PURPOSES (1) What was the survivorship at 5 and 10 years after proximal tibial hemiarthroplasty reconstruction in children with malignant tumors, using amputation, endoprosthesis removal, and revision surgery as the main endpoints of interest? (2) What was the Musculoskeletal Tumor Society Score-93 (MSTS-93) after reconstruction at a minimum of 2 years after the procedure? (3) What percentage of patients experienced a major complication (resulting in unplanned reoperation), and what percentage of patients experienced minor complications (resulting in nonoperative treatment)? (4) What factors were associated with knee subluxation, and what factors were associated with an LLD measuring ≥ 4 cm? METHODS This was a retrospective study performed by four consultant surgeons at a tertiary tumor referral center (Beijing Jishuitan Hospital, National Center for Orthopaedics, PR China) between 2005 and 2022. During that time, we generally recommended a tibial hemiarthroplasty to children with primary malignant tumors of the proximal tibia (Enneking stages IA, IB, and IIA and chemotherapy-responsive Stage IIB and IIIB tumors), as well as some metastatic tumors and some soft tissue sarcomas involving and surrounding the proximal tibia in children. We considered the ideal age range to be 9 to 13 years for males and 9 to 12 years for females, and we generally did not offer this procedure unless the surgeon believed that the neurovascular bundle was either uninvolved or could be separated during surgery. During that time, we considered 883 patients with primary malignant bone tumors to be potentially eligible. Of those, 781 were excluded because they underwent joint-preserving endoprosthetic reconstruction, recycled autografts, or extraarticular resection, leaving 110 who met the inclusion criteria for this study. Of those, 15% (16) of patients were lost to follow-up before the minimum follow-up of 2 years, leaving 85% (94) for analysis in this article at a mean ± SD follow-up time of 7 ± 4 years. The most common diagnoses were osteosarcoma (97% [91 of 94]) and Ewing sarcoma (3% [3 of 94]). The mean ± SD age was 11 ± 2 years; 57% (54 of 94) were male. At the last follow-up, 72% (68 of 94) of the patients had no evidence of disease, 9% (8) were alive with disease, 18% (17) had died of disease, and 1% (1) had died of other causes. Survivorship was estimated using the competing risk estimator, and data were presented at 5 and 10 years; outcome scores were derived from a longitudinally maintained institutional database. We reported on patients who developed major complications and underwent unplanned reoperation and minor complications that did not involve further surgery. Cox regression was used to evaluate the factors associated with knee subluxation and severe LLD (≥ 4 cm). RESULTS Five-year and 10-year survival of the surgically treated limb free of amputation for all patients was 96% (95% confidence interval [CI] 91% to 99%) and 90% (95% CI 81% to 96%), respectively. The 5-year endoprosthesis removal-free survival rate for all patients was 94% (95% CI 89% to 99%), and the 10-year survival rate was 85% (95% CI 75% to 94%). The 5-year endoprosthetic survivorship free from any revision surgery for all patients was 86% (95% CI 77% to 92%), and the 10-year endoprosthetic survivorship was 68% (95% CI 57% to 79%). The mean ± SD MSTS-93 score was 83% ± 7%. Twenty-eight percent (26 of 94) of patients underwent a total of 28 reoperations. Three percent (3 of 94) of patients underwent revision for knee subluxation (n = 1) and aseptic loosening (n = 2), and 11% (10 of 94) of patients underwent endoprosthesis revision surgery or amputation for local progression (n = 7) and infection (n = 3). No patient had an epiphysiodesis. After controlling for confounding variables such as gender, endoprosthetic type, and mesh reconstruction, multivariate analysis showed that previous surgery at the same site (cause-specific HR 10 [95% CI 5.2 to 59.0]; p < 0.001) and not using medial gastrocnemius flaps (cause-specific HR 7.1 [95% CI 1.4 to 33.0]; p = 0.02) were associated with the increased risk of knee subluxation, whereas age at operation ≤ 9 years was associated with the increased risk of severe LLD (≥ 4 cm) (cause-specific HR 7.3 [95% CI 3.7 to 25.0]; p = 0.002). CONCLUSION For skeletally immature patients with proximal tibial sarcomas, proximal tibial hemiarthroplasty appears to be a reasonable alternative to the standard rotating-hinge megaprosthesis, especially for pediatric patients age 10 years and older. This reconstruction can preserve the distal femoral epiphyseal growth capacity and thus potentially reduces final LLD. Moreover, patient age, skeletal maturity, implant availability, technical expertise, and surgeon preference should be considered when choosing a reconstructive approach after proximal tibial resection in children with osteosarcoma. This study did not compare pediatric patients treated with extendable prostheses. Future studies should consider direct comparisons between the two types of prosthetic reconstruction. LEVEL OF EVIDENCE Level IV, therapeutic study.