Q J Med 2001; 94:445± 448

Commentary
QJM
Salicylic acid: a link between aspirin, diet and
the prevention of colorectal cancer
J.R. PATERSON and J.R. LAWRENCE
From the Departments of Biochemistry and Medicine, Dumfries and Galloway Royal
Infirmary, Dumfries, UK

Summary
Aspirin was introduced into clinical practice more to reduce the risk of colorectal cancer. Serum
than 100 years ago. This unique drug belongs to a salicylic acid concentrations are greater in veget-
family of compounds called the salicylates, the arians than non-vegetarians, and there is overlap
simplest of which is salicylic acid, the principal between concentrations in vegetarians and those
metabolite of aspirin. Salicylic acid is responsible taking low-dose aspirin. We propose that the
for the anti-inflammatory action of aspirin, and cancer-preventive action of aspirin is due to its
may cause the reduced risk of colorectal cancer principal metabolite, salicylic acid, and that diet-
observed in those who take aspirin. Yet salicylic ary salicylates can have the same effect. It is also
acid and other salicylates occur naturally in fruits possible that natural salicylates contribute to the
and plants, while diets rich in these are believed other recognized benefits of a healthy diet.

Aspirin (acetylsalicylic acid) occupies a unique Salicylic acid is the principal metabolite of aspirin,
place in medicine. Since its clinical introduc- aspirin having a half-life of -30 min.7 Many of
tion in 1899, we have become familiar with this the salicylates share the same properties as aspirin,
drug and its many surprising effects, including although its anti-platelet action is specific. Extracts
reduced risk of cardiovascular disease and pos- of plants, such as the willow and meadowsweet
sibly colorectal cancer, as well as its analgesic, (which contain various compounds metabolized to
anti-inflammatory and anti-platelet actions. Aspirin salicylic acid), as well as salicylic acid prepared
is thought to reduce the risk of colorectal cancer, synthetically, pre-dated aspirin in the treatment
perhaps by as much as 40%, a property that is of inflammatory conditions. The occurrence of
shared by other non-steroidal anti-inflammatory `natural' salicylates, such as salicylic acid present
drugs (NSAIDS).1,2 Evidence for this effect comes in strawberries and other fruit, was discussed in the
from multiple epidemiological studies, most of Lancet of 1903,8 and the matter of whether `natural'
which have found that aspirin reduces the risk salicylates were superior to synthetic salicylates
of colorectal adenoma3 and carcinoma,4 as well was the subject of a JAMA editorial in 19139
as from animal models where aspirin inhibits (no superiority could be shown).
chemical-induced colonic carcinogenesis.5,6 Aspirin Today's healthy diet mantra of, `five servings
belongs to a family of compounds called the of fruit and vegetables a day', may well be good
salicylates, the simplest of which is salicylic acid. advice yet, other than in the most vague general

Address correspondence to Dr J.R. Paterson, Department of Biochemistry, Dumfries and Galloway Royal Infirmary,
Bankend Road, Dumfries DG1 4AP. e-mail: [email protected]
ß Association of Physicians 2001
446 J.R. Paterson and J.R. Lawrence

