KEY POINTS
- The study used planning and daily magnetic resonance imaging data from the first 10 patients receiving MR-guided adaptive pancreatic stereotactic ablative radiotherapy, providing 48 evaluable fractions.
- Clinical treatment prescribed 40 Gy in five fractions. The investigators retrospectively generated non-adaptive volumetric modulated arc therapy plans using gastrointestinal planning-organ-at-risk-volume margins of 2, 3 and 5 mm.
- None of the tested margins reliably encompassed interfraction motion. Stomach containment was only 8.3%, 14.6% and 27.1% with 2, 3 and 5 mm margins; corresponding duodenal containment was 0%, 2.3% and 13.6%.
- Without a planning-organ-at-risk-volume margin, median delivered D0.1 cm³ biological effective dose increased by 26.0% for the duodenum and 1.9% for the stomach relative to baseline planning.
- With 2, 3 and 5 mm margins, median duodenal D0.1 cm³ biological effective dose changed by −11.0%, −20.5% and −31.1%, respectively. Corresponding stomach changes were −11.4%, −16.9% and −22.8%.
- Protection of gastrointestinal organs came at the cost of target dose. Median planning target volume BED10 D70 decreased from 81.3 Gy in the baseline plan to 77.2, 74.2 and 64.9 Gy with 2, 3 and 5 mm margins.
- The minimum target-coverage criterion was met in 10 of 10, 8 of 10 and 5 of 10 plans using 2, 3 and 5 mm margins, respectively. Large individual organ-dose increases still occurred despite margins.
CLINICAL TAKEAWAY
A 2–3 mm gastrointestinal margin may offer a pragmatic compromise when online adaptation is unavailable, reducing average stomach and duodenal dose escalation without the target compromise seen with 5 mm. It is not a substitute for on-treatment assessment: the margins rarely encompassed actual motion, outliers persisted, and the study did not measure toxicity or tumour control.