Survival, fusion, and hardware failure after surgery for spinal metastatic disease

OBJECTIVE Decompression with instrumented fusion is commonly employed for spinal metastatic disease. Arthrodesis is typically sought despite limited knowledge of fusion outcomes, high procedural morbidity, and poor prognosis. This study aimed to describe survival, fusion, and hardware failure after decompression and fusion for spinal metastatic disease. METHODS The authors retrospectively examined a prospectively collected, single-institution database of adult patients undergoing decompression and instrumented fusion for spinal metastases. Patients were followed clinically until death or loss to follow-up. Fusion was assessed using CT when performed for oncological surveillance at 6-month intervals through 24 months postoperatively. Estimated cumulative incidences for fusion and hardware failure accounted for the competing risk of death. Potential risk factors were analyzed with univariate Fine and Gray proportional subdistribution hazard models. RESULTS One hundred sixty-four patients were identified. The mean age ± SD was 62.2 ± 10.8 years, 61.6% of patients were male, 98.8% received allograft and/or autograft, and 89.6% received postoperative radiotherapy. The Kaplan-Meier estimate of median survival was 11.0 months (IQR 3.5–37.8 months). The estimated cumulative incidences of any fusion and of complete fusion were 28.8% (95% CI 21.3%–36.7%) and 8.2% (95% CI 4.1%–13.9%). Of patients surviving 6 and 12 months, complete fusion was observed in 12.5% and 16.1%, respectively. The estimated cumulative incidence of hardware failure was 4.2% (95% CI 1.5–9.3%). Increasing age predicted hardware failure (HR 1.2, p = 0.003). CONCLUSIONS Low rates of complete fusion and hardware failure were observed due to the high competing risk of death. Further prospective, case-control studies incorporating nonfusion instrumentation techniques may be warranted.