Pancreatic cancer risk factors

Editor-In-Chief: C. Michael Gibson, M.S., M.D. [1]; Associate Editor(s)-in-Chief: Fahad Hasan, M.D.[2] Aravind Reddy Kothagadi M.B.B.S[3]

Pancreatic ductal adenocarcinoma (PDAC) accounts for approximately 90% of pancreatic neoplasms.^[1] Its risk factors fall into three groups: modifiable lifestyle and metabolic factors, non-modifiable demographic factors, and inherited (germline) susceptibility. Most individual factors carry only a modest relative risk. The factors with the most clinical leverage are:

• Cigarette smoking, the dominant modifiable cause (~20% of cases).
• New-onset diabetes mellitus in older adults, which often reflects occult PDAC rather than causing it.
• High-penetrance hereditary syndromes.^[2][1]

Approximately 5–10% of unselected patients with PDAC carry a pathogenic germline variant. At least half of these carriers have no suggestive family history, which is the rationale for universal germline testing of all patients with PDAC.^[3][4]

• Diabetes and PDAC are related in both directions. Long-standing type 2 diabetes is a modest causal factor. In contrast, new-onset diabetes (NOD) within ~3 years before diagnosis is increasingly viewed as a paraneoplastic manifestation of occult PDAC.^[2][6]
• Up to ~50% of patients have diabetes at the time of PDAC diagnosis. PDAC incidence is elevated for years before diagnosis, and the association is strongest within 1 year of diabetes onset (adjusted HR ~9).^[7]
• Absolute 3-year PDAC risk after NOD in adults ≥50 years is below 1%. This is too low to justify imaging every patient with NOD, but high enough to justify phenotype-based enrichment.^[6]
• Features that raise risk within NOD:^[6]
  • Older age
  • Unintentional weight loss
  • Rapidly escalating glycemia
• The ENDPAC (Enriching New-Onset Diabetes for Pancreatic Cancer) model uses age, weight change, and glucose change. A score ≥3 identified a subgroup with ~1.2% annual PDAC incidence.^[6]
• Post-pancreatitis diabetes mellitus carries higher PDAC risk than type 2 diabetes alone (~6.9-fold vs type 2 diabetes).^[8]

• Chronic pancreatitis:
  • Increases PDAC risk approximately 7-fold; some pooled analyses report standardized incidence ratios of ~13–23.
  • Absolute annual risk remains low, so sporadic chronic pancreatitis alone does not justify screening.^[9][1][3]
• Acute pancreatitis:
  • PDAC risk is highest within 1–2 years of an episode (OR >20).
  • This largely reflects reverse causality, because a tumor can cause obstructive pancreatitis.
  • Imaging is therefore reasonable after an unexplained episode of acute pancreatitis resolves.^[4][3]

• Age is the dominant non-modifiable factor. Incidence rises steeply after age 50 and peaks at 70–79 years.^[4]
• Male sex carries modestly higher incidence than female sex.^[4]
• Race/ethnicity: In the United States, incidence is highest among non-Hispanic Black individuals, followed by White, Hispanic, and Asian–Pacific Islander populations.^[3][4]
• Ashkenazi Jewish ancestry:
  • Elevated risk reflects largely a higher prevalence of founder BRCA1/BRCA2 variants.^[10]
  • In a prospective cohort of 49,410 men, Ashkenazi Jewish heritage alone showed only a trend toward higher risk (HR 1.29; 95% CI 0.94–1.77).
  • Risk was markedly elevated among Ashkenazi Jewish men with a family history of breast cancer (HR 2.61) or colorectal cancer (HR 1.92).^[11]
• Non-O ABO blood group is a reproducible, modest risk factor (OR ~1.3–1.7 vs group O; adjusted OR 1.37 in a population-based case–control study). The association is stronger in people aged ≤70 years.^[12][13]

• Approximately 5–10% of unselected patients with PDAC carry a pathogenic or likely pathogenic germline variant, most often in DNA damage-repair or homologous recombination genes.^[3][4]
• At least half of carriers report no family history and do not meet classic syndrome criteria.^[3][14]

Estimates are compiled from guideline and review sources.^{[15][16][17][3]}

• Hereditary pancreatitis lifetime risk: Older teaching cited lifetime PDAC risk as high as ~40–70%. More recent evidence suggests a lifetime risk below 10%, although relative risk remains very high.^[3]
• Familial pancreatic cancer (FPC) is defined as ≥2 affected first-degree relatives without an identified syndrome. It accounts for ~10% of cases.^[15][4]
• In FPC, relative risk rises with the number and age of affected relatives:
  • ~2.5-fold with one affected first-degree relative
  • ~4.5-fold with ≥2 affected first-degree relatives
  • ~6.9-fold when a relative was diagnosed before age 50^[4][18]
• In a prospective cohort, a family history of pancreatic cancer in a first-degree relative conferred an HR of 2.79 (95% CI 1.28–6.07).^[11]

• Early-onset GI cancer is typically defined as diagnosis before age 50.^[19] Some pancreatic cancer consortia instead define early-onset as before age 60 and very-early-onset as before age 45.^[20]
• Established risk factors are more strongly associated with earlier-onset disease than with later-onset disease. These include smoking, obesity, diabetes, tall height, and non-O blood group.
  • Carrying 3–5 of these factors conferred an HR of 9.24 in people aged ≤60 years, versus 1.46 in those aged >70 years.^[13]
• Alcohol intake >26 g/day (≈2 drinks) was associated with early-onset PDAC (OR 1.49), especially before age 45 (OR 2.18).^[20]
• Approximately 15–30% of early-onset GI cancers carry pathogenic germline variants, for example in BRCA1/2 or mismatch repair genes. All patients with early-onset GI cancer should therefore undergo genetic testing.^[19]

• Helicobacter pylori — evidence is inconsistent:
  • A case–control study linked CagA-negative H. pylori seropositivity with increased risk (OR 1.68).^[12]
  • A nested analysis of 5 prospective US cohorts found no association, regardless of CagA status (OR 0.83; 95% CI 0.65–1.06).^[21]
  • A 2025 meta-analysis found no significant overall association (OR 1.15; 95% CI 0.93–1.41).^[22]
  • Not currently an actionable risk factor.
• Alcohol as an independent cause — causal evidence remains incomplete because of confounding by smoking and pancreatitis.^[5]
• Moderate-penetrance carriers without a family history — guidelines disagree on whether BRCA1, ATM, or PALB2 carriers without a family history of PDAC should be considered high risk for surveillance purposes.^[17][10]

• Universal germline testing: Offer multigene panel germline testing to all patients diagnosed with PDAC, regardless of age or family history.^[10][15]
• Relatives: Refer first-degree relatives for genetic counseling. Use cascade testing when a pathogenic variant is identified.^[10]
• Screening: Screening unselected patients with diabetes or chronic pancreatitis is not supported, because absolute risk is too low. Enrichment by genotype, family history, or NOD phenotype is required.^[3][6]
• NOD evaluation: Consider occult PDAC in adults ≥50 years with new diabetes plus unexplained weight loss or rapidly worsening glycemia.^[6]
• Risk reduction: Counseling should prioritize smoking cessation, which carries the largest attributable benefit. Weight control, alcohol moderation, and metabolic optimization are also recommended.^[4]
• Gene-specific surveillance eligibility and start ages are covered in Pancreatic cancer screening.

Template:WH Template:WS