Air-containing space: significantly in predicting EGFR mutation for early lung adenocarcinoma appearing as subsolid nodule

Background: To demonstrate the interrelation between computed tomography (CT) characteristics and epidermal growth factor receptor (EGFR) mutation status in early-stage lung adenocarcinoma manifested as subsolid nodule. Method: A total of 202 patients with appearance as subsolid nodules on CT images and pathologically confirmed as early-stage lung adenocarcinoma were enrolled in this study. All of the patients divided in to EGFR mutation group and EGFR wild-type group. Correlation between clinical and radiological features and EGFR mutation status were analyzed with univariate and multivariate analysis. A receiver operating characteristic (ROC) curves were generated to evaluate the logistic regression model. Results: : EGFR mutation were detected in 137 of 202 patients. In univariate analysis, female gender, non-smokers, larger tumor size, lobulation, air-containing space and pleural retraction were more commonly indicated in EGFR mutated tumors than in EGFR wild-type tumors. Multivariate logistic regression analysis revealed that smoking history (OR, 0.275; 95%CI, 0.090–0.840), tumor size (OR, 4.263; 95%CI, 1.802–9.982) and air-containing space (OR, 3.055; 95%CI, 1.266–7.372) were independent predicting factor relevant to EGFR mutation. The AUC of ROC curve for predicting model reached 0.808. In addition, air-containing space with EGFR mutations were more commonly less than 5mm, solitary and mainly in an eccentric distribution. Conclusion: Our research indicated that smoking history, tumor size and air-containing space were independent factor for predicting EGFR mutation in early-stage lung adenocarcinoma appeared as subsolid nodule. Solitary air-containing space with≤5mm and eccentric distribution would help to determine EGFR mutation status.