术前薄层CT参数预测纯磨玻璃结节肺腺癌浸润性生长的效能

Efficiency of preoperative thin-section CT parameters for predicting invasive growth of pure ground-glass nodules in lung adenocarcinoma

目的探讨术前薄层CT参数预测纯磨玻璃结节(pGGN)肺腺癌浸润性生长的效能。方法回顾性分析150例CT表现为pGGN的肺腺癌患者的资料,并根据病理结果将患者分为惰性组95例(不典型腺瘤样增生10例、原位癌32例、微浸润性腺癌53例)和浸润组55例(浸润性腺癌)。比较两组薄层CT参数,采用多因素Logistic回归模型分析影响pGGN肺腺癌浸润性生长的因素,并应用受试者工作特征(ROC)曲线评价各薄层CT参数及其联合预测pGGN肺腺癌浸润性生长的价值。结果两组的病灶大小、CT值、形态为类圆形比例、毛刺征比例、分叶征比例、细支气管征比例差异有统计学意义(均P<0.05),而边界清晰比例、胸膜牵拉征比例、空泡征比例差异无统计学意义(均P>0.05)。多因素Logistic回归分析结果显示,病灶大小、CT值、分叶征、细支气管征为pGGN浸润性生长的影响因素(均P<0.05),且4个参数联合预测pGGN浸润性生长的效能最高,ROC曲线下面积为0.934,敏感性为96.36%,特异性为81.05%。结论病灶大小、CT值、分叶征、细支气管征是术前预测pGGN浸润性生长的因素,且这4个薄层CT参数联合可提高预测效能。

Objective To investigate the efficiency of preoperative thin-section CT parameters for predicting the invasive growth of pure ground-glass nodules(pGGN)in lung adenocarcinoma.Methods A retrospective analysis was carried out on the data of 150 lung adenocarcinoma patients manifesting pGGN on CT,and the patients were assigned to inertia group(95 cases,including 10 cases of atypical adenomatous hyperplasia,32 cases of adenocarcinoma in situ and 53 cases of minimally invasive adenocarcinoma)or invasive group(55 cases of invasive adenocarcinoma)according to their pathological findings.The parameters of thin-section CT were compared between the two groups,the multivariate Logistic regression model was used to analyze the factors influencing the invasive growth of pGGN in lung adenocarcinoma,and receiver operating characteristic(ROC)curve was employed to assess the value of various parameters of thin-section CT and their combination for predicting the invasive growth of pGGN in lung adenocarcinoma.Results There were statistically significant differences between the two groups in the lesion size,CT value,proportion of cases with suborbicular morphology,proportion of cases with spiculation sign,proportion of cases with lobulation sign,and proportion of cases with air bronchogram sign(all P<0.05),whereas no statistically significant differences were found in the proportion of cases with well-defined boundary,proportion of cases with pleural stretch sign,or proportion of cases with vacuole sign(all P>0.05).The results of multivariate Logistic regression analysis revealed that lesion size,CT value,lobulation sign,and air bronchogram sign were the influencing factors for the invasive growth of pGGN(all P<0.05);moreover,the combination of the four parameters had the highest efficiency for predicting the invasive growth of pGGN,with an area under ROC curve of 0.934,a sensitivity of 96.36%,and a specificity of 81.05%.Conclusion Lesion size,CT value,lobulation sign,and air bronchogram sign are the factors for preoperatively predicting the invasive growth of pGGN,and the combined application of these four parameters of thin-section CT can improve the predictive efficiency.