摘要
目的探讨高分辨率CT特征对于ⅠA期浸润性肺腺癌是否含有高级别组织学模式(微乳头或实体亚型)的预测价值。方法回顾性收集2023年1月至2024年9月本院收治的164例浸润性肺腺癌患者,依据病理分析结果将患者划分为不含HGP组与含HGP组。收集患者的临床信息与病灶CT特征。统计对比两组间特征差异,并利用多因素Logistic回归分析建立预测模型。结果单因素分析中含HGP组与不含HGP组之间年龄(P=0.014)、SIRI(P=0.029)、密度类型(P<0.001)、毛刺征(P<0.001)、血管集束征(P<0.001)、实性成分最大直径(P<0.001)、CTR(P<0.001)、平均CT值(P<0.001)差异有统计学意义。多因素Logistic回归分析显示毛刺征(OR=2.767;95%CI:1.144~6.690;P=0.024)、平均CT值(OR=1.011;95%CI:1.007~1.014;P<0.001)是HGP的独立预测因素。平均CT值诊断效能AUC为0.880(95%CI:0.826~0.935),灵敏度为83.6%,特异度为84.1%。毛刺征诊断效能AUC为0.708(95%CI:0.621~0.795),灵敏度为63.6%,特异度为78.0%。两者联合模型诊断效能AUC为0.892(95%CI:0.837~0.946),灵敏度87.3%,特异度85.3%。结论高分辨率CT特征有助于预测ⅠA浸润性肺腺癌中的高级别组织学模式,毛刺征、平均CT值是独立危险因素,两者的联合模型具有较好的诊断效能。
Objective To investigate the predictive value of high-resolution CT features for whether stage IA invasive lung adenocarcinoma contains a high-grade histologic pattern(micropapillary or solid subtype).Methods 164 patients with invasive lung adenocarcinoma admitted to our hospital from January 2023 to September 2024 were retrospectively collected,and the patients were classified into HGP-free and HGP-containing groups based on the results of pathological analysis.Clinical information and CT features of the lesions were collected from the patients.The differences in characteristics between the two groups were statistically compared,and a predictive model was established using multifactor logistic regression analysis.Results The differences in age(P=0.014),SIRI(P=0.029),density type(P<0.001),burr sign(P<0.001),vascular cluster sign(P<0.001),maximum diameter of the solid component(P<0.001),CTR(P<0.001),and the mean CT value(P<0.001)between the HGP-containing group and the HGP-free group in univariate analysis were statistically significant.Multifactorial logistic regression analysis showed that burr sign(OR=2.767;95%CI:1.144-6.690;P=0.024),and mean CT value(OR=1.011;95%CI:1.007-1.014;P<0.001)were independent predictors of HGP.The area under the curve(AUC)of predictive efficacy for burr sign was 0.708(95%CI:0.621-0.795),with a sensitivity of 63.6%and a specificity of 78.0%.The mean CT value had a predictive efficacy AUC of 0.880(95%CI:0.826-0.935),with a sensitivity of 83.6%and a specificity of 84.1%.The combined model had a predictive efficacy AUC of 0.892(95%CI:0.837-0.946),a sensitivity of 87.3%,and a specificity of 85.3%.Conclusion High-resolution CT features were helpful in predicting high-grade histologic patterns in IA invasive lung adenocarcinoma,and the burr sign and mean CT value were independent risk factors,and the combined model of the two had good diagnostic efficacy.
作者
黄子康
李思叶
于越洋
沈君
徐滨
孙琪
朱玉春
HUANG Zikang;LI Siye;YU Yueyang;SHEN Jun;XU Bin;SUN Qi;ZHU Yuchun(School of Medicine Jiangsu University,Zhenjiang,Jiangsu 212013,China;Gusu College,Nanjing Medical University,Suzhou,Jiangsu 215002,China;The Second People′s Hospital of Kunshan,Kunshan,Jiangsu 215300,China;Kunshan Hospital Affiliated to Jiangsu University,Kunshan,Jiangsu 215300,China)
出处
《临床肺科杂志》
2025年第9期1303-1309,共7页
Journal of Clinical Pulmonary Medicine
基金
江苏大学医教协同创新基金一般项目(JDYY2023060)
昆山市第一人民医院2022年度广仁基金科研课题(临床研究专项)重点项目(KRY-YN2022001)
昆山市重点研发计划(社会发展)指导性项目(KSZ2311)。
关键词
肺腺癌
高级别组织学模式
高分辨率CT
lung adenocarcinoma
high-grade histologic pattern
high-resolution CT
