Skip to content
Bending the Curve Trial by Trial
?

Pancreas (digestive system; ductal adenocarcinoma of the exocrine pancreas)

Pancreatic cancer

How surgery, chemotherapy, tumour genetics and RAS-targeted therapy shaped care of pancreatic ductal adenocarcinoma.

5-year relative survival · United States, SEER 8 registries

3.18%→14.44%

1975 (observed value; first year in the table) to 2018 (observed value; last observed year in the table). Read these with the caveats under “Then and now”.Source

25 milestones, by decade

  • In 1988 a laboratory study of 22 carcinomas of the exocrine pancreas found c-K-ras mutations at codon 12 in 21 of them, and its authors concluded that activation of this gene is a critical event in most, if not all, such cancers. In 2026 a phase 3 trial (RASolute 302; 500 people with previously treated metastatic pancreatic ductal adenocarcinoma) reported median overall survival of 13.2 months with daraxonrasib, an oral inhibitor of mutant and wild-type RAS, against 6.6 months with investigator's-choice chemotherapy in the RAS G12 population (hazard ratio 0.40), and the FDA approved the drug on 26 August 2026. The two reports are not comparable in design: the 1988 paper is a small tumour survey, and the 2026 trial was open-label.
  • In the SEER 21 registries (2016-2022), 51% of people with pancreatic cancer were diagnosed with distant-stage disease, with a five-year relative survival of 3.4%, and 15% with localized disease, with 43.6%. Over a longer period the SEER 8 series of observed five-year relative survival rises from 3.18% (1975) to 14.44% (2018), while the US age-adjusted death rate was 10.64 per 100,000 in 1975 and 11.31 in 2024 and the page describes death rates as stable over 2015-2024. The SEER 8 and SEER 21 series use different registries and are not chained here.
  • In the 1997 report of a randomised trial of gemcitabine against fluorouracil in 126 people with advanced symptomatic pancreatic cancer, the primary efficacy measure was clinical benefit response, a composite of pain, Karnofsky performance status and weight, not survival. Clinical benefit response was experienced by 23.8% with gemcitabine and 4.8% with fluorouracil (P = .0022), and median survival was 5.65 and 4.41 months (P = .0025); the authors described the survival advantage as modest.

Evidence current to October 2026 · Checked against primary sources to the extent possible

A guided tour of 12 milestones. You can pause or step through it yourself.

Filter milestonesshowing all 25
Treatment type
Setting

The age of rays

  1. 1935 · Treatment · SurgeryWhipple 1935: report on treating ampullary carcinomaIn 1935 the surgeon Allen Whipple and two colleagues published a report on treating cancer of the ampulla of Vater, where the bile and pancreatic ducts reach the bowel. A 2011 history describes it as covering three patients treated in two stages, and names Whipple, Halsted and Kausch as pioneers of pancreaticoduodenectomy.Whipple, Parsons and Mullins, Ann Surg 1935: "Treatment of carcinoma of the ampulla of Vater". PubMed lists no abstract, and no technique or outcome is taken from the paper itself. Are 2011 (HPB, historical review) describes it as a report of three patients managed with a two-stage radical resection of the duodenum and head of the pancreas, and names Halsted, Kausch and Whipple as pioneers of the operation.

Molecular origins

  1. 1982 · Understanding · UnderstandingMonoclonal antibodies detect an antigen in pancreatic cancer seraIn 1982 researchers reported that monoclonal antibodies, including one named 19-9, bind colorectal, gastric and pancreatic cancer cells. In a serum assay with another, 52a, 45 of 49 samples from people with pancreatic cancer were positive versus 2 of 108 healthy donors. A 1990 review called CA 19-9 the most useful blood test in pancreatic cancer.Herlyn 1982 (J Clin Immunol): three hybridoma antibodies, including 19-9 and 52a, bound colorectal, gastric and pancreatic carcinoma cells in culture. In a serum inhibition assay with antibody 52a, binding fell by more than 12% with 45 of 49 pancreatic carcinoma sera versus 2 of 108 healthy donors. Steinberg 1990 (review abstract): at 37 U/ml, sensitivity about 80%, specificity 90%.
  2. 1988 · Understanding · Understandingc-K-ras codon 12 mutations in 21 of 22 pancreatic carcinomasIn 1988 a laboratory study tested 22 carcinomas of the exocrine pancreas and found a change in the c-K-ras gene, at codon 12, in 21 of them. Where metastases were also tested (seven cases), they carried it too. Normal tissue did not. The authors judged it critical in most, if not all, such cancers.Almoguera 1988 (Cell): PCR amplification with RNase A mismatch cleavage, on frozen and formalin-fixed paraffin-embedded specimens. 21 of 22 exocrine pancreatic carcinomas had c-K-ras codon 12 mutations; in 7 cases tested, both primary tumour and metastasis carried it. None in normal tissue from the same patients or in 5 gallbladder carcinomas. The authors concluded somatic activation is critical in most, if not all, such cancers.
  3. 1996 · Treatment · ChemotherapyGemcitabine versus fluorouracil in advanced pancreatic cancerIn 1996 the US regulator approved gemcitabine for advanced pancreatic cancer. A trial of 126 people with symptomatic disease compared it with fluorouracil. About 24 in 100 people on gemcitabine and 5 in 100 on fluorouracil gained symptom relief, and median survival was about 5.7 versus 4.4 months.Randomised trial, 126 patients with advanced symptomatic pancreas cancer: gemcitabine (63) versus fluorouracil (63). Primary efficacy measure, clinical benefit response: 23.8% vs 4.8% (P=.0022). Median survival 5.65 vs 4.41 months (P=.0025); 12-month survival 18% vs 2%. FDA approved gemcitabine on 15 May 1996 for first-line treatment of advanced or metastatic pancreatic adenocarcinoma.

Targeted therapy

  1. 2001 · Treatment · ChemotherapyESPAC-1: adjuvant chemotherapy and chemoradiotherapy comparedIn ESPAC-1, people whose pancreatic cancer had been removed were randomly assigned to chemotherapy, chemoradiotherapy, both or observation, or to one of two simpler comparisons. The first report found no gain from radiotherapy and a possible gain from chemotherapy: median survival about twenty months with chemotherapy against fourteen without. The final report described chemoradiotherapy as deleterious.ESPAC-1 (Neoptolemos 2001): 541 eligible patients with resected PDAC, factorial or single-comparison randomisation. Median survival 19.7 vs 14.0 months with vs without chemotherapy (HR 0.66, 95% CI 0.52-0.83, p=0.0005) and 15.5 vs 16.1 months with vs without chemoradiotherapy (HR 1.18, 0.90-1.55, p=0.24). Final 2004 report: 5-year survival 21% vs 8% (chemotherapy) and 10% vs 20% (chemoradiotherapy).
  2. 2007 · Treatment · ChemotherapyCONKO-001: adjuvant gemcitabine versus observationIn CONKO-001, people whose pancreatic cancer had been removed received six months of gemcitabine or observation only. Cancer returned later with gemcitabine: median disease-free survival 13 months against 7 months. A 2013 follow-up reported longer overall survival too: about 21 in 100 alive at five years against 10 in 100.CONKO-001 (Oettle 2007, 2013): 368 randomised, 354 eligible; adjuvant gemcitabine for 6 months vs observation after R0/R1 resection. Median DFS 13.4 vs 6.9 months (P<.001); OS not different in 2007 (22.1 vs 20.2 months, P=.06). At 136 months median follow-up: DFS HR 0.55 (0.44-0.69); OS HR 0.76 (95% CI 0.61-0.95, P=.01); 5-year OS 20.7% vs 10.4%.
  3. 2008 · Understanding · UnderstandingSequencing reveals core pathways and subtypes of pancreatic cancerIn 2008 a team sequenced 24 pancreatic cancers and found about 63 gene changes per tumour on average, grouped into 12 core pathways altered in most tumours. In 2016 an analysis of 456 tumours described 10 pathways and four subtypes based on gene activity. Neither report describes a treatment result.Jones 2008 (Science): 23,219 transcripts sequenced in 24 pancreatic cancers plus copy-number arrays; average 63 genetic alterations, mostly point mutations, defining 12 core pathways each altered in 67 to 100% of tumours. Bailey 2016 (Nature): 456 PDACs, 32 recurrently mutated genes in 10 pathways; four expression subtypes (squamous, pancreatic progenitor, immunogenic, ADEX); squamous described as poor prognosis.
  4. 2010 · Reversal · TargetedBevacizumab, cetuximab, PEGPH20 added to chemotherapy: no gainThree large trials added a targeted drug to chemotherapy in advanced pancreatic cancer: bevacizumab, cetuximab and PEGPH20. None lengthened survival. Median survival with bevacizumab was about 5.8 months versus 5.9 with placebo, and in the PEGPH20 trial about 11.2 versus 11.5 months.Three phase 3 trials, each with overall survival as primary endpoint, were negative. CALGB 80303 (gemcitabine +/- bevacizumab, 602 enrolled): median OS 5.8 vs 5.9 months (P=.95). S0205 (gemcitabine +/- cetuximab, 745 eligible): 6.3 vs 5.9 months, HR 1.06 (0.91-1.23). HALO 109-301 (nab-paclitaxel-gemcitabine +/- PEGPH20, 492 ITT): 11.2 vs 11.5 months, HR 1.00 (0.80-1.27).

