P1.176. Cardiac Predictors of Esophageal Cancer Inoperability Using Coronary CT and Cardiac MRI
Abstract
Esophageal Cancer: Surgical Treatment of Esophageal Cancer Surgery remains a key part of multimodal esophageal cancer treatment but carries high morbidity and mortality. Early detection of patients unfit for major surgery is essential. We examined if thorough cardiac evaluation can predict surgical operability in esophageal tumor patients. We prospectively included patients with esophageal tumors during pre-treatment workup. Coronary CT angiography assessed Agatston calcium score (CaScore) and segment involvement score (SIS), while cardiac MRI measured left ventricular mass end-diastolic index (EDi) and ejection fraction (EF). Patients were deemed inoperable if a multidisciplinary team review found tumor progression (distant metastases or local irresectability) or poor general condition precluding major surgery. We compared operable vs. inoperable groups using non-parametric tests. A logistic regression model integrated SIS, EDi, and EF; performance was evaluated via 5-fold cross-validated AUC (discrimination), Brier score (calibration), and decision curve analysis (clinical utility). The primary endpoint was inoperability. Of 24 analyzed patients, 7 (29.2%) were inoperable. Inoperable cases had much higher CaScore—median 1016 (IQR 237–2387) vs. 7 (0–180), p=0.007—and greater plaque burden by SIS—9 (6–10.5) vs. 3 (2–4), p=0.004. MRI revealed higher LV mass EDi—60.7 g/m2 (51.6–65.0) vs. 46.7 g/m2 (41.8–53.0), p=0.009—and a trend toward lower EF—49.5% (44.5–56.9) vs. 60.5% (57.2–67.5), p=0.087. The SIS–EDi–EF model showed a cross-validated AUC of 0.74 and a Brier score of 0.137. Decision curve analysis demonstrated superior net benefit over treat-all or treat-none strategies across threshold probabilities ≈0.17–0.80. Cardiac markers from CCTA (CaScore, SIS) and MRI (EDi, EF) strongly correlate with inoperability. The simple SIS–EDi–EF model offers solid discrimination and meaningful clinical net benefit over a wide decision range, aiding preoperative risk stratification and patient selection for esophageal surgery.