OBJECTIVE: To analyze the FDG-PET and high-resolution computed tomography (HRCT) features of early lung adenocarcinoma manifesting as solitary ground-glass opacity nodules (GGNs), and to establish a new risk model for predicting the invasiveness of early lung adenocarcinoma. METHODS: We retrospectively analyzed the data of clinical stage IA lung adenocarcinoma patients who received preoperative PET/CT and HRCT examination. Patients were divided into invasive adenocarcinoma (IVA) group and preinvasive minimally invasive adenocarcinoma (MIA) group. The correlations between FDG-PET parameters, HRCT parameters and histopathological invasiveness, and their predictive efficacy were analyzed. A mathematical model for predicting histopathological invasiveness of early lung adenocarcinoma was established and assessed. RESULTS: This study enrolled 56 patients, 48 were in IVA group and 8 were in preinvasive MIA group. Compared with those in preinvasive MIA group, GGNs in IVA group showed larger diameter, higher ground-glass opacity (GGO) density and more pleural indentation signs (70.8%) on HRCT; they also showed higher maximum standardized uptake value (SUV) and SUV index on FDG-PET (P = 0.001-0.037). Logistic regression analysis found a risk model for predicting IVA of solitary GGNs that were established by CTGGO and SUV index. Receiver operating characteristic curves showed that this model had the highest area under the curve (AUC), sensitivity, specificity and accuracy (AUC, 0.948; sensitivity, 95.8%; specificity, 87.5%; accuracy, 94.6%). CONCLUSION: Using HRCT combined with FDG-PET to establish the corresponding mathematical prediction model has the potential to identify IVA in early lung adenocarcinoma preoperatively.
OBJECTIVE: To analyze the FDG-PET and high-resolution computed tomography (HRCT) features of early lung adenocarcinoma manifesting as solitary ground-glass opacity nodules (GGNs), and to establish a new risk model for predicting the invasiveness of early lung adenocarcinoma. METHODS: We retrospectively analyzed the data of clinical stage IA lung adenocarcinomapatients who received preoperative PET/CT and HRCT examination. Patients were divided into invasive adenocarcinoma (IVA) group and preinvasive minimally invasive adenocarcinoma (MIA) group. The correlations between FDG-PET parameters, HRCT parameters and histopathological invasiveness, and their predictive efficacy were analyzed. A mathematical model for predicting histopathological invasiveness of early lung adenocarcinoma was established and assessed. RESULTS: This study enrolled 56 patients, 48 were in IVA group and 8 were in preinvasive MIA group. Compared with those in preinvasive MIA group, GGNs in IVA group showed larger diameter, higher ground-glass opacity (GGO) density and more pleural indentation signs (70.8%) on HRCT; they also showed higher maximum standardized uptake value (SUV) and SUV index on FDG-PET (P = 0.001-0.037). Logistic regression analysis found a risk model for predicting IVA of solitary GGNs that were established by CTGGO and SUV index. Receiver operating characteristic curves showed that this model had the highest area under the curve (AUC), sensitivity, specificity and accuracy (AUC, 0.948; sensitivity, 95.8%; specificity, 87.5%; accuracy, 94.6%). CONCLUSION: Using HRCT combined with FDG-PET to establish the corresponding mathematical prediction model has the potential to identify IVA in early lung adenocarcinoma preoperatively.