Immunotherapy

  1. 2011 · Treatment · ChemotherapyFOLFIRINOX versus gemcitabine: PRODIGE 4 / ACCORD 11In a French trial of 342 people with spread pancreatic cancer who were well enough to be up and about, a four-drug regimen called FOLFIRINOX was compared with gemcitabine. Median survival was about 11 months with FOLFIRINOX and about 7 months with gemcitabine, with more side effects.PRODIGE 4 / ACCORD 11, phase 2/3: 342 patients with metastatic pancreatic cancer, ECOG 0-1, FOLFIRINOX vs gemcitabine. Median OS 11.1 vs 6.8 months, HR 0.57 (95% CI 0.45-0.73; P<0.001). Median PFS 6.4 vs 3.3 months, HR 0.47 (0.37-0.59). ORR 31.6% vs 9.4%. Febrile neutropenia 5.4% with FOLFIRINOX. Authors: an option for patients with good performance status.
  2. 2012 · Detection · Detection & preventionSurveillance of high-risk individuals: CAPS studies and consensusFrom 2012, studies of people at high inherited risk of pancreatic cancer tested imaging surveillance. Of 216 evaluable people screened once, about four in ten had a pancreatic abnormality, mostly small cysts. Longer follow-up found cancers or high-grade precancerous changes in a minority. A consensus statement advised surveillance for selected people in research settings.Canto 2012: 225 asymptomatic high-risk individuals enrolled, 216 evaluable, screened once by CT, MRI and EUS; 92 of 216 (42%) had a mass or dilated duct (84 cystic, 3 solid). Canto 2018 (n=354, median follow-up 5.6 years): neoplastic progression in 24 (7%), 14 PDAC and 10 high-grade dysplasia. Overbeek 2022 (n=366): 10 PDAC. Goggins 2020: Delphi consensus on 55 statements.
  3. 2013 · Treatment · ChemotherapyMPACT: nab-paclitaxel plus gemcitabine improves survivalIn 861 people with spread pancreatic cancer, adding nab-paclitaxel to gemcitabine lengthened median survival from about 6.7 to about 8.5 months. Nerve damage and low blood counts were more common. The US regulator approved the combination in September 2013.MPACT, phase 3: 861 patients with metastatic pancreatic cancer, KPS 70 or more; nab-paclitaxel plus gemcitabine vs gemcitabine. Median OS 8.5 vs 6.7 months, HR 0.72 (95% CI 0.62-0.83; P<0.001); 1-year survival 35% vs 22%. PFS 5.5 vs 3.7 months (HR 0.69). Grade 3 or higher neuropathy 17% vs 1%. Post hoc update: median OS 8.7 vs 6.6 months.
  4. 2017 · Treatment · ChemotherapyESPAC-4: gemcitabine plus capecitabine as adjuvant therapyESPAC-4 compared gemcitabine alone with gemcitabine plus the tablet capecitabine after pancreatic cancer surgery. Median survival was 28 months with the combination and 25.5 months with gemcitabine alone, with more severe side effects reported with the combination. The authors proposed the combination as the new standard of care.ESPAC-4 (Neoptolemos 2017): open-label phase 3, 730 of 732 enrolled patients analysed after R0/R1 resection of PDAC (366 gemcitabine, 364 gemcitabine plus capecitabine). Primary endpoint OS, ITT: median 28.0 months (95% CI 23.5-31.5) vs 25.5 months (22.7-27.9); HR 0.82 (95% CI 0.68-0.98), p=0.032. Reported at 458 of 480 target deaths at the data monitoring committee's request.
  5. 2017 · Treatment · ImmunotherapyMSI-H/dMMR tumours: pancreatic cases within tissue-agnostic evidenceIn 2017 the US regulator approved pembrolizumab for advanced solid tumours with a mismatch-repair defect, whatever the organ. Pancreatic cancer was among the most common types in a later 233-patient study, in which about one in three patients had a response. No pancreatic-only figure is given here.Tissue-agnostic approval (FDA accelerated approval, 23 May 2017) for MSI-H/dMMR solid tumours. KEYNOTE-158 noncolorectal cohort (phase 2, n=233, 27 tumour types; endometrial, gastric, cholangiocarcinoma and pancreatic most common): ORR 34.3% (95% CI 28.3-40.8), median PFS 4.1 months (2.4-4.9), median OS 23.5 months. Pancreatic-specific results are not in the abstract.
  6. 2017 · Understanding · UnderstandingBorderline resectable disease gets an international definitionA 2016 international symposium in Sendai proposed one shared definition of borderline resectable pancreatic cancer. It looks at three things: how the tumour touches blood vessels, signs that cancer may have spread, and the person's general fitness. Meeting any one of the three qualifies. The aim was to replace a definition that changed every year.Isaji 2018 (IAP consensus, Sendai 2016): BR-PDAC requires one or more of three dimensions. Anatomical: for example superior mesenteric or celiac artery contact under 180 degrees without stenosis or deformity. Biological: suspected but unproven metastases, nodal metastases diagnosed by biopsy or PET-CT, or CA 19-9 above 500 units/ml. Conditional: ECOG 2 or more. The NCCN adopted a definition in 2006 that changed yearly.
  7. 2018 · Treatment · ChemotherapyPRODIGE 24: adjuvant modified FOLFIRINOX versus gemcitabinePRODIGE 24 compared a four-drug chemotherapy, modified FOLFIRINOX, with gemcitabine after pancreatic cancer surgery. Median survival was about 54 months with the four-drug regimen and 35 months with gemcitabine. Severe side effects were more common with the four-drug regimen, in about three of four people against about half.PRODIGE 24 / CCTG PA6 (Conroy 2018): 493 patients with resected PDAC randomised to modified FOLFIRINOX or gemcitabine for 24 weeks. Primary endpoint DFS: median 21.6 vs 12.8 months (stratified HR 0.58, 95% CI 0.46-0.73). Median OS 54.4 vs 35.0 months (HR 0.64, 95% CI 0.48-0.86, P=0.003). Grade 3 or 4 adverse events 75.9% vs 52.9%. At 69.7 months, OS HR 0.68 (0.54-0.85).
  8. 2019 · Treatment · TargetedPOLO: maintenance olaparib for germline BRCA-mutated diseaseIn 154 people with spread pancreatic cancer and an inherited BRCA change whose disease had not worsened on platinum chemotherapy, the pill olaparib held the cancer back for a median of about 7 months versus about 4 months with placebo. The final analysis found no clear survival difference.POLO, double-blind phase 3: 154 patients with germline BRCA1/2-mutated metastatic pancreatic cancer, no progression on first-line platinum; maintenance olaparib (92) vs placebo (62). Median PFS (BICR) 7.4 vs 3.8 months, HR 0.53 (95% CI 0.35-0.82; P=0.004). Final OS 19.0 vs 19.2 months, HR 0.83 (0.56-1.22; P=.3487); estimated 3-year survival 33.9% vs 17.8%.

