IJE Advance Access originally published online on October 5, 2007
International Journal of Epidemiology 2007 36(5):960-962; doi:10.1093/ije/dym204
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Commentary: Preventing colorectal cancer with aspirin—what next?
Division of Epidemiology and Public Health, Queens Medical Centre, University of Nottingham, Nottingham NG7 2UH, UK.
E-mail: richard.logan{at}nottingham.ac.uk
Accepted 20 August 2007
Aspirin was first synthesized in 1897 and marketed 2 years later by Bayer, at that time a small German dyestuffs manufacturer.1 It is a simple drug consisting of only a benzene ring with an acetyl group and a carboxylic acid group attached at positions 1 and 2 on the ring. Over the past 100 years aspirin has had enormous usage throughout the world currently equivalent to 100 billion standard aspirin tablets a year. One might have imagined therefore that all important effects of taking aspirin would now be well established. So it is salutary that 2007 should see a report from two large randomized trials establishing that regular aspirin use will halve the incidence of colorectal cancer albeit 10 years later.2 What lessons, if any, are there from how this remarkable discovery was made?
The epidemiological trail appears to start with the paper by Kune et al.3 reproduced in this issue of IJE. This paper was one of a score that arose from the Melbourne Colorectal Cancer study for which data was collected in 1980–1981. At that time there had been few large case–control studies devoted to colorectal cancer and with over 700 cases, the Melbourne study was one of the largest. Other strengths were that the case series was population based and the controls were randomly selected from the community. Controls were identified by interviewers’ cold calling at selected dwellings and although not mentioned in this paper the response rate was 71%.4 The authors did not attempt to conceal the lack of specific hypotheses in relation to many of the exposures examined, which arguably was the major limitation of their paper. Indeed today it might seem surprising that specific questions on aspirin use were included. However, at this time regular use of aspirin and others over the counter analgesics was particularly prevalent in Australia with numerous papers on analgesic nephropathy resulting from the work of the nephrologist Priscilla Kincaid Smith in Melbourne. What is evident is that there was no particular hypothesis related to colorectal cancer and aspirin. While the authors appear to have collected some data on the frequency of aspirin use (daily or weekly) and on duration of use, in their results aspirin use is simply divided into use and non-use. Nevertheless, the inverse association they found was present in both sexes, evident for both colon and rectum and remained as strong in a multivariate analysis that adjusted for diet and other conditions including arthritis.
Striking the right balance in reporting a completely unexpected finding is a challenge. Give it too much prominence and one risks antagonising a journal's referees, weakening one's paper and even one's reputation. Give it too little and risk what might prove to be a seminal observation being overlooked. Kune and colleagues opted for mentioning the association with aspirin use as one of half a dozen positive findings in the paper's Abstract. Readers can make their own judgment whether this gives the inverse association with aspirin use sufficient prominence. Perhaps they would have given it greater prominence if they had been aware of the experimental work showing aspirin and other non-steroidal anti-inflammatories (NSAIDS) could inhibit the growth of colon tumours in animal models. Nonetheless the statement in the Discussion that ‘whatever the mechanism the finding has potential significance in colorectal cancer chemoprevention and merits early confirmation’ shows they did not underestimate its potential importance.
Early confirmation came 3 years later in 1991 with first the paper by Rosenberg et al.and then that by Thun et al.later the same year.5,6 Ironically the stimulus for the Rosenberg's study appears to have been the experimental animal studies rather than the Australian study which was not cited. While Rosenberg et al.'s paper appeared to confirm the findings of Kune and his colleagues regular aspirin use (use on 4 days a week for at least 3 months) appears to have been less common than in Melbourne and the association disappeared if aspirin or other NSAID use was discontinued a year or more before interview. Furthermore, there was no association evident at all if use was deemed to be not regular.
