ESTRO 2025 - Abstract Book

S1864

Clinical – Upper GI

ESTRO 2025

Chuong, M. D. et al Stereotactic MR-guided on-table adaptive radiation therapy (SMART) for borderline resectable and locally advanced pancreatic cancer: A multi-center, open-label phase 2 study. Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology , 191 , 110064 Michalet, M., et al. Stereotactic MR-Guided Radiotherapy for Pancreatic Tumors: Dosimetric Benefit of Adaptation and First Clinical Results in a Prospective Registry Study. Frontiers in oncology , 12 , 842402.

4200

Proffered Paper Integrating photon with proton dose-response data alters a pulmonary complication model-based patient selection in esophageal cancer Yuting Liu 1 , Pieter Populaire 1,2 , Camille Draguet 1,3 , Karin Haustermans 1,2 , Gilles Defraene 1 1 Laboratory of Experimental Radiotherapy, KU Leuven, Leuven, Belgium. 2 Department of Radiation Oncology, UZ Leuven, Leuven, Belgium. 3 Molecular Imaging Radiotherapy and Oncology (MIRO), UCLouvain, Brussels, Belgium Purpose/Objective: Proton therapy (PT), compared to photon-based radiotherapy (XT), delivers less dose to normal tissues, while ensuring target dose coverage. Muijs et al. suggested that normal tissue complication probability (NTCP) models developed on XT data overestimated postoperative complication risk of PT in esophageal cancer(EC) 1 . This study aims to investigate the impact of incorporating PT patient data into NTCP model development, using published lung dose-response data. Material/Methods: PubMed searches were conducted to identify studies describing postoperative pulmonary complications (PPC) after esophagectomy with neoadjuvant chemoradiotherapy for EC. Publications that specifically included mean lung dose (MLD) were included. Outcomes were compared between modalities using two-proportion Z-test. Logistic regression NTCP models were fitted to the study-level or dose-bin data of PPC versus MLD, using maximum log likelihood optimization and inverse-variance weighting, once including and once excluding PT data (respectively XT+PT model and XT model). Further, a simulation study was performed on 40 patients replanned with comparative XT and PT plans to 50.4 Gy in 28 fractions, adhering to the planning objectives of PROTECT trial. The probability of PPC for both modalities was estimated using NTCP models. A patient is assumably referred to PT when the gain in NTCP (∆NTCP) is 10% or larger. Results for the two models were statistically compared (paired t-test). Results: 2,026 publications were identified until 17th June 2024, with 11 non-overlapping studies meeting our selection criteria. The total sample size was 1,907, including 149 PT patients. 17.5% PT patients had PPC compared to 30.0% with XT(p=0.001). Both NTCP models revealed an increased PPC risk with MLD, and the XT model estimated a higher PPC rate than the XT+PT model in the low-dose region (absolute 4.7% at 1 Gy MLD) (Figure 1).