www.thelancet.com/oncology Vol 10 May 2009 501
Aspirin and non-steroidal anti-infl ammatory drugs for
cancer prevention: an international consensus statement
Jack Cuzick, Florian Otto, John A Baron, Powel H Brown, John Burn, Peter Greenwald, Janusz Jankowski, Carlo La Vecchia, Frank Meyskens,
Hans Jörg Senn, Michael Thun
Evidence clearly shows a chemopreventive eff ect for aspirin and other non-steroidal anti-infl ammatory drugs
(NSAIDs) on colorectal cancer and probably other cancer types; however, data on the risk–benefi t profi le for cancer
prevention are insuffi cient and no defi nitive recommendations can be made. Aspirin has emerged as the most likely
NSAID for use in chemoprevention because of its known cardiovascular benefi t and available safety and effi cacy data.
Other traditional NSAIDs, particularly sulindac, and selective COX-2 inhibitors are now given to patients at high risk
of colorectal cancer, although these drugs do not provide cardioprotection. More studies of aspirin and cancer
prevention are needed to defi ne the lowest eff ective dose, the age at which to initiate therapy, the optimum treatment
duration, and the subpopulations for which the benefi ts of chemoprevention outweigh the risks of adverse side-eff ects.
Although it might be possible to answer some of these questions with longer follow-up of existing clinical trials,
randomised controlled trials with new study designs will be needed. Future projects should investigate the eff ects of
aspirin treatment on multiple organ systems. Cancers of interest are colorectal, breast, prostate, lung, stomach, and
oesophageal. The main side-eff ect of aspirin is peptic ulcers; therefore coadministration of aspirin with a proton-pump
inhibitor is an attractive option and is under investigation in the AspECT trial.
Acetylsalicylic acid was fi rst synthesised in 1897 by Felix
Hoff mann and marketed by Bayer as Aspirin. After
100 years of use, new indications continue to be explored
(fi gure 1).1 At the fi fth International Conference on Cancer
Prevention in St Gallen, Switzerland (March 6–8, 2009),
an international expert group met to assess evidence of
the risks and benefi ts of aspirin and other NSAIDs as
treatments to prevent cancer. Of the potential chemo-
preventive drugs for gastrointestinal cancer, aspirin was
the only NSAID with suffi cient effi cacy and toxicity data
to qualify for a risk–benefi t analysis (except in very
high-risk populations such as familial adenomatous
polyposis). The panel planned to produce a consensus
statement on the use of aspirin and other NSAIDs for
cancer prevention; however, it became clear that gaps in
our understanding of appropriate dose, duration, and age
of use, would not support a formal risk–benefi t analysis.
Tables 1 and 22–10 summarise current knowledge of the
risks and benefi ts of aspirin. The focus of the meeting
shifted to the identifi cation of issues where further study
is needed. The topics discussed by the panel and their
conclusions are summarised in this Review.
Antitumour eff ect of aspirin and other NSAIDs
Much of the clinical evidence of a chemopreventive eff ect
of aspirin is from epidemiological studies.2,11 Only one
randomised controlled trial has specifi cally examined the
eff ect of aspirin on the incidence of cancer in healthy
people.12 Three randomised trials that assessed outcomes
from aspirin therapy have also reported cancer incidence,
with a focus on colorectal cancer (table 3).12–17 Several
randomised clinical trials have investigated the potential
of aspirin to prevent colorectal adenomas—the presumed
precursors of most colorectal cancers (table 3).12–17
Epidemiological evidence of a reduced risk of colorectal
adenomas with regular use of aspirin18–27 has been analysed
in great detail.2,11,28 In this Review, we focus on the clinical
implications of such evidence. NSAIDs delay or prevent
colorectal and mammary carcinoma in animal models of
tumorigenesis,29–34 but so far, no randomised clinical trials
of aspirin or other NSAIDs have used cancer mortality as
Lancet Oncol 2009; 10: 501–07
Cancer Research UK Centre for
and Statistics, Queen Mary
University of London, London
UK (Prof J Cuzick PhD); Tumor
Center ZeTuP, St Gallen,
Switzerland and Department
of Hematology and Oncology,
University Medical Center,
Freiburg, Germany (F Otto MD,
H J Senn MD); Dartmouth
Medical School, Hanover, NH,
USA (Prof J A Baron MD); Baylor
College of Medicine, Houston,
TX, USA (Prof P H Brown MD);
Newcastle University, Institute
for Human Genetics, Newcastle
upon Tyne, UK (Prof J Burn PhD);
Division of Cancer Prevention,
National Cancer Institute,
National Institutes of Health,
Bethesda, MD, USA
(P Greenwald MD); University of
Oxford, Oxford, UK
(Prof J Jankowski MD); Institute
