Aspirin Use and Pancreatic Cancer Mortality in a Large United States Cohort

Transcription

1 Aspirin Use and Pancreatic Cancer Mortality in a Large United States Cohort Eric J. Jacobs, Cari J. Connell, Carmen Rodriguez, Alpa V. Patel, Eugenia E. Calle, Michael J. Thun Background: Results from some epidemiologic studies have suggested that aspirin use may reduce risk of pancreatic cancer, but the evidence remains limited. Methods: We examined the association between aspirin use and pancreatic cancer mortality among U.S. adults in the Cancer Prevention Study II (CPS-II) cohort. CPS-II completed a self-administered questionnaire in 1982 and were followed for mortality through During follow-up, there were 4577 from pancreatic cancer (2434 in men and 2143 in women). We calculated rate ratios (RR) adjusted for age, sex, race, body mass index, diabetes, and cigarette smoking status. Results: was not associated with pancreatic cancer mortality. The rate ratio associated with aspirin use 30 or more times, compared with no use, was 0.97 (95% confidence interval [CI] 0.86 to 1.09). Even who reported both frequent aspirin use (>30 times ) and use for 20 or more years were not at decreased risk compared with nonusers (RR 0.96, 95% CI 0.69 to 1.33). We found no association between aspirin use and pancreatic cancer mortality in subgroup analyses by follow-up time, cigarette smoking status, or sex. Conclusion: Results from this large prospective study do not support an important effect of aspirin use on pancreatic cancer mortality. [J Natl Cancer Inst 2004;96:524 8] has been consistently associated with reduced risk of colon cancer in observational epidemiologic studies (1) and has recently been shown to reduce risk of colorectal polyp recurrence in two randomized trials (2,3). has also been associated with reduced risk of stomach and esophageal cancer in most epidemiologic studies (4). Inverse associations between aspirin use and a variety of other cancers have also been reported, although less consistently (4). A protective effect of aspirin use on pancreatic cancer is biologically plausible. Several different nonsteroidal antiinflammatory drugs (NSAIDs) have been shown to inhibit pancreatic cancer in a hamster model (5,6), and both sodium salicylate (7) and other NSAIDs (8) have been shown to inhibit the growth of human pancreatic cancer cell lines. To our knowledge, five epidemiologic studies (9 13) have specifically examined the association between aspirin use and pancreatic cancer risk. In the Iowa Women s Health Study cohort (including 80 cases of pancreatic cancer), frequent aspirin use was associated with substantially reduced risk (rate ratio [RR] 0.40, 95% confidence interval [CI] 0.20 to 0.82, for use 6 times per week compared with no use) (9). However, use of nonaspirin NSAIDs was not associated with pancreatic cancer risk (RR 1.28, 95% CI 0.68 to 2.43, for use 6 times per week compared with no use). In contrast, a recent analysis from the Nurse s Health Study cohort (which included 161 cases) found that regular aspirin use ( 2 tablets per week) for 20 or more years was associated with increased risk of pancreatic cancer (RR 1.58, 95% CI 1.03 to 2.43) (10). No association between aspirin use and pancreatic cancer risk was found in a U.S. hospital-based case control study (11) that included 194 cases or in a small U.S. cohort (12) that included 30 cases. A Danish study (13) that used pharmacy database information found no association between low-dose aspirin prescriptions and pancreatic cancer risk but could not examine standard aspirin dosages. Three additional studies (14 16) have examined the association between overall NSAID use and pancreatic cancer but did not examine aspirin use specifically. In a U.S. hospital-based case control study (14), there was a suggestion of decreased risk with regular continuing NSAID use for 5 or more years (odds ratio [OR] 0.6, 95% CI 0.4 to 1.1). In a Danish study (16) that used pharmacy database information, receipt of 10 or more nonaspirin NSAID prescriptions was not associated with pancreatic cancer risk (RR 0.9, 95% CI 0.6 to 1.3). In a case control study (15) in the United Kingdom that used pharmacy