ESTRO 2025 - Abstract Book

S1789

Clinical – Upper GI

ESTRO 2025

respectively). Comparing the density changes between T0 and T1, the mean density variation in the arterial phase was significantly lower in the responder group (-17.241%) compared to non-responders (2.895%) (p=0.043).

Conclusion: Our findings suggest that patient selection for SBRT should be based on pretreatment CT characteristics, including lesion size, volume, and contrast enhancement. These factors may be useful for predicting treatment response and guiding early post-treatment follow up management.

Keywords: Liver metastases, contrast enhanced CT, SBRT

2109

Digital Poster Real-world clinical outcomes of stereotactic MR-guided adaptive radiotherapy (SMART) for localized pancreatic cancer: A retrospective cohort analysis Rowena Cazalet 1 , Elena Moreno-Olmedo 1,2 , Killian Nugent 2 , Ben George 3 , Robert Owens 1,2 , Kasia Owczarczyk 4 , Andy Gaya 4 , Luis Aznar-Garcia 2 , Joe Drabble 2 , Tim Maughan 1,5,6 , James Good 2 , Somnath Mukherjee 1,2 1 Oncology, Oxford University Hospitals NHS Foundation, Oxford, United Kingdom. 2 Stereotactic and MR-guided Radiotherapy, Genesiscare UK, Oxford, United Kingdom. 3 Stereotactic and MR-guided Radiotherapy, GenesisCare UK, Oxford, United Kingdom. 4 Stereotactic and MR-guided Radiotherapy, Genesiscare UK, London, United Kingdom. 5 Oncology, University of Liverpool, Liverpool, United Kingdom. 6 Oncology, Oxford University, Oxford, United Kingdom Purpose/Objective: Stereotactic ablative radiotherapy (SABR) is an option for locally advanced pancreatic cancer (LAPC) . However, target motion and proximity to organs at risk (OAR) remain a challenge and severe toxicity has been reported in patients treated on CT-based platform. SMART may overcome these pitfalls, and a clinical trial has reported low toxicity with promising outcomes 1,2 . We have reported our initial experience 3. We now report outcomes from a large cohort of patients undergoing SMART with daily adaptation. Material/Methods: LPC (medically-inoperable/borderline-operable/locally-recurrent/locally advanced) who received SMART were included in this single-centre analysis . All delivered fractions were reoptimised. Acute and late toxicities (graded using the CTCAEv5.0 scale), local control (LC), local progression-free survival (LPFS), metastases-free survival (MFS) and overall survival (OS) were evaluated. Funding support (for SMART): John Black Charitable Foundation, GC foundation. Results: Between September/2020 and October/2023, 117 patients underwent SMART. Median age was 69 years (36-89), 54.7% were males and 100% had ECOG 0-1. The commonest tumour site was pancreas head (n=79; 67.5%); stage was predominantly inoperable (LAPC/local recurrent) (88.9%) and 84.6% underwent induction chemotherapy. The median time from diagnosis to SMART was 8 months (range 1-85). The median prescribed dose was 40 Gy (range 25 50 Gy) in 1-5 fractions and the most prevalent SABR regimen (64.5%) was 40Gy/5. All 532 fractions were successfully adapted. Two patients did not complete treatment. The cumulative incidence of acute G1 and G2 toxicity was 59.8%. Fatigue was the most prevalent acute side effect (48%). Acute G3 toxicity (fatigue, nausea, abdominal pain, diarrhoea, anorexia and cholangitis/biliary obstruction) was reported in 12.8% of patients and 13.7% showed late G3 toxicity. No early/late ³G4 toxicity was detected.

The median follow-up was 19 months (range 3-109 months) from diagnosis. At the time of analysis (missing data, n= 9), 36% of the patients have relapsed locally, 55.5% demonstrated distant progression and 36% were alive Key