肺部磨玻璃结节CT征象在微浸润腺癌与浸润腺癌病变的对比分析

发布时间：2018-08-19 06:26

【摘要】：目的:本文通过对术后病理证实为微浸润腺癌与浸润腺癌的肺部磨玻璃样结节(Ground glass nodules,GGNs)进行薄层CT(High-resolution CT,HRCT)征象的对比并分析,探讨肺磨玻璃结节在微浸润腺癌(Micro invasive adenocarcinoma,MIA)与浸润性腺癌(Invasive adenocarcinoma,IAC)之间CT征象的差异,以进一步认识两组磨玻璃结节的相关特性,准确判断两者的区别,从而更好的辅助临床鉴别诊断和指导治疗。方法:收集自2016年1月至2016年12月期间,就诊于大连医科大学附属第二医院胸外二科,并于本院行薄层计算断层扫描发现有肺部磨玻璃结节且经术后病理证实为微浸润腺癌和浸润腺癌的患者。影像图片均使用本院德国SOMATOM Definition AS 64型号机器,统一扫描范围均为由肺尖部至膈肌顶部。CT检查原始采集数据设定值:设备管电压设定值为120KV,管电流设定值为150m A,设定层厚8mm,层距8mm或层厚10mm,层距10mm。薄层扫描CT的重建参数设置层厚为1mm,层距为1mm。肺窗数据设置:窗宽为1200HU,窗位为-600HU;纵隔窗数据设置:窗宽为400HU,窗位为40HU。结果由两名高年资的影像科医师对CT影像资料的特征进行分析。术后切除标本送病理。行10%福尔马林(甲醛)以固定,给予石蜡行包埋并制作成载玻片标本,再行HE染色处理。必要时行免疫组化检查以确定诊断。最终病理诊断由两位病理科高年资医师进行判定(病理的结果依据肺腺癌新分类标准予以分类)。最后使用SPSS 17.0版本软件对所收集的数据进行分析。对所有病理结果诊断为微浸润腺癌与浸润性腺癌的CT征象:GGN性质、病灶位置、病灶大小、实性成分(实性成大小及实性成分比例)、形态、边界、分叶征、毛刺征、胸膜凹陷征、支气管充气征及空泡征进行统计学分析。相关计数资料运用χ2检验,期望值小于5则用Fisher检验,计量资料使用T检验。把有价值的影响因素行Logistic多因素分析,以寻找两组间的差异因素。对病灶大小、实性成分使用受试者工作特征曲线,既ROC曲线(receiver operating characteristic curve),确定最佳分界值,并算出曲线下的面积以评估其诊断价值。以P值0.05则表示微浸润腺癌与浸润性腺癌之间的差异具有显著的统计学意义。结果:73名患者中,共83例GGNs,病理结果为微浸润腺癌37例,占比例为44.6%;浸润腺癌46例,占比例为55.4%。单因素分析显示病灶大小(p=0.00)、GGN性质(p=0.003)、形态(p=0.037)、分叶征(p=0.049)、毛刺征(p=0.022)P值小于0.05,具有统计学意义。对两组中有意义的因素行多因素分析,病灶大小(P=0.000)、GGN性质(P=0.015)、毛刺征(P=0.022)具有统计学意义。对病灶大小进行ROC曲线分析:病灶大小曲线下面积0.823,病灶大小最佳截点11.50mm,敏感度78.3%,特异性73.0%。两组中m GGNs病灶实性成分进行比较,微浸润腺癌中平均实性成分大小(mm)2.50±1.008,实性成分比例(%)23.99±5.737,浸润性腺癌平均实性成分大小(mm)4.90±1.412,实性成分比例(%)35.55±3.861。m GGNs实性成分大小(P=0.00)及比例(P=0.00)在两组间有统计学意义。多因素分析实性成分大小(P=0.004)、实性成分比例(P=0.048)均具有统计学意义。对两组间实性成分大小、比例行ROC曲线分析:实性成分大小曲线下面积0.925,最佳截点3.25mm,敏感度88.6%,特异性77.8%;实性成分比例曲线下面积0.912,最佳截点28.1%,敏感度90.9%,特异性85.2%。结论:本研究通过对微浸润腺癌与浸润腺癌的磨玻璃样结节CT征象进行对比分析,分析结果示病灶大小、病灶性质、毛刺征在两组之间的差异显著,并具有统计学意义,这对于微浸润腺癌与浸润腺癌两者的鉴别诊断具有一定的价值所在,尤其对于病灶大小大于11.50mm,含实性成分,具有分叶征及毛刺征的病灶,更有可能为浸润性腺癌。而实性成分大小和实性成分比例在两组混合磨玻璃结节中的差异显著,具有统计学意义。实性成分大小及实性成分比例对以混合磨玻璃结节为表现的微浸润腺癌与浸润腺癌两者之间的相区别提供了参考,实性成分大小大于3.25mm,实性成分比例大于28.1%的病灶更倾向于浸润性腺癌。因此,通过磨玻璃的薄层扫描,可以为临床的鉴别诊断提供一定帮助及选择合理的术式,使患者受益。
[Abstract]:Objective: To investigate the role of ground glass nodules (GNs) in microinvasive adenocarcinoma (MIA) and invasive adenocarcinoma (Invasive adenocarcinoma) by comparing and analyzing the high resolution CT (HRCT) findings of GNs confirmed by pathology after operation. Methods: From January 2016 to December 2016, we collected the data from the Department of Thoracic Surgery, Second Affiliated Hospital of Dalian Medical University, and treated them in our hospital. Thin-slice computed tomography (TLCT) revealed ground glass nodules in the lungs, which were confirmed by postoperative pathology as microinvasive adenocarcinoma and invasive adenocarcinoma. The images were taken by our German SOMATOM Definition AS 64 machine. The uniform scan ranged from the tip of the lungs to the top of the diaphragm. The reconstruction parameters of thin-slice CT were set to 1 mm in thickness and 1 mm in interval. The