ESTRO 2024 - Abstract Book

S1604

Clinical - Lung

ESTRO 2024

One hundred and twelve patients with pathologically confirmed NSCLC III treated between 2015/10 and 2022/03 were eligible for this study. Patients received two cycles of platinum-based induction chemotherapy followed by high-dose radiotherapy (RT). As of 2017/09, durvalumab maintenance therapy was administered for one year. The clinical endpoints were based on the thresholds jointly published by the European Respiratory Society (ERS) and the American Thoracic Society (ATS). Pre-treatment DL CO of 60% was correlated to the incidence of pneumonitis. The post-treatment DL CO decline of 10%, which appears to be clinically relevant, was related to radiation dose.

Results:

Patients with a pre-treatment DL CO < 60% had a higher probability of pneumonitis (n = 98; r = 0.175; p-value = 0.042), which could be reproduced in the subgroup of patients who did not receive durvalumab (n = 40; r = 0.288; p value = 0.036). In these individuals, the decline in DL CO ≥ 10% depended significantly on the size of the lung volume receiving between 45% and 65% (V65 – 45%) of the total radiation dose (r = 0.354; p-value = 0.020) and V20 Total Lung (r = 0.466; corrected p-value = 0.042).

Conclusion:

The current analysis revealed that DL CO is a predictor for clinically relevant pneumonitis and a monitoring tool for post-treatment lung function as it correlates with radiation dose. This underlines the importance of peri-treatment lung function testing.

Keywords: durvalumab, immunotherapy, chemoradiotherapy

420

Digital Poster

Multicentric PET radiomics models to predict recurrence in lung cancer treated by radiation therapy

François Lucia 1 , Thomas Louis 2 , François Cousin 2 , Vincent Bourbonne 1 , Dimitris Visvikis 3 , Carole Mievis 4 , Nicolas Jansen 4 , Bernard Duysinx 5 , Romain Le Pennec 6 , Malik Nebbache 1 , Martin Rehn 1 , Mohamed Hamya 1 , Margaux Geier 7 , Pierre-Yves Salaun 6 , Ulrike Schick 1 , Mathieu Hatt 3 , Philippe Coucke 8 , Roland Hustinx 2 , Pierre Lovinfosse 2 1 CHU Brest, Radiation Therapy, Brest, France. 2 CHU Liège, Nuclear Medicine, Liège, Belgium. 3 University of Brest, LaTIM, Brest, France. 4 CHU Liège, Radiation Therapy, Liège, Belgium. 5 CHU Liège, Pulmonology, Liège, Belgium. 6 CHU Brest, Nuclear Medicine, Brest, France. 7 CHU Brest, Medical Oncology, Brest, France. 8 CHU Liège, rad, Liège, Belgium

Purpose/Objective:

To develop machine learning models to predict regional and/or distant recurrence in patients with early-stage non small cell lung cancer (ES-NSCLC) after stereotactic body radiation therapy (SBRT) using [18F]FDG PET/CT and CT radiomics combined with clinical and dosimetric parameters.