Nomogram prediction model for postoperative anterior pituitary dysfunction in patients with nonfunctioning pituitary adenomas: A multidimensional analysis and validation of Short-Term and Mid-to-Long-Term outcomes.

PURPOSE: To develop and validate a nomogram for predicting postoperative anterior pituitary dysfunction (APD) in patients with nonfunctioning pituitary adenomas (NFPAs).

METHODS

A total of 496 NFPA patients who underwent surgery at the Second Hospital of Lanzhou University between June 2018 and June 2024 were included, of whom 305 completed mid- to long-term postoperative follow-up. Patients were randomly assigned to training and validation cohorts (8:2 ratio) and classified into APD or non-APD groups. Clinical, radiological, and pathological variables were compared between groups. Multivariate logistic regression identified independent predictors of APD, new-onset APD, new-onset ACTH axis deficiency, and APD recovery at different postoperative stages. Model performance and clinical utility were evaluated using the area under the receiver operating characteristic curve (AUC), Hosmer-Lemeshow (H-L) test, and decision curve analysis (DCA).

RESULTS

(1) Independent risk factors for postoperative APD included male sex, low BMI, invasive tumor, preoperative APD, and low preoperative T4 and IGF-1 indices. New-onset APD was associated with sex, pituitary stalk compression, cavernous sinus invasion, low preoperative albumin, and low IGF-1. New-onset ACTH axis deficiency was linked to preoperative systemic inflammatory response index, uric acid (UA), phosphorus, and LDL-C. APD recovery was associated with BMI and the number of preoperative APD axes (all P < 0.05). (2) Independent predictors of APD were male sex, larger tumor volume, low preoperative sodium, high apolipoprotein B, low preoperative T3 and IGF-1 indices, short-term low IGF-1, and absence of hormone replacement. New-onset APD was associated with preoperative diastolic blood pressure, tumor volume, preoperative LDL-C, and short-term low IGF-1. New-onset ACTH axis deficiency was associated with tumor size, preoperative UA, low IGF-1 index, short-term low T4, and short-term low PRL. APD recovery correlated with preoperative sodium and IGF-1 index (all P < 0.05). (3) The nomograms demonstrated good predictive ability. For short-term outcomes, the AUCs for postoperative APD, new-onset APD, new-onset ACTH deficiency, and APD recovery were 0.78 (0.80), 0.72 (0.64), 0.80 (0.63), and 0.79 (0.74), respectively. For mid- to long-term outcomes, AUCs were 0.91 (0.82), 0.84 (0.86), 0.93 (0.69), and 0.82 (0.71), respectively. All models showed good calibration (H-L test, P > 0.05), and DCA confirmed clinical usefulness across a broad threshold range.

CONCLUSION

The proposed nomogram models reliably predict both the occurrence and recovery of APD after NFPA surgery, demonstrating strong clinical applicability.