for Cellular and Molecular
Sciences, Queen Mary’s College,
University of London, London
UK (Prof J Jankowski); Mario
Negri Institute for
Pharmacological Research and
University of Milan, Milan, Italy
(Prof C La Vecchia MD); Chao
Family Comprehensive Cancer
Center, University of California
Irvine Medical Center, Irvine,
Orange, CA, USA
(Prof F Meyskens MD); and
Epidemiology and Surveillance
Research, American Cancer
Society, Atlanta, GA, USA
(Prof M Thun MD)
Prof Jack Cuzick, Cancer Research
UK Centre for Epidemiology,
Mathematics, and Statistics,
Queen Mary University of
London, Barts & The London
School of Medicine and
Dentistry, Wolfson Institute of
London EC1M 6BQ, UK
Figure 1: Aspirin was fi rst synthesised by Felix Hoff mann in 1897
Since its initial use as an analgesic, aspirin has been proven benefi cial for
preventing myocardial infarction and stroke in high-risk individuals. Now the
drug is being assessed for reduction of cancer risk at several sites including the
colorectum, stomach, oesophagus, breast, ovary, and lung. Reproduced with
permission from Bayer AG.1
www.thelancet.com/oncology Vol 10 May 2009
a primary endpoint. Studies of the mechanisms by which
NSAIDs might inhibit carcinogenesis have not provided
conclusive evidence of pathways or molecular targets that
are clinically most relevant. The inducible cyclo-oxygenase
isoenzyme, COX-2, is overexpressed in precursor lesions
in breast (ductal carcinoma in situ), lung, and colorectum
(adenomatous polyps).35–41 Diff erent NSAIDs can inhibit
the activity of this enzyme to varying extents. Apart from
COX-2 selective inhibitors (coxibs), most NSAIDs are
non-selective and inhibit both COX-2 and the constitutively
expressed isoenzyme COX-1.41 Most cancers progress
through the action of multiple pathways that include
COX-2, Wnt–β-catenin, MAP kinase, cytokine, and
growth-factor signalling. Drugs that simultaneously block
several path ways might be particularly eff ective as
chemopreventive agents, if the clinical benefi ts outweigh
the toxic eff ects. Some laboratory evidence suggests that
aspirin works through several pathways.42 Further
assessment of clinical endpoints will take place in the
Most epidemiological studies report an inverse
association between the use of NSAIDs and incidence of
colorectal cancer or disease-related death;2,11,18,21,26,28,44–53 only
one cohort study reported a positive association.54 Both
cohort and case–control studies indicated that incidence
of colorectal cancer is about 40% lower in people who
take NSAIDs regularly than in those who do not. Such a
reduction in cancer incidence in the general population
would be an important achievement for public health. In
Western Europe and North America, risk of colorectal
cancer before age 75 is about 4–5% in men and 2·5–3·5%
in women.3 The dose and duration of treatment with
NSAIDs are not well defi ned in most reports. Retro-
spective and prospective studies show reductions in
incidence and size of adenomatous colorectal polyps with
the use of NSAIDs13,15,17,19–21 and coxibs.55,56 Despite the
consistency of this evidence, there are unresolved
questions about the lowest eff ective dose and duration of
NSAID treatment needed to achieve a meaningful
reduction in cancer risk.
Not all data support a chemopreventive eff ect of aspirin.
No preventive eff ect, for any cancer, was seen in two
randomised controlled trials with 5-year and 10-year
interventions. In the Women’s
39 876 healthy women, aged at least 45 years, were
randomly assigned to receive either aspirin (100 mg) or
placebo every other day for an average of 10 years. This
large, long-term study found that aspirin had no eff ect on
the incidence of any specifi c cancer.12 The Physicians’
Health Study13 included 22 071 male physicians randomly
assigned to receive aspirin (325 mg) or placebo on
alternate days. No reduction in the incidence of colorectal
cancer was seen in the intervention group versus the
control group after 5 years of treatment.13 This null
fi nding was confi rmed in a subsequent analysis of the
same study after a follow-up of 12 years.57 In addition, the
CAPP2 investigators58 recently reported no protective
Risk (%) to
age 74 (general
absolute benefi t†
to age 74
Hip or knee
6 10030·96 (0·91–1·02) ····
13 822 0·80 (0·73–0·87)
·· 1214 738
159 7602 0·74 (0·70–0·78) ····
5 0·74 (NA)14·0 (0·7% per
32350 0·75 (0·55–0·95) 2·0 (0·1% per
RCT=randomised clinical trial. NA=not available. *Baseline cancer rates are an average of US Surveillance, Epidemiology
and End Results (SEER) white rates and rates in England3 or from sources cited; averages do not include important risk
factors or dose eff ects. †Estimated potential benefi t for cancer prevention (sex-specifi c) is based on relative risk from
the case–control study.
