Wang Haiyu, Ding Yumin, Zhao Shujing, Li Kaixu, Li Dehong
School of Public Health, Gansu University of Chinese Medicine, Lanzhou, Gansu, China.
Department of Clinical Laboratory, Gansu Provincial Hospital, Lanzhou, Gansu, China.
Front Oncol. 2024 Nov 29;14:1463480. doi: 10.3389/fonc.2024.1463480. eCollection 2024.
Identifying high-risk populations and diagnosing gastric cancer (GC) early remains challenging. This study aimed to establish and verify a nomogram model for the early diagnosis of GC based on conventional laboratory indicators.
We performed a retrospective analysis of the clinical data of 2,770 individuals with first diagnosis of GC and 1,513 patients with benign gastric disease from January 2018 to December 2022. The cases were divided into the training set and validation set randomly, with a ratio of 7:3. Variable screening was performed by least absolute shrinkage and selection operator (LASSO) and logistic regression analysis. A nomogram was constructed in the training set to assist in the early diagnosis of GC.
There were 4283 patients included in the study, with 2998 patients assigned in the training set and 1285 patients in the validation set. Through LASSO regression and logistic regression analysis, independent variables associated with GC were identified, including CEA, CA199, LYM, HGB, MCH, MCHC, PLT, ALB, TG, HDL, and AFR. The nomogram model was constructed using the above 11 independent indicators. The AUC was 0.803 for the training set and 0.797 for the validation set, indicating that the model showed high clinical diagnostic efficacy. The calibration curves and decision curve analysis (DCA) of the nomogram presented good calibration and clinical application ability.
Based on the analysis of large sample size, we constructed a nomogram model with 11 routine laboratory indicators, which showed good discrimination ability and calibration.
识别高危人群并早期诊断胃癌（GC）仍然具有挑战性。本研究旨在建立并验证一种基于传统实验室指标的胃癌早期诊断列线图模型。
我们对2018年1月至2022年12月首次诊断为GC的2770例患者和1513例良性胃病患者的临床资料进行了回顾性分析。病例随机分为训练集和验证集，比例为7:3。通过最小绝对收缩和选择算子（LASSO）和逻辑回归分析进行变量筛选。在训练集中构建列线图以辅助GC的早期诊断。
本研究共纳入4283例患者，其中2998例分配到训练集，1285例分配到验证集。通过LASSO回归和逻辑回归分析，确定了与GC相关的独立变量，包括癌胚抗原（CEA）、糖类抗原199（CA199）、淋巴细胞（LYM）、血红蛋白（HGB）、平均红细胞血红蛋白含量（MCH）、平均红细胞血红蛋白浓度（MCHC）、血小板（PLT）、白蛋白（ALB）、甘油三酯（TG）、高密度脂蛋白（HDL）和甲胎蛋白（AFR）。使用上述11个独立指标构建列线图模型。训练集的曲线下面积（AUC）为0.803，验证集为0.797，表明该模型具有较高的临床诊断效能。列线图的校准曲线和决策曲线分析（DCA）显示出良好的校准和临床应用能力。
基于大样本量分析，我们构建了一个包含11个常规实验室指标的列线图模型，该模型具有良好的区分能力和校准性。
