Pancreatic Cancer Clinical Trial Timeline: An In-depth Report

Curated by Dual Brain Lab (csilab.net) Data cutoff: 2026-04 · Latest trial: RASolute-302 (ASCO GI 2026 topline)

1. One-sentence definition

This report maps the evolutionary logic and current decision landscape of landmark clinical trials in pancreatic ductal adenocarcinoma (PDAC) systemic therapy over the past 30 years (1994-2026), as cited in NCCN Pancreatic Adenocarcinoma V1.2026 (Principles of Systemic Therapy, PANC-G 13 pages, 52 references). The goal: give frontline clinicians a traceable panoramic map for “who, what, why” decisions at the 2026 time point.

Iron rule: every data point for every trial is traceable to PubMed (PMID) or ClinicalTrials.gov (NCT id) — each [PMID xxxxxxxx] in the text can be clicked to verify the PubMed original.

2. Longitudinal axis: six paradigm shifts over 30 years

PDAC systemic therapy has gone through six paradigm shifts in the past 30 years: adjuvant chemo upgraded from observation to gemcitabine, then to mFOLFIRINOX (mFFX, modified triplet), pushing OS from 18 to 54 months → advanced 1L formed the FOLFIRINOX / GnP chemo duopoly, rewritten a decade later by NAPOLI-3 → neoadjuvant / locally advanced three-path contention → advanced 2L from total failure to nal-IRI breakthrough → biomarker-matched sparse tiles (BRCA / MSI-H / NRG1 / NTRK) covering <10% of patients → KRAS three subtypes cracked open (G12C / G12D / pan-KRAS), with RASolute-302 in 2026 pushing 2L HR down to 0.40 for the first time.

Each shift rests on 1-3 phase III pivots. But compared to NSCLC’s five paradigm shifts, each PDAC shift is smaller in magnitude — the underlying biology differs: NSCLC runs on the “driver gene + immunotherapy” dual engine; PDAC faces a triple stranglehold of “90% KRAS trunk + stromal barrier + cold tumor”.

2.1 Adjuvant chemo evolution (1994-2023): observation → gemcitabine → mFOLFIRINOX

Story: before 1997 the PDAC post-op adjuvant standard was observation; CONKO-001 put gemcitabine on the standard shelf; ESPAC-4 won with combinations over monotherapy; PRODIGE-24 then used mFFX to push OS from 35 to 54.4 months — one real turnaround in 30 years. APACT delivered a lesson: what wins in the advanced setting (GnP) doesn’t necessarily win as adjuvant.

• ESPAC-1 [PMID 15028824] (Neoptolemos 2004 NEJM, N=289, European 2×2 factorial): adjuvant chemotherapy (5-FU + leucovorin) significantly improved OS (mOS 20.1 vs 15.5 months, HR 0.71, p=0.009); adjuvant chemoradiation was potentially harmful (HR 1.28, p=0.05 adverse trend). This shifted European adjuvant practice from the GITSG-era chemoradiation preference to “chemo-first / radiation with caution.”
• CONKO-001 [PMID 23982521] (Oettle 2013 JAMA long-term follow-up, N=368): gemcitabine vs observation. DFS median 13.4 vs 6.7 months (HR 0.55, p<0.001); 5-year OS 20.7% vs 10.4%, 10-year OS 12.2% vs 7.7%. First established adjuvant gemcitabine as the global standard, becoming the control for all subsequent adjuvant trials.
• ESPAC-3 [PMID 20823433] (Neoptolemos 2010 JAMA, N=1088): gemcitabine vs 5-FU + leucovorin. mOS 23.6 vs 23.0 months (HR 0.94, p=0.39), OS equivalent but gemcitabine had fewer serious adverse events (7.5% vs 14%). Gemcitabine’s safety advantage locked it as the subsequent comparator.
• ESPAC-4 [PMID 28129987] (Neoptolemos 2017 Lancet, N=732): gemcitabine + capecitabine vs gemcitabine monotherapy. mOS 28.0 vs 25.5 months (HR 0.82, p=0.032). Combination regimens beat monotherapy for the first time; from 2017 Europe widely adopted GemCap as adjuvant standard.
• JASPAC-01 [PMID 27275872] (Uesaka 2016 Lancet, N=385): S-1 vs gemcitabine adjuvant, Japanese multicenter. mOS 46.5 vs 25.5 months (HR 0.57, p<0.0001) — S-1 a major win. Asian PDAC adjuvant standard diverged from Western practice (S-1-associated GI toxicity is better tolerated in Asian populations).
• PRODIGE-24 / CCTG PA.6 [PMID 30575490] (Conroy 2018 NEJM, N=493): mFOLFIRINOX (oxaliplatin + irinotecan + leucovorin + 5-FU, modified doses) vs gemcitabine adjuvant. DFS 21.6 vs 12.8 months (HR 0.58, p<0.0001); mOS 54.4 vs 35.0 months (HR 0.64, p=0.003). This is the only time in 30 years of PDAC adjuvant trials that mOS was pushed past 4 years. From then on mFFX became the preferred adjuvant regimen for fit patients (ECOG 0-1).
• APACT [PMID 36521097] (Reni 2023 JCO, N=866): nab-paclitaxel + gemcitabine vs gemcitabine adjuvant. Independently assessed DFS missed the primary endpoint (HR 0.88, p=0.18). OS (secondary endpoint) HR 0.82, p=0.045 showed statistical significance but FDA held firm on primary endpoint, denying the adjuvant indication. APACT’s lesson: what wins in the advanced setting (GnP in MPACT) doesn’t necessarily win as adjuvant — the adjuvant scenario demands a higher effect-size threshold.
• PACT-15 [PMID 33301741] (Reni 2021 Lancet Gastro Hepatol, N=88, Italy): adjuvant PEXG (cisplatin+epirubicin+gemcitabine+capecitabine) vs gemcitabine. Phase IIb small-sample showed PEXG advantage, but did not enter global guidelines (insufficient sample size).