Table 1: Estimated benefi ts of long-term (about 20 years) aspirin use in the general population
www.thelancet.com/oncology Vol 10 May 2009 503
eff ect of 600 mg of enteric coated aspirin in a randomised
clinical trial involving 1071 carriers of Lynch syndrome—
the hereditary syndrome that predisposes individuals to
a range of cancers, particularly colorectal and endo-
metrial—despite therapy for up to 4 years (relative risk
1·0; 95% CI 0·7–1·4).
The results from the Women’s Health and Physicians’
Health randomised trials are diffi cult to reconcile with
epidemiological and preclinical data, or with results from
trials studying adenomatous polyps, unless the
chemopreventive benefi ts of aspirin become apparent
only after more than 10 years of treatment. Long-term
follow-up of the British
UK Transient Ischaemic Attack study, and several
observational studies suggest that a 10-year delay is
plausible with respect to chemoprevention of colorectal
cancer.14 Whether the duration of aspirin use or the
duration of follow-up is more important to see an eff ect
remains unclear. The 5-year duration of randomised
treatment in the Physicians’ Health Study, and the 10-year
treatment and follow-up in the Women’s Health Study
are not long enough to answer this question. A further
report from the CAPP2 trial is expected later this year. No
trials of aspirin, either for chemoprevention or other use,
have continued randomised treatment beyond 10 years.
On the basis of available evidence, the consensus panel
regarded the antitumour eff ect of aspirin and sulindac as
“very probable”, but that of other NSAIDs as “possible”
because of the paucity of specifi c data on these drugs.
Doctors study, the
Risk–benefi t ratio of aspirin and other NSAIDs
The side-eff ects of NSAIDs are well documented and
are, for the most part, attributed to inhibition of COX
activity. The risk of bleeding associated with use of
aspirin results from inhibition of COX-1 activity in
platelets, which prevents aggregation. Aspirin-induced
bleeding predominantly involves the gastrointestinal
and genitourinary tracts, but can cause intracranial
haemorrhage in rare cases. High doses of aspirin and
other non-selective COX-2 inhibitors block the
production of prostaglandin E2 in gastric epithelium
and increase the risk of gastric ulceration and bleeding.
Similarly, high doses of coxibs and non-selective
NSAIDs inhibit prostacyclin production by COX-2 in
vascular endothelium, increasing the risk of thrombotic
cardiovascular events in some people. The risk of
serious gastrointestinal bleeds over 10 years increases
from about 1% in untreated individuals to about 2–3%
in those who take regular aspirin.10 This risk increases
with age59 and dose.60,61 The incidence of gastric or
duodenal ulcers as a result of NSAID use increased
linearly from about 10% in those younger than 45 years
to 25% in those older than 75 years.59 By comparison,
the incidence of bleeding events increased from about
1% to 6% for people in the same age-groups who took a
placebo. Cranial haemorrhage is much rarer and
accurate risk estimates are not available, in part because
it is diffi cult to distinguish cranial haemorrhage from
occlusive cerebrovascular events, which are reduced by
aspirin. Standard contraindications for the use of
aspirin include previous history of gastrointestinal
ulcers, bleeding disorders, and allergic reactions to the
drug. Apart from these few indications, however,
identifi cation of the individuals most prone to adverse
eff ects is not currently possible.
A specifi c benefi t of aspirin over other NSAIDs is a
lowered risk of occlusive cardiovascular events.6 Use of
aspirin is widespread for the secondary prevention of
myocardial infarction and ischaemic stroke in patients
with pre-existing cardiovascular disease and for primary
prevention in high-risk groups. Current indications for
the prophylactic use of aspirin are based on cardiovascular
risk and side-eff ects, but do not take into account other
potential benefi ts. Prophylaxis is clearly indicated for
individuals with a high risk of cardiovascular events but
not for individuals with an average risk, for whom the
benefi t of aspirin treatment would be at least partially
off set by the adverse eff ect of bleeding. Age further
complicates the balance between risks and benefi ts. Risks
of serious bleeding and gastrointestinal ulceration
increase substantially after 60 years of age. Age-specifi c
changes in the risk–benefi t ratio of prophylactic treatment
with aspirin to prevent premalignant lesions or cancer
Because of uncertainties about the minimum dose
and duration of aspirin treatment needed to decrease
cancer incidence, and the mixed benefi cial and adverse
eff ects on the cardiovascular and other organ systems,
the panel concluded that further clinical studies were
needed to assess the risk–benefi t profi le of NSAIDs.
