Supplementary fixation enhances bone union in extracorporeal irradiated autograft reconstruction for malignant bone tumors: A competing-risk analysis from a tertiary referral center.

Abstract

PurposeExtracorporeal irradiation (ECRT) autograft reconstruction is an established option after limb-salvage resection of primary bone malignancy. Prior series have reported the proportion of junctions uniting rather than the rate of union, seldom accounting for oncologic mortality. We examined factors influencing union using competing-risk methods.MethodsOf 63 eligible patients undergoing wide resection and ECRT reconstruction between 2006 and 2024 at a single tertiary academic center, 50 (68 graft-host junctions) with complete data were reviewed retrospectively. Time to union was analyzed by Fine-Gray subdistribution hazard regression, with death, graft removal, and loss to follow-up as competing events. Eight covariates, including supplementary fixation, were entered into the multivariable model.ResultsOf 68 junctions, 57 (83.8%) united, six (8.8%) sustained a competing event before union, and five (7.4%) were censored. Median follow-up was 38 months (IQR 18-64). Supplementary fixation was independently associated with accelerated union (subdistribution hazard ratio 3.42, 95% CI 1.78-6.60, p < 0.001), whereas demographic and tumor-related variables were not. Cumulative incidence of union without versus with supplementary fixation was 63% (95% CI 49-74%) versus 100% at 12 months, and 78% (95% CI 64-87%) versus 100% at 24 months. Complications included late infection (13.2%), graft resorption (10.3%), and non-union requiring reoperation (1.5%). Local recurrence occurred in 9 patients (18.0%).ConclusionsSupplementary fixation was independently associated with faster union after ECRT autograft reconstruction. Fixation was applied selectively when intraoperative stability was judged insufficient, so confounding by indication cannot be excluded. These findings are hypothesis-generating and require confirmation in larger prospective cohorts.