Predicting invasiveness of subsolid nodules: a HRCT-based model for lung adenocarcinoma.

BACKGROUND: Subsolid nodules (SSNs) pose a diagnostic challenge in lung adenocarcinoma management. High-resolution computed tomography (HRCT) allows detailed characterization of SSNs, aiding in distinguishing pathological subtypes. OBJECTIVE: To investigate the correlation between HRCT features and pathological subtypes of SSNs, and their association with nodule size and morphological features. METHODS: Clinical and HRCT data from 84 patients with surgically confirmed lung adenocarcinoma were retrospectively analyzed. All patients underwent preoperative CT scans, with lesions measuring ≤3.0 cm and a ground-glass opacity component ≥50%. The evaluated CT characteristics included leison size, lobulation, spiculation, pleural indentation, and CT values. Pathological diagnosis were established according to the latest classification standards. RESULTS: Significant differences were observed among AIS, MIA, and IAC groups in age, lobulation, spiculation, and nodule size (all P<0.01). IAC showed larger size (90.9% between 8-10 mm) and more aggressive features than AIS (57.6% ≤8 mm). Advanced vascular and bronchial patterns were associated with invasive subtypes (P<0.001). EGFR+ tumors exhibited larger size and higher CT values. Multivariate analysis identified age ≥55, lesion diameter ≥8.51 mm, and bronchial pattern as significant predictors for distinguishing MIA from IAC. CONCLUSION: HRCT features effectively reflect pathological invasiveness of SSNs and can assist in differentiating lung adenocarcinoma subtypes, providing valuable information for diagnosis and treatment planning.