Takeaway: adjuvant PDAC evolved from “whether to treat” to “which combination.” In 2026 fit patients get mFFX (PRODIGE-24) as first choice; those who can’t tolerate or are older get GemCap (ESPAC-4) or gemcitabine monotherapy (CONKO-001); Asian populations may consider S-1 (JASPAC-01). Nab-paclitaxel + gemcitabine (APACT) is not recommended for adjuvant use.

2.2 Advanced 1L chemo duopoly (2011-2023): FOLFIRINOX / GnP coexist for a decade → NAPOLI-3 rewrites

Story: before 2011 advanced 1L PDAC had only gemcitabine monotherapy. PRODIGE-4 and MPACT, within two years, established two completely different toxicity-profile regimens (a triplet and a doublet), and no one dislodged them for a decade. Not until 2023, when NAPOLI-3 brought liposomal irinotecan (nal-IRI) into 1L, did the first challenger appear.

• PRODIGE-4 / ACCORD-11 [PMID 21561347] (Conroy 2011 NEJM, N=342): FOLFIRINOX (oxaliplatin+irinotecan+leucovorin+5-FU) vs gemcitabine. mOS 11.1 vs 6.8 months (HR 0.57, p<0.001), ORR 31.6% vs 9.4%. Toxicity notable: grade 3/4 neutropenia 45.7%, diarrhea 12.7%. First choice for fit patients (ECOG 0-1), but not suitable for >75 years / elevated bilirubin / poor performance status.
• MPACT [PMID 24131140] (Von Hoff 2013 NEJM, N=861): nab-paclitaxel + gemcitabine (GnP) vs gemcitabine. mOS 8.5 vs 6.7 months (HR 0.72, p<0.001), ORR 23% vs 7%. Relatively mild toxicity; patients with ECOG 2 or comorbidities may be considered. Became the other PDAC 1L pillar for a decade.
• GEST [PMID 23547081] (Ueno 2013 JCO, N=834, Japan): gemcitabine vs S-1 vs GS (gemcitabine+S-1). S-1 monotherapy non-inferior to gemcitabine on OS (mOS 9.7 vs 8.8 months, HR 0.88, non-inferiority boundary met). Evidence base for Asian 1L monotherapy regimens.
• CCTG PA.7 [PMID 36028483] (O’Callaghan 2022 Nature Communications, N=180): GnP ± durvalumab + tremelimumab. Primary endpoint OS negative (mOS 9.8 vs 8.8 months, HR 0.94, p=0.72). Reconfirmed that PDAC is a “cold tumor”; dual checkpoint blockade is ineffective in unselected populations.
• NAPOLI-3 [PMID 37708904] (Wainberg / O’Reilly 2023 Lancet, N=770): NALIRIFOX (liposomal irinotecan + oxaliplatin + leucovorin + 5-FU) vs GnP as 1L. mOS 11.1 vs 9.2 months (HR 0.83, p=0.036), mPFS 7.4 vs 5.6 months. FDA approved as 1L standard in February 2024. First regimen in ten years proven better than GnP in 1L — though HR 0.83 is a marginal win.

Takeaway: 2026 advanced PDAC 1L decision tree — ECOG 0-1 choose FOLFIRINOX or NALIRIFOX (the latter just FDA-approved, real-world accessibility and reimbursement still rolling out); ECOG 2 or heavy comorbidity choose GnP; monotherapy only when very frail (gemcitabine or S-1). Dual checkpoint blockade in unselected populations is not recommended.

2.3 Neoadjuvant and locally advanced (2013-2024): three paths, still no verdict

Story: for resectable / borderline resectable / locally advanced PDAC, the optimal strategy for timing of surgery and integration of chemoradiation has been the most controversial area in PDAC for ten years. PREOPANC-1 established feasibility of neoadjuvant chemoradiation; Alliance A021501 showed SBRT doesn’t add value; PREOPANC-2 brought mFFX into neoadjuvant; NORPACT-1 issued a warning: neoadjuvant is not good for everyone.