Precision & engineered

  1. 2020 · Treatment · ChemotherapyPREOPANC: preoperative chemoradiotherapy versus immediate surgeryPREOPANC randomised people with resectable or borderline resectable pancreatic cancer to chemoradiotherapy before surgery or surgery first. The 2020 report found no clear survival gain: median 16.0 against 14.3 months. A 2022 analysis of the same trial with longer follow-up reported better survival, with about 20 in 100 alive at five years against 6.5 in 100.PREOPANC (Versteijne 2020, 2022): 246 eligible patients, resectable or borderline resectable PDAC, preoperative gemcitabine chemoradiotherapy then surgery vs immediate surgery, both with adjuvant gemcitabine. 2020 (primary, ITT): median OS 16.0 vs 14.3 months, HR 0.78 (95% CI 0.58-1.05, P=.096); R0 resection 71% vs 40%. 2022 (59 months): HR 0.73 (0.56-0.96, P=.025); 5-year OS 20.5% vs 6.5%.
  2. 2022 · Reversal · ChemotherapyAPACT: adjuvant nab-paclitaxel plus gemcitabine, DFS not metAPACT compared nab-paclitaxel plus gemcitabine with gemcitabine alone after pancreatic cancer surgery. The main measure, disease-free survival judged by independent reviewers, did not differ: 19.4 against 18.8 months. Median survival was 41.8 against 37.7 months at five years of follow-up, which the authors treat as descriptive because the main measure was not met.APACT (Tempero 2023): 866 patients with resected PDAC, nab-paclitaxel plus gemcitabine vs gemcitabine, six cycles. Primary endpoint independently assessed DFS not met: 19.4 vs 18.8 months (HR 0.88, 95% CI 0.729-1.063, P=.18). Investigator-assessed DFS 16.6 vs 13.7 months (HR 0.82, P=.02). OS at 5-year follow-up 41.8 vs 37.7 months (HR 0.80, 95% CI 0.678-0.947). Grade 3 or worse events 86% vs 68%.
  3. 2022 · Treatment · TargetedKRAS G12C inhibitors in the small G12C-mutated subsetAbout one to two in every hundred pancreatic cancers carry a KRAS G12C change. In small single-group studies, 8 of 38 treated people responded to sotorasib, and 7 of 21 pancreatic patients to adagrasib. The US labels in the openFDA database in October 2026 list no pancreatic cancer indication for either drug.Single-group studies, previously treated KRAS G12C-mutated pancreatic cancer. Sotorasib (CodeBreaK 100, phase 1-2, pooled n=38): centrally confirmed ORR 21% (95% CI 10-37), median PFS 4.0 months (2.8-5.6), median OS 6.9 months (5.0-9.1). Adagrasib (KRYSTAL-1, mixed-tumour cohort): ORR 7/21 (33.3%) in pancreatic cancer. G12C occurs in about 1 to 2% of pancreatic cancers.
  4. 2023 · Treatment · ChemotherapyNAPOLI-3: NALIRIFOX versus nab-paclitaxel plus gemcitabineIn 770 people with untreated spread pancreatic cancer, a four-drug regimen called NALIRIFOX was compared with nab-paclitaxel plus gemcitabine. Median survival was about 11 months versus about 9 months. Severe (grade 3 or higher) side effects were about equally common. The US regulator approved it in February 2024.NAPOLI 3, open-label phase 3: 770 treatment-naive patients with metastatic PDAC, ECOG 0-1; NALIRIFOX (383) vs nab-paclitaxel-gemcitabine (387). Median OS 11.1 (95% CI 10.0-12.1) vs 9.2 months (8.3-10.6); HR 0.83 (95% CI 0.70-0.99; p=0.036). Grade 3 or higher treatment-emergent adverse events 87% vs 86%. FDA approval 13 February 2024.
  5. 2024 · Reversal · ChemotherapyNORPACT-1: neoadjuvant FOLFIRINOX versus upfront surgeryNORPACT-1, a randomised phase 2 trial in four Nordic countries, compared four cycles of FOLFIRINOX before surgery with surgery first in resectable pancreatic head cancer. About 60 in 100 were alive at 18 months with FOLFIRINOX first against 73 in 100 with surgery first. The authors did not find a survival benefit from the neoadjuvant approach.NORPACT-1 (Labori 2024): randomised phase 2, 140 patients with resectable pancreatic head cancer (77 neoadjuvant FOLFIRINOX, 63 upfront surgery), adjuvant chemotherapy in both. Primary endpoint OS at 18 months, ITT: 60% (95% CI 49-71) vs 73% (62-84), p=0.032. Median OS 25.1 vs 38.5 months, HR 1.52 (95% CI 1.00-2.33), log-rank p=0.050. Resection 82% vs 89%. Grade 3+ events 58% vs 40%.
  6. 2024 · Treatment · TargetedZenocutuzumab: pancreatic-cancer indication for NRG1 gene fusionsNRG1 gene fusions are rare drivers of pancreatic cancer. In a single-group study, 15 of 36 pancreatic cancer patients with measurable disease responded to the antibody zenocutuzumab. The US regulator gave it accelerated approval in December 2024 for NRG1 fusion-positive pancreatic and lung cancer after prior therapy.eNRGy, single-arm phase 2 (204 treated, 12 tumour types, NRG1 fusion-positive): zenocutuzumab 750 mg every 2 weeks. Pancreatic cancer ORR 15/36 (42%; 95% CI 25-59); all tumours ORR 30% (95% CI 23-37; n=158), median DOR 11.1 months, median PFS 6.8 months. FDA accelerated approval 4 December 2024 (NSCLC and pancreatic adenocarcinoma after prior systemic therapy).
  7. 2025 · Treatment · TargetedPANOVA-3: tumour treating fields plus chemotherapy, locally advancedIn 571 people with newly diagnosed locally advanced pancreatic cancer that could not be removed, adding a wearable electric-field device to chemotherapy lengthened median survival from about 14 to about 16 months. Skin reactions from the device were common. The US regulator approved the device in February 2026.PANOVA-3, open-label phase 3: 571 patients with unresectable locally advanced pancreatic adenocarcinoma; TTFields plus gemcitabine/nab-paclitaxel vs chemotherapy alone. Median OS 16.2 (95% CI 15.0-18.0) vs 14.2 months (12.8-15.4); HR 0.82 (0.68-0.99; P=.039). PFS, local PFS and ORR not improved. Device-related skin adverse events 76.3% (grade 3: 7.7%). FDA PMA decision 11 February 2026.
  8. 2025 · Treatment · ChemotherapyPREOPANC-2: neoadjuvant FOLFIRINOX versus chemoradiotherapyPREOPANC-2 compared FOLFIRINOX before surgery with gemcitabine chemoradiotherapy before surgery in resectable or borderline resectable pancreatic cancer. Median survival was 21.9 against 21.3 months, with no clear difference. The authors say both approaches may be considered.PREOPANC-2 (Janssen 2025): open-label phase 3, 375 randomised (188 FOLFIRINOX, 187 chemoradiotherapy), 369 in the modified ITT population. FOLFIRINOX x8 then surgery without adjuvant therapy vs gemcitabine plus 36 Gy then surgery and adjuvant gemcitabine. Median OS 21.9 vs 21.3 months, HR 0.88 (95% CI 0.69-1.13), p=0.32. Grade 3 or worse adverse events 67% vs 60% (p=0.20).
  9. 2026 · Treatment · TargetedDaraxonrasib: RAS(ON) inhibitor extends survival after prior therapyIn 500 people with spread pancreatic cancer already treated once, the pill daraxonrasib was compared with chemotherapy. Median survival was about 13 months versus about 7 months. The US regulator approved it in August 2026, and a European committee recommended it in October 2026.RASolute 302, open-label phase 3: 500 patients with previously treated metastatic PDAC randomised to daraxonrasib (248) or investigator's-choice chemotherapy (252). RAS G12 population: median OS 13.2 vs 6.6 months, HR 0.40; PFS 7.3 vs 3.5 months, HR 0.45 (P<0.001). Overall population OS 13.2 vs 6.7 months, HR 0.40 (95% CI 0.30-0.53; FDA page). Grade 3 or higher adverse events 61.8% vs 69.6%.

