孤立性肺结节患者肺腺癌相关危险因素的COX回归模型分析

目的：探讨孤立性肺结节患者肺腺癌相关的危险因素，并建立COX回归模型。方法：回顾性收集2018.01-2021.12期间在杭州市萧山区第三人民医院呼吸内科收治的首次发病经胸部CT发现孤立性肺结节，且经多次随访至最终确诊的患者为研究对象，应用COX回归模型对与肺腺癌相关的危险因素进行单因素、多因素分析，建立肺腺癌早期诊断模型。结果：经COX回归模型单因素和多因素分析发现，结节大小（HR=3.035，95%CI为1.478-6.297，P=0.001）、分叶征（HR=3.748，95%CI为1.399-9.586，P=0.013）、短毛刺（HR=0.332，95%CI为0.056-1.962，P=0.024）、CT值（HR=0.158，95%CI为0.049-0.513，P=0.002）、血管集束征（HR=2.279，95%CI为0.589-8.810，P=0.023）、磨玻璃密度影（HR=2.723，95%CI为1.020-7.274，P=0.046）、空泡（HR=7.027，95%CI为0.703-70.234，P=0.039）是肺腺癌相关的独立危险因素，经多因素COX回归模型建立的个体危险指数（HI）方程为：1.324结节大小+1.127分叶征+1.103短毛刺-1.846 CT值+0.824血管集束征+1.002磨玻璃密度影+1.950空泡。HI值越小，腺癌患病危险性就越低；HI值越大，腺癌患病危险性就越高。结论：结节大小、分叶征、短毛刺、CT值、血管集束征、磨玻璃密度影、空泡是孤立性肺结节患者肺腺癌的独立危险因素，初步成功建立了肺腺癌早期诊断模型及个体HI方程。

Cox Regression Model for Risk Factors Related of Lung Adenocarcinoma in Patients with Solitary Pulmonary Nodules

Objective To investigate the risk factors of lung adenocarcinoma in patients with solitary pulmonary nodules, and establish the Cox regression model. Methods The patients with solitary pulmonary nodules found by chest CT and followed up for many times to the final diagnosis in the Department of respiratory medicine of the third people''s Hospital of Xiaoshan District, Hangzhou from January 2008 to December 2021 retrospectively, were used Cox regression model to analyze the risk factors related to lung adenocarcinoma, and established the early diagnosis model of lung adenocarcinoma. Results Univariate and multivariate analysis of Cox regression model showed that nodule size (HR=3.035, 95% CI 1.478-6.297, P=0.001), lobulation sign (HR=3.748, 95% CI 1.399-9.586, P=0.013), short hairy thorn (HR=0.332, 95% CI 0.056-1.962, P=0.024), CT value (HR=0.158, 95% CI 0.049-0.513, P=0.002), vascular bundle sign (HR=2.279, 95% CI 0.589-8.810, P=0.023) Ground glass density (HR=2.723, 95% CI 1.020-7.274, P=0.046) and vacuole (HR=7.027, 95% CI 0.703-70.234, P=0.039) were independent risk factors associated with lung adenocarcinoma. The individual risk index (HI) of Cox regression model of adenocarcinoma was: 1.324 nodule size + 1.127 lobulation sign + 1.103 short hair thorn -1.846 CT value + 0.824 vascular bundle sign + 1.002 ground glass density shadow + 1.950 vacuole. The lower the hazard index was, the lower the risk of adenocarcinoma appeared, while the bigger the hazard index was, the higher the risk of adenocarcinoma appeared. Conclusion Nodule size, lobulation sign, long hairy thorn, CT value, vascular bundle sign, ground glass density shadow and vacuole were independent risk factors of lung adenocarcinoma in patients with solitary pulmonary nodules. The early diagnosis model of lung adenocarcinoma and individual hi equation were successfully established.