S2787
RTT - RTT contouring, target definition, and treatment planning
ESTRO 2026
which achieved dosimetric eligibility and were clinically deliverable. Conclusion: The AI dose prediction tool acted as a virtual member of the MDT, providing retrospective benchmarking indicating more patients could have received SIB- HypoRT during the pilot implementation phase. It provided RTTs with actionable insights to optimise and standardise local planning procedures and optimisation template. The AI tool supported quality improvement in the processes of prospective benchmarking, establishing pilot feasibility, and training and audit in RTT-led treatment planning. References: Study protocol of a randomised controlled trial of prostate radiotherapy in high-risk and node-positive disease comparing moderate and extreme hypofractionation (PRIME TRIAL). Murthy et al., BMJ Open 2020;10:e034623A Phase II Randomised Control Trial of a Conventional Versus Hypofractionationed Radiation Regimen in a Single Phase using IMRT Technique and Long Term Androgen Suppression Therapy in Patients with Solid Tumours: Advanced Technology Meets Clinical Outcomes(CHIRP Study) Keywords: AI Dose Prediction, Prostate, Benchmarking Comparative Dosimetric Analysis of SRS VMAT and HyperArc for Single and Multiple Brain Metastases Fernando M. Costa 1,2 , Patrícia Varzim 1 , Rosa Patrício 1,2 , Daniela Saraiva 1 , Raquel Rocha 1 , Anabela Gonçalves 1 , Gabriel Farinha 1 , Ana Rita Figueira 1 , Fátima Aires 1 , Pedro Soares 1 , Armanda Monteiro 1,2 , Lígia Osório 1 1 Radiation Oncology Department, São João Universitary Hospital Center, Porto, Portugal. 2 Medical Imaging and Radiotherapy Department, E2S School of Health, Polytechnic of Porto, Porto, Portugal Purpose/Objective: Brain metastases occur in approximately 20% of patients with malignant neoplasms. Radiosurgery has become a cornerstone in their management, with Linac-based SRS widely used, relying on a steep dose gradient beyond the target to spare healthy brain tissue.1,2,3 Plan quality indices (PQI)—including the Paddick Conformity Index (PCI), Selectivity Index (SI), Homogeneity Index (HI), and Gradient Index (GI)—are key parameters for comprehensive plan evaluation.1,3 This study aims to compare plan quality between SRS VMAT and HyperArc techniques for single (SBM) and multiple brain metastases (MBM) through PQI. Material/Methods: Proffered Paper 2916
prescription. Material/Methods:
The pilot involved 95 patients considered for SIB- HypoRT across three radiotherapy centres using two treatment planning systems (Eclipse, Monaco). Prescription-specific optimisation templates were applied and manually adjusted during RTT-led Manual Planning (MP). A subset of 51 cases was retrospectively analysed using an AI Dose Prediction tool (MVision Dose+), targeting prostate and nodal PTV coverage while sparing rectum and bladder. Dosimetrically superior AI Planning (AIP) was independently validated and the piloted treatment planning optimisation template was updated accordingly. Results: 50/95 (53%) pilot cases were dosimetrically eligible: 48/50 (96%) completed SIB-HypoRT; 2/50 (4%) underwent partial SIB-HypoRT.26/95 (27%) were dosimetrically ineligible due to target/OAR Dose Volume Constraints (DVCs), with several structures failing multiple DVCs: 17/26 (65%) failed bowel bag; 8/26 (31%) failed bladder; 6/26 (23%) failed rectal DVCs; 1/26 (4%) due to primary target coverage; 1/26 (4%) due to nodal target coverage.19/95 (20%) were independently ineligible for SIB-HypoRT: 9/19 (47%) due to updated radiotherapy prescription; 4/19 (21%) were clinically indicated for sequential phases; 4/19 (21%) were medically or surgically contra-indicated; 2/19 (11%) opted to delay or transfer treatment.
Excluding the 19/95 (20%) independently ineligible cases, 50/76 (66%) were dosimetrically eligible and 26/76 (34%) were dosimetrically ineligible. The AI subset of 51/76 (67%) cases improved a range of DVCs, increasing dosimetric eligibility to 44/51 (86%), from the pilot MP rate of 50/76 (66%).
For validation, a sample of six cases deemed dosimetrically ineligible by MP but eligible by AIP were replanned using a new AIP-informed protocol, all of
Made with FlippingBook - Share PDF online