Abstract: Objective To investigate the relationship between epidermal growth factor receptor (EGFR) gene mutation and computed tomography (CT) features and clinical features in non-small cell lung cancer (NSCLCs) patients. Methods The clinical data of 187 patients with NSCLCs admitted in our hospital from September 2014 to July 2016 were retrospectively analyzed. All the patients accepted the EGFR mutated gene detection, and they were divided into effective mutation group (n=67) and non-effective mutation group (n=120). The clinical data and lung CT imaging data were complete in the two groups. The univariate and multivariate Logistic regression analysis were used to analyze the differences of imaging and clinical features between the two groups. Results Comparing with the non-mutation group, there were higher proportion of women and lower smoking index in the EGFR effective mutation group. Lesions in the lung tissue showed a clear edge, leaf and burr, and containing ground-glass opacity (GGO) component, usually accompanied by airway bronchogram and pleural indentation, and associated with cancer lymphatic inflammation and lung metastasis in mutation group (all P < 0.05). There were no significant differences in the proportion of necrosis, cavitation, calcification, halo sign and vacuole sign between the two groups (P > 0.05). Logistic regression analysis showed that female, GGO containing composition, burr, air bronchogram, carcinomatous lymphangitis and lesion density showed obvious airway involvement were the predictive risk factors in patients with EGFR effective mutation. Conclusion The EGFR mutation occurs more oftern in female. The reliable predictive signs of CT include GGO composition, burr, airway bronchogram and carcinoid lymphangitis. In critical patients who are not easy to obtain clinical pathology, it has a guiding significance to radiographic assessment for EGFR effective mutation.

Key words: carcinoma, non-small-cell lung, tomography, X-ray computed, mutation, genes, erbB-1, epidermal growth factor receptor genes