Pedicle Involvement Is Associated with Preoperative Neurological Deficit in Metastatic Spinal Cord Compression: A Single-Institution Cohort Study.

Abstract

To evaluate whether pedicle involvement - tumor invasion of Weinstein-Boriani-Biagini (WBB) zones 3, 4, 9, or 10 - is associated with preoperative neurological deficit in metastatic spinal cord compression (MSCC), and to compare early recovery between surgical approaches in this subgroup. Thirty-nine patients with high-grade epidural spinal cord compression (ESCC grade ≥ 2) treated at a single tertiary center in Taiwan between September 2021 and December 2025 were reviewed retrospectively. The primary outcome was the American Spinal Injury Association motor score (AMS). Interobserver reliability was quantified by Cohen's kappa. Multiple linear regression was supported by variance inflation factors, bootstrap validation, and a prespecified incremental comparison against conventional radiographic measures. Early recovery (ΔAMS) was compared between corpectomy and laminectomy. Twenty-five patients (64%) had pedicle involvement and lower preoperative AMS (55 ± 17 vs. 83 ± 15, p < 0.001). Interobserver κ was 0.62. Pedicle involvement was the only independent correlate (β = -23.645, p < 0.001; n = 35; all variance inflation factors < 1.4). Adding it to conventional measures raised R² from 0.049 to 0.440 (ΔR² = 0.391, p < 0.001). Corpectomy showed greater early ΔAMS than laminectomy (median 14.5 vs. 4.5, p = 0.042) but longer operative time and greater blood loss; complication (47% vs. 17%) and 30-day mortality (16% vs. 0%) differences were non-significant but underpowered. In our practice, pedicle involvement was independently associated with lower preoperative AMS and added explanatory value beyond conventional radiographic measures. The surgical comparison was small, underpowered, and exploratory.