2295 - A Prospective Study of Multimodal Imaging in Primary Esophageal Cancer Target Delineation
Presenter(s)
R. Zeng, C. Yang, Z. Li, X. Y. Zhou, K. X. Ou, Y. N. Li, Y. W. Zha, T. Wang, M. LI, and J. Zhang; Department of Radiation Oncology, Zhongshan People's Hospital, Zhongshan, Guangdong, China
Purpose/Objective(s): Compare differences in volume, length, maximum diameter, Dice Index (DI), and Conformity Index (CI) of esophageal cancer primary target volumes (GTVp) delineated by localization CT alone (GTVp-ct), CT-PET fusion (GTVp[ct+pet]), and CT-MRI-T2WI fusion (GTVp[ct+mri]) to evaluate the value of CT-MRI fusion in GTVp delineation versus CT-PET.
Materials/Methods: Thirty-one locally advanced esophageal squamous cell carcinoma patients (upper/middle/lower lesions) undergoing radical/neoadjuvant radiotherapy (Dec 2021–Dec 2023) underwent pretreatment localization CT, PET-CT, and chest MRI under identical immobilization. Images were rigidly registered in the Varian system. GTVp-ct, GTVp[ct+pet], and GTVp[ct+mri] were delineated. Lower/middle esophageal cases utilized 4D-CT for motion management; upper esophageal cases used 3D-CT.
Results:
Volume: Lower esophageal GTVp[ct+pet] volumes (60.62 cm³) exceeded GTVp-ct (52.9 cm³; Z=2.340, P<0.05). Length: Overall, GTVp[ct+pet] (8.22 cm) and GTVp[ct+mri] (8.38 cm) were longer than GTVp-ct (7.65 cm; t=2.134, P<0.05). Lower esophageal GTVp[ct+pet] length (10.14 cm) exceeded GTVp-ct (8.89 cm; t=3.204, P<0.05). Maximum diameter: GTVp[ct+pet] (3.62 cm) and GTVp[ct+mri] (3.77 cm) had smaller transverse diameters than GTVp-ct (3.95 cm; P<0.05), with no significant inter-modality differences. CI: GTVp-ct vs. GTVp[ct+pet] CI=0.69 vs. GTVp-ct vs. GTVp[ct+mri] CI=0.68 (t=2.835, P=0.0038). GTVp[ct+pet] vs. GTVp[ct+mri] CI=0.69 (t=2.244, P=0.025). DI: GTVp[ct+pet] DI (0.79) exceeded GTVp[ct+mri] DI (0.77; t=-2.801, P<0.05). Lower esophageal GTVp[ct+pet] DI (0.87) surpassed GTVp[ct+mri] DI (0.84; t=-2.534, P<0.05).Conclusion: CT-MRI and CT-PET fusion demonstrated comparable GTVp metrics (volume, length, diameter), suggesting MRI may substitute PET for esophageal target delineation. However, directional discrepancies (?Y displacement) and variable DI/CI between modalities, compounded by limited sample size, warrant further validation of CT-MRI fusion’s clinical utility relative to CT-PET.