In retrospect it is difficult to judge how much attention the paper by Rosenberg et al. attracted particularly as it was so soon followed by the Thun's paper.6 Undoubtedly it was this paper appearing in the New England Journal of Medicine which really grabbed people's attention and ignited the interest of researchers. The paper was based on data collected from two-thirds of a million participants enrolled across the United States in a prospective mortality study for the American Cancer Society. Thus a major strength of this study was its sheer size which ensured that even at the highest level of aspirin use (more than 16 standard tablets per month) there were substantial numbers of users. Conversely, it was limited by being restricted to mortality as the outcome. Thus conceivably the reduction in risk of fatal colon cancer in aspirin users might have been the result of an effect of aspirin on survival with colon cancer or might have been because aspirin use led in some way to earlier diagnosis of colon cancer. Despite these criticisms and some others made in an accompanying editorial Thun et al.'s results were impressive with a convincing and highly statistically significant dose–response relationship for both sexes and across various subgroups. In addition, the reduction in colon cancer risk of about 40–50% for the people taking 16 or more standard aspirin tablets a month was very similar to that reported in the two earlier studies.
Not surprisingly these papers created enormous interest right across the research spectrum. There has been a flood of further observational studies. At the last count there were 20 separate case–control studies and 12 cohort studies examining aspirin and NSAID use and colorectal cancer with the vast majority finding some reduction in cancer risk.2 Much of the key information has come from the Nurses Health Study cohort which suggested early, on that aspirin taken in the low doses, used for cardiovascular disease was unlikely to reduce cancer risk and also that aspirin use for 10 years or more might be necessary.7,8 These observations almost certainly explain why the randomised trials with low-dose aspirin (<100 mg/day) have failed to find any reduction in incidence of colorectal cancer in the aspirin takers.9,10
For the basic scientists interest has focussed on mechanisms.11 While it seemed likely that aspirin's effect was mediated through inhibition of cyclooxygenase, the enzyme catalyzing the conversion of arachidonic acid to prostaglandins and related eicosanoids, aspirin has several other effects that are unrelated to cyclooxygenase activity and might be involved. The evidence that it is indeed a cyclooxygenase mediated effect has been much strengthened by another outstanding study from the Nurses Health Study cohort which showed that aspirin use seems only to reduce the risk of colorectal cancer in those cancers that over-express cyclooxygenase 2.12
For the pharmaceutical industry, which was already investigating heavily in developing cyclo-oxygenase 2 enzyme inhibitors (Coxibs), interest focussed on the possibility of using these new agents in cancer chemoprevention. The rationale for developing Coxibs was that by not inhibiting the constitutively expressed cyclo-oxygenase enzyme (COX-1), they would be free of the important side effects of peptic ulceration and bleeding seen with aspirin and traditional NSAIDs.13 Subsequent clinical trials have found Coxibs to be highly effective in reducing the recurrence rate of colorectal adenomas, the precursors of cancers.14,15 However, it was during these same trials that the serious adverse cardiovascular effects of Coxibs (around 2-fold increased risk of heart attack) were clearly demonstrated, dashing hopes that Coxibs might be a safer alternative to aspirin.16,17
So the inevitable question arises as to what exactly has been the contribution of the Australian paper to this line of research? While it was the first observational study published and by 2007 had been cited 400 times, Rosenberg and her colleagues appear to have been unaware of it when their paper (435 citations) was published. This paper was the stimulus to the second US study by Thun et al. and it was only as these authors were revising their paper with 1102 citations that they became aware of the Australian study (Thun, personal communication). Accordingly the accolade for the seminal publication should rest with Rosenberg and colleagues while that for primacy of publication with Kune and colleagues. It is also worth noting that at a time when there is so much emphasis on hypothesis driven research that in these three studies the data on aspirin and NSAID use was collected ‘hypothesis free’.
Finally what place now for aspirin in colorectal cancer prevention? It is too early to say whether aspirin itself will have a significant role. Side effects such as gastrointestinal bleeding and haemorrhagic stroke make aspirin unsuitable for primary prevention,18,19 but trials of its use in secondary prevention in groups with an increased bowel cancer risk such as those with previous colorectal cancer or adenomas have been encouraging.20–22 So far little is known as to the effect of aspirin or other NSAIDs in treated colorectal cancer patients, but aspirin may also have a role in tertiary prevention analogous to that of tamoxifen in breast cancer.
 |
References |
|---|
 Top References |
|---|
1 Jack DB. One hundred years of aspirin. Lancet (1997) 350:437–39.[CrossRef][Web of Science][Medline]
2 Flossmann E, Rothwell PM, on behalf of the British Doctors Aspirin Trial and the UK-TIA Aspirin Trial. Effect of aspirin on long-term risk of colorectal cancer: consistent evidence from randomised and observational studies. Lancet (2007) 369:1603–13.[CrossRef][Web of Science][Medline]
3 Kune GA, Kune S, Watson LF. Colorectal cancer risk, chronic illnesses, operations, and medications: case control results from the Melbourne Colorectal Cancer Study. Cancer Research (1988) 48:4399–404. (Reprinted Int J Epidemiol 2007;36:951–57).
