Multi-slice Computed Tomography Diagnosis Of Ground Glass Nodule Lung Adenocarcinoma

Abstract one Multi-slice computed tomography diagnosis of ground glass nodule lung adenocarcinomaSection one:The ordinary CT features of GGNsObjectiveTo investigate the ordinary CT features of GGNs and its value in diagnosis.Materials and methodsOne hundred and sixty-nine GGNs(diameter were less than2cm and ground glass opacities areas were more than50%) were reviewed. GGNs were divided into3groups, benign group include33cases, pre-invasive group37cases, adenocarcinomas group99cases. All cases were conformed by clinical and pathology. The lesion’s distribution, size, density, shape, spiculation, lobulation, borders, vacuoles, indentation, content of solid components was observed. SPPSS17.0software was used for statistical analysis, measurement data using indepentant sample t test; count data using Pearson x2test, when the theoretic number less than1or the probability close to the size of a test, Fisher’s exact test were used and Spearman rank correlation were used for correlation analysis, and when P<0.05it consider to be statistically significant difference.ResultsThe locations of GGNs were no significant difference in all3groups (x2=15.192, P=0.056). There was significant difference between adenocarcinomas and benign lesions, between adenocarcinomas and pre-invasive lesions in size(t=-6.514, P=0.000; t=-8.821, P=0.000). And there was no significant difference between benign lesions and pre-invasive lesions in size(t=1.799, P=0.076).Three-dimensional ratios of benign lesions were bigger than pre-invasive lesions and adenocarcinomas(P<0.05), but there was no significant difference between adenocarcinomas and pre-invasive lesions(P>0.05). And the differences of two-dimensional ratio between benign lesions and adenocarcinomas was significant(P<0.05). There were significant difference between benign lesions and pre-invasive lesions in shape and border(P<0.05). There were no significant difference among three groups in spiculation, lobulation, vacuoles, indentation(P>0.05).There were significant difference between pre-invasive lesions and adenocarcinomas in spiculation, lobulation, vacuoles, indentation(P<0.05), and no significant difference in borders of lesion(P>0.05). In group of adenocarcinomas, there were significant difference between IACs and MIAs in spiculation, vacuoles and indentation(P<0.05), and no significant difference in lobulation and border(P>0.05).ConclusionThe morphological features of GGNs may help to the diagnosis of GGNs. Section two:The relationships between pulmonary ground glass nodules and bronchiObjectiveTo investigate the relationships between GGNs and bronchi, and its value in diagnosis.Materials and methodsOne hundred and sixty-nine GGNs(diameter were less than2cm and ground glass opacities areas were more than50%) were reviewed. GGNs were divided into3groups, benign group include33cases, pre-invasive group37cases, adenocarcinoma group99cases. All cases were confirmed by clinical and pathology.The lesion’s distribution, content of ground glass opacity components and relationships between GGNs and bronchi both in axial images and multiplanar reformation(MPR) images were observed. SPPSS17.0software was used for statistical analysis, measurement data using indepentant sample t test; count data using Pearson x2test, when the theoretic number less than1or the probability close to the size of a test, Fisher’s exact test were used, Spearman rank correlation were used for correlation analysis, and when P<0.05it consider to have significant statistically difference.ResultsThe relationships between GGNs and bronchi among adenocarcinoma, before invasive lesions and benign lesions were significant difference (x2=17.269, P=0.000).The relationships between GGNs and bronchi were divided into five types. Type I:The bronchi were "cut off’in the solid components. Type II:The bronchi were distortion and extension in the solid components. Type Ⅲ:The bronchi were distortion and extension in the GGO components. Type IV:The bronchi were normal course in the GGO components. Type V:The bronchi were circumvent GGN lesions. From type I to type V,there were15、37、28、16、8cases respectively. There were significant significance among3groups in all types (P=0.000). The relationships between GGNs and bronchi of adenocarcinomas most manifest as type Ⅰ、Ⅱ、Ⅲ. The relationships between GGNs and bronchi of benign lesions most manifest as type IV and type Ⅴ.There were only4cases of pre-invasive lesions has relationships between GGNs and bronchi. In the group of adenocarcinomas, from type Ⅰ～Ⅴ, there were13、18、5、3、0cases of IACs and2、16、19、3、1cases of MIAs respectively, and there was significant difference between them (x2=18.468,P=0.001). The relationships between GGNs and bronchi of IACs most manifest as type Ⅰ、Ⅱ, and MIAs most manifest as type Ⅱ、 Ⅲ.It was divided into3categories according to different GGO components in CT images. Category A:Lesions were pGGN, Category B:Lesions were mGGN which contains76%-99%GGO area, Category C:Lesions were GGN which contains50%-75%GGO area. There were19cases were category A, and28,62cases were category B, C respectively. The difference of GGO percentage among different categories with the relationships between GGNs and bronchi were correlation,with statistical significance(r=-0.343,P=0.000).ConclusionThe relationships between GGNs and bronchi,and the content of GGO area of the lesion may help to diagnosis of GGNs. Section three:The relationships between pulmonary ground glass nodules and vesslesObjectiveTo investigate the relationships and types between pulmonary ground glass nodules(GGNs) and vessels, and explore its value in diagnosis.Materials and methodsOne hundred and sixty-nine GGNs(diameter were less than2cm and ground glass opacities areas