• LAP07 [PMID 27139057] (Hammel 2016 JAMA, N=449): locally advanced PDAC induction gemcitabine (± erlotinib) for 4 months, then stable/responders randomized to continued chemo vs capecitabine concurrent chemoradiation. Primary endpoint OS negative (mOS 16.5 vs 15.2 months, HR 1.03, p=0.83). Chemoradiation added to locally advanced PDAC showed no OS benefit, but LRPF (locoregional progression-free interval) extended by 7.6 months.
• SCALOP [PMID 23474363] (Mukherjee 2013 Lancet Oncol, N=74): locally advanced PDAC induction gemcitabine + capecitabine for 12 weeks, then concurrent chemoradiation phase compared capecitabine vs gemcitabine as radiosensitizer. mOS 15.2 vs 13.4 months (adjusted HR 0.39, p=0.012) — capecitabine significantly better. From then on capecitabine became the preferred radiosensitizer in the locally advanced PDAC chemoradiation phase.
• SCALOP-2 [PMID 34048677] (Mukherjee 2021 JCO, N=170): locally advanced PDAC post-induction ± nelfinavir (HIV protease inhibitor as radiosensitizer). Primary endpoint negative. The chemoradiation + sensitizer approach failed.
• Alliance A021501 [PMID 35834226] (Katz / O’Reilly 2022 JAMA Oncology, N=126): resectable / borderline resectable PDAC neoadjuvant mFOLFIRINOX 8 cycles vs mFOLFIRINOX 7 cycles + SBRT. SBRT arm had worse 18-month OS (47.3% vs 66.7%). SBRT boost did not improve outcomes, and may even have harmed them. The borderline-resectable neoadjuvant SBRT concept was rejected.
• PREOPANC-1 [PMID 35188492] (Versteijne 2022 JCO, N=246, long-term follow-up): borderline resectable / resectable PDAC neoadjuvant gemcitabine concurrent chemoradiation vs upfront surgery. 5-year OS 20.5% vs 6.5% (HR 0.73, p=0.025). Neoadjuvant chemoradiation showed OS benefit in long-term follow-up.
• ESPAC-5F [PMID 37103886] (Ghaneh 2023 JAMA Oncology, N=90, pilot): borderline resectable PDAC neoadjuvant FOLFIRINOX / GemCap / chemoradiation vs upfront surgery. 1-year OS neoadjuvant chemo arm (any regimen) 77% vs upfront surgery 40%. Pilot sample small but direction clear — for borderline resectable, neoadjuvant chemo beats upfront surgery.
• PREOPANC-2 [PMID 39500336] (Janssen 2024 Lancet, N=375): borderline resectable / resectable PDAC neoadjuvant mFOLFIRINOX 8 cycles vs neoadjuvant gemcitabine concurrent chemoradiation. Primary endpoint OS equivalent (HR 0.87, p=0.28). The two neoadjuvant paths are evenly matched; neither dominates.
• NORPACT-1 [PMID 38237621] (Lassen 2024 Lancet Gastroenterology & Hepatology, N=140, Nordic): resectable pancreatic head cancer neoadjuvant FOLFIRINOX 4 cycles vs upfront surgery. Primary endpoint OS negative (mOS 25.1 vs 38.5 months, HR 1.52, p=0.0468 — direction “unfavorable”). For resectable (not borderline) PDAC, neoadjuvant FOLFIRINOX may delay surgery and harm outcomes. A warning that “neoadjuvant” is not suitable for all surgical candidates.
• Alliance A021806 [PMID 39048905] (Katz 2024 design/update paper, ongoing): resectable PDAC perioperative mFOLFIRINOX vs adjuvant mFOLFIRINOX head-to-head. Primary not yet readout — this will be the first phase III to systematically answer “pre-op vs post-op chemo in resectable PDAC.”

Takeaway: 2026 neoadjuvant / locally advanced PDAC has no global consensus. Borderline resectable leans toward neoadjuvant chemo (PREOPANC-1/2 support, ESPAC-5F points the same way); for resectable pancreatic head cancer NORPACT-1 warns against casually giving neoadjuvant FFX; locally advanced chemoradiation serves locoregional control only (LAP07 OS neutral); SBRT boost neoadjuvant Alliance A021501 rejected; sensitizer approach SCALOP-2 rejected.

2.4 Advanced 2L and beyond (2014-2016): from total failure to nal-IRI breakthrough

Story: before 2016 advanced PDAC 2L had no standard regimen. CONKO-003 and NAPOLI-1 within two years moved 2L from “wait to die” to “have options.” PANCREOX, meanwhile, reminded us via FOLFOX’s failure in Western cohorts that “East and West results may differ.”

• CONKO-003 [PMID 24982456] (Pelzer 2014 JCO, N=168, Germany): after gemcitabine failure, OFF (oxaliplatin + 5-FU + leucovorin) vs 5-FU + leucovorin. mOS 5.9 vs 3.3 months (HR 0.66, p=0.010). First positive 2L RCT in Europe.
• NAPOLI-1 [PMID 26615328] (Wang-Gillam 2016 Lancet, N=417): after gemcitabine failure, liposomal irinotecan (nal-IRI) + 5-FU/LV vs 5-FU/LV vs nal-IRI monotherapy. nal-IRI+5FU/LV arm mOS 6.1 vs 4.2 months (HR 0.67, p=0.012). First 2L regimen globally approved by FDA; nal-IRI entered the PDAC essential arsenal from then on.
• PANCREOX [PMID 27621395] (Gill 2016 JCO, N=108, Canada): after gemcitabine failure, mFOLFOX6 vs 5-FU/LV. mOS 6.1 vs 9.9 months (HR 1.78, p=0.02) — the FOLFOX arm was worse. Suggests CONKO-003’s OFF result could not be reproduced in Western cohorts; FOLFOX is not recommended in Western 2L.

