上海口腔医学

上海口腔医学 ›› 2025, Vol. 34 ›› Issue (6): 595-600.doi: 10.19439/j.sjos.2025.06.006

• 论著 • 上一篇    下一篇

舍格伦综合征患者发生淋巴瘤的风险预测模型建立与效能评价

范怡嘉1, 陆欣悦1, 吕中静1,2   

  1. 1.徐州医科大学口腔医学院, 江苏 徐州 221004;
    2.徐州医科大学附属医院 口腔科, 江苏 徐州 221006
  • 收稿日期:2024-10-09 修回日期:2024-11-14 发布日期:2025-12-30
  • 通讯作者: 吕中静,E-mail: Zhongjing_lv2012@163.com
  • 作者简介:范怡嘉(1998-),女,硕士研究生,E-mail: 3154138854@qq.com
  • 基金资助:
    徐州市科技计划基金项目（KC22217）; 江苏省肿瘤生物治疗重点实验室课题基金（XZSYSKF2023032）

Establishment and efficacy evaluation of a risk prediction model for lymphoma in patients with Sjögren’s syndrome

Fan Yijia1, Lu Xinyue1, Lyu Zhongjing1,2   

  1. 1. School of Stomatology, Xuzhou Medical University. Xuzhou 221004;
    2. Department of Stomatology, Affiliated Xuzhou Municipal Hospital of Xuzhou Medical University. Xuzhou 221006, Jiangsu Province, China
  • Received:2024-10-09 Revised:2024-11-14 Published:2025-12-30

PDF (PC)

11

摘要： 目的:探讨原发性舍格伦综合征（primary Sjögren's syndrome,pSS）患者发生淋巴瘤的风险因素,构建风险预测模型并评判其预测效能。方法:收集2019年2月—2024年8月于徐州医科大学附属医院口腔科就诊并明确诊断为pSS的患者202例,根据是否发生淋巴瘤分为发生组（23例）和未发生组（179例）。采用SPSS 26.0软件包中的单因素分析和多因素logistic回归筛选风险因素,构建预测模型,绘制列线图。通过受试者工作特征（receiver operating characteristic, ROC）曲线,校准曲线对模型进行评价,并绘制临床决策曲线。结果:单因素分析显示,淋巴结病变、腺体肿大、中性粒细胞减少、血小板减少、淋巴细胞减少、高IgG水平、间质性肺炎、肾脏受累等因素在淋巴瘤发生组和未发生组间有统计学差异（P<0.05）。多因素logistic回归分析显示,淋巴结病变、中性粒细胞减少、淋巴细胞减少和高IgG水平是淋巴瘤发生的独立危险因素。ROC曲线下面积（area under the curve, AUC）值为0.931。校准曲线表明模型的预测性能较好,决策曲线在5%~95%的阈值概率区间内,预测模型的净收益较高。结论:淋巴结病变、中性粒细胞减少、淋巴细胞减少和高IgG水平可作为pSS患者发生淋巴瘤的预测因子。该模型有助于评估pSS患者发生淋巴瘤的风险,进而积极干预,减少淋巴瘤的发生。

关键词: 舍格伦综合征, 淋巴瘤, 风险因素, 模型预测

Abstract: PURPOSE: To investigate the risk factors for lymphoma in patients with primary Sjögren's syndrome (pSS), construct a risk prediction model, and evaluate its predictive efficacy. METHODS: A total of 202 patients diagnosed with pSS at the Department of Stomatology, Affiliated Hospital of Xuzhou Medical University from February 2019 to August 2024 were collected and divided into two groups: lymphoma occurrence group(n=23) and non-occurrence group(n=179). Univariate analysis and multivariate logistic regression were performed to screen risk factors with SPSS 26.0 software package, and a risk prediction nomogram model was established. The model was evaluated by the receiver operator characteristic curve(ROC) and calibration curve(CC). Then, the decision curve analysis(DCA) was plotted. RESULTS: Univariate analysis showed that lymphadenopathy, swollen glands, neutropenia, thrombocytopenia, lymphocytopenia, high IgG levels, interstitial pneumonia, and renal damages had significant differences between two groups (P<0.05). Multivariate logistic regression analysis showed that lymphadenopathy, neutropenia, lymphocytopenia, and high IgG levels were independent risk factors. The area under the curve(AUC) was 0.931. The calibration curve indicated that the model had good predictive performance. The decision curve showed that the prediction model had a high net benefit within the threshold probability range of 5%-95%. CONCLUSIONS: Lymphadenopathy, neutropenia, lymphocytopenia, and high IgG levels can be used as predictive factors for lymphoma occurrence in pSS. This risk prediction model is helpful for assessing the risk of lymphoma occurrence in pSS, thereby actively intervening to reduce the occurrence of lymphoma.

Key words: Primary Sjögren's syndrome, Lymphoma, Risk factors, Prediction model

中图分类号: