经蝶窦垂体神经内分泌肿瘤切除术后迟发性低钠血症预测模型的建立和验证  

Establishment and validation of a predictive model for delayed hyponatremia after resection of transsphenoidal pituitary neuroendocrine tumors

作　　者：章启铭 肖爵贤 于万里 罗海涛 程祖珏[1] Zhang Qiming;Xiao Juexian;Yu Wanli;Luo Haitao;Cheng Zujue(Department of Neurosurgery,the Second Affiliated Hospital of Nanchang University,Nanchang 330000,China)

机构地区：[1]南昌大学第二附属医院神经外科,南昌330000

出　　处：《中华神经外科杂志》2025年第2期158-164,共7页Chinese Journal of Neurosurgery

基　　金：中国中医科学院自主选题研究项目(ZZ15-WT-04)。

摘　　要：目的探讨经蝶窦垂体神经内分泌肿瘤(PitNET)切除术后迟发性低钠血症的潜在危险因素,建立预测模型并验证。方法回顾性分析2017年8月至2023年5月南昌大学第二附属医院神经外科行经蝶窦手术(TSS)治疗的386例PitNET患者的临床资料。按7∶3的比例将患者随机分配至建模队列(271例)和验证队列(115例)。根据是否发生迟发性低钠血症分为迟发性低钠血症组(简称低钠组)与无迟发性低钠血症组(对照组)。建模队列中,低钠组33例,对照组238例;验证队列中,低钠组16例,对照组99例。采用单因素分析和多因素logistic回归模型分析建模队列资料,评估术后发生迟发性低钠血症的重要影响因素,筛选模型的预测因子。应用编程软件R 4.3.2制作基于logistic回归模型的列线图。在建模队列和验证队列中,通过受试者工作特征(ROC)曲线评估列线图的预测性能,采用校准曲线评估预测模型与理想模型的拟合程度,采用决策曲线(DCA)评估其在临床中的应用价值。结果建模队列中,与对照组比较,低钠组患者术前肿瘤最大径较大,肿瘤侵袭海绵窦、鞍上者占比较高,术后第1天血钠浓度较低、术后低钾血症者占比较高、术后并发尿崩症者占比较低、术前垂体柄偏转角较大,术后垂体柄偏转角差值和鞍膈下沉深度均较大,差异均具有统计学意义(均P<0.05)。多因素logistic回归模型分析结果显示,术后并发低钾血症(OR=5.92,95%CI:2.22~15.77),术前肿瘤最大径(OR=1.08,95%CI:1.01~1.14)、垂体柄偏转角差值(OR=1.05,95%CI:1.01~1.10)和鞍膈下沉深度(OR=1.09,95%CI:1.01~1.19)较大,均为术后发生迟发性低钠血症的独立危险因素(均P<0.05);而术后并发尿崩症(OR=0.13,95%CI:0.04~0.39)、术后第1天血钠浓度较高(OR=0.77,95%CI:0.66~0.91)均为保护性因素(均P<0.05)。预测模型在建模和验证队列中的ROC曲线下面积(AUC)分别为0.90(95%CI:0.85~0.96)和0.88(95%CI:0.81~0.96)。HosmObjectiveTo investigate the potential risk factors of delayed hyponatremia after transsphenoidal pituitary neuroendocrine tumor(PitNET)resection,and to establish and validate a prediction model.MethodsThe clinical data of 386 PitNET patients who underwent transsphenoidal surgery(TSS)at the Department of Neurosurgery of the Second Affiliated Hospital of Nanchang University from August 2017 to May 2023 were retrospectively analyzed.Patients were randomly assigned in a 7∶3 ratio to either the modeling cohort(271 patients)or the validation cohort(115 patients).The patients were divided into delayed hyponatremia group(hyponatremia group)and non-delayed hyponatremia group(control group)according to whether delayed hyponatremia occurred.In the modeling cohort,there were 33 cases in the hyponatremia group and 238 cases in the control group.In the validation cohort,there were 16 patients in the hyponatremia group and 99 patients in the control group.Univariate analysis and multivariate logistic regression model were used to analyze the data of the modeling cohort to evaluate the important influencing factors of postoperative delayed hyponatremia and to screen the predictive factors of the model.The programming software R(4.3.2)was used to make the nomogram based on logistic regression model.In the modeling cohort and validation cohort,the receiver operating characteristic(ROC)curve was used to evaluate the predictive performance of the nomogram,the calibration curve was used to evaluate the fit degree of the prediction model to the ideal model,and the decision curve analysis(DCA)was used to evaluate its clinical application value.ResultsIn the modeling cohort,compared with the control group,the patients in the hyponatremia group had a larger preoperative tumor diameter,a higher proportion of tumors invading the cavernous sinus and suprasellar sinus,a lower serum sodium concentration on the first day after surgery,a higher proportion of patients with postoperative hypokalemia,a lower proportion of patients with postopera

关 键 词：垂体肿瘤 蝶窦 显微外科手术 低钠血症 影响因素分析 列线图 

分 类 号：R73[医药卫生—肿瘤]