S2484
Physics - Radiomics, functional and biological imaging, and outcome prediction
ESTRO 2026
Digital Poster 4165 Lymphocyte loss during CRT of rectal cancer: prediction with a dynamic flow model. Celia Juan-Cruz 1 , Jérémy Godart 2,3 , Sander C Kuipers 2,3 , Bas Schipaanboord 1 , Alice Couwenberg 1 , Tomas Janssen 1 1 Department of Radiation Oncology, The Netherlands Cancer Institute, Amsterdam, Netherlands. 2 Department of Radiotherapy, Erasmus MC Cancer Institute-University Medical Center Rotterdam, Rotterdam, Netherlands. 3 Department of Medical Physics & Informatics, HollandPTC, Delft, Netherlands Purpose/Objective: Lymphocytes are immune system cells that play a fundamental role in the success of anti-cancer therapies. Radiation-induced lymphopenia (RIL), however, is a known side-effect of radiotherapy [1]. Reducing RIL might improve treatment outcomes [2], which requires proper understanding of the underlying dose-effect relationship. Lymphocytes are no conventional organ-of-interest but rather a system- of-interest, being dynamically distributed throughout the body and continuously produced. These dynamics can be captured in a compartmental model to predict absolute lymphocyte counts (ALC), as previously developed for cervical cancer [3]. This work aims to optimize this model on locally-advanced rectal cancer (LARC) patients undergoing concurrent chemoradiation (CRT), compare its performance to a simpler PTV volume-based prediction, and assess its potential to explore RIL mitigation strategies. Material/Methods: Data from LARC patients treated between 2012-2024 with CRT (25x2Gy+capecitabine) were retrospectively collected. Patients with available baseline ALC and recorded nadir ALC after week 3 of treatment were selected. The compartmental model describes the lymphocyte flow through seven compartments: skin, spleen, blood [4], bone marrow, intestines, lymph nodes, and other tissue. Contours and subsequent dose-volume histograms were tailored for our cohort following [3] and dose-probability distributions were derived for each patient as model input. Model intercept (predicted ALC at baseline) was optimized per-patient to account for patient ALC variability and measurement uncertainty. Lymphocyte radiosensitivity was also optimized per-patient. Parameters were optimized via grid search to minimize root-mean squared errors (RMSEs) between measured and predicted ALC. To benchmark performance, RMSE of ALC predictions at nadir was compared with that of a linear regression model predicting nadir with PTV volume. To assess potential RIL reduction, planned doses were synthetically contracted to simulate 10mm PTV margin reduction
exploratory analysis aimed to characterize the relationship between dosimetric parameters and clinically relevant adverse events in pediatric patients with MB and EPN undergoing salvage re-irradiation. Material/Methods: A retrospective review identified 49 pediatric patients (aged 3-18 years) with recurrent MB (n=24) or EPN (n=25) who received Re-RT between January 2010 and December 2023. Dosimetric data from treatment planning systems were retrieved for analysis. Clinicopathological variables, radiation parameters, and clinical outcomes were extracted from institutional databases. The Mann-Whitney U testing and Fisher's Exact test were performed to evaluate associations between dose-volume parameters and RN. Cumulative doses were converted to equivalent dose in 2-Gy fractions (EQD2, α / β =2 Gy). Results: The median patient age at presentation and recurrence were 7 years (IQR 4-9.5years) and 10 years (IQR 7-13 years) respectively. Initial radiotherapy median dose was 55.8 Gy (IQR 54-56 Gy), while re- irradiation median dose was 50.4 Gy (IQR 45-54 Gy). The Median interval between treatments was 29.2 months (IQR 24.3-49.2 months). The cumulative median EQD2 was 101.9 Gy (IQR 99-104.8 Gy). Radiation necrosis occurred in 3 patients (6.1%), with 2 symptomatic patients requiring hospitalization and prolonged corticosteroid therapy. The median time to necrosis development post Re-RT was 3 months (IQR 2-13.5 months). The median cumulative EQD2 in cases of symptomatic radiation necrosis post Re-RT was 101.9 Gy (IQR 99-104.8 Gy). Statistical analysis revealed no significant correlation between cumulative prescription dose, target volumes, or organ-at-risk doses with radiation necrosis development (p>0.05). Conclusion: Children appear to tolerate high dose Re RT well, with low rates of symptomatic radiation necrosis, with putative associations with superior neuroplasticity and repair mechanisms in the pediatric brain. The underlying biological mechanisms and coherent dose volume associations need to be elucidated via more granular analysis of larger datasets and representative laboratory studies. References: 1. Ajithkumar T, Avanzo M, Yorke E, et al. Brain and Brain Stem Necrosis After Reirradiation for Recurrent Childhood Primary Central Nervous System Tumors: A PENTEC Comprehensive Review. Int J Radiat Oncol Biol Phys. 2024;119(2):655-6682. Bentzen SM, Constine LS, Deasy JO, et al. Quantitative Analyses of Normal Tissue Effects in the Clinic (QUANTEC): An Introduction to the Scientific Issues. Int J Radiat Oncol Biol Phys. 2010;76(3 Suppl):S3-S9. Keywords: Re-irradiation, Radiation Necrosis, Ependymoma,
Made with FlippingBook - Share PDF online