# Run the ENDPAC score on every new diabetes diagnosis after 50 and scan the high scorers

Source: https://onco.cc/ideas/idea-pdac-new-onset-diabetes-risk-score-pathway/  
OnCo record `idea-pdac-new-onset-diabetes-risk-score-pathway` (Idea). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

About 1 in 100 people who develop diabetes after 50 has a pancreatic cancer behind it. A score built from weight change, blood sugar change and age, calculable from records already in primary care, picks out a group where the rate is nearer 1 in 30; the proposal is to scan that group rather than wait for symptoms.

## Summary

Chari's population cohort found pancreatic cancer in 0.85 percent of 2,122 new diabetics aged 50 or over within three years, an observed-to-expected ratio of 7.94, with 10 of 18 diagnosed within six months of the diabetes. Sharma's ENDPAC model (weight change, glucose change, age at onset) had an area under the curve of 0.87; in validation a score of 3 or more identified 78 percent of cancers at 85 percent specificity, with a prevalence of 3.6 percent among high scorers, and a score of 0 or below in 49 percent of patients marked extremely low risk. The proposal is a pragmatic pathway, not a new test: an automated ENDPAC flag in the primary care record at diabetes diagnosis, contrast CT or MRI for scores of 3 or more, and CA 19-9 with awareness that Lewis-negative patients (Tempero 1987) cannot make it. It is distinct from the existing OnCo idea of a blood-based multi-cancer test in the same population and can precede it, because imaging is available now. The test is a prospective cohort with stage at diagnosis against contemporaneous controls; the Hungarian NODES cohort (2,522 estimated) and the US NOD cohort are the nearest running studies. The UK, where diabetes diagnoses are coded in a single primary care system, is an unusually good place to run it.

## Fields

- Kind: Idea
- Last checked: 2026-09-24
- Tags: pancreatic-evidence
- Hypothesis: An automated ENDPAC score at new diabetes diagnosis after 50, with imaging for scores of 3 or more, detects pancreatic cancer at a resectable stage in a higher proportion than symptomatic presentation and finds one cancer for every 30 to 40 people scanned.
- Rationale: The score needs no new assay, the prevalence in high scorers is high enough for a scan to be justified, and 75 percent of cases in the derivation cohort were flagged more than six months before diagnosis.
- Proposed test: Prospective pragmatic cohort in a primary care network: ENDPAC computed at diagnosis, imaging for high scorers, three-year cancer ascertainment; primary outcome stage distribution and resection rate versus matched contemporaneous cases; secondary outcomes scans per cancer found and incidental findings.
- Maturity: early-clinical
- Actor: clinic

## Sources

- Sharma et al.: ENDPAC model (Gastroenterology 2018): https://europepmc.org/article/MED/29775599
- Chari et al.: probability of pancreatic cancer following diabetes (Gastroenterology 2005): https://europepmc.org/article/MED/16083707
- ClinicalTrials.gov NCT04164602: https://clinicaltrials.gov/study/NCT04164602

## Connected records

- ideas: [A fast-track pathway from suspicion to treatment for pancreatic cancer in the NHS, measured from first scan to first treatment](https://onco.cc/ideas/idea-pdac-uk-fast-track-diagnosis-to-treatment-pathway/), [Blood-based pancreatic cancer detection in new-onset diabetes](https://onco.cc/ideas/idea-mced-new-onset-diabetes/)
- roadmaps: [Early detection roadmap: organ screening → blood tests for many cancers](https://onco.cc/roadmaps/early-detection-roadmap/), [Pancreatic cancer roadmap: from Whipple's operation to gemcitabine, FOLFIRINOX, adjuvant chemotherapy, PARP inhibition, KRAS inhibition, vaccines and the surveillance question](https://onco.cc/roadmaps/pancreatic-roadmap/)
- cancers: [Pancreatic ductal adenocarcinoma](https://onco.cc/cancers/pancreatic/)
- technologies: [CT (computed tomography)](https://onco.cc/technologies/ct/), [High-risk pancreatic surveillance (CAPS / PRECEDE)](https://onco.cc/technologies/pancreatic-surveillance/), [MRI](https://onco.cc/technologies/mri/)
- terms: [CA 19-9](https://onco.cc/terms/ca19-9/), [Positive predictive value (PPV)](https://onco.cc/terms/ppv/)
- bottlenecks: [Biomarkers are not validated or standardised](https://onco.cc/bottlenecks/b-biomarker-validation/), [The hardest cancers are found late](https://onco.cc/bottlenecks/b-early-detection/)
- key papers: [Lead-Time Trajectory of CA19-9 as an Anchor Marker for Pancreatic Cancer Early Detection](https://onco.cc/key-papers/paper-fahrmann-ca19-9-lead-time-gastroenterology-2021/), [Model to Determine Risk of Pancreatic Cancer in Patients With New-Onset Diabetes](https://onco.cc/key-papers/paper-sharma-endpac-model-new-onset-diabetes-gastroenterology-2018/), [Probability of pancreatic cancer following diabetes: a population-based study](https://onco.cc/key-papers/paper-chari-pancreatic-cancer-following-diabetes-gastroenterology-2005/), [Relationship of carbohydrate antigen 19-9 and Lewis antigens in pancreatic cancer](https://onco.cc/key-papers/paper-tempero-ca19-9-lewis-antigens-cancer-res-1987/)

---
JSON: https://onco.cc/api/v1/entities/idea-pdac-new-onset-diabetes-risk-score-pathway.json