Risk factor identification and nomogram-based predictive model development for recurrence of pituitary adenomas following transsphenoidal surgery: a dual-center retrospective study

Objective To identify independent risk factors for pituitary adenoma recurrence after transsphenoidal surgery and construct and validate a nomogram for recurrence risk quantification. Methods We retrospectively collected clinical data of 469 patients with pituitary adenomas from Fuzhou 900 Hospital and Huashan Hospital of Fudan University. LASSO regression was applied to screen potential predictive factors, followed by multivariate Logistic regression to identify independent prognostic factors. Nomograms were subsequently constructed to evaluate the model's discriminative ability, calibration, and clinical utility. Results Recurrence occurred in 64 cases (13.65%). Multivariate analysis confirmed modified Knosp classification (grades 3B–4), clival invasion, and subtotal resection as independent risk factors ( P < 0.05). The model had AUC 0.905 [95%CI (0.8254, 0.9847)], C-index 0.905, sensitivity 85%, specificity 86.4%, with good calibration and clinical net benefits (DCA: 0.05–0.90 threshold). Conclusion The nomogram, which is the first to integrate clival invasion with modified Knosp classification and extent of resection, effectively predicts pituitary adenoma recurrence after transsphenoidal surgery and shows promising clinical utility.