4 Kune GA, Kune S, Watson LF. Children, age at first birth, and colorectal cancer risk: data from the Melbourne Colorectal Cancer Study. Am J Epidemiol (1989) 129:533–42.
5 Rosenberg L, Palmer JR, Zauber AG, Warshauer ME, Stolley PD, Shapiro S. A Hypothesis: nonsteroidal anti-inflammatory drugs reduce the incidence of large-bowel cancer. J Natl Cancer Inst (1991) 83:355–58.
6 Thun MJ, Namboodiri MH, Heath CW. Aspirin use and reduced risk of fatal colon cancer. N Engl J Med (1991) 325:1593–96.[Abstract]
7 Giovannucci E, Egan KM, Hunter DJ, et al. Aspirin and the risk of colorectal cancer in women. N Engl J Med (1995) 333:609–14.
8 Chan AT, Giovannucci EL, Meyerhardt JA, Schernhammer ES, Curhan GC, Fuchs CS. Long-term use of aspirin and nonsteroidal anti-inflammatory drugs and risk of colorectal cancer. JAMA (2005) 294:914–23.
9 Sturmer T, Glynn RJ, Lee I-M, Manson JE, Buring JE, Hennekens CH. Aspirin use and colorectal cancer: post-trial follow-up data from the Physicians’ Health Study. Ann Intern Med (1998) 128:713–20.
10 Cook NR, Lee I-M, Gaziano JM, et al. Low-dose aspirin in the primary prevention of cancer: The Women's Health Study: a randomized contrlled trial. JAMA (2005) 294:47–55.
11 Wang D, DuBois RN. Prostaglandins and cancer. Gut (2006) 55:115–22.
12 Chan AT, Ogino S, Fuchs CS. Aspirin and the risk of colorectal cancer in relation to the expression of COX-2. N Engl J Med (2007) 356:2131–42.
13 Hawkey CJ. COX-2 chronology. Gut (2005) 54:1509–14.
14 Bertagnolli M, Eagle C, Zauber AG, et al. Celecoxib for the prevention of sporadic colorectal adenomas. N Engl J Med (2006) 355:873–84.
15 Arber N, Eagle CJ, Spicak J, et al. Celecoxib for the Prevention of Colorectal Adenomatous Polyps. N Engl J Med (2006) 355:885–95.
16 Bresalier RS, Sandler RS, Quan H, et al. Cardiovascular events associated with Rofecoxib in a colorectal adenoma chemoprevention trial. N Eng J Med (2005) 352:1092–102.
17 Solomon SD, McMurray JJV, Pfeffer MA, et al. Cardiovascular risk associated with celecoxib in a clinical trial for colorectal adenoma prevention. N Engl J Med (2005) 352:1071–80.
18 Dubè C, Rostom A, Lewin G, et al. The use of aspirin for primary prevention of colorectal cancer: a systematic review prepared for the U.S. Preventive Services Task Force. Ann Intern Med (2007) 146:365–75.
19 U.S. Preventive Services Task Force. Routine aspirin or nonsteroidal anti-inflammatory drugs for the primary prevention of colorectal cancer: U.S. Preventive Services Task Force Recommendation Statement. Ann Intern Med (2007) 146:361–64.
20 Baron JA, Cole BF, Sandler RS, et al. A randomized trial of aspirin to prevent colorectal adenomas. N Engl J Med (2003) 348:891–99.
21 Sandler RS, Halabi S, Baron JA, et al. A randomized trial of aspirin to prevent colorectal adenomas in patients with previous colorectal cancer. N Engl J Med (2003) 348:883–90.
22 Logan RFA, Grainge MJ, Shepherd VC, Armitage NC, Muir KR. Aspirin and folic acid for the prevention of recurrent colorectal adenomas. Gastroenterology (2007) (in press).
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||