terms, we are unable to explain its basis. Nutritional

research has already examined, and continues
to assess, various plant constituents. However, the
responsible constituent(s) for producing better
health remains, as yet, elusive. One compound
which we believe should be more fully considered
is salicylic acid, a compound which plays a cent-
ral role in the development of local and systemic
disease resistance to pathogen infection in plants.10
Salicylic acid is present in fruits and vegetables,
with herbs and spices being a particular rich
source.11,12
Janssen et al.13 have concluded that a normal
diet provides only 0±6 mg of salicylates daily, and
no measurable aspirin.12 The estimated dietary
intake of salicylates was based on urinary analysis
in 17 volunteers eating a wide variety of diets, and
it was suggested that the intake was probably too
low to affect disease risk. Urinary salicylate con-
centrations, however, provide little information on
blood or tissue concentrations, since salicylic acid Figure 1. Hypothesis: salicylic acid, an anti-inflammat-
is extensively metabolized, and its renal excretion ory drug which reduces the risk of colorectal cancer and
is influenced by factors such as urinary pH and which is common to both aspirin and a plant-based diet.
flow, and the presence of other organic acids.7 In a
study of 10 subjects given 40.5 mg aspirin, the
mean peak plasma salicylic acid concentration was inflammation, but which may be also present
11.8 mmol/l, with a SD of 8.18 mmol/l, indicating constitutively in some tissues.19 COX-2 expression
a large inter-individual variability in the salicylic in animal models is associated with tumour pro-
acid concentrations after the same dose of aspirin.14 gression.20 Aspirin and the other NSAIDS are
In addition, no studies have investigated whether believed to reduce the risk of colorectal cancer,
a dietary salicylate intake of a few milligrams has at least in part, by inhibiting COX-2 activity.21
health benefits or not. Paterson et al.15 identified The anti-inflammatory activity of aspirin is due
salicylic acid and two other salicylates as normal to its major metabolite, salicylic acid,22 yet salicylic
constituents of serum in individuals not taking sali- acid is inactive against COX in either broken cells
cylate drugs. Salicylates were found to be present in or purified enzyme preparations.23 It was, however,
every serum sample analysed. The same group went found to be a weak inhibitor of both COX isoforms
on to show that higher serum concentrations of in intact cells. How then does salicylic acid exert
salicylic acid were present in vegetarians than in its anti-inflammatory action? Salicylic acid appears
non-vegetarians, and that there was overlap in the to inhibit the transcription of the COX-2 gene,24,25
serum concentrations between vegetarians and inhibition occurring at concentrations found in
those taking aspirin (75 mg daily).16 those taking low-dose aspirin. The salicylic acid
We postulate that dietary salicylates have concentration which inhibited COX-2 transcription
beneficial properties because of their effect on by 50% was estimated to be 5000 nmol/l, although
the `inflammatory process', a concept that would even concentrations as low as 100 nmol/l appeared
explain why both aspirin and a diet rich in fruits to have some effect.25 The median serum concen-
and vegetables help prevent colorectal cancer tration of salicylic acid in a group of vegetarians
(Figure 1), and probably other inflammatory dis- not taking salicylate drugs was 107 nmol/l, with
eases. Inflammatory processes are involved in the highest concentration being 2468 nmol/l.16 We
carcinogenesis and cancer growth.17 Most human propose that serum concentrations of salicylic acid,
colorectal cancers express high levels of cyclooxy- arising at least in part from dietary plant sources, are
genase-2 (COX-2), a key enzyme catalysing the sufficiently high in some cases to reduce COX-2
conversion of arachidonic acid into prostaglandins, gene transcription. This proposed action of dietary
contributing to the inflammatory response.18 There salicylates does not exclude the possibility that
are two isoforms of COX; COX-1 is constitutively other components of fruits and vegetables have
expressed in platelets and other tissues, and similar properties, or that the salicylates have
COX-2 is an enzyme induced by various growth actions in addition to their inhibition of COX-2,
factors, interleukins and lipopolysaccharides in as there is evidence that NSAIDS may have a
 Salicylic acid 447

chemopreventive effect through COX-independent to different endogenous salicylic acid concentra-