Risk (%) to age 74
absolute risk to
age 74 (per 1000)
Serious gastrointestinal bleed9
2·0 (0·1% per year) 24
No relation with dose
12·1 (1·42% over
1583192 0·70 (0·65–0·76)
4 5971·03 (0·87–1·21) 6·0 (0·3%
4 851·36 (0·88–2·1) 0·8 (0·04%
RCT=randomised clinical trial. *Baseline cancer rates are an average of US Surveillance, Epidemiology and End Results (SEER)
white rates and rates in England3 or from sources cited; averages do not include important risk factors or dose eff ects.
Table 2: Estimated risks of long-term (about 20 years) aspirin use in the general population
www.thelancet.com/oncology Vol 10 May 2009
Without a clear indication of the overall eff ects on
health, evidence was deemed insuffi cient to make a
recommendation on the use of aspirin or other NSAIDs
for cancer prevention. Therefore, the panel sought to
identify necessary research.
Most of the physiological eff ects of aspirin and other
NSAIDs are dose dependent. Low-dose aspirin (100 mg
or less and possibly as low as 30–50 mg per day) reduces
the risk of thrombotic cardiovascular events, with only a
small increase in risk of bleeding. Low-dose aspirin
selectively inhibits thromboxane synthesis in platelets
that pass through the portal circulation. Only small
amounts of the drug survive passage through the liver to
reach the systemic circulation. High doses of aspirin and
other NSAIDs reach the systemic circulation and inhibit
prostaglandin E2 production in gastric epithelium and
prostacyclin synthesis in vascular endothelial cells.
Because of its selective eff ect on platelets, low-dose
aspirin prevents thrombotic cardiovascular events with
minimum adverse eff ects in the gastric epithelium.
By contrast with low-dose aspirin for cardiovascular
benefi t, the optimum dose of aspirin for cancer
prevention is not well defi ned. Baron and colleagues15
report that 81 mg might be more eff ective than 325 mg in
preventing colorectal adenomas, whereas other data
(largely observational) suggest that doses of 325 mg or
more might be necessary to achieve the maximum
chemopreventive eff ect.18 Two trials of aspirin doses of
100 mg and 325 mg every other day for prevention of
colorectal cancer reported negative results;12,57 however,
follow-up was less than 10 years. Two other randomised
trials of 300 mg, 500 mg, or 1200 mg daily showed a
decrease in cancer incidence compared with placebo.14
Uncertainties about the lowest eff ective dose for cancer
prevention need to be resolved, because the incidence of
peptic ulcers and bleeding increases with higher doses.60,61
Randomised clinical trials are needed to identify the best
prophylactic dose for multiple endpoints.
Data suggest that aspirin should be taken for several
years for substantial protection against cancer; however,
whether it should be taken continuously or if stopping
treatment after 10 years might have a continued benefi t is
Invasive colorectal cancer†
Women’s Health Study12
100 mg qod
Physicians’ Health Study13
325 mg qod
British Doctors Aspirin
UK Transient Ischaemic
Attack Aspirin Trial14
300 mg or 1200 mg
Baron et al 200315
Sandler et al 200316
160 mg or 300 mg
Women aged ≥45 years
Male physicians aged 40–84 years
Physicians aged ≤80 years, highest chemopreventive eff ect
after 10–19 years of follow-up
Age >40 years, history of transient ischaemic attack, highest
chemopreventive eff ect after 10–19 years of follow-up
Recent history of colorectal adenomas
History of colorectal cancer
History of colorectal adenomas
qod=every other day. *Risk relative to placebo (reference) group. †Incidence of colorectal cancer in randomised primary prevention trials assessing aspirin. ‡Incidence of
colorectal adenomas in randomised secondary prevention trials assessing aspirin.
Table 3: Incidence of colorectal cancer and adenomas in trials of aspirin for cancer prevention
www.thelancet.com/oncology Vol 10 May 2009 505
unclear. Two randomised studies reported a decrease in
the incidence of colon cancer that lasted for more than
10 years after stopping of aspirin treatment.14 This eff ect
was not seen in other (observational) studies.62 How long
the carry-over eff ect will continue in trials of high-dose
aspirin and whether low-dose aspirin will have a benefi t
in post 10-year follow-up are unknown. In view of the
long follow-up time needed to answer both questions,
priority should be given to continuing the follow-up of
current trials of aspirin dose.
Observational evidence of a chemopreventive eff ect of
aspirin and other NSAIDs has been reported for
oesophageal, gastric, lung, breast, prostate, and colorectal
cancer. Most of these cancers develop after age 60 years,
but the best age to start NSAIDs for cancer prevention is
unknown. Given the apparent delay in the chemo-
preventive eff ect of NSAIDs (about 10 years), optimum
treatment might start at age 40–50 years. Most individuals
who develop premalignant lesions do so in their 50s and
60s, several years before the appearance of cancer, so this
age range might be the best time for cancer prevention.
