Background
Pancreatic cancer remains one of the deadliest cancers worldwide, with a 5-year survival rate of approximately 10%. Therapeutic options are still limited today and preventive strategies are hindered by insufficient knowledge of modifiable risk factors; epidemiological data on these factors remain scarce.
Objective
In this study, we focused on pesticides, specifically investigating the association between exposure to triazole fungicides and pancreatic cancer risk using a mechanistic-oriented rationale based on Adverse Outcome Pathway.
Methods
We analyzed data from the French AGRICAN cohort, which includes 181,842 participants involved in agriculture or related sectors, recruited between 2005 and 2007. Individual pesticide exposure was assessed using the crop-exposure matrix PESTIMAT, which provides exposure parameters for pesticides used in France since 1950. Triazoles (N=26, including epoxiconazole, prothioconazole, tebuconazole…) are systemic fungicides widely used in agriculture to control fungal diseases by inhibiting ergosterol biosynthesis. We considered exposures as both categorical (exposed/unexposed but exposed to other pesticides/unexposed to any pesticide) and continuous cumulative exposure scores. Main analyses were restricted to participants who had worked in agriculture. The reference group consisted of farmers unexposed to any pesticides.  We used Cox proportional hazards models stratified by diabetes status and adjusted for sex, body mass index, educational level and smoking (pack-years) to estimate hazard ratios (HRs) with 95% confidence intervals (CIs). Effect modification by sex was assessed using interaction terms between sex and exposure variables.
Results
We included 181,618 participants from the AGRICAN cohort (54.3% men, median age: 67 years, 86.6% farmers). During a 1,798,150 person-year follow-up (median: 10.9 years, maximum: 12.6 years), we observed 783 incident pancreatic cancer cases. Among cases with specified location (N=618), 68.6% were located in the head, 14.2% in the body ,and 17.2% in the tail, which was more frequent in women (18.4%) than in men (10.5%).
Among farmers, 27,505 (22.1%) were considered exposed to at least one triazole fungicide, with substantially lower prevalence among women farmers (3.4%) compared to men. The median number of individual triazoles substances to which participants were exposed was 13 (14 in men and 6 in women).
Overall, no significant increase in pancreatic cancer risk associated with triazoles exposure was observed in farmers analysed globally. However, elevated, though non-significant, HRs were found for some specific triazoles such as difenoconazole (HR=1.36 (0.95 - 1.92), 59 exposed cases). A significant effect modification by sex was identified (p-value= 0.001). In women, an increase of one standard deviation in the cumulative exposure score was associated with a higher risk (HR= 2.09 (1.35 - 3.25)). Several individual triazole substances showed increased risks in women only. Comparing women exposed to specific triazoles versus unexposed to any pesticides, HRs ranged from 2.97 (1.38 - 6.41) for triadimefon (7 exposed cases) to 7.70 (2.42- 24.50) for metconazole (3 exposed cases). 
Conclusions
These findings suggest a potential increased risk of pancreatic cancer associated with triazole fungicide exposure in women. Sex-related differences may be explained by hormonal and metabolic factors influencing susceptibility. Further analyses, including sensitivity analysis are ongoing to assess the robustness of these results.