database information, receipt of seven or more NSAID prescriptions during the 13- to 36-month interval before the date of diagnosis was associated with an increased risk of pancreatic cancer (RR 1.49, 95% CI 1.02 to 2.18). However, NSAID use before this interval was not examined, and the increased risk associated with recent NSAID use may have been a result of NSAID use to relieve pain from undiagnosed cancer. We examined the association between aspirin use and pancreatic cancer mortality in the Cancer Prevention Study II (CPS- II), a large cohort of U.S. adults. In this cohort, aspirin use has been previously demonstrated to be associated with reduced risk of colon, stomach, and esophageal cancer mortality (17,18). SUBJECTS AND METHODS Study Cohort and Follow-up Subjects in this analysis were drawn from the ( men and women) in the CPS-II. Participants were enrolled in 1982 by American Cancer Society (ACS) volunteers in all 50 U.S. states, the District of Columbia, and Puerto Rico, as previously described (19). Participants completed a four-page baseline self-administered questionnaire in 1982 that included information on demographic characteristics Affiliation of authors: Department of Epidemiology and Surveillance Research, American Cancer Society, Atlanta, GA. Correspondence to: Eric J. Jacobs, PhD, Department of Epidemiology and Surveillance Research, American Cancer Society, National Home Office, 1599 Clifton Rd. NE, Atlanta, GA ( eric.jacobs@cancer.org). See Note following References. DOI: /jnci/djh084 Journal of the National Cancer Institute, Vol. 96, No. 7, Oxford University Press 2004, all rights reserved. 524 ARTICLES Journal of the National Cancer Institute, Vol. 96, No. 7, April 7, 2004

2 and various behavioral, environmental, occupational, and dietary factors. The vital status of study was determined through December 31, 2000, using two approaches. ACS volunteers made personal inquiries in September 1984, September 1986, and September 1988 to determine whether the they had enrolled were alive or dead and to record the date and place of all. Reported were then verified by obtaining death certificates (20). At completion of the 1988 follow-up obtained from ACS volunteer reports, vital status was known for 98.2% of the cohort (20). Automated linkage using the National Death Index then extended follow-up of the entire cohort from September 1988 through December 31, 2000, and also identified among the lost to follow-up between 1982 and 1988 (20). At the completion of follow-up in December 2000, (28.1%) had died, (71.7%) were alive, and 2896 (0.2%) had follow-up truncated on September 1, 1988, because of insufficient data to perform linkage with the National Death Index. Death certificates or codes for cause of death have been obtained for 98.7% of all known. The underlying cause of death was coded according to the International Classification of Diseases, 9 th and 10 th revisions (ICD-9 and ICD-10) (21). Pancreatic cancer were defined as ICD-9 codes and ICD-10 codes C25.0 C25.9. All aspects of the CPS-II study have been approved by the Emory University School of Medicine Human Investigations Committee. Informed consent to participate was implied by the return of a completed self-administered questionnaire in All analyses excluded who at enrollment reported a history of cancer other than nonmelanoma skin cancer (n ) or who had incomplete or uninterpretable data on aspirin use (n ) or cigarette smoking status (n ). A total of ( men and women) remained for analysis, of whom 4577 died from pancreatic cancer. We also performed a subgroup analysis in the CPS-II Nutrition Cohort, a subset of the larger CPS-II cohort. The CPS-II Nutrition Cohort (hereafter referred to simply as the Nutrition Cohort) consists of CPS-II living in 21 states who in returned a mailed self-administered enrollment questionnaire providing updated or new information about nutrition, aspirin use, cigarette smoking status, and other health-related factors (22). In the Nutrition Cohort subgroup analysis, we were able to examine the association between consistent aspirin use at two time points (1982 and ) and pancreatic cancer mortality. Of the CPS-II included