Additionally, because the risk of serious side-eff ects of
aspirin increases steeply after age 60 years, long-term
treatment before this age might be needed to avoid
adverse eff ects.
Treatment of many people to prevent a few cancers is
not an effi cient approach to chemoprevention. In this
context, even rare side-eff ects can adversely aff ect the
overall risk–benefi t profi le. A better approach is to focus
preventive interventions for individuals with a substantial
risk of cancer. Although age, lifestyle, environment, and
particularly a history of high-risk adenomas contribute to
an individual’s risk of cancer, genetic factors are likely to
be important determinants. Additional molecular and
genetic studies are needed to better identify individuals
who are at risk of cancers that respond to aspirin
treatment, and to identify those most likely to experience
serious toxic eff ects. Advances in risk prediction will
improve the overall risk–benefi t profi le of interventions.
A large body of evidence supports an antitumour eff ect
of aspirin and other NSAIDs on colorectal cancer.
Additionally, aspirin might also protect against
oesophageal, gastric, lung, breast, and prostate
tumours.2,52 Evidence in these organ systems is sparse
and will need to be further assessed in large controlled
trials. Trial design is a challenge because many studies of
aspirin stop when cardiovascular effi cacy is proven but
before the duration of treatment is suffi cient to assess
A key factor in trial design is the duration of treatment
and follow-up needed to fi nd signifi cant diff erences in
the incidence of cancer between treatment groups.
Long trial times and the large number of participants
needed to achieve adequate power means that
chemoprevention trials are expensive and laborious.
Therefore, we should maximise insights from trials that
already exist. In view of the delayed protective eff ect of
aspirin on colorectal cancer, and similar fi ndings for
tamoxifen on breast cancer,63,64 chemopreventive
treatments might be most eff ective for premalignant
lesions that progress over decades, so that lengthy
follow-up is needed to see the full benefi t of treatment.
The proven benefi ts of aspirin for cardiovascular disease
(although small for individuals at average risk) give it
an advantage over other NSAIDs for further study of
cancer chemoprotection. However, other NSAIDs such
as sulindac also warrant investigation.
Trials of chemoprevention could be improved by better
identifi cation of individuals at high risk for tumours; a
higher event rate would allow diff erences to be seen more
clearly. Another challenge for this type of trial is the choice
of a suitable endpoint. In clinical settings, the follow-up of
precancerous lesions until cancer develops is unethical, so
new approaches are needed. Assessment of colon cancer
as an endpoint is nearly impossible in trials in which
patients in both the control and intervention groups have
endoscopic surveillance and removal of high-risk polyps.
A study design with the identifi cation of colorectal
adenomas as an intermediate endpoint allows for smaller,
shorter trials. If colorectal adenomas are the primary
endpoint, however, only adenomas with a high risk for
progression to cancer should be included as endpoints.
More work is needed to reliably identify high-risk
adenomas. By comparison with colorectal cancer, the
identifi cation of good surrogate markers for other cancers
is at an even earlier stage, but mammographic density
might be such an endpoint for breast cancer.
Another challenge in cancer prevention is the
development of an infrastructure where high-risk
individuals are identifi ed and are given counselling and
% of cancer budget
UK research and development
UK National Cancer Research Institute
USA National Cancer Institute
Figure 2: Categories of cancer spending (%) in the UK and USA
Adapted from a thesis by Jankowski J, Oxford University, 2007.65
www.thelancet.com/oncology Vol 10 May 2009
support to adopt prevention measures that are eff ective.
Inadequate funding currently limits progress in cancer
prevention (fi gure 2).65 About 10% of cancer funding in
the USA, and less than 5% in the UK, is spent on
prevention. Within this small proportion, chemo-
prevention receives only 25%.
Only treatment with aspirin combines the benefi t of
protection against cardiovascular disease with the
potential to reduce the risk of some types of cancer.
Aspirin might eventually be useful for the primary
prevention of some cancers in patients who already
qualify for prophylactic antiplatelet therapy on the basis
of cardiovascular criteria. Aspirin or other NSAIDs might
also prove eff ective for secondary chemoprevention of
gastrointestinal cancers in patients with no antecedent
risk of gastrointestinal bleeding. Upper gastro intestinal
bleeding, a common side-eff ect of aspirin therapy, is
eff ectively prevented with a proton-pump inhibitor, so
coadminis tration of aspirin and proton-pump inhibitors
is an attractive option in this setting, and is currently
being studied in the AspECT study of esomeprazole and
aspirin in patients with Barrett’s oesophagus.43,66 Coxibs
and other NSAIDs are unlikely to be useful for cancer
prevention beyond their current use in young patients
with familial adenomatous polyposis. Large-scale studies
are needed to assess whether long-term aspirin treatment
can prevent gastrointestinal and other cancers. Future
studies should address outstanding questions about
dose, the best age to begin treatment, and duration of
therapy. Randomised clinical trials will be essential to
establish the effi cacy and safety of a clearly-defi ned
All authors participated in the panel discussion and supplied material
for the manuscript. The manuscript was drafted by JC and FO and all
authors contributed to the revision and fi nal approval.