Takeaway: 2026 advanced PDAC 2L — nal-IRI + 5FU/LV (NAPOLI-1) is the only globally approved regimen; OFF (CONKO-003) is an option in Europe; FOLFOX is not recommended in Western populations (PANCREOX). The 2L OS ceiling still sits at 6 months — this is the floor that the KRAS revolution aims to break through.

2.5 On the eve of precision therapy: four tiles don’t make a wall (2019-2025)

Story: before KRAS was cracked, PDAC had four “rare but actionable” biomarkers accessible via basket trials: BRCA (POLO), MSI-H (KEYNOTE-158), NRG1 (eNRGy), NTRK (NAVIGATE / STARTRK / TRIDENT-1). Each covers <2-5% of patients, totaling under 10%. “First tile” points the way but is far from a wall.

• POLO [PMID 31157963 / 35834777] (Golan 2019 NEJM / Kindler 2022 JCO final OS, N=154): gBRCA1/2-mutant advanced PDAC (~6% of patients), ≥16 weeks of 1L platinum chemo without progression → olaparib maintenance vs placebo. mPFS 7.4 vs 3.8 months (HR 0.53, p=0.004), but mOS 19.0 vs 19.2 months (HR 0.83, p=0.35) negative. First FDA-approved biomarker-driven treatment in PDAC; but the PFS-OS mismatch (PFS significant / OS neutral) exposed the clinical-significance limits of maintenance therapy.
• KEYNOTE-158 (PDAC cohort) [PMID 31682550] (Marabelle / Diaz 2020 JCO, PDAC n=22): MSI-H / dMMR advanced solid tumor basket, pembrolizumab monotherapy. PDAC subgroup ORR 18.2%. MSI-H PDAC is only 1-2% of patients; sample tiny but supports the tumor-agnostic approval.
• eNRGy (PDAC cohort) [PMID 39908431] (Schram 2025 NEJM, PDAC n=36): NRG1-fusion advanced solid tumor basket, zenocutuzumab (HER2×HER3 bispecific antibody). PDAC subgroup ORR 42% (15/36), mDoR 11.1 months. NRG1 incidence in PDAC ~0.5-1.5% (enriched in KRAS wild-type PDAC). FDA 2024-12 accelerated approval — first non-chemo non-PARP biomarker-matched targeted approval in PDAC.
• larotrectinib (NAVIGATE / pooled) [PMID 32105622] (Hong 2020 Lancet Oncology): NTRK-fusion tumor-agnostic, PDAC n=3, responses seen in small samples. NTRK in PDAC <0.5%.
• entrectinib (STARTRK-2 integrated) [PMID 31838007] (Doebele 2020 Lancet Oncology): another NTRK-fusion tumor-agnostic path; PDAC subpopulation also has n=3-level small data.
• TRIDENT-1 (PDAC cohort) (Drilon et al., ROS1/NTRK integrated analysis, 2025 ongoing manuscript): repotrectinib (next-gen NTRK/ROS1 TKI). NCT04094610. PDAC subgroup case count very small; awaiting full publication.

Takeaway: in 2026 all newly diagnosed PDAC should undergo comprehensive molecular profiling (covering at least KRAS subtype / BRCA1/2 / MSI / NRG1 / NTRK / HER2 / BRAF) — this is an explicit NCCN V1.2026 recommendation. But adding up the “four tiles,” only 5-10% of patients can enter biomarker-matched therapy. 90%+ of PDAC is still waiting for KRAS or another trunk breakthrough.

2.6 KRAS cracked open (2023-2026): G12C → G12D → pan-KRAS, three subtypes

Story: KRAS mutation occurs in 90%+ of PDAC; “undruggable” was clinical dogma for 30 years. Starting in 2021 G12C inhibitors won in lung cancer first (CodeBreaK 200 / KRYSTAL-1 NSCLC); in 2023 baskets extended to PDAC; in 2024 Revolution Medicines’ RMC-6236 (pan-KRAS G12X) phase 1 PDAC data shook ASCO GI / ESMO; in April 2026 RASolute-302 phase 3 topline pushed PDAC 2L HR down to 0.40 for the first time — an effect size unseen in 30 years.

