Abstract
BACKGROUND: Cystitis glandularis (CG) is a chronic inflammatory condition of the bladder characterized by a high recurrence rate, imposing a substantial burden on patients. The mechanisms underlying recurrence remain unclear. OBJECTIVES: This study aims to identify markers associated with CG recurrence and develop a predictive model for recurrence risk. DESIGN: Retrospective cohort study of patients with confirmed recurrence based on outpatient visits or readmissions were included in this study, which was subsequently divided into training and test set. METHODS: Patients diagnosed with CG from four hospitals between 2013 and 2023 were retrospectively included and followed for one year. Recurrence was defined as the appearance of new nonneoplastic lesions on cystoscopy after complete resection of the primary disease. A total of 161 patients were divided into a training set (n = 98) from XiangYa Hospital and a test set (n = 63) from Shaoyang Central Hospital, the Second Affiliated Hospital of South China University, and the First People's Hospital of Changde City. Cox regression analysis was performed in the training set to identify serological indicators associated with recurrence, which were further validated at the histological level by immunohistochemistry. A prognostic model was then constructed using LASSO regression, and its predictive performance was evaluated using receiver operating characteristic (ROC) curves. A nomogram was also developed for clinical application. RESULTS: Among 161 patients followed for 12 months, the recurrence rate was 49.6% (n = 80). Univariate and multivariate Cox regression analyses revealed that serological eosinophil and basophil counts were significantly associated with CG recurrence, with histological validation confirming their relevance. The LASSO-based risk model demonstrated good predictive ability, with an area under the ROC curve exceeding 0.75. CONCLUSION: Serological indicators, specifically eosinophil and basophil counts, are closely linked to CG recurrence. A risk score model based on these markers was developed, providing effective prediction of recurrence in clinical practice.