A predictive nomogram for progression-free survival in radioiodine-refractory papillary thyroid carcinoma after reoperation: a longitudinal study in Vietnam.

Radioiodine-refractory (RAI-R) papillary thyroid carcinoma (PTC) with disease progression is associated with poor prognosis. This study aimed to develop a predictive nomogram for progression-free survival (PFS) in RAI-R PTC by integrating clinical, pathological, and mutational features. In this longitudinal study, 145 patients with RAI-R PTC who underwent reoperation for locoregional recurrence were followed for disease progression. A Bayesian model averaging (BMA) approach was employed to identify the optimal model for predicting PFS. Model performance was assessed using time-dependent area under the receiver operating characteristic curve (AUC), Brier scores, and calibration plots. A nomogram for PFS prediction was then developed based on parameters from the optimal model. During a median follow-up of 29.7 months, 38 patients (26.2%) experienced disease progression after reoperation. Among eight candidate variables, BMA identified an optimal model that included four key predictors: treatment response, extrathyroidal/extranodal extension, mitotic count, and coexisting BRAFV600E and TERT promoter mutations (Dmut). The hazard ratios (95% confidence intervals) were 3.86 (1.54-9.72) for biochemical incomplete response, 9.39 (3.35-26.3) for structural incomplete response, 3.40 (1.69-6.85) for extrathyroidal/extranodal extension, 1.34 (1.13-1.58) for mitotic count, and 6.08 (2.79-13.2) for Dmut. The optimal model demonstrated strong discriminative performance and stability across time points, with AUCs (95% confidence intervals) of 0.905 (0.801-1.000), 0.919 (0.857-0.981), and 0.921 (0.861-0.981) at 12, 24, and 36 months, respectively, and showed good calibration. This study demonstrates that integrating treatment response, extrathyroidal/extranodal extension, mitotic count, and Dmut into a predictive nomogram offers a clinically relevant tool for stratifying progression risk in patients with RAI-R PTC.