Text

Objectives and questions: To determine thresholds of anatomical complexity where navigation enhances pedicle screw accuracy by identifying predictors of malposition in oncologic spine surgery.

Material and methods: In 56 patients with 444 pedicle screws undergoing posterior spinal instrumentation for primary or metastatic spinal tumors, pedicle screw accuracy was retrospectively analyzed, comparing intraoperative navigation with fluoroscopy-guided techniques. Screw placement was assessed on postoperative computed tomography using the Gertzbein-Robbins scale (GRS) and the pedicle scoring system (PSS) according to Zdichavsky et al. Tumor-related anatomical complexity was quantified using the Spinal Instability Neoplastic Score (SINS) and the Epidural Spinal Cord Compression (ESCC) scale. Screw-level mixed effects logistic regression models were applied to account for clustering within patients.

Results: Overall malposition rates were 27.7% (GRS ≥ C) and 8.8% (PSS ≥ IIa). Navigation independently reduced clinically relevant malposition (with ASIA C–E impact) across both grading systems (PSS: OR 0.15, p=0.020; GRS: OR 0.55, p=0.031). In multivariable analysis, neoadjuvant radiotherapy (OR 2.40, p=0.002) and ESCC grade 2-3 (OR 1.42, p=0.009) were independent predictors of malposition. The protective effect of navigation was most pronounced in patients with high anatomical complexity particularly those with advanced spinal cord compression ESCC 2-3. Neurological improvement was observed in 57%, with no postoperative deficits associated with navigation.

Discussion and conclusions: Thresholds of tumor-related anatomical complexity, especially in the presence of ESCC 2-3 and after neoadjuvant radiotherapy, guide selective navigation use in oncologic spine stabilization, allowing to be reserved for anatomically complex, high-risk constructs.