Study Of CT Features Of Pre-invasive Lesion Versus Invasive Pulmonary Adenocarcinomas(IPAs) Appearing As Ground Glass Nodule(GGN)

OBJECTIVE: To evaluate the role of computed tomography(CT) in differentiating preinvasive and invasive pulmonary adenocarcinomas(IPAs) appearing as ground glass nodule(GGNs) retrospectively.MATERIALS AND METHODS: This study included the chest CT scans of the 122 patients appearing as ground glass nodules(GGNs),of which 27 were male and 85 were female, the age between 26~78 with average 55.80±9.951 years old. Of which 57 were pure GGNs and 65 were part solid GGNs. Pathologically proven pre-invasive lesions and invasive pulmonary adenocarcinomas(IPAs) are 71 and 51 respectively. Among 71 pre-invasive lesions, 13 lesions are confirmed as atypical adenomatous hyperplasia(AAH) whereas 58 lesions were adenocarcinoma in situ(AIS). Similarly, among 51 invasive lesions 39 lesions are diagnosed as minimally invasive adenocarcinoma(MIA) while 12 lesions were invasive pulmonary adenocarcinoma(IPAs). CT machines used in our study are from different hospitals namely GEHD 750 CT scan, Somatotom definition, & Sensation 16 siemens with 120 k Vp, 100-200 m As,pitch of 0.85 – 1.4 and slice thickness of 1-2.5mm. CT physical parameters such as window width and window level also varies from 1000-1200 HU and-600 to-700 HU respectively from different hospitals. For statistical analysis, the level of significance was considered to be p value less than 0.05.RESULTS: In our study, we found that lesion size and lobulation are two important predicting factors to differentiate pre-invasive from invasive pulmonary adenocarcinomas(p<0.05) in case of pure GGNs. Pure GGNs with lobulation is 5.65 times more likely to be invasive pulmonary adenocarcinomas compared to pre-invasive lesions and also pure GGNs with 1 unit increase in size are 1.22 times more likely to be invasive pulmonary adenocarcinomas and optimal cut-off value for pre-invasive lesions is less than 11.82 mm.For part solid GGN- lesion size, solid proportion grade, lobulation, speculation and pleural retraction are statistically significant in bivariate analysis(p< 0.05). However multivariate analysis showed that only large lesion size and presence of speculations are statistically significant in predicting invasive pulmonary adenocarcinomas( p< 0.05). One unit increase in size increases the likelihood of IPAs by 1.15 times. The presence of speculation in part solid GGN increases the likelihood of being IPAs by 15.53 times. From the ROC curve, the optimal cut-off value for lesion size in invasive pulmonary adenocarcinomas(IPAs) is found to be more than 12.5 mm.CONCLUSION: Absence of lobulation and smaller lesion size less than 11.82 mm are specific for pre-invasive lesions in case of pure GGN. Presence of speculation and larger lesion size more than 12.5 mm are specific for invasive pulmonary adenocarcinomas(IPAs) in partsolid GGN.