PREOPANC-2

Study design, population and eligibility

Treatment and comparator

Neoadjuvant FOLFIRINOX: Eight 14-day cycles: oxaliplatin 85 mg/m² IV, irinotecan 180 mg/m² IV, leucovorin 400 mg/m² IV, 5-FU 400 mg/m² IV bolus followed by 2,400 mg/m² continuous IV infusion over 46 hours. Surgery followed; no protocol adjuvant chemotherapy.

Gemcitabine-based chemoradiotherapy: Three 28-day cycles of gemcitabine 1,000 mg/m² IV. During cycles 1 and 3, gemcitabine was given on days 1 and 8; during cycle 2, days 1, 8 and 15, with hypofractionated RT 36 Gy in 15 fractions during cycle 2. Surgery followed by four cycles of adjuvant gemcitabine.

Endpoints

Primary endpoint: Overall survival in the modified intention-to-treat population. Secondary outcomes included progression-free survival, resection and margin status, pathological response, recurrence and safety. Numerical PFS/recurrence estimates are not reproduced in the accessible primary abstract and are not invented here.

Key efficacy and clinical outcomes

Median OS was 21.9 months with FOLFIRINOX versus 21.3 months with gemcitabine-based chemoradiotherapy (HR 0.88; 95% CI 0.69–1.13; p=0.32). The primary endpoint did not demonstrate a statistically significant difference.

In a predefined surgical-complication analysis, 280/369 patients (74.7%) underwent pancreatic resection: 142 after FOLFIRINOX and 138 after chemoradiotherapy. Major postoperative complications (Clavien–Dindo grade ≥3) were 26.8% versus 27.5% (p=0.884). Ninety-day postoperative mortality was 1.4% (2/142) versus 2.9% (4/138; p=0.442).

Source-reported outcome visualization

Primary endpoint median OS in months and overall grade ≥3 adverse-event proportions, shown as separate measures.

Median survival bars are not fixed-time survival probabilities. They are not used to calculate NNT. The grade ≥3 AE bars are the separate safety comparison.

Absolute effect and NNT/NNH

No statistically significant OS difference was demonstrated (HR 0.88; 95% CI 0.69–1.13; p=0.32). The median OS difference of 0.6 months is not converted into NNT. The accessible primary abstract does not provide fixed-time OS rates suitable for a landmark NNT. This was a superiority comparison; failure to demonstrate superiority does not establish equivalence or non-inferiority.

Trial-specific toxicity, grade 3–5 harms and treatment burden

Grade ≥3 adverse events: 67% with FOLFIRINOX versus 60% with chemoradiotherapy (p=0.20). Serious adverse events: 49% versus 43% (p=0.26). Most frequent grade 3–4 events: neutropenia 25% versus 22%; diarrhoea 23% versus 1%; leukopenia 8% versus 15%. Treatment-related deaths: 2 (1%) in FOLFIRINOX (multi-organ failure and intestinal mucositis) and 1 (1%) in CRT (upper GI haemorrhage).

Major postoperative complications were similar (26.8% vs 27.5%). FOLFIRINOX involves eight intensive biweekly cycles; the CRT strategy involves three gemcitabine cycles, 15 RT fractions, surgery and four adjuvant gemcitabine cycles. Total treatment burden differs substantially between arms.

Quality of life

Subgroups and applicability

Maturity, limitations and critical interpretation

Interpretation: Open-label phase III study; primary OS superiority hypothesis not met. Confidence interval includes clinically relevant benefit and harm. The result does not prove equivalence or non-inferiority.

Applicability: Dutch multicentre cohort, WHO PS 0–1, resectable/borderline resectable PDAC. Results should not be extrapolated to metastatic or locally advanced unresectable disease.

Outcome reporting: The primary abstract reports OS and key toxicity but not numerical fixed-time OS, PFS or recurrence rates. The separate surgical-complication analysis provides postoperative safety data. Race and ethnicity were not collected. No NNT is calculable from the reported median OS values.

Clinical interpretation: PREOPANC-2 did not show that neoadjuvant FOLFIRINOX improves OS over gemcitabine-based chemoradiotherapy followed by adjuvant gemcitabine in fit patients with resectable or borderline resectable PDAC. Both strategies remain possible approaches; selection should consider resectability, fitness, toxicity, treatment duration, local expertise and multidisciplinary judgement. This is a negative superiority result, not a non-inferiority demonstration.

Primary sources, protocol and follow-up reports

1. PREOPANC-2 primary phase III report, Lancet Oncology 2025 (PMID 40945523)
2. PREOPANC-2 DOI
3. Surgical complications analysis, 2025 (PMID 40792645)

Numbers not reported in the accessible primary source are not imputed. Median survival differences are never converted into NNT.