mechanisms.21 tions, with the so-called baseline possibly reflecting
In plants subjected to pathogen attack, salicylic salicylic acid concentration as well as other
acid contributes to the containment of infection, the influences. There has been one randomized trial
activation of cell death and the induction of local of aspirin (American Physicians Study) which
and systemic disease resistance.26 Salicylic acid reported no difference in colorectal cancer inci-
achieves these effects through increasing defence dence between aspirin (325 mg on alternate days)
gene expression, potentiating cell death and alter- and placebo.37 This trial, however, was primarily
ing the expression or activity of various enzymes. designed to examine the effect of aspirin on
Many of these actions occur in plants at salicylic cardiovascular disease and involved only a 5-year
acid concentrations comparable to those present period of continuous use, with follow-up for
in patients who take low-dose aspirin. It is pos- 12 years, periods of time which may have been
sible that some genes common to plants and insufficient to observe an effect.21
animals which govern ancient conserved proteins We believe that our hypothesis can be tested
(or regions thereof ), are modulated by salicylic to establish its validity. Low serum salicylic acid
acid.27 One of the major problems in salicylate concentrations should increase the risk of develop-
research, however, is that salicylic acid affects ing colorectal adenomas and adenocarcinomas.
many different biological systems when it is present The administration of small doses of salicylic acid
at concentrations of mmol/lÐconcentrations much should be investigated to determine whether this
above those normally found in patients taking low- compound has an effect on COX-2 gene transcrip-
dose aspirin.28 Our understanding of the actions tion or any other anti-inflammatory effect. Serum
of aspirin and salicylic acid is not necessarily or urine measurements of salicylates would be
enhanced by studies involving such concentrations. better undertaken than dietary intakes in such
In a broader context, dietary salicylates, like studies, since there is significant variability between
aspirin, may have benefits in regard to other individuals in the metabolism and excretion of
`inflammatory' pathologies in which the COX-2 salicylic acid. In addition, there may be variability
gene is induced. COX-2 gene induction occurs in in the salicylate content of dietary plants, deter-
some oesophageal and gastric cancers,29,30 as well mined in part through their differing exposure to
as in monocytes, macrophages and fibroblasts,31 pathogen attack. Assessment of diet also needs
cells which are involved in atherosclerosis, now to incorporate an examination of the different
recognized as a chronic inflammatory disease.32 types of plant foods, such as spices and herbs.
However, the evidence for the putative anti- Our basic hypothesis is that the chemopreventive
inflammatory effects of aspirin in these conditions action of aspirin is due primarily to its principal
is less well established. In fact, it is possible that metabolite, salicylic acid, and dietary salicylates
inhibition of COX-2 in cases of congestive cardiac can have the same effect (Figure 1). It is also emin-
failure may have deleterious effects.31 ently possible that natural salicylates contribute to
Any hypothesis which purports to explain how the other recognized benefits of a healthy diet.
a particular component of our diet helps reduce
the risk of colorectal cancer, must include recogni-
tion of its limitations, as well as its strengths. Serum Acknowledgements
concentrations of salicylic acid after aspirin admin- We thank Professor J. Little and Drs M. Murphy,
istration are higher than those observed in people M. McMahon and F. Toolis for their comments,
not taking salicylate drugs. This suggests that, even Mr C. Murray for his assistance with the figure
if the chemopreventive action of aspirin is mainly and Mrs V. Reid for typing the paper.
dependent on the formation of salicylic acid,
dietary salicylates may reduce the risk of colo-
rectal cancer much less than aspirin. No one as yet References
knows what dose of aspirin (or salicylic 1. Thun MJ, Namboodiri MM, Calle EE, Flanders WD,
acid) is required to produce chemoprevention of Heath Jr. CW. Aspirin use and risk of fatal cancer. Cancer
cancer.33 Aspirin at a dose of 81 mg was found Res 1993; 53:1322±7.
to reduce colorectal prostaglandin concentrations 2. International Agency for Research on Cancer. IARC
or formation in two studies14,34 but not in another.35 Handbooks of Cancer Prevention: Non-Steroidal
Anti-Inflammatory Drugs 1997; 1:1±202.
In these and another study,36 examination of
the baseline tissue prostaglandin E2 concentrations 3. Garcia Rodriguez LA, Huerta-Alvarez C. Reduced incidence
of colorectal adenoma among long-term users of non-
before aspirin administration reveals great vari- steroidal antiinflammatory drugs: a pooled analysis of
ability.14,34±36 Some of the variability in tissue pros- published studies and a new population-based study.
taglandin concentrations may be due to exposure Epidemiology 2000; 11:376±81.
448 J.R. Paterson and J.R. Lawrence