Confl icts of interest
FM is co-founder of Cancer Prevention Pharmaceuticals. JJ is a
consultant to AstraZeneca. PHB receives funding for basic science from
Lilly Pharmaceuticals. JAB is a consultant for Bayer.
1 Zundorf U. Aspirin 100 years: the future has just begun (cover
illustration). Leverkusen: BayerAG, 1997.
2 Bosetti C, Gallus S, La Vecchia C. Aspirin and cancer risk: an
updated quantitative review to 2005. Cancer Causes Control 2006;
3 Parkin DM, Whelan SL, Ferlay J, et al. Cancer Incidence in Five
Continents. IARC Scientifi c Publications; 8: 155. Lyon, France, 2002.
4 Neugat AI, Jacobson JS, Rella VA. Prevalence and incidence of
colorectal adenomas and cancer in asymptomatic persons.
Gastrointest Endosc Clin N Am 1997; 7: 387–99.
Peto R, Darby S, Deo H, et al. Smoking, smoking cessation and
lung cancer in the UK since 1950: combination of national statistics
with two case-control studies. BMJ 2000; 321: 323–29.
Antithrombotic Trialists’ Collaboration.Collaborative meta-analysis
of randomised trials of antiplatelet therapy for prevention of death,
myocardial infarction, and stroke in high risk patients. BMJ 2002;
Hart RG, Halperin JL, McBride R, et al. Aspirin for the primary
prevention of stroke and other major vascular events. Arch Neurol
2000; 57: 326–32.
Pulmonary Embolism Prevention (PEP) Trial Collaborative Group.
Prevention of pulmonary embolism and deep vein thrombosis with
low dose aspirin. Lancet 2000; 355: 1295–1302.
Garcia Rodriguez LA, Hernandez-Diaz S, de Abajo FJ. Association
between aspirin and upper gastroinstestinal complications:
systematic review of epidemiologic studies. Br J Clin Pharmacol
2001; 52: 563–71.
10 Derry S, Loke YK. Risk of gastrointestinal haemorrhage with long
term use of aspirin: meta-analysis. BMJ 2000; 321: 1183–87.
11 Bosetti C, Gallus S, La Vecchia C. Aspirin and cancer risk: a
summary review to 2007. Recent Results Cancer Res 2009;
12 Cook NR, Lee IM, Gaziano JM, et al. Low-dose aspirin in the
primary prevention of cancer: the Women’s Health Study:
a randomized controlled trial. JAMA 2005; 294: 47–55.
13 Gann PH, Manson JE, Glynn RJ, Buring JE, Hennekens CH.
Low-dose aspirin and incidence of colorectal tumors in a
randomized trial. J Natl Cancer Inst 1993; 85: 1220–24.
14 Flossmann E, Rothwell PM. Eff ect of aspirin on long-term risk of
colorectal cancer: consistent evidence from randomised and
observational studies. Lancet 2007; 369: 1603–13.
15 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.
16 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.
17 Benamouzig R, Deyra J, Martin A, et al. Daily soluble aspirin and
prevention of colorectal adenoma recurrence: one-year results of the
APACC trial. Gastroenterology 2003; 125: 328–36.
18 Chan AT, Giovannucci EL, Meyerhardt JA, et al. Long-term use of
aspirin and nonsteroidal anti-infl ammatory drugs and risk of
colorectal cancer. JAMA 2005; 294: 914–23.
19 Chan AT, Giovannucci EL, Schernhammer ES, et al. A prospective
study of aspirin use and the risk for colorectal adenoma.
Ann Intern Med 2004; 140: 157–66.
20 Garcia Rodriguez LA, Huerta-Alvarez C. Reduced incidence of
colorectal adenoma among long-term users of nonsteroidal
antiinfl ammatory drugs: a pooled analysis of published studies and
a new population-based study. Epidemiology 2000; 11: 376–81.
21 Giovannucci E, Rimm EB, Stampfer MJ, et al. Aspirin use and the
risk for colorectal cancer and adenoma in male health professionals.
Ann Intern Med 1994; 121: 241–46.
22 Greenberg ER, Baron JA, Freeman Jr DH, et al. Reduced risk of
large-bowel adenomas among aspirin users. The Polyp Prevention
Study Group. J Natl Cancer Inst 1993; 85: 912–16.
