Abstract

10559 Background: Non-O blood type has been associated with increased risk of pancreatic ductal adenocarcinoma (PDAC), the third leading cause of cancer death in the United States (US), in largely white populations. We aimed to explore PDAC risk across blood types in the racially diverse Veteran’s Health Administration (VA), the largest integrated healthcare system in the US. Methods: We included all individuals in the VA from 2002 - 2022 with at least one inpatient or outpatient encounter and a non-missing race. We defined the index date as 1.5 years after the initial VA encounter and excluded individuals with prior PDAC. We defined the exposure by non-O (A, B, AB) or type O blood type. We defined PDAC based on a previously validated algorithm of ≥2 sources (cancer registry, hospitalizations, or outpatient oncology encounters) with 95% positive predictive value. Using time to PDAC as the outcome in a multivariable Cox regression, we evaluated the association between blood type and PDAC risk, adjusting for age, sex, BMI, smoking, alcohol use, Hepatitis C (HCV) and diabetes. We censored at time of death, 1.5 years after the last VA encounter, or the end of the study period. We performed a subgroup analysis to evaluate the association of blood type by race. Results: Of the 2.5 million individuals with an available blood type, 1.35 million (52.6%) had a non-O blood type (Table) and over a quarter of our cohort (28.2%, 725,878) were over baseline age 50. Median time to PDAC diagnosis was 8.6 years. There were 877 PDAC cases (0.03%), with 530 (60.4%) of those having a non-O blood type. Compared to those with type O blood, risk of PDAC for individuals with non-O blood type was higher overall [adjusted HR = 1.37 (95% CI 1.20-1.57)], and higher among Whites (n=1,689,092, HR 1.52, 95% CI 1.27-1.82) than Black patients (n=525,666, HR 1.15, 95% CI 0.90-1.48), though this was limited by smaller sample size. Conclusions: Non-O blood type was associated with higher risk of PDAC in a diverse population. Given the potential differences between races, we will further explore the impact of neighborhood-level socioeconomic data on PDAC risk. [Table: see text]