in the main analyses, were in the Nutrition Cohort. Of these, were excluded because they provided incomplete aspirin use information or reported a history of cancer on the questionnaire. The Nutrition Cohort subgroup analysis included the remaining ( men and women), of whom 311 died from pancreatic cancer. Ascertainment of Aspirin Use at enrollment in the CPS-II cohort was ascertained from the 1982 questionnaire, which included a section about use of medications and vitamin supplements. Participants were asked to fill in two boxes, the first box reporting the number of times in the last month they used aspirin and the second box reporting the number of years of use. Participants were instructed to write 1 2 in the times box if they used aspirin only occasionally. No information was collected on aspirin dose or on aspirin use that had stopped before study enrollment. No information was collected on use of NSAIDs other than aspirin. For the Nutrition Cohort subgroup analysis, aspirin use was also ascertained from the Nutrition Cohort enrollment questionnaire, which asked if they took aspirin regularly and, if so, for how many days per month. The Nutrition Cohort enrollment questionnaire also included similar questions on the use of ibuprofen and other NSAIDs. Statistical Analysis We used Cox proportional hazards modeling (23) to calculate rate ratios for pancreatic cancer mortality associated with aspirin use, while adjusting for other potential risk factors. The time axis used was follow-up time since enrollment in We found no evidence that the proportional hazards assumption was violated when we modeled interaction terms between measures of aspirin use and follow-up time. We categorized aspirin use as unquantified occasional, 1 14 times, times, or 30 or more times. We selected these categories because use 30 or more times could include daily use of aspirin, while use 15 or more times could include use every other day. All Cox models were adjusted for age, sex, and several additional factors associated with risk of pancreatic cancer in the CPS-II cohort and other study populations (race [white, black, other/unknown], cigarette smoking status [never, former, current], body mass index [weight in kg/height in m 2 ], and diabetes [yes, no]). All covariates except age were modeled as dummy variables using the categories shown in Table 1. We adjusted for age by stratifying on exact year of age at enrollment within each Cox model (24). For the Nutrition Cohort subgroup analysis, updated information reported in was used to adjust for body mass index, cigarette smoking status, and diabetes. Body mass index and cigarette smoking status were defined using information reported in 1982 when information on these variables was incomplete or uninterpretable on the questionnaire. We further examined potential confounding by including a more detailed cigarette smoking variable incorporating number of pack-years smoked, family history of pancreatic cancer, physical activity, and history of gallstones, cholecystectomy, heart disease, and arthritis. However, we did not adjust for these factors in the final models because such adjustment had negligible effects on our results (data not shown). We examined whether the association between aspirin use and pancreatic cancer mortality varied by cigarette smoking status, body mass index, and diabetes by modeling multiplicative interaction terms between frequency of aspirin use (continuous) and variables for cigarette smoking status (never, former, current), diabetes (yes, no), and body mass index ( 25 kg/m 2, 25 kg/m 2 ). Two-sided P values for interaction were calculated using the likelihood ratio statistic (25). Journal of the National Cancer Institute, Vol. 96, No. 7, April 7, 2004 ARTICLES 525

