ESTRO 2025 - Abstract Book

S1300

Clinical - Lung

ESTRO 2025

maximum ACC=90.1%. Our study shows that predictive models generated with features from pCT without any dosimetric information is not inferior to its counterpart. This makes early adjustment in treatment planning decisions in advance possible. Conclusion: We had actionable insight for developing better models to predict radiation pneumonitis after lung tumor SBRT, by identifying predictive values of different meningful regions in the lung.

Keywords: radiomics, prediction model, radiation pneumonitis

525

Digital Poster Suitability of Metastatic Lung Cancer for PET-Guided Radiotherapy: A Representative Sampling of NSCLC Patients at Staging Ottar Schmitz 1 , Angela J Da Silva 2 , Karine A Al Feghali 3 , Dorine de Jong 3 , Arjun Maniyedath 3 , Shervin M Shirvani 3 1 Product Strategy, RefleXion Medical, Hayward, USA. 2 Research & Development, RefleXion Medical, Hayward, USA. 3 Medical Affairs, RefleXion Medical, Hayward, USA Purpose/Objective: SCINTIX ® biology-guided radiotherapy uses positron emissions from the tumor after injection of a radiotracer to direct intra-fractional treatment 1 . “Biology guidance” is particularly interesting for treatment of tumors subject to motion in multiple planes such as lung tumors. To assess the benefit of SCINTIX therapy in patients with more than one treatment target 2 , a representative dataset from patients diagnosed with Non-Small Cell Lung Cancer (NSCLC) was constructed to illustrate prevalence and location of lesions in patients with multiple targets in a SBRT-like fractionation schedule. Material/Methods: A 355-patient staging dataset from The Cancer Imaging Archive (TCIA) was interrogated to identify NSCLC patients with more than one lesion on diagnostic FDG PET/CT at the time of staging. For patients meeting this criterion, all lesions were segmented using a Mirada Medical RTx workstation. Eligibility for SCINTIX therapy was determined for each lesion based on three criteria: meeting RTOG 0813 inclusion criteria for SBRT ascertained in consultation with a radiation oncologist; having sufficiently strong FDG avidity defined as an SUVmax of 6 or greater; and being located at least 2 cm from any other FDG-avid structure. Results: A total of 29 patients with more than one lesion on PET at the time of staging were identified. Of these patients, 21 (72%) had at least one SCINTIX eligible target, 11 (38%) had more than two SCINTIX eligible targets, and 7 (24%) had more than three potential targets. A total of 207 lesions were segmented, with 160 lesions (77%) having FDG-avidity of at least SUVmax of 6. After applying RTOG 0813 criteria, the number of SCINTIX eligible targets with both strong FDG-avidity and location appropriate for RTOG 0813 was 47 (23%). Of these criteria, the location of the potential target had the greatest impact on whether SCINTIX or any SBRT therapy could be applied per RTOG 0813, with 84 targets (53%) excluded based on anatomical location. Conclusion: Despite a conservative application of RTOG 0813 inclusion/exclusion criteria, the number of patients that could be considered for SCINTIX therapy in this metastatic cohort is relatively high. We observed that the largest contributor to exclusion from any SBRT-like therapy was target location within the central or ultra-central region of the thorax. Hence, a therapy that could be safely and effectively applied in these regions may have a significant impact on patient inclusion in SCINTIX or any SBRT-like radiotherapy.