lung window data were set to 1200 HU in width and - 600 HU in position. The mediastinal window data were set to 400 HU in width and 40 HU in position. The CT features were analyzed. The specimens were removed and sent to pathology. The specimens were fixed with 10% formaldehyde, embedded with paraffin and stained with HE. Immunohistochemistry was performed if necessary to confirm the diagnosis. The final pathological diagnosis was determined by two senior pathologists. Results were classified according to the new classification criteria of lung adenocarcinoma. The data were analyzed using SPSS 17.0 software. All pathological findings were diagnosed as micro-invasive adenocarcinoma and invasive adenocarcinoma on CT: GGN nature, location, size of the lesion, solid component (solid to solid ratio), shape, boundary, lobulation sign. _2 test was used for the related counting data, Fisher test was used for the expectation value less than 5, and T test was used for the measurement data. The ROC curve, the receiver operating characteristic curve, was used to determine the best boundary value and calculate the area under the curve to evaluate the diagnostic value. Univariate analysis showed lesion size (p = 0.00), GGN property (p = 0.003), morphology (p = 0.037), lobulation sign (p = 0.049), burr sign (p = 0.022) P value was less than 0.05, with statistical significance. The ROC curve analysis showed that the area under the curve of lesion size was 0.823, the best cut-off point of lesion size was 11.50m m, the sensitivity was 78.3%, and the specificity was 73.0%. The sex component ratio (%) was 23.99 (%) and the average solid component size (m m) was 4.90 (%) and the solid component ratio (%) was 35.55 (%) 3.861.m GGNs (P = 0.00) and the ratio (P = 0.00) were statistically significant between the two groups. ROC curve analysis showed that the area under the solid component size curve was 0.925, the best cut-off point was 3.25 mm, the sensitivity was 88.6%, and the specificity was 77.8%; the area under the solid component size curve was 0.912, the best cut-off point was 28.1%, the sensitivity was 90.9%, and the specificity was 85.2%. CT findings of nodules were compared and analyzed. The results showed that the size, nature and burr sign of the nodules were significantly different between the two groups, and the difference was statistically significant. The difference in the size and proportion of solid components between the two groups of mixed ground-glass nodules was statistically significant. For reference, lesions with solid components larger than 3.25 mm and solid components larger than 28.1% are more likely to be invasive adenocarcinoma. Therefore, thin-layer scanning with ground-glass can provide some help for clinical differential diagnosis and choose a reasonable surgical method to benefit patients.
【学位授予单位】：大连医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R734.2

【参考文献】