• KRYSTAL-1 (PDAC cohort) [PMID 37099736] (Bekaii-Saab 2023 JCO, PDAC n=21): KRAS G12C PDAC (~1-2% of PDAC patients), adagrasib 600 mg BID. ORR 33% (7/21), mPFS 5.4 months, mOS 8.0 months. First evidence that KRAS G12C inhibitors reproduce the lung-cancer response pattern in PDAC.
• CodeBreaK 100 (PDAC cohort) [PMID 36546651] (Strickler 2023 NEJM, PDAC n=38): KRAS G12C PDAC, sotorasib 960 mg QD. ORR 21% (8/38), mPFS 4.0 months. The second drug in the G12C class, equivalent to KRYSTAL. Together they turned KRAS G12C PDAC from “untreatable” into a ~25% response 2L+ standard.
• MRTX1133 phase I (NCT05737706): KRAS G12D-selective inhibitor (Mirati / BMS). 2023 dose-escalation started, but in 2024 due to PK issues and formulation constraints, clinical development terminated. The G12D story appeared interrupted.
• RMC-9805 (zoldonrasib) phase I (NCT06040541): Revolution Medicines’ KRAS G12D-selective next-gen molecule. First public data at EORTC-NCI-AACR 2024 late-breaking abstract: PDAC G12D 2L+ ORR ~30% (safety and tolerability data better than MRTX1133). FDA Breakthrough Designation. G12D story continues with a handoff.
• RMC-6236 (daraxonrasib) phase I/Ib (NCT05379985): Revolution Medicines’ pan-KRAS G12X “tri-complex RAS(ON) inhibitor”, simultaneously covering G12D/G12V/G12C/G13D (~85% of KRAS PDAC). 2024 ESMO and 2025 ASCO GI released the PDAC 2L cohort (n≈60-80): ORR ~36%, mPFS ~8.5 months, mOS ~14.5 months — versus historical PDAC 2L mOS of 6 months, an unprecedented effect size.
• RASolute-302 (phase III, NCT06625320): RMC-6236 monotherapy vs investigator’s choice (FOLFIRINOX / GnP) in 2L+ KRAS G12X PDAC. April 2026 ASCO GI topline: mOS 13.2 vs 6.7 months, HR 0.40, p<0.001. This is an HR magnitude unseen in 30 years of PDAC 2L (previous largest 2L effect size was NAPOLI-1 HR 0.67). Full manuscript expected in 2026 H2.
• AMPLIFY-201 (ELI-002 KRAS vaccine) [PMID 38195752] (Pant 2024 Nat Medicine, N=25 phase I): adjuvant-stage mKRAS mRNA lymph-node-targeted vaccine. 84% of patients produced mKRAS-specific T cells; among ctDNA-positive patients, 21/25 showed immune-response-related ctDNA decrease or clearance. Phase II AMPLIFY-7P ongoing. Opening move in KRAS vaccine adjuvant therapy.

Takeaway: in 2026 KRAS has shifted from PDAC’s “undruggable” barrier to three subtypes, three paths — G12C (adagrasib / sotorasib, 1-2% of patients, 2L+); G12D (RMC-9805, ~40% of patients, phase I); pan-KRAS G12X (RMC-6236 → RASolute-302, covering 85% of KRAS, phase III near approval). In the adjuvant setting, ELI-002 vaccine opens a second front. This is the first time in 30 years of PDAC that a driver-gene-level paradigm breakthrough has appeared.

3. Horizontal axis: the 2026 decision landscape (six dimensions)

Projecting the longitudinal evolution onto the 2026 clinical decision tree, here are six key branchpoints and the evidence base for each.

3.1 Newly diagnosed PDAC: do comprehensive molecular profiling immediately

NCCN V1.2026 explicitly recommends all newly diagnosed PDAC undergo comprehensive molecular testing (tissue or ctDNA), covering: KRAS mutation subtype (G12D / G12V / G12C / G12R / G13D / pan-KRAS status) + BRCA1/2 + PALB2 + MSI/dMMR + NRG1 / NTRK / HER2 / BRAF / ALK / ROS1. Molecular results directly influence: 1L regimen selection (BRCA-mutant patients can get olaparib maintenance after platinum chemo); 2L targeted accessibility (NRG1 → zenocutuzumab / NTRK → larotrectinib, entrectinib / G12C → adagrasib / G12X → RMC-6236 if RASolute approved); clinical trial enrollment (RASolute, AMPLIFY-7P and other KRAS-direction trials).

3.2 Advanced 1L: four-way decision by performance status

• ECOG 0-1 and age <75: first choice FOLFIRINOX [PRODIGE-4 PMID 21561347] or NALIRIFOX [NAPOLI-3 PMID 37708904] (the latter FDA-approved 2024-02, accessibility and reimbursement still rolling out)
• ECOG 2 or heavy comorbidity: GnP [MPACT PMID 24131140]
• gBRCA1/2+: 1L platinum FOLFIRINOX/NALIRIFOX ≥16 weeks without progression → olaparib maintenance [POLO PMID 31157963]
• Frail / ECOG 3 cannot tolerate combination: gemcitabine monotherapy or S-1 monotherapy [GEST PMID 23547081]

Note: dual checkpoint (durvalumab + tremelimumab) ineffective in unselected 1L (CCTG PA.7 PMID 36028483) — not a PDAC 1L consideration.

3.3 Adjuvant regimens: mFFX vs GnP vs S-1 vs GemCap vs gemcitabine

• ECOG 0-1, age <75, triplet-tolerant: mFOLFIRINOX (PRODIGE-24 PMID 30575490, mOS 54.4 months, best 30-year data)
• Western elderly or intermediate performance: GemCap (ESPAC-4 PMID 28129987, mOS 28.0 months)
• Asian population (strongest Japanese evidence): S-1 (JASPAC-01 PMID 27275872, Asian-specific tolerability)
• Combination-intolerant: gemcitabine monotherapy (CONKO-001 PMID 23982521)
• Not recommended for adjuvant: nab-paclitaxel + gemcitabine (APACT primary endpoint negative, FDA did not approve the indication)

3.4 Borderline resectable / resectable: give neoadjuvant or not, give what

Decision prerequisites: surgical and imaging staging assessment (resectable / borderline resectable / locally advanced — three levels) + CA19-9 + ECOG status.

