Dai Zhihao, Jiang Jin, Chen Qianping, Bai Minghua, Sun Quanquan, Feng Yanru, Liu Dong, Wang Dong, Zhang Tong, Han Liang, Ng Litheng, Zheng Jun, Zou Hao, Mao Wei, Zhu Ji
School of Public Health, Nanjing Medical University, Nanjing, 211166, China; Department of Radiation Oncology, Zhejiang Cancer Hospital, Hangzhou, 310000, Zhejiang, China; Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou 310000, China.
Department of Radiation Oncology, Zhejiang Cancer Hospital, Hangzhou, 310000, Zhejiang, China; Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou 310000, China; Department of Oncology, Affiliated Hospital of Jiaxing University, The First Hospital of Jiaxing, Jiaxing, 31400, China.
Transl Oncol. 2025 Jan;51:102190. doi: 10.1016/j.tranon.2024.102190. Epub 2024 Nov 13.
Gastric cancer (GC) is a common malignant tumor, and early diagnosis significantly improves patient survival rates. This study aimed to investigate the diagnostic value of ring finger protein 180 (RNF180) and secreted frizzled protein 2 (SFRP2) in GC.
MATERIALS & METHODS: A total of 165 healthy individuals, 34 patients with precancerous gastric lesions, and 104 patients with confirmed GC were divided into training and validation sets; methylated RNF180 and SFRP2 were detected in circulating DNA from blood samples. Six models, including those based on logistic regression, Naive Bayes, K-nearest neighbor algorithm, glmnet, neural network, and random forest (RF) were built and validated. Area under the curve (AUC), sensitivity, specificity, positive predictive value, and negative predictive value were determined.
In the training set, the RF model with RNF180 and SFRP2 (R + S) had an AUC of 0.839 (95 % CI: 0.727-0.951), sensitivity of 60.3 %, and specificity of 85.5 % for diagnosing GC. The RF model with R + S+ Tumor markers had an AUC of 0.849 (95 % CI: 0.717-0.981), sensitivity of 62.8 %, and specificity of 87.1 %. In the validation set, the RF model with R + S had an AUC of 0.844 (95 % CI: 0.774-0.923), sensitivity of 87.8 %, and specificity of 69.2 %. The RF model with R + S + Tumor markers had an AUC of 0.858 (95 % CI: 0.781-0.939), sensitivity of 85.4 %, and specificity of 76.9 %.
Our results suggest that RNF180 and SFRP2 could serve as diagnostic biomarkers for GC when using the RF model.
胃癌（GC）是一种常见的恶性肿瘤，早期诊断可显著提高患者生存率。本研究旨在探讨环状泛素连接酶180（RNF180）和分泌型卷曲相关蛋白2（SFRP2）在胃癌中的诊断价值。
将165名健康个体、34例胃癌前病变患者和104例确诊胃癌患者分为训练集和验证集；检测血液样本循环DNA中RNF180和SFRP2的甲基化情况。构建并验证了包括逻辑回归、朴素贝叶斯、K近邻算法、广义线性模型、神经网络和随机森林（RF）在内的六种模型。测定曲线下面积（AUC）、灵敏度、特异性、阳性预测值和阴性预测值。
在训练集中，基于RNF180和SFRP2（R+S）的RF模型诊断胃癌的AUC为0.839（95%CI：0.727-0.951），灵敏度为60.3%，特异性为85.5%。基于R+S+肿瘤标志物的RF模型AUC为0.849（95%CI：0.717-0.981），灵敏度为62.8%，特异性为87.1%。在验证集中，基于R+S的RF模型AUC为0.844（95%CI：0.774-0.923），灵敏度为87.8%，特异性为69.2%。基于R+S+肿瘤标志物的RF模型AUC为0.858（95%CI：0.781-0.939），灵敏度为85.4%，特异性为76.9%。
我们的结果表明，在使用RF模型时，RNF180和SFRP2可作为胃癌的诊断生物标志物。
Zotero 插件
