Risk factors for hardware failure after total spondylectomy in patients with spinal tumors

Introduction Total spondylectomy for spinal tumors provides optimal local control and is associated with a high risk of implant instability. The objective was to determine risk factors for implant instability after spondylectomy in patients with neoplastic lesions of the spine. Material and methods A retrospective cohort study included patients with spinal tumors treated with tumor resection between 2007 and 2023. Inclusion criteria were spondylectomy and vertebral body replacement, thoracic or lumbar spine localization, follow-up period ≥ 12 months. LASSO regression and Random Forest methods and multivariate analysis were used to identify instability predictors. Results Implant instability was observed in 16 patients (18.4 %). Risk factors included the use of bone cement instead of allograft (OR = 0.125, p = 0.014), contact surface mismatch > 10° (OR = 0.214, p = 0.026), prosthesis subsidence > 2 mm at 3 months (OR = 4.497, p = 0.023). Discussion The risk factors identified had a great clinical role for the prevention of implant instability. The use of bone graft instead of cement, precise matching of contact surfaces and control of early prosthetic subsidence can significantly reduce the risk of metal construct failure. Careful preoperative planning and regular postoperative monitoring are essential for the outcome. Conclusion Three independent risk factors for implant instability after spondylectomy identified in patients with spinal tumor lesions included the use of bone cement instead of allograft, a discrepancy between the contact prosthetic surfaces of more than 10°, and an implant subsidence of more than 2 mm after 3 months. These factors are important for planning of the surgical intervention and postoperative monitoring to prevent metal construct instability.