Abstract
This study aimed to develop and validate a nomogram for predicting postoperative hypernatremia risk in male nonfunctioning pituitary adenoma (NFPA) patients undergoing transsphenoidal surgery. A retrospective analysis was performed on the clinical data of 376 male patients with NFPAs who underwent transsphenoidal surgery at Fuzhou 900 Hospital and Anhui Provincial Hospital between January 2022 and December 2024. Patients from Fuzhou 900 Hospital were divided into the training set (a total of 275 cases, among which 51 cases were hypernatremia) for model development, while patients from Anhui Provincial Hospital were used as the external validation set (a total of 101 cases, among which 17 cases were hypernatremia). Univariate analysis and least absolute shrinkage and selection operator regression were used for preliminary screening of candidate risk factors for postoperative hypernatremia, followed by multivariate logistic regression analysis to identify independent predictors. A predictive nomogram was constructed based on these independent predictors using R software, and the performance of the nomogram was systematically evaluated in 3 domains: discriminative ability, calibration, and clinical utility. Univariate analysis and least absolute shrinkage and selection operator regression identified 7 candidate variables, and multivariate logistic regression confirmed 4 independent predictors of postoperative hypernatremia: preoperative cortisol level (odds ratio [OR] = 0.153, 95% confidence interval [CI]: 0.057-0.406, P = .002), intraoperative cerebrospinal fluid leak (OR = 8.40, 95% CI:1.36-51.9, P = .022), pituitary stalk morphological change (OR = 6.47, 95% CI: 1.09-38.60, P = .040), and pituitary stalk deflection angle difference (OR = 2.31, 95% CI: 1.22-4.40, P = .001). The constructed nomogram showed excellent discriminative ability in the training set (area under the receiver operating characteristic curve = 0.8645, 95% CI: 0.7499-0.9791) and good performance in the external validation set (area under the receiver operating characteristic curve = 0.7983, 95% CI: 0.6596-0.9371). Calibration plots and the Hosmer-Lemeshow test (P > .05) indicated good calibration, and decision curve analysis and the clinical impact curve confirmed significant clinical utility. A predictive nomogram for postoperative hypernatremia in male NFPA patients was established, integrating preoperative cortisol levels, intraoperative cerebrospinal fluid leak status, and pituitary stalk dynamic imaging parameters to inform evidence-based, personalized perioperative management.