Then and now

  • 5-year relative survival · United States, SEER 8 registries

    • 3.18%

      1975 (observed value; first year in the table)

    All races, both sexes; SEER Cancer Stat Facts category 'pancreatic cancer' (the page does not separate histologies, so ductal adenocarcinoma, other exocrine tumours and neuroendocrine tumours are not distinguished). SEER 8 five-year relative survival series, 1975-2018. Age range is not stated for this series.

    Modeled trend value for 1975: 1.84%. The observed series moves from year to year (for example 1.46% in 1977 and 3.40% in 1983). Relative survival excludes deaths from other causes. 'Observed' is the table's label for the actual yearly value, as opposed to the Joinpoint modeled-trend value in the adjacent column. The page headings and chart note ('SEER 8 5-Year Relative Survival Percent from 1975-2018') do not label the 'Year' column as year of diagnosis. The page does not describe changes over time in stage mix, imaging, case definition or treatment and does not apportion any change or attribute it to a cause; the page says only that stage at diagnosis has a strong influence on the length of survival, and it shows the US death rate as a separate series without linking the two, so read this beside the mortality values. Compare only with other values from this same SEER 8 series, not with the SEER 21 values.

    Source

  • 5-year relative survival · United States, SEER 8 registries

    • 14.44%

      2018 (observed value; last observed year in the table)

    All races, both sexes; SEER Cancer Stat Facts category 'pancreatic cancer' (the page does not separate histologies, so ductal adenocarcinoma, other exocrine tumours and neuroendocrine tumours are not distinguished). SEER 8 five-year relative survival series, 1975-2018. Age range is not stated for this series.

    Modeled trend value for 2018: 14.94%; the table shows no observed five-year value for 2019 onward, only modeled trend values (for example 18.82% for 2023), which are projections from the fitted trend and are not entered here. Relative survival excludes deaths from other causes. 'Observed' is the table's label for the actual yearly value, as opposed to the Joinpoint modeled-trend value in the adjacent column. The page headings and chart note ('SEER 8 5-Year Relative Survival Percent from 1975-2018') do not label the 'Year' column as year of diagnosis. The page does not describe changes over time in stage mix, imaging, case definition or treatment and does not apportion any change or attribute it to a cause; the page says only that stage at diagnosis has a strong influence on the length of survival, and it shows the US death rate as a separate series without linking the two, so read this beside the mortality values. Compare only with other values from this same SEER 8 series, not with the SEER 21 values.

    Source

  • 5-year relative survival · United States, SEER 21 (excluding IL)

    • 13.7%

      2016-2022

    Pancreatic cancer, SEER 21 (excluding IL), 2016-2022, as shown in the page headline. The page does not label the headline figure with race, sex or age range (the stage table beneath it is labelled All Races, Both Sexes). Histologies are not separated.

    A different registry set and period from the SEER 8 series above; do not chain or subtract them. Relative survival excludes deaths from other causes.

    Source

  • 5-year relative survival · United States, SEER 21 (excluding IL)

    • 43.6%

      2016-2022; localized stage at diagnosis (15% of cases)

    All races, both sexes; pancreatic cancer, SEER Combined Summary Stage 'localized' (confined to the primary site).

    The page says stage at diagnosis has a strong influence on the length of survival and gives no further interpretation of stage-specific values. SEER Combined Summary Stage, not AJCC TNM. Same SEER 21 series as the 13.7% figure above; do not chain it to the SEER 8 values or subtract. The page text gives the localized share as 15.4% and the table as 15%.

    Source

  • 5-year relative survival · United States, SEER 21 (excluding IL)

    • 17%

      2016-2022; regional stage at diagnosis (28% of cases)

    All races, both sexes; pancreatic cancer, SEER Combined Summary Stage 'regional' (spread to regional lymph nodes).

    The page says stage at diagnosis has a strong influence on the length of survival and gives no further interpretation of stage-specific values. SEER Combined Summary Stage, not AJCC TNM. Same SEER 21 series as the 13.7% figure above; do not chain it to the SEER 8 values or subtract. The page text gives the localized share as 15.4% and the table as 15%.

    Source

  • 5-year relative survival · United States, SEER 21 (excluding IL)

    • 3.4%

      2016-2022; distant stage at diagnosis (51% of cases)

    All races, both sexes; pancreatic cancer, SEER Combined Summary Stage 'distant' (cancer has metastasized).

    The page says stage at diagnosis has a strong influence on the length of survival and gives no further interpretation of stage-specific values. SEER Combined Summary Stage, not AJCC TNM. Same SEER 21 series as the 13.7% figure above; do not chain it to the SEER 8 values or subtract. The page text gives the localized share as 15.4% and the table as 15%.

    Source

  • Age-standardised mortality · United States (U.S. Mortality data shown on the SEER page; not a SEER registry series)

    • 10.64per 100,000

      1975 (observed value; first year in the table)

    All races, both sexes; US death rate for pancreatic cancer (age-adjusted), the 'Death Rate - U.S.' column of the SEER Cancer Stat Facts table, which draws on U.S. Mortality data (not SEER registries).

    Age-adjusted rate per 100,000 (the page uses 'age-adjusted'; the schema label is 'age-standardised'). Shown beside survival; the page reports the death rate and survival as separate series and does not link them. The page states death rates were stable over 2015-2024; the observed values in the table range from 10.45 (1995) to 11.35 (2021). Modeled-trend value in the adjacent column for this year: 10.73. Compare only with other values from this same death-rate series.

    Source

  • Age-standardised mortality · United States (U.S. Mortality data shown on the SEER page; not a SEER registry series)

    • 11.28per 100,000

      2018 (observed value; the same year as the last observed SEER 8 survival value)

    All races, both sexes; US death rate for pancreatic cancer (age-adjusted), the 'Death Rate - U.S.' column of the SEER Cancer Stat Facts table, which draws on U.S. Mortality data (not SEER registries).

    Age-adjusted rate per 100,000 (the page uses 'age-adjusted'; the schema label is 'age-standardised'). Shown beside survival; the page reports the death rate and survival as separate series and does not link them. The page states death rates were stable over 2015-2024; the observed values in the table range from 10.45 (1995) to 11.35 (2021). Modeled-trend value in the adjacent column for this year: 11.29. Compare only with other values from this same death-rate series.

    Source

  • Age-standardised mortality · United States (U.S. Mortality data shown on the SEER page; not a SEER registry series)

    • 11.31per 100,000

      2024 (observed value; last year in the table)

    All races, both sexes; US death rate for pancreatic cancer (age-adjusted), the 'Death Rate - U.S.' column of the SEER Cancer Stat Facts table, which draws on U.S. Mortality data (not SEER registries).

