肺部磨玻璃结节浸润性腺癌病理亚型MSCT特征及对贴壁生长亚型的鉴别价值

MSCT Features of Pathological Subtypes of Infiltrative Adenocarcinoma with Ground Glass Nodules in the Lung and the Value of Differentiation of the Adnexal Growth Subtype

目的探讨肺部磨玻璃结节(GGN)浸润性腺癌(IA)病理亚型多层螺旋CT(MSCT)特征,并分析MSCT对贴壁生长亚型的鉴别价值。方法回顾2020年8月~2022年7月本院收治的154例GGN患者的临床资料,病理证实为IA,均接受MSCT检查,术后病理检查明确结节的病理亚型,比较各病理亚型结节的MSCT特征。采用多因素Logistic回归分析法评估鉴别贴壁生长亚型与非贴壁生长亚型的主要因素,并据此构建Logistic回归模型,利用受试者工作特征(ROC)曲线分析评估建立的Logistic回归分析模型对贴壁生长亚型的诊断效能。结果贴壁生长型31例(20.13%)、腺泡型95例(61.69%)、乳头型18例(11.69%)、微乳头型7例(4.55%)、实体型3例(1.95%)。各病理亚型结节在结节密度、边界模糊、深分叶征、毛刺征、空泡征、支气管改变、血管集束征占比及结节长径、实性占比、CT值方面比较,差异有统计学意义(P<0.05)。经Logistic回归分析显示,支气管改变、毛刺征、实性占比偏高、CT值偏低是鉴别IA患者GGN结节为非贴壁生长亚型的独立危险因素(P<0.05)。利用支气管改变、毛刺征、实性占比、CT值建立Logistic回归分析模型,ROC分析结果显示,Logistic回归分析模型对GGN贴壁生长亚型预测的AUC(95%CI)为0.870(0.807~0.919),分别高于无支气管改变[0.635(0.554~0.711)]、无毛刺征[0.647(0.566~0.722)]、实性占比[0.797(0.725~0.857)]、CT值[0.668(0.588~0.742)]单项的预测效能(均P<0.05)。结论MSCT特征能帮助鉴别肺部IA患者GGN的不同病理亚型,其中支气管改变、毛刺、实性占比及CT值与贴壁生长亚型密切相关,据此构建的Logistic回归模型对贴壁生长亚型有较高的鉴别价值。

Objective To investigate the pathological subtype multilayer spiral CT(MSCT)features of pulmonary ground glass nodules(GGN)invasive adenocarcinoma(IA)and to analyse the value of MSCT in identifying the adnexal growth subtype.Methods Clinical data of 154 patients with GGN admitted to our hospital with pathologically confirmed IA from August 2020 to July 2022 were reviewed and all underwent MSCT.The pathological subtypes of nodules were clarified by postoperative pathological examination,and the MSCT characteristics of nodules of each pathological subtype were compared.Multi-factor logistic regression analysis was used to evaluate the main factors that distinguish the adherent growth subtypes from non-adherent growth subtypes.The diagnostic efficacy of the established logistic regression analysis models for the adherent growth subtypes was assessed using subject operating characteristic(ROC)curve analysis.Results There were 31 cases(20.13%)of adnexal growth type,95 cases(61.69%)of alveolar type,18 cases(11.69%)of papillary type,7 cases(4.55%)of micropapillary type and 3 cases(1.95%)of solid type.The differences were statistically significant(P<0.05)when comparing the nodules of each pathological subtype in terms of nodule density,blurred borders,deep lobar sign,burr sign,vacuolar sign,bronchial changes,percentage of vascular collection sign and nodule length diameter,percentage of solidity and CT value.Logistic regression analysis showed that bronchial changes,burr sign,high percentage of solidity and low CT values were independent risk factors for identifying GGN nodules as non-adherent growth subtypes in IA patients(P<0.05).Logistic regression analysis models were developed using bronchial changes,burr sign,solid percentage,and CT values.The results of ROC analysis showed that the AUC(95%CI)predicted by the logistic regression analysis model for the GGN apposed growth subtype was 0.870(0.807-0.919),which was higher than that for no bronchial change[0.635(0.554-0.711)],no burr sign[0.647(0.566-0.722)],solidity proportion[0.797(0.725-0.857)],and CT value[0.668(0.588-0.742)]alone were predictive of efficacy(all P<0.05).Conclusion Conclusion MSCT features can help identify different pathological subtypes of GGN in lung IA patients,among which bronchial changes,burr,solid percentage and CT values are closely related to the adnexal growth subtype,and the logistic regression model constructed accordingly has a high value for the identification of the adnexal growth subtype.