ESMO Congress 2026

Even after resection, locoregional recurrence within a year is common in pancreas cancer, and local progression causes biliary obstruction, gastric-outlet obstruction and pain. AGITG MASTERPLAN isolated that question in a randomised phase 2: high-risk resectable, borderline or locally advanced disease, modern chemotherapy plus pancreatic SBRT (40 Gy in 5 fractions). The 2021 protocol planned 120 patients, 2:1, preferred mFOLFIRINOX; primary endpoint 12-month locoregional control, with OS, PFS and R0 secondary.^[1]

The open contemporaneous randomised comparison is Alliance A021501: borderline-resectable PDAC, preoperative mFOLFIRINOX ± hypofractionated radiotherapy. The radiotherapy arm closed at interim (R0 10/30 vs 17/30); median OS in evaluable patients 17.1 vs 29.8 months.^[2] Different trial, different population, different primary. MASTERPLAN’s secondary OS does not overturn A021501, and A021501 does not negate MASTERPLAN’s local-control readout.

Floor abstract 3136RO reports n=74, smaller than the protocol. Floor numbers: 12-month locoregional recurrence 52.0% vs 20.2% (P=0.022); median PFS 6.3 vs 11.9 months (P=0.002); OS 14 vs 22 months (P=0.006). Grainne O'Kane: about 90% received mFOLFIRINOX, more R0 on the SBRT arm. Sydney Clinical Trials Centre: the result is in The Lancet Oncology and larger studies are needed. Shankar Siva’s post writes OS hazard ratio 0.45; the floor background wrote HR not reported. The two sit side by side and are not merged.^[3]

What this adds is a positive primary on local control in the mFOLFIRINOX era, with secondary survival in the same direction. Phase 2, 74 patients, OS secondary: not enough to change the standard of care. Confidence intervals, strata, resection rates and toxicity wait on the Lancet Oncology paper and the slides; this note has not read that table.