ESTRO 2025 - Abstract Book

S414

Clinical - Biomarkers

ESTRO 2025

Purpose/Objective: The translational task of the RADIOSA phase II randomized trial (NCT03940235) investigates the identification of predictive and prognostic biomarkers in hormone-sensitive oligorecurrent PCa. The aim is to assess whether a liquid biopsy-based mutational signature can differentiate between true oligometastatic disease and cases likely to progress to polymetastatic disease and need treatment intensification. Material/Methods: Oligorecurrent PCa patients with an upper threshold of three bone or lymph node metastases, were randomized to receive: SBRT alone (ArmA) or SBRT+6-month ADT (ArmB). Ten milliliters (10 ml) of blood were collected at one of three predefined timepoints: (i) randomization, (ii) follow-up (3-, 6-month and yearly thereafter) visits, (iii) eventual clinical progression. Next-generation sequencing (NGS) was performed on the Illumina NextSeq550 platform. Variants were included in the analysis if they had a minimum coverage depth of 500×, a quality score≥20, and a variant allele frequency≥1%. Clinical PFS (cPFS) was the primary endpoint, defined as the time from randomization until the presence of new local, regional, or distant metastatic lesions at imaging or death from any cause. We defined a high-risk mutational (HRM) signature as a pathogenic somatic mutation within ATM, BRCA1/2, Rb1, and TP53 1 , in HIF-1α/Notch1 2 , JAK2 3 , DNMT3A 4 , ESR1 5 , KRAS and EGFR 6 . Correction for clonal hematopoiesis of indeterminate potential alterations was not performed on this preliminary data and it is planned on the entire cohort. Results: Between August 1 st 2019 and April 30th 2023, 105 patients were enrolled in the trial. A total of 92 blood samples were collected (46 in each arm). This preliminary data refers to 47 plasma samples (19 armA, 28 armB), balanced in terms of cPFS events. A total of 73 mutations were found in 34 genes, of those 47/73 (64%) were pathogenic/likely pathogenic, and 26/73 (36%) were variants of uncertain significance/likely neutral. Patients with an HRM in armA were 14/19 (73%) and 17/28 in armB (60%). When patients were stratified according to the presence of an HRM, the median PFS in the HRM group was 19mo (95% CI, 13.0-32.0) compared with 35mo (95% CI, 18.0-46.0) in those without an HRM ( Figure1 ).

Conclusion: Our preliminary seems to offer a valuable tool to be integrated with next-generation imaging to identify patients who deserve treatment intensification. Completion of the ongoing analyses with the whole cohort samples and with tissue-based confirmation of the results will hopefully provide a stronger indication of the mutational landscape in the oligorecurrent PCa scenario.

Keywords: oligometastatic state, liquid biopsy, ctDNA