3 Table 1. Demographics and potential pancreatic cancer risk factors in Cancer Prevention Study II by aspirin use at enrollment* Men Women No aspirin use (n ) (n ) (n ) No aspirin use (n ) (n ) (n ) Age, y Race White Black Other/unknown Body mass index, kg/m Unknown Diabetes Cigarette smoking status Never Former Current *As reported at enrollment in Percentages were adjusted to the age distribution of the entire study population. RESULTS Table 1 compares reporting no aspirin use, aspirin use fewer than 30 times, or use 30 or more times with respect to several factors known to be associated with risk of pancreatic cancer. Most were white and middle-aged or elderly, regardless of aspirin use. Compared with nonusers and less frequent aspirin users ( ), frequent aspirin users ( ) were somewhat older and slightly more likely to be white. Frequent aspirin users did not differ substantially from nonusers or less frequent aspirin users with respect to cigarette smoking status, body mass index, or diabetes., even frequent use, was not associated with pancreatic cancer mortality (for use compared with no use, RR 0.97, 95% CI 0.86 to 1.09) (Table 2). Results were similar for men and for women, and all further analyses are presented for men and women combined. We found no apparent differences in the association between frequency of aspirin use and pancreatic cancer mortality by body mass index, diabetes, or cigarette smoking status (results not shown). There was no association between aspirin use and pancreatic cancer mortality, even for who reported both frequent aspirin use ( per month) and use for 20 or more years (RR 0.96, 95% CI 0.69 to 1.33) (Table 3). Because the aspirin use pattern of some undoubtedly changed over the 18-year follow-up period, we hypothesized that any association between recent aspirin use and pancreatic cancer mortality might be more apparent during the first few years of follow-up. Therefore, we analyzed the association between aspirin use and pancreatic cancer mortality with follow-up stratified into three approximately 6-year intervals (Table 4). However, we found no association between aspirin use and pancreatic cancer mortality during any of the follow-up intervals. We had no updated information on aspirin use during the approximately 18 years of follow-up in the overall CPS-II co- Table 2. Association of frequency of aspirin use with pancreatic cancer mortality in the Cancer Prevention Study II, *, times Men Women Men and women No use (referent) (referent) (referent) Occasional use (0.92 to 1.10) (0.87 to 1.06) (0.92 to 1.05) (0.76 to 0.97) (0.89 to 1.15) (0.85 to 1.01) (0.75 to 1.24) (0.76 to 1.27) (0.81 to 1.16) (0.83 to 1.15) (0.80 to 1.14) (0.86 to 1.09) 526 ARTICLES Journal of the National Cancer Institute, Vol. 96, No. 7, April 7, 2004

4 Table 3. Association of frequency and duration of aspirin use with pancreatic cancer mortality in the Cancer Prevention Study II, * Duration of aspirin use, times per month 10 years years 20 years No use (referent) (referent) (referent) Occasional use (1.02 to 1.53) (0.89 to 1.32) (0.83 to 1.02) (0.82 to 1.16) (0.77 to 1.19) (0.79 to 1.04) (0.87 to 1.49) (0.52 to 1.27) (0.63 to 1.27) (0.81 to 1.10) (0.76 to 1.30) (0.69 to 1.33) *Excludes with unknown duration of aspirin use. Rate ratios (RRs) were adjusted for age, sex, race, smoking status, body mass index, and diabetes. CI confidence interval. The referent group for all duration categories is all nonusers of aspirin. Table 4. Association of frequency of aspirin use with pancreatic cancer mortality by follow-up period in the Cancer Prevention Study II, *, times per month Follow-up period, Follow-up period, Follow-up period, hort. However, we could evaluate changes in aspirin use during follow-up among who completed a second questionnaire in at enrollment in the Nutrition Cohort (n ). In this subgroup, 60% of frequent aspirin users (use ) in 1982 took aspirin at least 15 days per month in In comparison, 21% of who reported no aspirin use in 1982 took aspirin at least 15 days per month in Table 5 shows the association between pancreatic cancer mortality and aspirin use among in the CPS-II Nutrition Cohort. We observed no association between aspirin use and pancreatic cancer mortality, even among who appeared to be consistent long-term aspirin users, that is, those who reported aspirin use 15 or more times in both 1982 and in (RR 1.27, 95% CI 0.78 to 2.08). There was also no association between frequency of nonaspirin NSAID use and pancreatic cancer mortality in the Nutrition Cohort (RR 1.08, 95% CI 0.75 to 1.55 for use of nonaspirin NSAIDs 15 