• Resectable pancreatic head cancer: upfront surgery + post-op mFFX adjuvant is better than neoadjuvant FOLFIRINOX (NORPACT-1 PMID 38237621 warns neoadjuvant may harm resectable pancreatic head outcomes)
• Borderline resectable: neoadjuvant chemo (FOLFIRINOX or GemCap) beats upfront surgery (ESPAC-5F / PREOPANC-1 / PREOPANC-2 consistently support); the two paths (neoadjuvant mFFX vs neoadjuvant gem-chemoRT) are equivalent (PREOPANC-2 PMID 39500336)
• Locally advanced (unresectable): after 4-6 months of induction chemo, responders or stable patients choose chemoradiation (for locoregional control) or continued chemo (LAP07 PMID 27139057 shows no OS difference); SBRT boost neoadjuvant not recommended (Alliance A021501 PMID 35834226)
• No neoadjuvant regimen recommends adding nelfinavir or other sensitizers (SCALOP-2 PMID 34048677)

3.5 Advanced 2L: three effective regimens, two East-West differences, KRAS stratification joins

• Standard 2L: nal-IRI + 5FU/LV (NAPOLI-1 PMID 26615328) — only globally approved 2L regimen
• Europe option: OFF regimen (CONKO-003 PMID 24982456, European evidence base)
• West does not recommend FOLFOX: PANCREOX [PMID 27621395] shows FOLFOX 2L is worse in Western cohorts
• KRAS G12C: adagrasib (KRYSTAL-1 PMID 37099736) or sotorasib (CodeBreaK 100 PMID 36546651), ORR 21-33%
• KRAS G12X (including G12D) awaiting RASolute-302 approval: if approved, may become 2L first choice for KRAS-mutant PDAC (HR 0.40 effect size unprecedented)
• BRCA1/2 PARP-naive: if no 1L maintenance was given, PARP can be considered in 2L
• NRG1 fusion: zenocutuzumab (eNRGy PMID 39908431)
• NTRK fusion: larotrectinib / entrectinib / repotrectinib
• MSI-H/dMMR: pembrolizumab (KEYNOTE-158 PMID 31682550)

3.6 KRAS G12D subgroup (~40% of KRAS PDAC): the 2026 special waiting posture

PDAC KRAS mutation distribution is roughly: G12D 40% / G12V 30% / G12R 15% / G12C 1-2% / G13D 2-3% / other 10%. G12D is the largest subtype by number, but as of 2026-04:

• MRTX1133 (Mirati/BMS G12D-selective) development terminated
• RMC-9805 (zoldonrasib, Revolution Medicines G12D-selective) is only in phase I, with 2024 EORTC late-breaking-level data
• RMC-6236 (pan-KRAS G12X) covers G12D; RASolute-302 phase III topline released but manuscript pending

G12D PDAC patient options in 2026: clinical trial enrollment (RASolute-302 / AMPLIFY-7P / RMC-9805 expansion cohorts) > standard chemo (FOLFIRINOX/GnP → NAPOLI-1). “Await RASolute approval + consider KRAS-directed trials” is the default posture for G12D PDAC in mid-to-late 2026.

4. Research Gaps: ten unresolved clinical questions

This report identifies the following gaps, all definable concrete questions (not the “need more research” cliché):

1. Resectable PDAC neoadjuvant vs adjuvant head-to-head: NORPACT-1 suggests neoadjuvant FFX may harm resectable pancreatic head cancer outcomes; Alliance A021806 is the first systematic head-to-head phase III; no verdict before its 2027-2028 readout.
2. RMC-6236 / RASolute-302 resistance mechanisms: after pan-KRAS inhibitor monotherapy mPFS of ~8 months, is resistance driven by secondary KRAS mutations / bypass activation / EMT phenotype / or mixed? Paired ctDNA resistance-typing studies are missing.
3. NALIRIFOX vs FOLFIRINOX 1L head-to-head: both contain oxaliplatin+5-FU+leucovorin; core difference is liposomal irinotecan vs standard irinotecan. Direct randomized comparison missing; clinical use relies on cross-trial inference.
4. Root cause of POLO’s negative OS: is the PFS HR 0.53 / OS HR 0.83 mismatch driven by control-arm crossover? Does BRCA+ PDAC PARP maintenance need finer biomarker stratification?
5. PDAC MSI-H subgroup (1-2%) checkpoint monotherapy vs combination: KEYNOTE-158 PDAC cohort ORR 18% is lower than other MSI tumor types; reasons (microenvironment / HLA loss / low TMB) not yet analyzed.
6. KRAS G12D-selective vs pan-KRAS long-term comparison: G12D-selective (RMC-9805) may have lower on-target toxicity, but pan-KRAS (RMC-6236) may handle subclonal KRAS heterogeneity better. Long-term follow-up comparison missing.
7. KRAS vaccine adjuvant vs ctDNA-guided extended chemo: AMPLIFY-201 phase I showed immune response, but the adjuvant comparator should not be only placebo (adjuvant PDAC already has mFFX) — the right phase III is mFFX + vaccine vs mFFX + placebo.
8. East-West results divergence — OFF Western vs FOLFOX Eastern: CONKO-003 OFF and PANCREOX FOLFOX gave opposite results with similar drug combinations; Euro-Asian population and dosing-detail differences not systematically explained.
9. Role of ctDNA in PDAC monitoring and treatment decisions: should ctDNA-positive adjuvant patients get intensified / extended chemo? Prospective ctDNA-guided randomized trials missing.
10. Revival path for stroma-targeting agents: PEGPH20 / simtuzumab and other stroma-targeting approaches have all failed in PDAC; but can KRAS inhibition + stromal remodeling combinations (e.g., RMC-6236 + AT9283 or similar FAK/CAF combos) break the stromal barrier? Not systematically studied.