were more than50%) were reviewed. GGNs were divided into3groups, benign group include33cases, pre-invasive group37cases, adenocarcinoma group99cases. All cases were confirmed by clinical and pathology. The lesion’s content of solid components and relationships between GGNs and and vessels both in axial images and multi planar reformation (MPR) images were observed. SPPSS17.0software was used for statistical analysis, measurement data using indepentant sample t test; count data using Pearson x2test, when the theoretic number less than1or the probability close to the size of a test, Fisher’s exact test were used, Spearman rank correlation were used for correlation analysis, and when P<0.05it consider to have significant statistically difference.ResultsIn benign group about60.61%GGNs have vessels go through them and75.68%and90.91%in preinvasive group and invasive adenocarcinoma group respectively. There were significant differences among three different pathological groups of GGNs whether there were vessels go though the lesions (x2=11.349, P=0.003).The relationships between GGNs and vascular were divided into four types according to the results of observation and analysis. Type I:the vessels were circumvent GGN lesions (31cases); Type II:the vessels go though the lesions without any morphological changes(89cases); Type Ⅲ:the vessel in the lesion were distortion or rigidity but without increase in amount or converge toward the lesion (30cases) Type IV:the vessels go through GGN and it’s lumen broadened or vessel gathered (19cases).There were significant significance among3different pathological groups in all types (P=0.000). Different pathological groups of GGN among different categories with the relationships between GGNs and vessels were correlation, with statistical significance (r=0.427, P=0.000). In the group of adenocarcinomas, from type I-IV, there were7、33、9、3cases of IACs and2、14、16、15cases of MIAs respectively, and there were significant difference between them(x2=20.218, P=0.000). The vessels in the lesions were divided into3categories. Category A:Only pulmonary artery go through the lesion; Category B: Only pulmonary vein go through the lesion; Category C:both pulmonary artery and pulmonary vein go through the lesion. There was no significant difference and no correlation significance difference between different category of vascular and the different pathologic groups of GGNs (x2=4.088, P=0.394; P=0.152).ConclusionThe relationships between GGNs and vessels may help to diagnosis of GGNs. Abstract two The preoperative diagnostic value of solid components for lung adenocarcinomaObjectiveTo investigate the diagnositic value of solid components for preoperative diagnosis of lung adenocarcinoma by CT examination. Materials and methodsOne hundred and seventy-seven GGN lesions which confiremed by operating pathology, were divided into A, B two groups randomly, than each group were divided into preinvasive lesion, minimal invasive adenocarcinoma and invasive adenocarcinoma subgroups by operating pathology. Group A127cases, the size of the lesion, the proportion of GGO composition of the lesion, the long diameter and the size of the solid component in the lesion were measured on the CT imaging, in order to establish the CT diagnosis standard of lung adenocarcinoma which represented as GGN on CT imaging. Group B50cases were used to evaluated the accuracy of the above CT diagnosis standard.SPPSS17.0software was used for statistical analysis. Count data using Pearson x2; measurement data using independent t test, and rendering the ROC curve of the solid component in the lesion, then calculating the area under the curve (AUC), if the AUC were in the range of0.50to0.70, it was considered of low diagnostic value, in the range of0.70to0.90, medium diagnositic value, then more than0.90was high diagnostic value, Spearman rank correlation were used for correlation test. When P value was less than0.05it was considered to be significant difference.ResultIn group A, pre-invasive lesions all represented as pGGN (28/28);31.58%(12/38) cases of MIA represented as pGGN and68.42%(26/38) cases represented as mGGN;97.98%(97/99) cases of IACs represented as mGGN and2.02%(2/99) cases represented as pGGN.There were significant significance between group of preinvasive lesion and group of MIA, group of MIA and group of IAC in every index of diagnose (P<0.05); the size of the lesion, the proportion of GGO components of the lesion, the long diameter, and the size of the solid component in the lesion, these all have correlation with the pathologic type of lesion, with statistical significance(.P=0.000). By rendering the ROC curve, between group of preinvasive lesion and group of MIA, every index of diagnose were medium diagnositic value, in the range of0.70to0.90; between group of MIA and group of IAC, the size of the lesion had medium diagnositic value, in the range of0.70to0.90, the long diameter of the solid component in the lesion, the size of the solid component in the lesion and the proportion of GGO components of the lesion had high diagnostic value, AUC was more than0.90. This CT diagnosis standard got from group A, were used to analysis the pathologic type of group B, and there was no significant difference between the CT preoperative diagnosis and the operative pathologic diagnosis (P>0.05). The correct diagnosis rate of the size of the lesion, the proportion of GGO composition of the lesion, the long diameter, and the size of the solid component in the lesion were46.00%,70.00%,80.00%and72.00%respectively.ConclusionPreoperative CT inspection can be used to diagnosis of the pathological types of lung adenocarcinoma, based on the the size of the lesion, the proportion of GGO composition of the lesion, the long diameter and the size of the solid component in the lesion.