RESUMO
OBJECTIVE: To assess whether quantitative computed tomography (CT) can help predict histological invasiveness of pulmonary adenocarcinoma appearing as pure ground glass nodules (pGGNs). METHODS: A total of 110 pulmonary pGGNs were retrospectively evaluated, and pathologically classified as pre-invasive lesions, minimally invasive adenocarcinoma (MIA) and invasive pulmonary adenocarcinoma (IPA). Maximum nodule diameters, largest cross-sectional areas, volumes, mean CT values, weights, and CT attenuation values at the 0th,2th,5th, 25th, 50th,75th, 95th, 98th and100th percentiles on histogram, as well as 2th to 98th, 5th to 95th, 25th to 75th,and 0th to 100thslopes, respectively, were compared among the three groups. RESULTS: Of the 110 pGGNs, 50, 28, and 32 were pre-invasive lesions, MIA, and IPA, respectively. Maximum nodule diameters, largest cross-sectional areas, andmass weights were significantly larger in the IPA group than in pre-invasive lesions. The 95th, 98th, 100th percentiles, and 2th to 98th, 25th to 75th, and 0th to 100thslopes were significantly different between pre-invasive lesions and MIA or IPA. Logistic regression analysis showed that the maximum nodule diameter (OR=1.21, 95%CI: 1.071-1.366, p<0.01) and 100th percentile on histogram (OR=1.02, 95%CI: 1.009-1.032, p<0.001) independently predicted histological invasiveness. CONCLUSIONS: Quantitative analysis of CT imaging can predict histological invasiveness of pGGNs, especiallythe maximum nodule diameter and 100th percentile on CT number histogram; this can instruct the long-term follow-up and selective surgical management.