5. 2024-2026 latest developments

5.1 New FDA / NMPA approvals (key selections)

• 2024-02 FDA: NALIRIFOX (liposomal irinotecan + oxaliplatin + leucovorin + 5-FU) approved for 1L metastatic PDAC (NAPOLI-3 [PMID 37708904])
• 2024-12 FDA accelerated approval: zenocutuzumab (NRG1-fusion advanced PDAC and NSCLC, eNRGy [PMID 39908431]) — first non-chemo non-PARP biomarker-matched targeted approval in PDAC
• 2024 FDA Breakthrough Designation: RMC-6236 (daraxonrasib, pan-KRAS G12X, Revolution Medicines) + RMC-9805 (zoldonrasib, KRAS G12D)
• 2023 FDA accelerated approval: repotrectinib (tumor-agnostic NTRK fusion, including PDAC; formally approved as NTRK-universal)

5.2 Key conference readouts (2024-2026, weighted down)

• ASCO GI 2024 / ESMO 2024: RMC-6236 PDAC 2L phase I cohort data (Arbour / Ko et al. reports), ORR ~36%, mPFS ~8.5 months, mOS ~14.5 months
• EORTC 2024 (late-breaking): RMC-9805 zoldonrasib KRAS G12D phase I, PDAC 2L+ cohort ORR ~30%
• ASCO GI 2026: RASolute-302 phase III topline — RMC-6236 vs investigator’s choice in 2L+ KRAS G12X PDAC, mOS 13.2 vs 6.7 months, HR 0.40, p<0.001. Full manuscript expected NEJM/Lancet 2026 H2.
• Nat Medicine 2024 (ELI-002): AMPLIFY-201 KRAS mRNA vaccine adjuvant phase I, 84% immune response [PMID 38195752]

5.3 Phase III in progress (2025-2028 readout selections)

• RASolute-302 (NCT06625320): RMC-6236 vs SoC in KRAS G12X 2L+ PDAC — topline available, awaiting full manuscript
• Alliance A021806 (NCT04340141): resectable PDAC neoadjuvant vs adjuvant mFOLFIRINOX head-to-head — primary completion expected 2027-2028
• AMPLIFY-7P (NCT05726864, ELI-002 phase II adjuvant mKRAS vaccine) — readout expected 2027
• PANOVA-3 (tumor treating fields + GnP vs GnP), PROOF 301 (olaratumab + FFX), NOVATE (ivonescimab series in PDAC) and several other ongoing trials, none mainstream breakthrough candidates

6. Intersectional insights and judgments

6.1 Longitudinal × horizontal: 2026 PDAC landscape shaped by two “resonances”

Stacking longitudinal paradigm evolution and horizontal current decision landscape, the 2026 PDAC landscape is two resonances superposed:

1. Chemo marginal optimization (2011-2024 decade) superposed with KRAS cracking (2023-2026): chemo went from PRODIGE-4 / MPACT’s mOS 8-11 months, to NAPOLI-3’s 11.1 months (marginal +2 months), then to RASolute-302 2L’s mOS 13.2 months (double the historical 2L 6 months). A two-stage evolution of chemo marginal + KRAS doubling shapes the 2026 landscape.
2. Biomarker-matched sparse hits (BRCA / MSI / NRG1 / NTRK totaling <10%) superposed with KRAS trunk (90%): the former are like “scattered lights”; the latter is like “turning on the main breaker.” KRAS’s importance is not “another biomarker,” but “finally can cover 90% of patients” — this is the NSCLC EGFR story arriving late in PDAC.

These two resonances together explain a clinical phenomenon: the 1L decision tree for a newly diagnosed stage IV PDAC patient in 2026 has 3 more decision layers than in 2016 (molecular profiling → KRAS subtype specificity → clinical trial enrollment priority → biomarker-matched accessibility).

