Oral Disease and Risk of Pancreatic Cancer

Abstract

To the Editor: Pancreatic cancer is characterized by a largely unknown etiology and poor survival, stressing the need to identify preventable risk factors. The few established risk factors (ie, tobacco smoking, chronic pancreatitis, diabetes, insulin resistance, and obesity) are only moderately or weakly associated with this tumor. A reported association between oral disease and pancreatic cancer remains to be confirmed in large and well-designed epidemiologic studies.1–5 Therefore, we conducted a large-scale case-control study, nested in a cohort comprising the nationwide Swedish Patient Register,6 during 1968–2008. Details of the study have been described elsewhere.7 All new cases of pancreatic cancer diagnosed during 1973–2008 with no prior cancer (except nonmelanoma skin cancer) were identified in the Swedish Cancer Register. Controls were frequency-matched using density-based sampling on the variables age, sex, calendar year, and year of the first record in the Patient Register. Hospitalization for a diagnosis of a defined disease of the oral cavity (ICD-8/9 codes 520–529 or ICD-10 codes K00–K14) since 1968, at least 5 years prior to inclusion, represented the study exposure. Estimates of risk of pancreatic cancer were calculated as odds ratios (ORs) with 95% confidence intervals (CIs), using unconditional logistic regression analysis with multivariable adjustment for confounding. Data on potential confounding factors were collected from diagnoses recorded in the Patient Register: alcoholism or alcohol-related disease; chronic obstructive pulmonary disease or bronchitis; obesity; diabetes; gastric or duodenal ulcer; and pancreatitis. The occurrence of disease of the oral cavity was slightly higher among the 29,218 pancreatic cancer cases than among the 99,992 control participants. The covariates pancreatitis, alcohol-related diseases, and diabetes were also overrepresented among cases compared with controls (data not shown). The adjusted odds ratio among patients with a history of oral disease was 1.2 (95% CI = 1.0–1.4) (Table).TABLE: Disease of the Oral Cavity and the Risk of Pancreatic Cancer in Sweden During the Period 1973–2008These data provide support for a modest association between diseases of subsequent oral cavity and the risk of pancreatic cancer. Strengths include the large sample size, complete nationwide coverage of hospitalizations for oral cavity disease and of all cancers, and adjustments for confounding. Moreover, the nationwide register-based design removes possible recall or selection bias. By considering only oral disease that occurred at least 5 years before inclusion, we also reduce bias from reverse causation. Among weaknesses, only the most severe forms of oral cavity disease are captured because the exposure is based only on in-hospital care. Tobacco use is the most consistent risk factor for pancreatic cancer, and a main risk factor for periodontal disease.8 The use of hospitalization for tobacco-related diseases (chronic obstructive pulmonary disease and bronchitis) as surrogate assessment of tobacco use is crude, and leaves open the possibility of residual confounding. However, adjustment for known tobacco-related diseases had no effect on the estimate (Table). The finding of an increased risk of pancreatic cancer after oral disease is supported by some previous research,3–5 whereas other studies have found no association.2 While more research is needed to establish an association, possible biologic mechanisms include an increased production of nitrosamines due to an altered oral flora or systemic inflammation. Rickard Ljung Upper Gastrointestinal Research Department of Molecular Medicine and Surgery Karolinska Institutet Stockholm, Sweden National Board of Health and Welfare Stockholm, Sweden [email protected] Tomas Sjöberg Bexelius Lena Martin Upper Gastrointestinal Research Department of Molecular Medicine and Surgery Karolinska Institutet Stockholm, Sweden Jesper Lagergren Upper Gastrointestinal Research Department of Molecular Medicine and Surgery Karolinska Institutet Stockholm, Sweden Division of Cancer Studies King's College London, United Kingdom