times compared with no use) (data not shown in table). DISCUSSION No use (referent) (referent) (referent) Occasional use (0.83 to 1.09) (0.84 to 1.06) (0.93 to 1.15) (0.92 to 1.28) (0.71 to 0.97) (0.80 to 1.05) (0.59 to 1.28) (0.59 to 1.14) (0.90 to 1.51) (0.84 to 1.30) (0.74 to 1.11) (0.79 to 1.18) We found no suggestion of any association between aspirin use and pancreatic cancer mortality in this large prospective study, despite being able to examine both frequent and longduration use. This result differs somewhat from the decreased risk associated with frequent aspirin use in the Iowa Women s Health Study (9) and the increased risk associated with longduration aspirin use in the Nurse s Health Study (10). We know of no clear reason, other than chance, for these differing results. A limitation of this study is that we could examine only frequency of aspirin use, rather than daily dose, because we did not have information on the dose or number of aspirin tablets taken each time aspirin was used. In addition, the amount of measurement error for self-reported aspirin use is not known. Measurement error would be expected to bias the results toward the null, potentially obscuring a true increase or decrease in risk of mortality from pancreatic cancer associated with aspirin use. However, frequency of aspirin use has been shown to be associated with substantially decreased risk of mortality from colon Table 5. Association of frequency of aspirin use at two time points with pancreatic cancer mortality in the Cancer Prevention Study II Nutrition Cohort, * No use (referent) Inconsistent or infrequent use (0.79 to 1.36) Use 15 times in both 1982 and (0.78 to 2.08) No use reported in either 1982 or Some aspirin use reported in either 1982 or , but not regular use ( 15 times ) at both time points. Journal of the National Cancer Institute, Vol. 96, No. 7, April 7, 2004 ARTICLES 527

5 cancer, esophageal cancer, and stomach cancer in this cohort (17,18), consistent with results from several other studies (4). These results suggest that any similarly important effect of aspirin use on pancreatic cancer mortality could have been detected in this cohort. An additional limitation of our study is that information on aspirin use was not updated during the 18-year follow-up period in the main CPS-II cohort, and a substantial proportion of the cohort may have either started or stopped regular aspirin use at some point during follow-up. However, there was no suggestion of an effect of aspirin in analyses restricted to the first 6 years of follow-up. In addition, we observed no association between consistent use of aspirin 15 or more times and pancreatic cancer mortality in analyses that incorporated updated aspirin use information in the Nutrition Cohort. However, this subgroup was too small to allow us to examine the risk associated with consistent daily use of aspirin. It should be noted that this analysis focused on aspirin use and does not rule out effects of nonaspirin NSAIDs, although no association with nonaspirin NSAIDs was seen in the Nutrition Cohort. In addition, this analysis examined pancreatic cancer mortality rather than pancreatic cancer incidence. However, our results are likely to be generalizable to pancreatic cancer incidence because pancreatic cancer is usually fatal (5-year survival 5%) (26). An important strength of this analysis is its unusually large size, resulting in risk estimates with narrow confidence intervals that excluded a strong association between aspirin use and pancreatic cancer mortality. Our results therefore provide evidence against an important effect of aspirin use on risk of pancreatic cancer. REFERENCES (1) Thun MJ, Henley SJ, Patrono C. Nonsteroidal anti-inflammatory drugs as anticancer agents: mechanistic, pharmacologic, and clinical issues. J Natl Cancer Inst 2002;94: (2) Baron JA, Cole BF, Sandler RS, Haile RW, Ahnen D, Bresalier R, et al. A randomized trial of aspirin to prevent colorectal adenomas. N Engl J Med 2003;348: (3) Sandler RS, Halabi S, Baron JA, Budinger S, Paskett E, Keresztes R, et al. A randomized trial of aspirin to prevent colorectal adenomas in patients with previous colorectal cancer. N