Etiology and early detection of pancreatic cancer

Pancreatic cancer is one of the most dismal malignancies. Although its incidence and prevalence are relatively low, the disease burden is still an important health issue, especially in terms of mortality. Primary prevention is of utmost importance for such detrimental disease, however, the underlying etiology remains largely unknown. Moreover, the poor survival is attributed to a lack of non- or minimally-invasive methods for early cancer detection. In this thesis, we explored several biological factors, including poor dental health, diagnosis of inflammatory bowel disease (IBD), and gastric mucosal abnormality, which could enrich our understanding of the etiology of pancreatic cancer. In addition, we took the challenge to identify a panel of plasma protein biomarkers for detecting pancreatic ductal adenocarcinoma (PDAC) at an early stage, when therapies are most likely to be successful. In Study I, we retrieved the information on dental health status (healthy condition, caries, root canal infection, mild inflammation, and periodontitis) and the number of teeth of 5.9 million residents from the Swedish Dental Health Register between 2009 and 2016. Multivariate- adjusted hazard ratios (HRs) were derived from Cox proportional-hazards regression models. Individuals aged younger than 50 years and with compromised dental health status had a higher risk of pancreatic cancer, than those with good dental health status; among individuals aged between 50-70 years, only those with periodontitis had a 20% higher risk of pancreatic cancer. Moreover, individuals with fewer teeth tended to have a higher risk of pancreatic cancer. In Study II, we investigated the association between IBD and the risk of pancreatic cancer, by using population-based cohorts from Norway (1987-2015) and Sweden (1987-2016). Standardized incidence ratio (SIR) was calculated to estimate the relative risk. We found a 1.3-fold higher risk of pancreatic cancer in patients with IBD, compared with the general population. Furthermore, IBD patients with primary sclerosing cholangitis (PSC) had a pronounced increase of pancreatic cancer risk (SIR = 9.0, 95% CI 6.3-12.6), compared with IBD patients without PSC (SIR = 1.2, 95% CI 1.0-1.3). In Study III, patients with gastric biopsies from the Swedish histopathology registers during 1979-2011 were included and followed up until 2014. The gastric biopsies were categorized according to Correa’s cascade: normal mucosa, minor changes, superficial gastritis, and atrophic gastritis/intestinal metaplasia/dysplasia [AG/IM/Dys]). We found an extremely higher risk of pancreatic cancer across all groups shortly after undergoing gastric biopsies compared with the general population, which was very likely driven by reverse causality and confounding by indication. After the first three years of follow-up, SIRs dropped dramatically, with a 20%-30% increased risk across all groups of gastric biopsies. However, no significant excess risk was observed when comparing minor changes, superficial gastritis, and AG/IM/Dys with the normal gastric mucosa (HRs were closed to 1). In Study IV, we conducted a case-control study including 71 early PDAC cases and 72 healthy controls, based on 93 candidate plasma protein biomarkers, and three variables of sex, age, and smoking. An eight-protein panel was identified and used for prediction modeling for discriminating patients with early PDAC from healthy controls; the value of the area under the receiver-operating characteristic curve (AUC) was 0.85 (95% CI 0.78-0.91). This panel was further validated in a Spanish population of 37 early PDAC cases and 36 healthy controls; the AUC value was 0.81 (95% CI 0.70-0.92). In conclusion, this thesis found that individuals with poor dental health, diagnosis of IBD, and recent gastric biopsies were at an increased risk of pancreatic cancer, providing evidence to identify high-risk groups for further surveillance. The eight-protein panel might propose a promising approach to detect PDAC at an early stage. These findings need to be replicated in other populations and require a thorough evaluation on cost-effectiveness. Taken together, this thesis enriches our knowledge of the etiology of pancreatic cancer, and provides a promising panel for pancreatic cancer detection at an early stage