Abstract
BACKGROUND: Subsidence of an interbody implant, defined as its sinking into adjacent bone, is a common complication that can compromise spinal fusion outcomes and may require revision surgery. Custom 3D-printed titanium vertebral bodies are increasingly used for anterior spinal reconstruction. However, their reported subsidence incidence (15–30%) remains comparable to conventional cages. Early identification of high-risk patients is therefore crucial. This study aimed to identify preoperative imaging predictors of 3D-printed vertebral body subsidence and develop a nomogram for individualized risk prediction. METHODS: We retrospectively analyzed 102 patients (2020–2024) who underwent single-level anterior corpectomy and fusion with a custom 3D-printed titanium vertebral body. Preoperative CT and MRI metrics—including lower adjacent vertebral Hounsfield units (HU), endplate concavity depth, endplate angle, and Modic changes—were assessed as candidate risk factors. Subsidence was defined as implant sinking ≥ 3 mm at the prespecified 3-month postoperative follow-up. Independent predictors were identified by multivariate logistic regression, and a nomogram was constructed. Model performance was evaluated by area under the ROC curve (AUC), along with calibration and decision curve analysis, and internally validated via bootstrap resampling. RESULTS: Subsidence occurred in 32 of 102 patients (31.4%). Multivariate logistic regression identified lower adjacent vertebral HU, greater endplate concavity depth, and presence of Modic changes as independent predictors of subsidence. The predictive model showed excellent discrimination (AUC = 0.896, 95% CI 0.821–0.970) and remained robust after internal validation (bootstrap-corrected AUC ~ 0.886). It also demonstrated good calibration and a positive net clinical benefit on decision curve analysis. CONCLUSIONS: The major risk factors for 3D-printed vertebral body subsidence (poor bone quality and weak endplate integrity) are similar to those for conventional cages. Our imaging-based nomogram provides accurate risk stratification to guide personalized surgical strategies—such as bone density optimization or use of larger implants in high-risk cases—while avoiding unnecessary interventions in low-risk patients. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12891-026-09676-2.