    Age-adjusted rate per 100,000 (the page uses 'age-adjusted'; the schema label is 'age-standardised'). Shown beside survival; the page reports the death rate and survival as separate series and does not link them. The page states death rates were stable over 2015-2024; the observed values in the table range from 10.45 (1995) to 11.35 (2021). Modeled-trend value in the adjacent column for this year: 11.25. Compare only with other values from this same death-rate series. The page's headline death rate for deaths in 2020-2024 is 11.3 per 100,000.

    Source

Survival figures are affected by earlier detection (lead time) and overdiagnosis, and by changes in who is diagnosed. For screened cancers, read them alongside mortality.

Evidence chains

A lane follows one defined population and setting, where each new trial tests against what was then standard. Compare only within a lane, and read every comparison as illustrative.

Resected PDAC, adjuvant chemotherapy: overall survival

Adults after R0 or R1 resection of pancreatic ductal adenocarcinoma, treatment-naive for this cancer (CONKO-001, ESPAC-4, PRODIGE 24, APACT) · Adjuvant · endpoint: OS

Results are not comparable across trials.

Controls differ: CONKO-001 compared gemcitabine with observation, while ESPAC-4, PRODIGE 24 and APACT used gemcitabine as the control, so the trials do not extend one another on a common axis. Eras, eligibility (for example APACT required a CA 19-9 below 100 U/mL; PRODIGE 24 enrolled people aged 18 to 79) and follow-up differ, and the primary endpoint was overall survival in ESPAC-4 but disease-free survival in CONKO-001, PRODIGE 24 and APACT. CONKO-001 is drawn from its 2007 report, which found no significant survival difference; its 2013 long-term update reported a hazard ratio and 5- and 10-year survival without medians, so it is described in the milestone and not drawn. PRODIGE 24 is drawn from its 5-year report and APACT from its 5-year follow-up update; their earlier analyses are described in the milestones. ESPAC-1 (fluorouracil-based regimens, overlapping factorial comparison groups) is not drawn. Compare only each trial's own two arms; the chart is not a pooled or network estimate.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

0102030405060OS, median (months)2007201720182022CONKO-001ESPAC-4PRODIGE 24APACT
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 4 trials in this lane. CONKO-001 (2007): experimental 22.1 months vs control 20.2 months; ESPAC-4 (2017): experimental 28 months vs control 25.5 months; PRODIGE 24 (2018): experimental 53.5 months vs control 35.5 months; APACT (2022): experimental 41.8 months vs control 37.7 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
CONKO-0012007OS20.2 months22.1 months
ESPAC-42017OS25.5 months28 months
PRODIGE 242018OS35.5 months53.5 months
APACT2022OS37.7 months41.8 months

Resected PDAC, adjuvant chemotherapy: disease-free survival

Adults after R0 or R1 resection of pancreatic ductal adenocarcinoma, treatment-naive for this cancer (CONKO-001, PRODIGE 24, APACT) · Adjuvant · endpoint: DFS

Results are not comparable across trials.

Disease-free survival was the primary endpoint of all three trials, but assessment differs: APACT's drawn value is the independently assessed primary endpoint (its investigator-assessed DFS is described in the milestone), and recurrence surveillance, eras and follow-up differ. CONKO-001 compared gemcitabine with observation; PRODIGE 24 and APACT used gemcitabine as the control. APACT did not meet its primary endpoint. PRODIGE 24 is drawn from its 5-year report. ESPAC-4 reported no disease-free survival result and is not in this lane. Compare only each trial's own two arms; the chart is not a pooled or network estimate.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

0510152025DFS, median (months)200720182022CONKO-001PRODIGE 24APACT
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
DFS, median, across 3 trials in this lane. CONKO-001 (2007): experimental 13.4 months vs control 6.9 months; PRODIGE 24 (2018): experimental 21.4 months vs control 12.8 months; APACT (2022): experimental 19.4 months vs control 18.8 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
CONKO-0012007DFS6.9 months13.4 months
PRODIGE 242018DFS12.8 months21.4 months
APACT2022DFS18.8 months19.4 months

Resectable and borderline resectable PDAC, neoadjuvant strategies

Adults with resectable or borderline resectable pancreatic ductal adenocarcinoma without prior treatment (PREOPANC and PREOPANC-2 enrolled both groups; NORPACT-1 enrolled resectable pancreatic head cancer only) · Neoadjuvant · endpoint: OS

Results are not comparable across trials.

Definitions of borderline resectable disease vary by trial (the 2017 consensus milestone describes one definition). Controls differ: PREOPANC and NORPACT-1 compared a neoadjuvant strategy with upfront surgery, whereas PREOPANC-2 compared neoadjuvant FOLFIRINOX with neoadjuvant chemoradiotherapy and had no upfront-surgery arm. Adjuvant treatment differed between and within trials. NORPACT-1 was a randomised phase 2 trial whose primary endpoint was overall survival at 18 months; its drawn value is the median from the intention-to-treat analysis. PREOPANC is drawn from its 2022 long-term analysis of the same randomisation; its 2020 primary analysis (hazard ratio 0.78, P = .096) is described in the milestone and the two analyses are not counted as separate trials. These people differ from the adjuvant lanes, in which surgery came first. Compare only each trial's own two arms; the chart is not a pooled or network estimate.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

010203040OS, median (months)202020242025PREOPANCNORPACT-1PREOPANC-2
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 3 trials in this lane. PREOPANC (2020): experimental 15.7 months vs control 14.3 months; NORPACT-1 (2024): experimental 25.1 months vs control 38.5 months; PREOPANC-2 (2025): experimental 21.9 months vs control 21.3 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
PREOPANC2020OS14.3 months15.7 months
NORPACT-12024OS38.5 months25.1 months
PREOPANC-22025OS21.3 months21.9 months

Metastatic PDAC, first-line chemotherapy

Adults with previously untreated metastatic pancreatic ductal adenocarcinoma (PRODIGE 4 / ACCORD 11, MPACT, NAPOLI-3) · Metastatic · endpoint: OS

Results are not comparable across trials.

Controls differ: gemcitabine in PRODIGE 4 and MPACT, nab-paclitaxel plus gemcitabine in NAPOLI-3. Performance-status limits differ (ECOG 0 or 1 in PRODIGE 4 and NAPOLI-3; Karnofsky 70 or more in MPACT), PRODIGE 4 was a phase 2/3 trial, and the trials span more than a decade of imaging, supportive care and later-line therapy. FOLFIRINOX and nab-paclitaxel plus gemcitabine were not compared directly in these trials. MPACT is drawn from its primary analysis; a post hoc long-term update is described in the milestone. HALO 109-301, a hyaluronan-high metastatic population on a nab-paclitaxel plus gemcitabine backbone, is not drawn here (see the gemcitabine-era lane). Compare only each trial's own two arms; the chart is not a pooled or network estimate.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

02468101214OS, median (months)201120132023PRODIGE 4 / ACCORD 11MPACTNAPOLI 3
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 3 trials in this lane. PRODIGE 4 / ACCORD 11 (2011): experimental 11.1 months vs control 6.8 months; MPACT (2013): experimental 8.5 months vs control 6.7 months; NAPOLI 3 (2023): experimental 11.1 months vs control 9.2 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
PRODIGE 4 / ACCORD 112011OS6.8 months11.1 months
MPACT2013OS6.7 months8.5 months
NAPOLI 32023OS9.2 months11.1 months

Advanced PDAC, first line: gemcitabine-era comparisons

Adults with previously untreated advanced pancreatic cancer, described as advanced symptomatic (Burris), advanced (CALGB 80303) or unresectable locally advanced or metastatic (SWOG S0205); the share with locally advanced versus metastatic disease is not recorded for these abstract-sourced trials · Locally advanced · endpoint: OS

Results are not comparable across trials.

