NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Use of Aspirin and NSAIDs to Prevent Colorectal Cancer

Use of Aspirin and NSAIDs to Prevent Colorectal Cancer

Evidence Syntheses, No. 45

Investigators: , MD, MSc, FRCPC, , MD, MSc, FRCPC, and , MD.

University of Ottawa Evidence-based Practice Center at The University of Ottawa, Ottawa, Canada
Rockville (MD): Agency for Healthcare Research and Quality (US); .
Report No.: 07-0596-EF-1

Structured Abstract


The purpose of this report was to systematically review the literature on the effectiveness of aspirin (ASA), non-aspirin nonsteroidal anit-inflammatory drugs (non-ASA NSAIDs), and cyclooxygenase-2 inhibitors (COX-2 inhibitors) for the chemoprevention of colorectal cancer (CRC) and CRC-related mortality in average-risk individuals. The review also assessed the harms associated with the use of these agents and their cost effectiveness.

Data Sources:

To review the effectiveness of ASA/NSAIDs, Medline 1966 to week 3, November 2004; Embase 1980 to week 47, 2004, and CENTRAL, the Cochrane Collaboration's registry of clinical trials (Issue 4, 2004) were searched. To identify recent systematic reviews of NSAIDs that address harms, Medline (2003 to week 3, November 2004), the Cochrane Database of Systematic Reviews (CDSR), and DARE (Cochrane Library, 3rd Quarter 2004) were searched. Additional material potentially relevant to the economic analysis question was sought in Medline (1966 to week 3, November 2004), HealthStar (1987 to November 2004), Embase (1980 to 2004 Week 50), NHS EED, and HTA databases of The Cochrane Library (4th Quarter 2004). The TRIP ( database was also searched (December 14, 2004).


RCTs, case-control, and cohort studies were sought for the effectiveness of ASA, NSAIDs, and COX-2 inhibitors to prevent colorectal adenomas (CRAs,), CRC, and mortality. Systematic reviews were sought for the harms of these agents, and cost-effectiveness analyses were sought for each of the agents. Multilevel screening by two independent reviewers was conducted to identify studies to be included based on predefined inclusion criteria. Data from included studies were abstracted and their quality assessed. Included studies were grouped based on an a priori defined hierarchy, and statistical pooling was conducted only if clinically and statistically appropriate.

Results: Effectiveness for Chemoprevention:

Regular use of ASA appears to be effective at reducing the incidence of CRA. Two of the four RCTs demonstrated statistically significant relative risk reductions (RRR) in CRAs (relative risk [RR]=0.44, and 0.58), whereas the remaining two RCTs, including the Physicians Health Study (PHS), showed no benefit. The pooled estimate, however, was statistically significant (RR=0.82; 95% CI: 0.7–0.95). Pooled estimates for the case-control (RR=0.87; 95% CI: 0.77–0.98) and cohort studies (RR=0.72; 95% CI: 0.61–0.85) also showed a statistically significant relative risk reduction in CRAs with ASA. The use of non-ASA NSAIDs appeared to be associated with somewhat higher RRRs (pooled estimate - RR= 0.43; 95% CI: 0.26–0.70), and based on a very limited number of studies, the RRRs are likely higher still for higher-risk individuals than for those at average risk. The regular use of ASA was associated with RRRs of 15% to 40% for CRC incidence. The pooled RR for cohort studies was 0.78 (95% CI: 0.63–0.97). The RCT data of the effect of ASA on CRC incidence was, however, negative, both in the PHS and the newly published Women's Health Study (WHS). The pooled estimates for non-ASA NSAIDs suggest somewhat greater RRRs, on the order of 30% to 40%, and longer duration of use of ASA/NSAIDs and higher doses also appear to offer greater protection.

Only two observational studies considered the effect of ASA/NSAIDs on CRC mortality. Among the observational studies, one study reported that CRC mortality was reduced by about 40% with ASA use for more than 15 years, whereas, the other found nonsignificant trends towards increased standardized mortality ratios for bowel and rectal cancers with ibuprofen. The recently published WHS found no benefit of ASA on CRC mortality.


The use of ASA is associated with an increased incidence of important ulcer complications with RRs of 1.5 to 3.0. The annualized incidence of these events for non-ASA NSAIDs, as a group, is approximately 1.5% to 2.0% in average-risk individuals with arthritis. As a “class,” COX-2 inhibitors are associated with fewer endoscopic ulcers and clinically important ulcer complications, when compared with non-ASA NSAIDs overall, with pooled RRRs of about 50% for important ulcer complications.

