Differential diagnosis of the MDCT features between lung adenocarcinoma preinvasive lesions and minimally invasive adenocarcinoma appearing as ground-glass nodules.
- Author:
Jia LIU
1
;
Wenwu LI
2
;
Yong HUANG
3
;
Dianbin MU
;
Haiying YU
;
Shanshan LI
Author Information
- Publication Type:Journal Article
- MeSH: Adenocarcinoma; diagnostic imaging; pathology; Diagnosis, Differential; Humans; Lung Neoplasms; diagnostic imaging; pathology; Multidetector Computed Tomography; Neoplasm Invasiveness; ROC Curve; Retrospective Studies; Sensitivity and Specificity
- From: Chinese Journal of Oncology 2015;37(8):611-616
- CountryChina
- Language:Chinese
-
Abstract:
OBJECTIVEThe aim of this study was to retrospectively investigate the multi-detector computed tomography (MDCT) features of preinvasive lesions and minimally invasive adenocarcinoma (MIA) appearing as ground-glass nodules (GGNs), and to analyze their significance in differential diagnosis.
METHODSThe pathological data and MDCT images of 111 GGNs in 93 patients were reviewed and analyzed retrospectively, to identify the differentiating CT features between preinvasive lesions and MIA and to evaluate their differentiating accuracy.
RESULTSIn the 93 patients included in the study, there were 27 cases with preinvasive lesions (38 GGNs) and 66 cases with MIA (73 GGNs). No statistically significant difference was observed in terms of the gender, age and number of lesions between the two groups. There were significant differences (P<0.05) in the size of lesion, size of solid portion, content of solid portion, and morphological characteristics of the lesion edge between preinvasive lesions and MIA. ROC curve analysis showed that the optimal cut-off value of lesion size for differentiating preinvasive lesions from MIA was 13.0 mm (sensitivity, 83.0%; specificity, 80.0%), and that of solid portion size was 2.0 mm (sensitivity, 90.0%; specificity, 97.0%) and that of solid proportion was 12.0% (sensitivity, 88.0%; specificity, 97.0%). The analysis of CT morphological features showed that there were significant differences in the terms of lesion nature (pGGO, mGGO), presence or absence of lobulated sign and spiculated sign (P<0.05) between preinvasive lesions and MIA, but there were no significant differences in terms of the lesion edge, the presence or absence of vacuole sign, bubble lucency and pleural retraction (P>0.05).
CONCLUSIONSPreinvasive lesions can be accurately distinguished from MIA by the size of lesion, size of solid portion,solid proportion and morphological characteristics of the lesion edge. The size of lesion, size of solid portion, content of solid proportion and morphological characteristics of the lesion edge are of significance in the differential diagnosis of preinvasive lesions and minimally invasive adenocarcinoma of the lung.