Populations include locally advanced as well as metastatic disease, so this lane is kept apart from the metastatic-only lane. The Burris trial's primary efficacy measure was clinical benefit response, not survival; its drawn median survival is a further reported measure, and its control was fluorouracil. CALGB 80303 (bevacizumab, placebo-controlled) and SWOG S0205 (cetuximab) added an agent to gemcitabine and reported no survival gain; values come from abstracts. HALO 109-301 is described in the same milestone but not drawn: it enrolled a hyaluronan-high metastatic population on a nab-paclitaxel plus gemcitabine backbone and was reported in 2020, which would place it out of step with the milestone's 2010 position on this timeline. Compare only each trial's own two arms; the chart is not a pooled or network estimate.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

0246810OS, median (months)199620102010gemcitabine-vs-fluorouracil-burrisCALGB 80303S0205
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 3 trials in this lane. gemcitabine-vs-fluorouracil-burris (1996): experimental 5.65 months vs control 4.41 months; CALGB 80303 (2010): experimental 5.8 months vs control 5.9 months; S0205 (2010): experimental 6.3 months vs control 5.9 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
gemcitabine-vs-fluorouracil-burris1996OS4.41 months5.65 months
CALGB 803032010OS5.9 months5.8 months
S02052010OS5.9 months6.3 months

Locally advanced unresectable PDAC, tumour treating fields

Adults with newly diagnosed unresectable locally advanced pancreatic adenocarcinoma (PANOVA-3) · Locally advanced · endpoint: OS

Results are not comparable across trials.

Only one randomised comparison with a drawn value is in this lane, so it shows a single pair, not a trend. PANOVA-3 added a device (tumour treating fields) to gemcitabine and nab-paclitaxel; it was open-label with no sham device, and progression-free survival and response rate were not improved. It is not comparable with the metastatic or gemcitabine-era lanes, whose populations and backbones differ. Compare only the trial's own two arms.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

05101520OS, median (months)2025PANOVA-3
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 1 trial in this lane. PANOVA-3 (2025): experimental 16.2 months vs control 14.2 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
PANOVA-32025OS14.2 months16.2 months

Metastatic PDAC, after earlier systemic therapy

Adults with previously treated metastatic pancreatic ductal adenocarcinoma (RASolute 302) · Relapsed / refractory · endpoint: OS

Results are not comparable across trials.

Only one randomised comparison is in this lane, so it shows a single pair, not a trend. The control was investigator's-choice chemotherapy rather than one regimen, and the trial was open-label. The drawn value is the overall population (500 randomised); the dual primary endpoints were assessed in the RAS G12 population (91.8% of participants), whose values are described in the milestone. Follow-up is short. Compare only the trial's own two arms.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

05101520OS, median (months)2026RASolute 302
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 1 trial in this lane. RASolute 302 (2026): experimental 13.2 months vs control 6.7 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
RASolute 3022026OS6.7 months13.2 months

Standard of care, then and now

How the usual sequence of treatment changed for one setting. Choose a point in time; each step links to the milestone behind it.

Metastatic PDAC: systemic regimens tested in randomised trials

Adults with metastatic pancreatic ductal adenocarcinoma: first-line regimens, maintenance in germline BRCA-mutated disease, and treatment after earlier therapy · Metastatic

  1. Chemotherapy1GemcitabineGemcitabine versus fluorouracil in advanced pancreatic cancer (1996)
  2. Chemotherapy2FOLFIRINOX (good performance status)FOLFIRINOX versus gemcitabine: PRODIGE 4 / ACCORD 11 (2011)
  3. Chemotherapy3Nab-paclitaxel plus gemcitabineMPACT: nab-paclitaxel plus gemcitabine improves survival (2013)
  4. Chemotherapy4NALIRIFOXNAPOLI-3: NALIRIFOX versus nab-paclitaxel plus gemcitabine (2023)
  5. Targeted5Olaparib maintenance (germline BRCA)POLO: maintenance olaparib for germline BRCA-mutated disease (2019)
  6. Targeted6Daraxonrasib (after earlier therapy)Daraxonrasib: RAS(ON) inhibitor extends survival after prior therapy (2026)
2026. Based only on the trials cited; local practice varies. Daraxonrasib was compared with investigator's-choice chemotherapy after earlier treatment; the FDA approved it on 26 August 2026.
Text version
  • 2000: Gemcitabine
  • 2012: Gemcitabine → FOLFIRINOX (good performance status)
  • 2014: Gemcitabine → FOLFIRINOX (good performance status) → Nab-paclitaxel plus gemcitabine
  • 2024: Gemcitabine → FOLFIRINOX (good performance status) → Nab-paclitaxel plus gemcitabine → NALIRIFOX → Olaparib maintenance (germline BRCA)
  • 2026: Gemcitabine → FOLFIRINOX (good performance status) → Nab-paclitaxel plus gemcitabine → NALIRIFOX → Olaparib maintenance (germline BRCA) → Daraxonrasib (after earlier therapy)

Resected PDAC: adjuvant chemotherapy regimens tested in randomised trials

Adults after R0 or R1 resection of pancreatic ductal adenocarcinoma · Adjuvant

  1. Chemotherapy1Fluorouracil plus folinic acidESPAC-1: adjuvant chemotherapy and chemoradiotherapy compared (2001)
  2. Chemotherapy2GemcitabineCONKO-001: adjuvant gemcitabine versus observation (2007)
  3. Chemotherapy3Gemcitabine plus capecitabineESPAC-4: gemcitabine plus capecitabine as adjuvant therapy (2017)
  4. Chemotherapy4Modified FOLFIRINOX (more toxicity)PRODIGE 24: adjuvant modified FOLFIRINOX versus gemcitabine (2018)
2018. Based only on the trials cited; local practice varies. ESPAC-4 and PRODIGE 24 each used gemcitabine as the control; PRODIGE 24 reported longer survival with modified FOLFIRINOX, with more serious adverse events.
Text version
  • 2002: Fluorouracil plus folinic acid
  • 2008: Fluorouracil plus folinic acid → Gemcitabine
  • 2018: Fluorouracil plus folinic acid → Gemcitabine → Gemcitabine plus capecitabine → Modified FOLFIRINOX (more toxicity)

Reversals and dead ends

Approaches that were abandoned or withdrawn, and trials whose results were later questioned, are part of the story.

  • 2010

    Bevacizumab, cetuximab, PEGPH20 added to chemotherapy: no gain

    Three large trials added a targeted drug to chemotherapy in advanced pancreatic cancer: bevacizumab, cetuximab and PEGPH20. None lengthened survival. Median survival with bevacizumab was about 5.8 months versus 5.9 with placebo, and in the PEGPH20 trial about 11.2 versus 11.5 months.

    Three phase 3 trials, each with overall survival as primary endpoint, were negative. CALGB 80303 (gemcitabine +/- bevacizumab, 602 enrolled): median OS 5.8 vs 5.9 months (P=.95). S0205 (gemcitabine +/- cetuximab, 745 eligible): 6.3 vs 5.9 months, HR 1.06 (0.91-1.23). HALO 109-301 (nab-paclitaxel-gemcitabine +/- PEGPH20, 492 ITT): 11.2 vs 11.5 months, HR 1.00 (0.80-1.27).

  • 2022

    APACT: adjuvant nab-paclitaxel plus gemcitabine, DFS not met

    APACT compared nab-paclitaxel plus gemcitabine with gemcitabine alone after pancreatic cancer surgery. The main measure, disease-free survival judged by independent reviewers, did not differ: 19.4 against 18.8 months. Median survival was 41.8 against 37.7 months at five years of follow-up, which the authors treat as descriptive because the main measure was not met.

