No Association Between Nonsteroidal Anti-inflammatory Drug Use and Pancreatic Cancer Incidence and Survival
Pancreatic cancer is a very fatal disease with few known risk factors. Currently approved drugs confer modest survival advantage.1 Pancreatic cancer is associated with chronic inflammation, and preclinical studies have demonstrated that cyclooxygenase 2 expression is elevated in pancreatic cancer tissue compared with normal tissue, raising the possibility that cyclooxygenase 2 inhibition can reduce pancreatic cancer risk and mortality.2 Nevertheless, the association of anti-inflammatory drugs with pancreatic cancer risk are controversial and association with survival even less well understood.3 A recent meta-analysis of mainly case–control studies reported that aspirin but not nonaspirin nonsteroidal anti-inflammatory drugs (NSAID) use is associated with reduced pancreatic cancer risk.3 Only 2 studies have evaluated the association of NSAID use with pancreatic cancer mortality and neither observed any protective effect. Our objective in this study is to investigate the association of NSAID use with pancreatic cancer risk and survival in a large prospective study.
The study population consisted of participants in The NIH-AARP Diet and Health study (N = 334,907) that completed a questionnaire on NSAID use.4 In brief, the NIH-AARP study included men and women who were 50 to 71 years old AARP members residing in 6 states (California, Florida, Louisiana, New Jersey, North Carolina, and Pennsylvania) and 2 metropolitan areas (Atlanta, Georgia; and Detroit, Michigan) in 1995–1996. After excluding individuals who reported a history of cancer at baseline, had questionnaires filled by proxies, or lacked data on NSAID, our study subcohort consisted of 297,525 participants of which 1048 developed exocrine pancreatic adenocarcinoma (PDAC) after study enrollment. Pancreatic cancer was defined as an incident case of PDAC (International Classification of Diseases for Oncology, Third Edition [code C250–C259]). Cancer incidence was ascertained through linkage to state cancer registries.5 Almost 90% of cancer cases in the NIH-AARP cohort were accurately captured by the cancer registries.4 Mortality was ascertained by linkage to the National Death Index Plus. Frequency of NSAID use was categorized into the following 4 groups: no use, monthly, weekly, and daily.
Cox proportional hazards regression model adjusted for confounders was used to investigate the associations of NSAID use with pancreatic cancer risk and survival. The proportional hazards assumption was tested and satisfied through the use of time-dependent covariate method. Age, smoking, diabetes status, and body mass index were included in the multivariable analyses for incidence analyses. Follow-up time for incidence began from the date of questionnaire receipt to pancreatic cancer diagnosis or December 31, 2006. Survival among PDAC cases (N = 1048) was calculated from the day of pancreatic cancer diagnosis to the day of death or December 2011, whichever came first. In addition to the adjusted confounding variables for incidence analyses, cancer stage was also included in the survival analyses. Hazard ratios (HRs) and 95% confidence intervals (95% CI) were calculated by categories of NSAID use. We conducted P-trend tests by coding NSAID use as ordinal and treating it as linear in Cox models.
We observed no statistically significant associations between NSAID use and pancreatic cancer incidence or survival (Table 1). The HRs for pancreatic cancer incidence associated with daily use of aspirin and nonaspirin NSAID, compared with no use, were 1.07 (95% CI, 0.90–1.27, P trend = 0.70) and 0.87 (95% CI, 0.70–1.09, P trend = 0.26), respectively. The HRs for pancreatic cancer survival associated with daily use of aspirin and nonaspirin NSAID were 1.03 (95% CI, 0.86–1.23, P trend = 0.54) and 0.99 (95% CI, 0.79–1.26, P trend = 0.55). We observed similar null results in analyses stratified by smoking status, history of diabetes, and sex, although aspirin seemed to be associated with longer survival among women (HR, 0.81; 95% CI, 0.59–1.11).