Association Between Unenhanced CT Features and Perioperative Outcomes in Bronchogenic Cyst Resection

Background: Reliable predictors of perioperative risk beyond symptomatology in bronchogenic cyst (BC) resection are lacking. We aimed to compare perioperative outcomes across different imaging features to identify perioperative risk factors. Methods: We conducted a retrospective cohort study of 57 patients who underwent pathologically confirmed BC resection at a single tertiary center from 2004 to 2023. Preoperative unenhanced CT images were retrospectively reviewed for four prespecified imaging features. Only unenhanced images were used for imaging-feature classification and Hounsfield-unit measurement. Patients were stratified into two groups based on the presence (n = 19) or absence (n = 38) of abnormal features. The prespecified primary perioperative outcomes were peri-cystic adhesion, operative time, estimated intraoperative blood loss, and intensive care unit (ICU) length of stay. Secondary outcomes included chest-tube duration, postoperative hospital length of stay, major intraoperative complications, 30-day postoperative complications, and recurrence. Results: The abnormal image-feature group had a higher prevalence of male patients, symptomatic presentation, and larger cysts. Nineteen patients (33%) had at least one abnormal CT feature. Patients with abnormal CT features had significantly higher peri-cystic adhesion rates (68% vs. 34%; p = 0.015), longer mean operative times (128 ± 60 vs. 87 ± 35 min; p = 0.011), greater blood loss (100 ± 150 vs. 4 ± 24 mL; p = 0.012), and prolonged intensive care unit stays (0.84 ± 1.21 vs. 0.05 ± 0.23 days; p = 0.011) than those with normal imaging. No significant differences were observed in outcomes of chest-tube duration, postoperative hospital length of stay, or recurrence. The presence of abnormal features and cyst size remained significant in multivariate models. Conclusions: Abnormal imaging features on non-contrast CT and cyst size are associated with more challenging perioperative courses in BC resection. Incorporating these imaging features into preoperative assessment may enhance risk stratification and guide surgical approach decisions in the CT screening era.