Salvage Posterior C1-C2 Fusion for Odontoid Nonunion After Failed Nonoperative Management: A Propensity Score-Matched...

Background/Objectives: Posterior C1-C2 fusion is commonly used for unstable traumatic odontoid injuries, but it is less commonly used for patients who initially undergo nonoperative management and later require salvage fusion. This study compared hospital length of stay, short-term complications, and postoperative radiographic alignment between salvage posterior C1-C2 fusion after failed nonoperative management and primary posterior C1-C2 fusion. Materials and Methods: A retrospective cohort study was performed of 106 adult patients who underwent posterior C1-C2 instrumented fusion for traumatic cervical spine injuries from 2011 to 2023. Patients were stratified into the salvage fusion group after radiographic nonunion following attempted nonoperative management with external immobilization or the primary fusion group, who underwent initial surgical management. The primary outcome was hospital length of stay. Secondary outcomes included postoperative radiographic alignment, screw loosening, hardware failure, revision surgery, and 30-day emergency department visits. Propensity score matching and full-cohort augmented inverse probability weighting were used to account for baseline differences between groups. Results: Twenty-seven patients underwent salvage fusion and 79 underwent primary fusion. Propensity score matching produced 25 matched pairs. In the matched cohort, salvage fusion was associated with significantly shorter length of stay than primary fusion, with a median of 2 versus 5 days, respectively (p < 0.001). This remained significant in the full-cohort augmented inverse probability weighting analysis, where salvage fusion was associated with a 2.41-day reduction in length of stay (95% CI, -3.63 to -1.19; p < 0.001). Short-term complications were uncommon in both groups, and no clear sign of increased screw loosening, hardware failure, revision surgery, or 30-day emergency department visits was observed in the salvage cohort. Salvage fusion was also associated with lower postoperative C2-C7 lordosis and a greater C1 lamina-occiput distance. Conclusions: Salvage posterior C1-C2 fusion for radiographic nonunion after attempted nonoperative management was not associated with higher short-term complication rates compared with primary fusion. While surgical-admission length of stay was shorter in the salvage cohort, this difference should be interpreted cautiously because salvage and primary fusion occur in different admission contexts and do not reflect the total episode-of-care burden. Early postoperative alignment differences were observed, but these were not correlated with clinical outcomes or longitudinal imaging, and their long-term significance remains unclear. Future multicenter studies should evaluate total healthcare utilization, fusion status, longitudinal alignment, and patient-reported outcomes after salvage C1-C2 fusion.