    APACT (Tempero 2023): 866 patients with resected PDAC, nab-paclitaxel plus gemcitabine vs gemcitabine, six cycles. Primary endpoint independently assessed DFS not met: 19.4 vs 18.8 months (HR 0.88, 95% CI 0.729-1.063, P=.18). Investigator-assessed DFS 16.6 vs 13.7 months (HR 0.82, P=.02). OS at 5-year follow-up 41.8 vs 37.7 months (HR 0.80, 95% CI 0.678-0.947). Grade 3 or worse events 86% vs 68%.

  • 2024

    NORPACT-1: neoadjuvant FOLFIRINOX versus upfront surgery

    NORPACT-1, a randomised phase 2 trial in four Nordic countries, compared four cycles of FOLFIRINOX before surgery with surgery first in resectable pancreatic head cancer. About 60 in 100 were alive at 18 months with FOLFIRINOX first against 73 in 100 with surgery first. The authors did not find a survival benefit from the neoadjuvant approach.

    NORPACT-1 (Labori 2024): randomised phase 2, 140 patients with resectable pancreatic head cancer (77 neoadjuvant FOLFIRINOX, 63 upfront surgery), adjuvant chemotherapy in both. Primary endpoint OS at 18 months, ITT: 60% (95% CI 49-71) vs 73% (62-84), p=0.032. Median OS 25.1 vs 38.5 months, HR 1.52 (95% CI 1.00-2.33), log-rank p=0.050. Resection 82% vs 89%. Grade 3+ events 58% vs 40%.

Today and tomorrow

Standard of care today

Screening
Guideline documents read at abstract level only. The USPSTF reaffirmation statement (JAMA 2019, abstract read) recommends against screening for pancreatic cancer in asymptomatic adults (D recommendation); its abstract states that the task force found no evidence that screening or treatment of screen-detected disease improves disease-specific morbidity or mortality, or all-cause mortality. Surveillance of defined high-risk individuals is addressed by the CAPS international consensus (Gut 2020), cited by identifier only; its recommendations were not read here, so none is attributed to it. Whether the USPSTF statement has since been updated was not checked beyond a PubMed guideline search.
US Preventive Services Task Force, Owens DK, Davidson KW, Krist AH, Barry MJ, Cabana M, Caughey AB, et al 2019Goggins M, Overbeek KA, Brand R, Syngal S, Del Chiaro M, Bartsch DK, et al 2020
Adjuvant
Resected pancreatic ductal adenocarcinoma. No FDA or EMA document for an adjuvant pancreatic indication was read. Trial evidence for adjuvant modified FOLFIRINOX (PRODIGE 24), gemcitabine plus capecitabine (ESPAC-4) and gemcitabine plus nab-paclitaxel (APACT, whose primary endpoint was not met) is on the milestone timeline. Guideline documents are cited by identifier only and their recommendations were not read, so no regimen choice is attributed to them: the ESMO Clinical Practice Guideline (Ann Oncol 2023, abstract not available in PubMed) and the NCCN Guidelines for Pancreatic Adenocarcinoma, Version 4.2026, through its JNCCN summary article (abstract read: it describes the guideline's scope and says nothing about specific regimens).
Conroy T, Pfeiffer P, Vilgrain V, Lamarca A, Seufferlein T, O'Reilly EM, et al 2023Ko AH, Malafa MP, Aguilera T, Basturk O, Benson AB 3rd, Cardin DB, et al 2026Conroy T, Hammel P, Hebbar M, Ben Abdelghani M, Wei AC, Raoul JL, et al 2018Neoptolemos JP, Palmer DH, Ghaneh P, Psarelli EE, Valle JW, Halloran CM, et al 2017
Neoadjuvant
Resectable and borderline resectable disease. No FDA or EMA document for a neoadjuvant pancreatic regimen was read. Randomised evidence on preoperative treatment is on the milestone timeline: PREOPANC (neoadjuvant gemcitabine-based chemoradiotherapy against upfront surgery, in resectable and borderline resectable disease), NORPACT-1 (a randomised phase 2 trial of neoadjuvant FOLFIRINOX against upfront surgery, in resectable pancreatic head cancer) and PREOPANC-2 (a randomised phase 3 trial of neoadjuvant FOLFIRINOX against neoadjuvant gemcitabine-based chemoradiotherapy, in resectable and borderline resectable disease); the three differ in design, population and comparator and are not compared here. An international consensus defines borderline resectable disease (Pancreatology 2018). Phase 3 comparisons of perioperative against adjuvant chemotherapy are in progress (see horizon). Guideline text was not read for this setting.
Isaji S, Mizuno S, Windsor JA, Bassi C, Fernández-Del Castillo C, Hackert T, et al 2018Versteijne E, van Dam JL, Suker M, Janssen QP, Groothuis K, Akkermans-Vogelaar JM, et al 2022Labori KJ, Bratlie SO, Andersson B, Angelsen JH, Biörserud C, Björnsson B, et al 2024Janssen QP, van Dam JL, van Bekkum ML, Bonsing BA, Bos H, Bosscha KP, et al 2025
Locally advanced
Regulatory document read: the FDA premarket approval (PMA) database record for Optune Pax (P250034, Novocure GmbH; a device, not a drug) gives a decision date of 11 February 2026 and the approval order statement 'intended for treatment of adult patients with locally advanced pancreatic cancer, concomitant with gemcitabine and nab-paclitaxel'. The record's page was last updated 19 October 2026. The pivotal PANOVA-3 trial is on the milestone timeline. No guideline text was read for this setting.
US Food and Drug Administration 2026Babiker HM, Picozzi V, Chandana SR, Melichar B, Kasi A, Gang J, et al 2025
Metastatic

First-line treatment of metastatic pancreatic adenocarcinoma, FDA documents read. Gemcitabine:

  • approval letter of 15 May 1996 (scanned package
  • the letter text read gives first-line treatment of advanced or metastatic adenocarcinoma of the pancreas). Erlotinib with gemcitabine: supplement approval letter of 2 November 2005, for first-line treatment of locally advanced, unresectable or metastatic pancreatic cancer. Nab-paclitaxel with gemcitabine: 6 September 2013. Liposomal irinotecan with oxaliplatin, fluorouracil and leucovorin (NALIRIFOX): 13 February 2024. No accelerated-approval wording was seen in the letters or approval packages. FOLFIRINOX is supported by the PRODIGE 4 trial (milestone timeline)
  • no FDA or EMA document for the regimen was read. Guideline documents are cited by identifier only: the ESMO Clinical Practice Guideline (Ann Oncol 2023) and the ASCO metastatic pancreatic cancer guideline updates (J Clin Oncol 2018 and 2020
  • only the abstracts were read, and they address second-line therapy and biomarker testing and predate NALIRIFOX and daraxonrasib), so no first-line recommendation is attributed to any of them
US Food and Drug Administration 1996US Food and Drug Administration 2005US Food and Drug Administration 2013US Food and Drug Administration 2024Conroy T, Desseigne F, Ychou M, Bouché O, Guimbaud R, Bécouarn Y, et al 2011Conroy T, Pfeiffer P, Vilgrain V, Lamarca A, Seufferlein T, O'Reilly EM, et al 2023Sohal DPS, Kennedy EB, Khorana A, Copur MS, Crane CH, Garrido-Laguna I, et al 2018Sohal DPS, Kennedy EB, Cinar P, Conroy T, Copur MS, Crane CH, et al 2020
Maintenance
Germline BRCA-mutated metastatic pancreatic adenocarcinoma, FDA letter read. Olaparib (NDA 208558/S-010; replacement letter, effective approval date 27 December 2019): maintenance treatment of adults with a deleterious or suspected deleterious germline BRCA mutation whose disease has not progressed on at least 16 weeks of first-line platinum-based chemotherapy, with patient selection by an FDA-approved companion diagnostic. The letter has no accelerated-approval wording. The POLO trial is on the milestone timeline. No guideline text was read for this setting.
US Food and Drug Administration 2019Golan T, Hammel P, Reni M, Van Cutsem E, Macarulla T, Hall MJ, et al 2019
Relapsed / refractory