23 Kahn HS, Tatham LM, Thun MJ, et al. Risk factors for self-reported
colon polyps. J Gen Intern Med 1998; 13: 303–10.
24 Logan RF, Little J, Hawtin PG, et al. Eff ect of aspirin and
non-steroidal anti-infl ammatory drugs on colorectal adenomas:
case–control study of subjects participating in the Nottingham
faecal occult blood screening programme. BMJ 1993; 307: 285–89.
25 Morimoto LM, Newcomb PA, Ulrich CM, et al. Risk factors for
hyperplastic and adenomatous polyps: evidence for malignant
potential? Cancer Epidemiol Biomarkers Prev 2002; 11: 1012–18.
26 Suh O, Mettlin C, Petrelli NJ. Aspirin use, cancer, and polyps of
the large bowel. Cancer 1993; 72: 1171–77.
27 Tangrea JA, Albert PS, Lanza E, et al. Non-steroidal
anti-infl ammatory drug use is associated with reduction in
recurrence of advanced and non-advanced colorectal adenomas
(United States). Cancer Causes Control 2003; 14: 403–11.
Search strategy and selection criteria
This is not a comprehensive review of the literature, which is
available in other reports.2,11,29 This paper is a focused
discussion of the key outstanding issues. All relevant studies
were identifi ed by checking recent comprehensive reviews
and in consultation with the panellists.
www.thelancet.com/oncology Vol 10 May 2009 507
28 Dubé C, Rostom A, Lewin G, et al. The use of aspirin for primary
prevention of colorectal cancer: a systematic review prepared for
the US Preventive Services Task Force. Ann Intern Med 2007;
29 Harris RE, Alshafi e GA, Abou-Issa H, et al. Chemoprevention of
breast cancer in rats by celecoxib, a cyclooxygenase 2 inhibitor.
Cancer Res 2000; 60: 2101–03.
30 Howe LR, Subbaramaiah K, Patel J, et al. Celecoxib, a selective
cyclooxygenase 2 inhibitor, protects against human epidermal
growth factor receptor 2 (HER-2)/neu-induced breast cancer.
Cancer Res 2002; 62: 5405–07.
31 Kawamori T, Rao CV, Seibert K, et al. Chemopreventive activity of
celecoxib, a specifi c cyclooxygenase-2 inhibitor, against colon
carcinogenesis. Cancer Res 1998; 58: 409–12.
32 Oshima M, Dinchuk JE, Kargman SL, et al. Suppression of
intestinal polyposis in Apc delta716 knockout mice by inhibition
of cyclooxygenase 2 (COX-2). Cell 1996; 87: 803–09.
33 Reddy BS, Rao CV, Rivenson A, et al. Inhibitory eff ect of aspirin
on azoxymethane-induced colon carcinogenesis in F344 rats.
Carcinogenesis 1993; 14: 1493–97.
34 Reddy BS, Rao CV, Seibert K. Evaluation of cyclooxygenase-2
inhibitor for potential chemopreventive properties in colon
carcinogenesis. Cancer Res 1996; 56: 4566–69.
35 Brown JR, DuBois RN. COX-2: a molecular target for colorectal
cancer prevention. J Clin Oncol 2005; 23: 2840–55.
36 Eberhart CE, Coff ey RJ, Radhika A, et al. Up-regulation of
cyclooxygenase 2 gene expression in human colorectal adenomas
and adenocarcinomas. Gastroenterology 1994; 107: 1183–88.
37 Half E, Tang XM, Gwyn K, et al. Cyclooxygenase-2 expression in
human breast cancers and adjacent ductal carcinoma in situ.
Cancer Res 2002; 62: 1676–81.
38 Hwang D, Scollard D, Byrne J, et al. Expression of cyclooxygenase-1
and cyclooxygenase-2 in human breast cancer. J Natl Cancer Inst
1998; 90: 455–60.
39 Zhang H, Sun XF. Overexpression of cyclooxygenase-2 correlates
with advanced stages of colorectal cancer. Am J Gastroenterol 2002;
40 Wolff H, Saukkonen K, Anttila S, et al. Expression of
cyclooxygenase-2 in human lung carcinoma. Cancer Res 1998;
41 Thun MJ, Henley SJ, Patrono C. Nonsteroidal anti-infl ammatory
drugs as anticancer agents: mechanistic, pharmacologic, and
clinical issues. J Natl Cancer Inst 2002; 94: 252–66.
42 Jankowski J, Anderson M. Management of oesophageal
adenocarcinoma: control of acid, bile and infl ammation in
intervention strategies of Barrett’s oesophagus.
Aliment Pharmacol Ther 2004; 20 (suppl 5): 71–80.
43 Jankowski JA, Moayyedi P. Re: cost-eff ectiveness of aspirin
chemoprevention for Barrett’s esophagus (correspondence).
