慢性阻塞性肺疾病合并肺癌的“证候基因环境”预测模型的构建

Construction of a “TCM syndrome-gene-environment” model for predicting lung cancer risk among patients with chronic obstructive pulmonary disease

关键词:  慢性阻塞性肺疾病  肺癌  证候  基因多态性  预测模型

Key words:  chronic obstructive pulmonary disease  lung cancer  traditional Chinese medicine syndrome  gene polymorphisms  prediction model

摘要:

目的 分析慢性阻塞性肺疾病（chronic obstructive pulmonary disease, COPD）合并肺癌的中医证候、易感基因多态性和环境因素的特点，探索相关的危险因素并建立预测模型。方法 研究对象来源于2020年1月—2022年12月同济大学附属上海市肺科医院收治的COPD（无肺癌）患者155例、COPD合并肺癌患者155例，分别设为COPD组、COPD合并肺癌组。收集所有研究对象的16种中医证候要素信息、8个候选基因的11个单核苷酸多态性位点（single nucleotide polymorphism, SNP）信息、环境因素及临床参数信息，使用单因素及多因素Logistic回归分析筛选出独立危险因素，构建COPD合并肺癌的预测模型并绘制受试者工作特征（receiver operating characteristic, ROC）曲线。结果 环境因素和临床参数的单因素分析结果显示COPD合并肺癌组的BMI水平低于COPD组，吸烟包年数水平和肺气肿发生率均高于COPD组（P<0.05）；中医证候要素的单因素分析结果显示COPD合并肺癌组的血瘀证、痰证的出现比例较COPD组显著增加（P<0.05）；基因多态性位点的频率分布和关联分析结果显示携带HHIP基因rs1489759基因型AA或GA、携带CYP2A6基因rs56113850基因型CC或TC的个体发生COPD合并肺癌的风险分别较携带GG和TT的个体更高。多因素Logistic回归分析筛选出痰证、rs1489759（基因型AA或GA）、rs56113850（基因型CC或TC）、BMI偏低、吸烟包年数偏高是COPD合并肺癌的独立危险因素，计算回归方程并构建“证候-基因-环境”的预测模型，其ROC曲线下面积（area under curve, AUC）为0.742，拟合优度较高。结论 本研究开发出一种用于COPD患者中肺癌风险筛查的“证候-基因-环境”中西医多维度危险因素的预测模型，具有一定的预测价值，旨在为COPD人群的肺癌风险筛查提供评估和参考。

Abstract:

Objective To develop a prediction model for lung cancer risk among patients with chronic obstructive pulmonary disease(COPD) based on traditional Chinese medicine(TCM) syndrome, susceptibility gene polymorphisms and environmental factors. Methods The study subjects, including 155 patients with COPD(COPD group) and 155 COPD patients complicated with lung cancer(COPD-LC group) were enrolled from Shanghai Pulmonary Hospital affiliated to Tongji University from January 2020 to December 2022. The information of 16 types of TCM syndrome elements, the 11 single nucleotide polymorphisms(SNP) in 8 candidate genes, environmental factors and clinical parameters of all subjects were collected. Univariate and multivariate Logistic regression analyses were applied to select the independent risk factors and to construct the prediction model of lung cancer risk in COPD patients, and the performance of the model was evaluated with the receiver operating characteristic(ROC) curve. Results Univariate analysis of environmental factors and clinical parameters showed that the BMI level was lower, the smoking index(pack-years) and the incidence of emphysema in COPD-LC group were higher than those in the COPD group(P<0.05). Univariate analysis of TCM syndrome elements showed that the distribution proportion of blood stasis syndrome and phlegm syndrome in COPD-LC group increased significantly compared with COPD group(P<0.05). For all SNP, the genotype frequency distribution and association analysis showed that subjects with rs1489759 genotype AA or GA in HHIP gene, rs56113850 genotype CC or TC in CYP2A6 gene had a higher risk of lung cancer than those with GG or TT. Multivariate Logistic regression analysis showed that the phlegm syndrome, rs1489759(genotype AA or GA), rs56113850(genotype CC or TC), low BMI and high smoking index were independent risk factors for lung cancer in COPD patients. The regression equation was calculated and a TCM syndrome-Gene-Environment prediction model was constructed, with an area under ROC curve(AUC) of 0.742 and a high goodness-of-fit. Conclusion A TCM syndrome-gene-environment prediction model for lung cancer risk has been developed in the study, which provides a useful tool for evaluation and screening of lung cancer risk for COPD patients.