Development and validation of a nomogram for predicting recurrence risk in pediatric ameloblastoma treated with conservative surgery.

BACKGROUND: Ameloblastoma (AM) constitutes a locally invasive, non-malignant odontogenic neoplasm with significant recurrence potential. In pediatric patients, treatment must balance effective tumor control with the preservation of normal jaw development, leading to a common preference for conservative surgical approaches such as marsupialization or local curettage. However, even with conservative treatment, a subset of children remains at a heightened risk of recurrence. Currently, there is a lack of dedicated clinical tools for predicting recurrence risk within pediatric patients. METHODS: In this retrospective study, 115 pediatric patients with AM who underwent conservative surgical treatment at the Hospital of Stomatology, Sun Yat-sen University between November 2005 and August 2025 were enrolled. Researchers gathered information from clinical examinations, radiographic images, and pathological findings. To pinpoint the key elements contributing to recurrence, both single-variable and multi-variable logistic regression analyses were carried out. Building on these significant predictors, a nomogram forecasting model was then developed. The model’s performance was evaluated by the area under the receiver operating characteristic curve (AUC), the Hosmer–Lemeshow goodness-of-fit test, calibration curve, and decision curve analysis (DCA). RESULTS: Recurrence occurred in 39 of the 115 patients (33.9%). Multivariate analysis identified female gender (OR = 0.325, P = 0.034) and the surgical approach of curettage and fenestration decompression (CFD) (OR = 0.231, P = 0.021) as protective factors against recurrence. In contrast, a honeycomb radiographic pattern (OR = 16.980, P = 0.018) and involvement of the inferior alveolar nerve (OR = 3.680, P = 0.036) proved to be a major red flag for recurrence. When we put these pieces together in a nomogram, the predictive power was quite impressive, with an AUC of 0.818 (95% CI: 0.734–0.902). The calibration curve indicated excellent agreement between ideal and observed outcomes. CONCLUSIONS: This research marks the creation and verification of the inaugural predictive nomogram designed to estimate the likelihood of recurrence in pediatric ameloblastoma cases managed through conservative surgical approaches. The model exhibits good discriminative ability and calibration, potentially serving as a practical tool to assist clinicians in optimizing surgical planning and tailoring treatment strategies, thereby improving long-term outcomes for affected children.