J Natl Cancer Inst 2004; 96: 885–87.
44 Slattery ML, Samowitz W, Hoff man M, et al. Aspirin, NSAIDs, and
colorectal cancer: possible involvement in an insulin-related
pathway. Cancer Epidemiol Biomarkers Prev 2004; 13: 538–45.
45 Friis S, Sørensen HT, McLaughlin JK, et al. A population-based
cohort study of the risk of colorectal and other cancers among users
of low-dose aspirin. Br J Cancer 2003; 88: 684–88.
46 Garcia-Rodriguez LA, Huerta-Alvarez C. Reduced risk of colorectal
cancer among long-term users of aspirin and nonaspirin
nonsteroidal antiinfl ammatory drugs. Epidemiology 2001; 12: 88–93.
47 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.
48 Juarranz M, Calle-Purón ME, González-Navarro A, et al. Physical
exercise, use of Plantago ovata and aspirin, and reduced risk of
colon cancer. Eur J Cancer Prev 2002; 11: 465–72.
49 Kune GA, Kune S, Watson LF. Colorectal cancer risk, chronic
illnesses, operations, and medications: case control results from the
Melbourne Colorectal Cancer Study. Cancer Res 1988; 48: 4399–404.
50 La Vecchia C, Negri E, Franceschi S, et al. Aspirin and colorectal
cancer. Br J Cancer 1997; 76: 675–77.
51 Reeves MJ, Newcomb PA, Trentham-Dietz A, et al. Nonsteroidal
anti-infl ammatory drug use and protection against colorectal cancer
in women. Cancer Epidemiol Biomarkers Prev 1996; 5: 955–60.
52 Schreinemachers DM, Everson RB. Aspirin use and lung, colon,
and breast cancer incidence in a prospective study. Epidemiology
1994; 5: 138–46.
53 Thun MJ, Namboodiri MM, Heath Jr CW. Aspirin use and reduced
risk of fatal colon cancer. N Engl J Med 1991; 325: 1593–96.
54 Paganini-Hill A. Aspirin and colorectal cancer: the Leisure World
cohort revisited. Prev Med 1995; 24: 113–15.
55 Bertagnolli MM, Eagle CJ, Zauber AG, et al. Celecoxib for the
prevention of sporadic colorectal adenomas. N Engl J Med 2006;
56 Baron JA, Sandler RS, Bressalier RS, et al. A randomised trial of
refcoxib for the chemoprevention of colorectal adenomas.
Gastroentrerology 2006; 131: 2003–05.
57 Stürmer T, Glynn RJ, Lee IM, et al. Aspirin use and colorectal
cancer: post-trial follow-up data from the Physicians’ Health Study.
Ann Intern Med 1998; 128: 713–20.
58 Burn J,Bishop DT, Mecklin JP, et al. Eff ect of aspirin or resistant
starch on colorectal neoplasia in the Lynch syndrome. N Engl J Med
2008; 359: 2567–78.
59 Boers M, Tangelder MJ, van Ingen H, et al. The rate of
NSAID-induced endoscopic ulcers increases linearly but not
exponentially with age: a pooled analysis of 12 randomised trials.
Ann Rheum Dis 2007; 66: 417–18.
60 Serebruany VL, Steinhubl SR, Berger PB, et al. Analysis of risk of
bleeding complications after diff erent doses of aspirin in 192,036
patients in 31 randomised controlled trials. Am J Cardiol 2005;
61 Patrono C, Baigent C, Hirsh J, et al. Antiplatelet drugs: American
College of Chest Physicians evidence-based clinical practice
guidelines (8th Edition). Chest 2008; 133: 199–233S.
62 Chan AT, Giovannucci EL, Meyerhardt JA, et al. Aspirin dose and
duration of use and risk of colorectal cancer in men.
Gastroenterology 2008; 134: 21–28.
63 Cuzick J, Forbes JF, Sestak I, et al. Long-term results of tamoxifen
prophylaxis for breast cancer: 96-month follow-up of the
randomized IBIS-I trial. J Natl Cancer Inst 2007; 99: 272–82.
64 Powles TJ, Ashley S, Tidy A, et al. Twenty-year follow-up of the
Royal Marsden randomized, double-blinded tamoxifen breast
cancer prevention trial. J Natl Cancer Inst 2007; 99: 283–90.
65 Jankowski J. Reducing gastrointestinal cancer: chemoprevention as
a model strategy. MSc thesis, University of Oxford, 2006: 95.
66 Jankowski J, Hunt R. Cox-2s, confi rmation of cardiovascular and
cancer risk/benefi t ratio: better the devil you know.
Cancer Epidemiol Biomarkers Prev 2008; 17: 1858–61.