基于常规血液标志物的早期胰腺癌诊断模型

Diagnostic Model for Early Pancreatic Cancer Based on Routine Blood Biomarkers

摘要: 目的：胰腺导管腺癌(PDAC)早期诊断困难，现有诊断方法耗时且侵入性强。本研究旨在构建一种基于常规血液标志物的非侵入性诊断模型，以提高PDAC早期诊断效率。方法：本研究为回顾性病例对照研究，纳入2018年6月至2024年3月青岛大学附属医院677例胰胆疾病患者(PDAC组210例，对照组467例)。采用倾向性评分匹配(PSM)平衡基线特征，最终纳入410例(每组205例)。暴露变量为术前血液标志物(CA19-9、CEA、DBIL、CYS-C、COUNT评分)，结局变量为PDAC诊断，协变量包括年龄、性别、身高、体重、肝肾功能及炎症营养指标等。结果：最终模型纳入CA19-9、CEA、DBIL、CYS-C和COUNT评分五个变量，ROC曲线下面积(AUC)为0.86 (95%CI: 0.82~0.90)，准确率为77% (95%CI: 0.72~0.82)。多因素分析显示，CEA (OR = 1.16, 95%CI: 1.02~1.32)和DBIL (OR = 1.02, 95%CI: 1.01~1.02)为独立危险因素，CYS-C (OR = 0.06, 95%CI: 0.01~0.36)和COUNT评分 = 2 (OR = 0.21, 95%CI: 0.08~0.56)为保护因素。结论：本研究构建的五维血液标志物模型可有效鉴别PDAC，具有较高的诊断准确性和临床应用潜力，有望成为PDAC早期诊断的非侵入性工具。

Abstract: Objective: Pancreatic ductal adenocarcinoma (PDAC) is challenging to diagnose early, and current methods are invasive and time-consuming. This study aimed to develop a non-invasive diagnostic model based on routine blood biomarkers to improve early PDAC detection. Methods: A retrospective case-control study was conducted at Qingdao University Hospital from June 2018 to March 2024, including 677 patients with pancreatobiliary diseases (210 PDAC, 467 controls). Propensity score matching (PSM) balanced baseline characteristics, resulting in 410 matched cases (205 per group). Exposure variables were preoperative blood biomarkers (CA19-9, CEA, DBIL, CYS-C, COUNT score), outcome variable was PDAC diagnosis, and covariates included age, sex, height, weight, liver/ kidney function, inflammatory indicators and nutritional assessment indices. Results: The final model incorporated five variables: CA19-9, CEA, DBIL, CYS-C, and COUNT score, achieving an area under ROC curve (AUC) of 0.86 (95%CI: 0.82~0.90) and accuracy of 77% (95%CI: 0.72~0.82). Multivariate analysis identified CEA (OR = 1.16, 95%CI: 1.02~1.32) and DBIL (OR = 1.02, 95%CI: 1.01~1.02) as independent risk factors, while CYS-C (OR = 0.06, 95%CI: 0.01~0.36) and COUNT score = 2 (OR = 0.21, 95%CI: 0.08~0.56) were protective factors. Conclusion: The five-dimensional blood biomarker model constructed in this study effectively distinguishes PDAC with high diagnostic accuracy and clinical application potential, showing promise as a non-invasive tool for early PDAC diagnosis.