In individuals with low-to-average cardiovascular (CV) risk (i.e., the primary prevention population), ASA significantly reduced the incidence of total CV events, but had no effect on coronary heart disease mortality, fatal and nonfatal stroke events, or all cause mortality. In high-risk CV patients (i.e., secondary prevention), the use of ASA significantly reduces all-cause mortality, and CV mortality. There is a paucity of RCT data on the CV harm of non-ASA NSAIDs, but non-naproxen NSAIDs appear to offer no cardioprotection in observational studies, and may actually increase the risk of CV events. Knowledge regarding the CV harms associated with non-ASA NSAIDs and COX-2 inhibitors is in a rapid state of flux. COX-2 inhibitors appear to be associated with an increased risk of CV events.


In average-risk populations, and in the context of regular endoscopic screening for CRC, NSAID chemoprevention is presently not cost-effective because of the relatively large costs associated with their adverse effects, as well as their relative inefficacy compared with colonoscopy. To be cost-effective, daily ASA use would have to decrease the CV mortality by 0.1% or more, and it would have to decrease CRC mortality by at least 30%. Additionally, chemoprevention with COX-2 inhibitors, independent of their newly recognized cardiotoxicity, is expensive and their use as an adjunct to colonoscopy is economically acceptable (i.e. ICER less than $100,000/LY saved) if they can prevent CRC mortality by at least 60% and their cost be reduced by at least 75%. In higher-risk groups, the use of COX-2 inhibitors for chemoprevention of CRC is both less effective and considerably more costly than screening protocols, which are in themselves cost effective by all criteria—their use as an adjunct to screening is economically acceptable if their current cost is considerably reduced and if their efficacy as chemopreventive agents is of at least 50%. These results do not account for any potential CV harms of COX-2 inhibitor use.


ASA and non-ASA NSAIDs appear to be effective at reducing the incidence of CRAs and CRC. However, the data on CRC incidence is inconsistent with observational studies, which tend to be positive, whereas, two large RCTs showed no benefit for low-dose ASA every second day on CRC incidence. The effect of ASA/NSAIDs on CRC mortality is also mixed, with one positive and one negative cohort study, and the negative findings of the WHS. There are well-defined GI risks associated with ASA and NSAIDs when used daily for months. There are no quantitative data on GI or CV risk for chronic multiyear use of daily NSAIDs. We found no information regarding the effectiveness of COX-2 inhibitors on these outcomes in average-risk individuals. Available data on COX-2 inhibitors suggest that an absolute risk increase of over 1% for CV events can be anticipated from only 2 to 3 years use, and higher risks may accrue over longer periods. Further, the results of the economic evaluations consistently reveal that chemoprevention is not cost-effective. In the case of ASA, the costs of complications are significant; in the case of COX-2 inhibitors, independently of the recently reported CV toxicity, drug costs are great. Lastly, in addition to emerging CV toxicity, non-ASA NSAIDs are associated with significant GI harms. Arguments can be made for the use of ASA chemoprevention, particularly if used in populations that may benefit from its CVS harms prevention. However, since observational studies suggest that higher doses and prolonged use improve chemopreventative efficacy, more information is required to clarify the optimal dose, starting age, and duration of use of ASA. In addition, clarification of its effect on CRC incidence and mortality, particularly given the evidence that in patients at average CV risk, all-cause mortality is not reduced with the use of ASA.

Alexander Tsertsvadze, MD, Msc, Nicholas Barrowman, PhD, Catherine Code, MD, FRCPC, Margaret Sampson, MILS, and David Moher, PhD, Director, University of Ottawa Evidence-based Practice Center

This report is based on research conducted by the University of Ottawa Evidence-based Practice Center (EPC), Ottawa, Canada under contract to the Agency for Healthcare Research and Quality (AHRQ)1 , Rockville, MD (Contract No. 290-02-0021). Funding was provided by the Centers for Disease Control and Prevention.

Suggested citation:

Rostom A, Dube C, Lewin G. Use of Aspirin and NSAIDs to Prevent Colorectal Cancer: An Evidence Synthesis. Prepared for the Agency for Healthcare Research and Quality by the University of Ottawa Evidence-based Practice Center at the University of Ottawa, Ottawa Canada, under Contract No. 290-02-0021. AHRQ Publication No. 07-0596-EF-1. Rockville, Maryland: Agency for Healthcare Research and Quality. March 2007.

The findings and conclusions in this document are those of the author(s), who are responsible for its content, and do not necessarily represent the views of AHRQ. No statement in this report should be construed as an official position of AHRQ or of the U.S. Department of Health and Human Services.

The information in this report is intended to help clinicians, employers, policymakers, and others make informed decisions about the provision of health care services. This report is intended as a reference and not as a substitute for clinical judgment.


U.S. Department of Health and Human Services, 540 Gaither Road, Rockville, MD 20850. www​

Bookshelf ID: NBK33464PMID: 20722142


Similar articles in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...