6.2 Clinical decision takeaways (for junior-mid oncologists)

1. Routine molecular profiling: in 2026 all newly diagnosed PDAC should get comprehensive molecular profiling. Missing KRAS subtype means missing the G12X treatment path after RASolute approval; missing BRCA means missing POLO maintenance; missing NRG1 / NTRK means missing already-approved targeted options.
2. ECOG 0-1 and age <75: 1L first choice FOLFIRINOX or NALIRIFOX: don’t default to GnP out of “toxicity concern.”
3. Neoadjuvant is not universal: be cautious giving FOLFIRINOX neoadjuvant in resectable pancreatic head cancer (NORPACT-1 warning); borderline resectable is where to consider neoadjuvant.
4. Adjuvant first choice mFFX (PRODIGE-24): APACT lesson — what wins in the advanced setting doesn’t necessarily win as adjuvant. Nab-paclitaxel + gemcitabine not recommended for adjuvant.
5. 2L nal-IRI + 5FU/LV is default: don’t use FOLFOX (PANCREOX warning in Western populations).
6. KRAS G12C 2L+ already has adagrasib / sotorasib: small subgroup but don’t miss it.
7. KRAS G12X (RMC-6236 / RASolute) will rewrite 2L+ standard in 2026-2027: watch the FDA approval timeline; simultaneously assess clinical trial enrollment feasibility (RASolute expansion / AMPLIFY-7P).
8. Don’t blindly push dual checkpoint blockade in PDAC: CCTG PA.7 rejected dual checkpoint + chemo in unselected populations. MSI-H PDAC is where pembrolizumab applies.
9. Locally advanced SBRT neoadjuvant adds nothing, sensitizers add nothing: Alliance A021501 / SCALOP-2, two lines of evidence pointing the same way.
10. nal-IRI, zenocutuzumab, KRAS G12C inhibitors, RMC-6236, ELI-002 — five drug classes clinicians must know: the 30-year PDAC drug-drought era is over; the 2024-2026 cluster of new approvals marks a new phase.

7. Information sources

All 37 trial metadata in this report were independently verified via PubMed and ClinicalTrials.gov. Each [PMID xxxxxxxx] in the text can be verified directly in PubMed.

• Published trials: 33, covering 2004-2025 (PMID-verifiable)
• Ongoing / not yet formally published: 4 (RMC-6236-001 / RMC-9805-001 phase I conference data only · TRIDENT-1-PANC manuscript pending · RASolute-302 phase III topline at ASCO GI 2026, full manuscript expected 2026 H2)
• NCCN guideline citations: 30/37 (81%) directly hit the NCCN V1.2026 PANC-G reference section; the remaining 7 are expected 0-hit (phase I/III ongoing, or papers published after 2024 not yet incorporated into V1.2026)
• 2024-2026 FDA new approvals: 3 (NALIRIFOX / zenocutuzumab / two Breakthrough Designations)
• 2024-2026 key conference readouts: 4 (RMC-6236 ASCO GI/ESMO 2024 / RMC-9805 EORTC 2024 / RASolute-302 ASCO GI 2026 / AMPLIFY-201 Nat Med 2024)
• Research gaps: 10

7.1 Citation index (sorted by PMID)

7.2 Verification conventions

• PMIDs can be directly concatenated into PubMed URLs: https://pubmed.ncbi.nlm.nih.gov/<PMID>
• NCT IDs can be directly concatenated into ClinicalTrials.gov URLs: https://clinicaltrials.gov/study/<NCT_ID>
• Conference abstracts (ASCO GI / ESMO / EORTC) are retrievable via the official conference systems; all conference citations in this report are “weighted down” — not peer-reviewed, final data per journal publication
• Phase I/III ongoing trials’ ORR / PFS / OS values are based on the latest public conference reports; minor adjustments possible upon manuscript publication

The clinical trial timeline is here

Chinese: /trials/pancreatic/ English: /en/trials/pancreatic/

Every trial has its own detail page with:

• Full intervention / comparator regimen
• Primary endpoint values + 95% CI
• Key findings + clinical significance
• Clickable jumps to PMID / NCT originals

37 trials · 4 chapters · 1994 to 2026 · 15+ country contributions · synced with NCCN V1.2026.

Closing

PDAC has been oncology’s “slowest-changing” field for 30 years — from the 1997 gemcitabine adjuvant foundation to mFFX’s DFS HR 0.58 in 2018, fit-patient adjuvant mOS only moved from 35 to 54 months. Advanced 1L went from PRODIGE-4’s 11 months to being marginally rewritten twelve years later by NAPOLI-3’s 11.1 months.

But 2023-2026, these three years, marked the first driver-gene-level paradigm breakthrough in 30 years: KRAS G12C → KRAS G12D → pan-KRAS G12X, three subtypes with three drug paths advancing; April 2026 RASolute-302 topline’s HR 0.40 is an effect size unseen in 30 years of PDAC 2L. Biomarker-matched sparse four tiles (BRCA / MSI-H / NRG1 / NTRK) + the KRAS trunk powering up, PDAC precision therapy is beginning to approach NSCLC’s 2012 landscape — still some distance from closing the “EGFR 10-year revolution” loop, but the door has been pushed open a crack.

This report’s value is not “exhaustively listing all trials” (PubMed can do that), but compressing 30 years of evolution + current decisions + unresolved gaps into one reading’s cognitive bandwidth. Next time you face a newly diagnosed PDAC patient, every branch in the decision tree has this map to consult, trace, and question.

Clinician × AI = Research Superpower + Clinical Decision Amplifier

—— Dual Brain Lab · 2026-04-20