Diagnostic value of solid component for lung adenocarcinoma shown as ground-glass nodule on computed tomography

Abstract

To explore the feasibility of making a preoperative diagnosis of lung adenocarcinoma shown as ground-glass nodule (GGN) on computed tomography (CT). A total of 143 GGN lesions proved pathologically were divided randomly into A and B groups. Then each group was further divided pathologically into preinvasive lesion, minimal invasive adenocarcinoma (MIA) and invasive adenocarcinoma (IAC) subgroups. Group A (n = 101), size of lesion, proportion of ground glass opacity (GGO) composition of lesion, long diameter, longest diameter and size of solid component in lesion were measured on CT so as to establish the CT diagnostic standard of lung adenocarcinoma shown as GGN on CT. Group B (n = 42) was employed to evaluate the accuracy of the above CT diagnostic standard. SPSS 17.0 software was used for statistical analysis. Significant statistic significance existed in all parameters among all groups (P < 0.05). All parameters were correlated the pathologic type of lesion. The differences were statistically significant (P = 0.000). Through the receiver operating characteristic (ROC) curve, between groups of preinvasive lesion and MIA, each parameter had a medium diagnostic value of 0.70-0.90; between groups of MIA and IAC, size of lesion and long diameter of solid component in lesion had a medium diagnostic value of 0.70-0.90, longest diameter of solid component, size of solid component in lesion and proportion of GGO composition of lesion had a high diagnostic value with an AUC of >0.90. The CT diagnostic standard, derived from group A, was used to analyze the pathologic type of group B. And t no significant statistic significance existed between CT preoperative diagnosis and operative pathologic diagnosis (P > 0.05) . The correct diagnosis rates of size of lesion, proportion of GGO composition of lesion, long diameter, longest diameter and size of solid component in lesion were 71.43%, 76.19%, 90.05%, 90.05% and 88.10% respectively. Based upon size of lesion, proportion of GGO composition of lesion, long diameter, longest diameter and size of solid component in lesion, preoperative CT examination may be used to determine the pathological types of lung adenocarcinoma shown as GGN.