Silent corticotroph adenoma (SCA) exhibits high invasiveness and recurrence; thus, accurate prediction of postoperative recurrence is crucial. To predict recurrence-free survival (RFS) in SCA using habitat analysis (subregion radiomics) and to develop a comprehensive nomogram integrating habitat scores, clinical, radiological, and pathological features. This retrospective study included 325 SCA patients, randomly assigned to training and testing cohorts in a 7:3 ratio. Radiomics features were extracted from tumor subregions clustered via K-means based on CE-T1WI and T2WI. Feature selection involved t-tests, Pearson / Spearman correlation, and the least absolute shrinkage and selection operator (LASSO) regression, and the habitat score was calculated based on the selected subregional radiomics features and their corresponding coefficients. Kaplan–Meier and Cox regression analyses were used to identify prognostic factors. A nomogram incorporating independent predictors was constructed and validated for RFS prediction. Predictive performance was evaluated through the Harrell C-index, calibration curves, and decision curve analysis (DCA). Habitat scores significantly stratified patients by RFS ( p < 0.001). Multivariate Cox analysis identified habitat score, Ki-67 index, surgical method, and postoperative gamma knife radiotherapy as independent predictors. Calibration and DCA curves confirmed good agreement and clinical utility. Habitat scores derived from subregional radiomics provide good prognostic value for RFS prediction in SCA. The proposed nomogram enables individualized recurrence risk assessment, supporting postoperative decision-making.
Zhao et al. (Tue,) studied this question.