Previously treated metastatic pancreatic adenocarcinoma, regulatory documents read. FDA, 22 October 2015:

  • liposomal irinotecan (Onivyde) in combination with fluorouracil and leucovorin for metastatic adenocarcinoma of the pancreas that has progressed following gemcitabine-based therapy, with a limitation of use that it is not indicated as a single agent. FDA, 26 August 2026: daraxonrasib (Rasonque, NDA 220910) for adults with metastatic pancreatic adenocarcinoma who have received at least one prior systemic therapy or who are not candidates for multiagent systemic therapy
  • the letter has no accelerated-approval wording. EMA: the CHMP meeting highlights of 12-15 October 2026 (page dated 16 October 2026) list a positive opinion for Rasonque for adults with metastatic pancreatic adenocarcinoma who have received at least one previous treatment, with the European Commission decision pending when the page was read. The RASolute 302 trial is on the milestone timeline. An ESMO Clinical Practice Guideline Express Update on daraxonrasib (Ann Oncol 2026) is cited by identifier only
  • PubMed gives no abstract and its text was not read
US Food and Drug Administration 2015US Food and Drug Administration 2026European Medicines Agency 2026O'Reilly EM, Wainberg ZA, Hendifar AE, Borad MJ, Pietrantonio F, Pant S, et al 2026Conroy T, Ducreux M 2026
Any setting

Biomarker-selected disease, FDA documents read. Pembrolizumab for unresectable or metastatic MSI-H or mismatch-repair-deficient solid tumours after prior treatment (a tissue-agnostic indication, not specific to pancreatic cancer):

  • accelerated approval on 23 May 2017
  • the supplement letter of 28 March 2023 states 'SUBPART E FULFILLED' (fulfilment of postmarketing requirements) and does not use the words conversion or regular approval. Larotrectinib for solid tumours with an NTRK gene fusion (also tissue-agnostic): accelerated approval on 26 November 2018
  • the supplement letter of 9 April 2025 states that it supports conversion from accelerated to regular approval and carries a 'SUBPART H FULFILLED' section. Zenocutuzumab-zbco (Bizengri), accelerated approval on 4 December 2024, includes advanced, unresectable or metastatic pancreatic adenocarcinoma harbouring an NRG1 gene fusion with disease progression on or after prior systemic therapy. The openFDA indication sections of the sotorasib and adagrasib labels read list no pancreatic indication. Pancreatic-specific results for the tissue-agnostic indications are not given on this page
  • see the milestone timeline
US Food and Drug Administration 2017US Food and Drug Administration 2023US Food and Drug Administration 2018US Food and Drug Administration 2025US Food and Drug Administration 2024US Food and Drug Administration (openFDA drug label API) 2025US Food and Drug Administration (openFDA drug label API) 2026

Trials to watch

  • Alliance A021806 NCT04340141

    In resectable pancreatic adenocarcinoma, does perioperative chemotherapy (registry title: the usual chemotherapy before and after surgery) differ in overall survival from surgery followed by adjuvant chemotherapy? Registry record: randomised phase 3 (Alliance for Clinical Trials in Oncology); status active, not recruiting; enrolment 358 (actual); overall survival is the primary outcome; estimated primary completion December 2028; no results posted. A PubMed search on 21 October 2026 found a design paper (Ann Surg Oncol 2024) and no efficacy report.

  • PREOPANC-3 NCT04927780

    In resectable pancreatic ductal adenocarcinoma, does perioperative modified FOLFIRINOX differ in overall survival from adjuvant modified FOLFIRINOX? Registry record: randomised phase 3 (Erasmus Medical Center); status active, not recruiting; enrolment 378 (estimated); overall survival is the primary outcome; estimated primary completion January 2027; no results posted. A PubMed search on 21 October 2026 found a protocol paper (BMC Cancer 2023) and no efficacy report.

  • RASolute 303 NCT07491445

    In first-line metastatic pancreatic adenocarcinoma, how do daraxonrasib alone and daraxonrasib with gemcitabine and nab-paclitaxel compare with gemcitabine plus nab-paclitaxel? Registry record: randomised, open-label, three-arm phase 3 (Revolution Medicines); status recruiting; enrolment 900 (estimated); primary outcomes are progression-free survival and overall survival; estimated primary completion June 2028; no results posted.

  • RASolute 304 NCT07252232

    After resection and perioperative multi-agent chemotherapy for pancreatic ductal adenocarcinoma, does adjuvant daraxonrasib differ in disease-free survival from observation? Registry record: randomised, open-label phase 3 (Revolution Medicines); status recruiting; enrolment 500 (estimated); the primary outcome is investigator-assessed disease-free survival; estimated primary completion May 2029; no results posted.

  • RASolute 305 NCT07621718

    In first-line metastatic KRAS G12D-mutated pancreatic adenocarcinoma, does adding zoldonrasib to investigator's-choice chemotherapy (modified FOLFIRINOX or gemcitabine plus nab-paclitaxel) differ from placebo plus chemotherapy? Registry record: randomised, double-blind, placebo-controlled phase 3 (Revolution Medicines); status recruiting; enrolment 670 (estimated); primary outcomes are progression-free survival and overall survival; estimated primary completion December 2028; no results posted.

  • PRISM-1 NCT06608927

    In treatment-naive metastatic pancreatic ductal adenocarcinoma, does adding quemliclustat to gemcitabine plus nab-paclitaxel differ in overall survival from placebo plus chemotherapy? Registry record: randomised, double-blind, placebo-controlled phase 3 (Arcus Biosciences); status active, not recruiting; enrolment 610 (estimated); overall survival is the primary outcome; estimated primary completion November 2030; no results posted. A PubMed search on 21 October 2026 found no phase 3 efficacy report.

References

Show all 77 references

Loading references… Open the list as a page

Every figure was checked against its primary source to the extent possible. Items that could not be fully confirmed are marked “needs review”. Educational history, not medical advice.

Report an error

Links open the primary source. A flag marked ⚑ needs review means a value or claim has not yet been independently confirmed.

2008UnderstandingUnderstanding

majorAny setting

Sequencing reveals core pathways and subtypes of pancreatic cancer

In 2008 a team sequenced 24 pancreatic cancers and found about 63 gene changes per tumour on average, grouped into 12 core pathways altered in most tumours. In 2016 an analysis of 456 tumours described 10 pathways and four subtypes based on gene activity. Neither report describes a treatment result.

Jones 2008 (Science): 23,219 transcripts sequenced in 24 pancreatic cancers plus copy-number arrays; average 63 genetic alterations, mostly point mutations, defining 12 core pathways each altered in 67 to 100% of tumours. Bailey 2016 (Nature): 456 PDACs, 32 recurrently mutated genes in 10 pathways; four expression subtypes (squamous, pancreatic progenitor, immunogenic, ADEX); squamous described as poor prognosis.

Why it mattered

The 2008 authors state that dysregulation of the core pathways can explain the major features of pancreatic tumorigenesis. The 2016 authors state that the subtype differences identify opportunities for therapeutic development.

Caveats

  • Descriptive genomics: neither abstract reports a treatment outcome, and the abstracts do not say whether the subtypes guide therapy.
  • The 2008 cohort was 24 tumours; the 2016 analysis integrated 456 tumours. Subtype names are the authors' own, and their definitions rest on expression methods in the full paper, which was not read.
  • Different cohorts and methods; the two reports are not a single time series.