Aspirin Raised Pancreatic Ca Risk in Large Nurses’ Study

PHOENIX, Arizona—Chronic regular aspirin use may dramatically increase women’s risk of pancreatic cancer, according to results from a prospective study of 88,378 participants in the Nurses’ Health Study, one of the largest US studies of major risk factors for chronic disease.

The women, all nurses who did not initially have cancer, were surveyed regarding aspirin use beginning in 1980, and were followed for 18 years. During this period, 161 pancreatic cancer cases occurred. The researchers found that pancreatic cancer risk was 58% greater in women who had used aspirin regularly for 20 or more years (two or more standard tablets a week) and 86% greater in those who took 14 or more aspirin tablets a week.

While chronic aspirin use appears to be associated with increased risk, the researchers emphasized that their findings need to be investigated further, and they do not advocate that women stop taking aspirin altogether.

The findings are surprising given that aspirin has been demonstrated to be effective in lowering the risk of colorectal cancer, and other research has suggested its promise in the prevention of breast and even pancreatic cancer. The investigators, from Harvard School of Public Health, the Dana-Farber Cancer Institute, Channing Laboratory, and the National Cancer Institute, found no significant association with current use of aspirin, but concluded their data suggest that a longer duration of aspirin use may increase the risk of pancreatic cancer.

Eva Schernhammer, MD, DrPH, an epidemiologist at Harvard School of Public Health, presented the group’s report at the American Association for Cancer Research (AACR) Second Annual Conference on Frontiers in Cancer Prevention Research (abstract 1321).

Pancreatic cancer is "rapidly fatal" and understudied, Dr. Schernhammer said, "but apart from smoking, only a few risk factors have been identified." Large studies of chemopreventive agents against pancreatic cancer are also lacking.

The Nurses’ Health Study was begun in 1976 at Brigham and Women’s Hospital by Frank Speizer, MD, professor of environmental science, Harvard School of Public Health, with funding from the National Institutes of Health, to investigate potential long-term consequences of oral contraceptive use.

The questionnaire-based study targeted married female registered nurses aged 35 to 55 years in 1976 and living in the 11 most populous states (California, Connecticut, Florida, Maryland, Massachusetts, Michigan, New Jersey, New York, Ohio, Pennsylvania, and Texas).

A total of 121,700 women responded to the baseline questionnaire and were enrolled, responding every 2 years to follow-up questionnaires about all major risk factors for chronic diseases such as cardiovascular disease and cancer, lifestyle factors, and health-related topics such as smoking, hormone use, and menopausal status. Over the years, additional topics have been addressed, such as diet and quality of life. Biomarkers are also used to assess participants’ risk for a variety of chronic diseases.

In 1980, Dr. Schernhammer said, a question on frequency of aspirin use was added. There were also questions about duration of aspirin use prior to 1980, as well as on other nonsteroidal anti-inflammatory drugs (NSAIDs), but the data on NSAIDs were not as complete as on aspirin use. Since 1980, aspirin use has been assessed every 2 years, with a 90% questionnaire response rate.

Of the 161 pancreatic cancer patients, 34 had taken two or more aspirin tablets weekly for 20 years or more, and 20 had taken 14 or more tablets a week. "We found no significant association with current aspirin use," she said, "but over long periods of time, the more aspirin a woman took, the higher was her risk."

Current regular aspirin use, which the investigators defined as taking two tablets per week, was associated with a nonsignificant relative risk (RR) of 1.20 (95% CI 0.87 to 1.65). However, regular aspirin use over long periods of time was associated with a significant increase in pancreatic cancer risk: Women with 20 or more years of regular aspirin use had a 58% greater risk of developing pancreatic cancer (RR 1.58; 95% CI, 1.03 to 2.43, P for trend = .01).

Among women who reported aspirin use on at least two of three consecutive biennial questionnaires, compared with consistent nonusers of aspirin, the risk of pancreatic cancer rose with increased use (P for trend = .02) (see Table).

In 1990, Dr. Schernhammer said, as a validation study, 100 of the women who reported taking aspirin were surveyed as to why. "Especially with the higher dosages, the major reason was for headache or muscular or skeletal pain," she said. "About 8% to 10% were taking it for prevention of cardiovascular disease."

Implications of the Findings

"The results were highly unexpected and have very important implications," said Raymond N. DuBois, MD, PhD, chair of the AACR conference and director, Division of Gastroenterology, Vanderbilt University Medical Center. However, he concurred with the investigators that the results are preliminary: "The information that we obtain from well-designed large cohort studies of people looking at what they actually take—aspirin or other agents—provides the basis for hypothesis generation to take to the next level of randomized, controlled clinical trials."

He stressed that the cautionary finding needs to be put into the context of what else we know about aspirin and how it works in pancreatic cancer prevention. "The vast majority of studies of aspirin in cancer have found either no effect or a protective effect," Dr. DuBois said. "One of the reasons that large, well-designed epidemiological studies like this . . . are not definitive is that it’s very difficult to control for the reason people have certain habits, such as taking aspirin, and it could be the reason that people are taking aspirin that is posing the risk."

Dr. Schernhammer added, "Clearly, the underlying biology for pancreatic cancer genesis isn’t settled." In addition to the current findings , she said, "there have been a couple of case studies and observational studies that showed aspirin use was associated with pancreatitis, which we know is associated with an increased risk of pancreatic cancer."

Further, aspirin and other NSAIDs block lipoxygenase (LOX), Dr. Schernhammer explained, "and there is anti- and procarcinogenic LOX; there have been studies showing that in pancreatic cancer tissue, 5-LOX is overexpressed, which is the procarcinogenic LOX." Noting that, similar to other recent research, the Nurses’ Health Study found an inverse association between aspirin use and colorectal cancer, she said, "it may well be that while aspirin protects in some tissues, it causes damage in others."