Engl J Med 2003;348: (4) Baron JA. Epidemiology of non-steroidal anti-inflammatory drugs and cancer. Prog Exp Tumor Res 2003;37:1 24. (5) Takahashi M, Furukawa F, Toyoda K, Sato H, Hasegawa R, Imaida K, et al. Effects of various prostaglandin synthesis inhibitors on pancreatic carcinogenesis in hamsters after initiation with N-nitrosobis(2-oxopropyl)amine. Carcinogenesis 1990;11: (6) Schuller HM, Zhang L, Weddle DL, Castonguay A, Walker K, Miller MS. The cyclooxygenase inhibitor ibuprofen and the FLAP inhibitor MK886 inhibit pancreatic carcinogenesis induced in hamsters by transplacental exposure to ethanol and the tobacco carcinogen NNK. J Cancer Res Clin Oncol 2002;128: (7) Perugini RA, McDade TP, Vittimberga FJ, Duffy AJ, Callery MP. Sodium salicylate inhibits proliferation and induces G1 cell cycle arrest in human pancreatic cancer cell lines. J Gastrointest Surg 2000;4: (8) Yip-Schneider MT, Sweeney CJ, Jung SH, Crowell PL, Marshall MS. Cell cycle effects of nonsteroidal anti-inflammatory drugs and enhanced growth inhibition in combination with gemcitabine in pancreatic carcinoma cells. J Pharmacol Exp Ther 2001;298: (9) Anderson KE, Johnson TW, Lazovich D, Folsom AR. Association between nonsteroidal anti-inflammatory drug use and the incidence of pancreatic cancer. J Natl Cancer Inst 2002;94: (10) Schernhammer E, Kang JH, Chan AT, Michaud DS, Skinner HS, Giovannucci E, et al. A prospective study of aspirin use and risk of pancreatic cancer in women. J Natl Cancer Inst 2004;96:22 8. (11) Menezes RJ, Huber KR, Mahoney MC, Moysich KB. Regular use of aspirin and pancreatic cancer risk. BMC Public Health 2002;2:18. (12) Schreinemachers DM, Everson RB. and lung, colon, and breast cancer incidence in a prospective study. Epidemiology 1994;5: (13) Friis S, Sorensen HT, McLaughlin JK, Johnsen SP, Blot WJ, Olsen JH. A population-based cohort study of the risk of colorectal and other cancers among users of low-dose aspirin. Br J Cancer 2003;88: (14) Coogan PF, Rosenberg L, Palmer JR, Strom BL, Zauber AG, Stolley PD, et al. Nonsteroidal anti-inflammatory drugs and risk of digestive cancers at sites other than the large bowel. Cancer Epidemiol Biomarkers Prev 2000; 9: (15) Langman MJ, Cheng KK, Gilman EA, Lancashire RJ. Effect of antiinflammatory drugs on overall risk of common cancer: case-control study in general practice research database. BMJ 2000;320: (16) Sorensen HT, Friis S, Norgard B, Blot WJ, McLaughlin JK, Ekbom A, et al. Risk of cancer in a large cohort of nonaspirin NSAID users: a population-based study. Br J Cancer 2003;88: (17) Thun MJ, Namboodiri MM, Heath CW. and reduced risk of fatal colon cancer. N Engl J Med 1991;325: (18) Thun MJ, Namboodiri MM, Calle EE, Flanders WD, Heath CW. Aspirin use and risk of fatal cancer. Cancer Res 1993;53: (19) Stellman SD, Garfinkel L. Smoking habits and tar levels in a new American Cancer Society prospective study of 1.2 million men and women. J Natl Cancer Inst 1986;76: (20) Calle EE, Terrell DD. Utility of the national death index for ascertainment of mortality among Cancer Prevention Study II. Am J Epidemiol 1993;137: (21) World Health Organization. International classification of diseases: manual of the international statistical classification of diseases, injuries, and causes of death. Vol 1. 9th rev. Geneva (Switzerland): World Health Organization; (22) Calle EE, Rodriguez C, Jacobs EJ, Almon ML, Chao A, McCullough ML, et al. The American Cancer Society Cancer Prevention Study II Nutrition Cohort: rationale, study design, and baseline characteristics. Cancer 2002; 94: (23) Cox DR. Regression models and life tables. J R Stat Soc Series B 1972; 34: (24) Kleinbaum DG. Survival analysis: a self-learning text. New York (NY): Springer-Verlag; p (25) Hosmer DW, Lemeshow S. Applied logistic regression. New York (NY): John Wiley and Sons; p (26) Ries LA, Eisner MP, Kosary CL, Hankey BF, Miller BA, Clegg L, et al., editors. SEER Cancer Statistics Review, , National Cancer Institute. Bethesda (MD); Available at: csr/1973_1999/. NOTE Manuscript received September 11, 2003; revised January 30, 2004; accepted February 9, ARTICLES Journal of the National Cancer Institute, Vol. 96, No. 7, April 7, 2004