Association between thoracic CT scan findings and EGFR mutation subtypes in patients with pulmonary adenocarcinoma

Abstract Background Pulmonary adenocarcinoma is one of the most common types of lung cancer and accounts for approximately 12.6% of all cancer-related deaths. Epidermal growth factor receptor (EGFR) mutation subtypes are clinically relevant for determining prognosis and guiding first-line EGFR tyrosine kinase inhibitor (EGFR-TKI) therapy. Since molecular EGFR testing requires invasive and costly techniques, chest computed tomography (CT) has been explored as a noninvasive alternative to predict EGFR mutation subtypes based on morphological imaging features. Materials and methods This analytical observational study with a cross-sectional design retrospectively reviewed the medical records of 92 patients with histopathologically confirmed pulmonary adenocarcinoma. These patients underwent contrast-enhanced chest CT scans and EGFR mutation testing at Dr. Saiful Anwar General Hospital, East Java, Indonesia, from January 2018 to December 2023. Two blinded thoracic radiologists assessed CT morphological features, which were then correlated with EGFR mutation subtypes (exon 19 deletion, exon 20 insertion, exon 21 L858R, and wild-type). Associations between CT features and EGFR mutations were analyzed using Chi-square and Kruskal–Wallis tests, with odds ratios calculated for significant findings. Statistical significance was set at p < 0.05. Results In this study involving 92 patients with pulmonary adenocarcinoma, EGFR wild-type tumors were significantly associated with the absence of circumscribed margins and were 3.7 times more likely to exhibit lobulated margins. Tumors harboring EGFR exon 19 mutations showed an 8.2-fold increased likelihood of calcification and a 4.3-fold increased likelihood of atelectasis. No significant associations were found between EGFR subtypes and tumor size, air bronchogram, pleural effusion, or cavitation. Additionally, tumors with exon 21 mutations did not demonstrate any distinctive CT characteristics. Conclusion Specific thoracic CT morphological features are associated with distinct EGFR mutation subtypes. The presence of lobulated margins and the absence of circumscribed margins suggest a wild-type EGFR status, whereas calcification and atelectasis are more frequently observed in tumors with exon 19 mutations. These findings support the potential use of chest CT imaging as a noninvasive predictor of EGFR mutation subtypes in pulmonary adenocarcinoma.