4. Garcia Rodriguez LA, Huerta-Alvarez C. Reduced risk of 22. Higgs GA, Salmon JA, Henderson B, Vane JR.
colorectal cancer among long-term users of aspirin and Pharmacokinetics of aspirin and salicylate in relation
nonaspirin nonsteroidal antiinflammatory drugs. to inhibition of arachidonate cyclooxygenase and anti-
Epidemiology 2001; 12:88±93. inflammatory activity. Proc Natl Acad Sci USA 1987;
5. Craven PA, DeRubertis FR. Effects of aspirin on 84:1417±20.
1,2-dimethylhydrazine-induced colonic carcinogenesis. 23. Mitchell JA, Akarasereenont P, Thiemermann C, Flower RJ,
Carcinogenesis 1992; 13:541±6. Vane JR Selectivity of nonsteroidal antiinflammatory drugs
6. Reddy BS, Rao CV, Rivenson A, Kelloff G. Inhibitory effect as inhibitors of constitutive and inducible cyclooxygenase.
of aspirin on azoxymethane-induced colon carcinogenesis Proc Natl Acad Sci USA 1994; 90:11 693±7.
in F344 rats. Carcinogenesis 1993; 14:1493±7. 24. Wu KK, Sanduja R, Tsai A-L, Ferhanoglu B,
7. Needs CJ, Brooks PM. Clinical pharmacokinetics of the Loose-Mitchell DS. Aspirin inhibits interleukin 1-induced
salicylates. Clinical Pharmacokinetics 1985; 10:164±77. prostaglandin H synthase expression in cultured
endothelial cells. Proc Natl Acad Sci USA 1991;
8. Anon. Salicylic acid in strawberries. Lancet 1903;
88:2384±7.
April 25:1187.
25. Xu X-M, Sansores-Garcia L, Chen X-M, Matijevic-Aleksic N,
9. Anon. `Natural' and synthetic salicylates. JAMA 1913;
Du M, Wu KK. Suppression of inducible cyclooxygenase
Sept 20:968. 2 gene transcription by aspirin and sodium salicylate.
10. Dangl J. Plants just say NO to pathogens. Nature 1998; Proc Natl Acad Sci USA 1999; 96:5292±7.
394:525±6.
26. Dempsey D'MA, Shah J, Klessig DF. Salicylic acid and
11. Robertson GL, Kermode WJ. Salicylic acid in fresh and disease resistance in plants. Crit Rev Plant Sci 1999;
canned fruit and vegetables. J Sci Food Agric 1981; 18:547±75.
32:833±6.
27. Eastwood MA. A molecular biological basis for the
12. Venema DP, Hollman PCH, Janssen PLTMK, Katan MB. nutritional and pharmacological benefits of dietary plants.
Determination of acetylsalicylic and salicylic acid in foods, Q J Med 2001; 94:45±8.
using HPLC with fluorescence detection. J Agric Food Chem 28. Wu KK. Aspirin and salicylate. An old remedy with a new
1996; 44:1762±7. twist. Circulation 2000; 102:2022±3.
13. Janssen PLTMK, Hollman PCH, Reichman E, Venema DP,
29. Ratnasinghe D,Limburg P, Tangrea JA. Cyclooxygenase-2,
van Staveren WA, Katan MB. Urinary salicylate excretion in non-steroidal anti-inflammatory drugs and cancer of the
subjects eating a variety of diets shows that amounts of esophagus. Gastroenterol Int 1999; 12:178±81.
bioavailable salicylates in foods are low. Am J Clin Nutr
1996; 64:743±7. 30. Ristimaki A, Honkanen N, Jankala H, Sipponen P,
Harkonen M. Expression of cyclooxygenase-2 in
14. Ruffin IV MT, Krishnan K, Rock CL, et al. Suppression of human gastric carcinoma. Cancer Res 1997;
human colorectal mucosal prostaglandins: determining the
57:1276±80.
lowest effective aspirin dose. J Natl Cancer Inst 1997;
89:1152±60. 31. Wu KK. Cyclooxygenase-2 induction in congestive heart
failure. Friend or foe. Circulation 1998; 98:95±6.
15. Paterson JR, Blacklock CJ, Campbell G, Wiles D, Lawrence
JR. The identification of salicylates as normal constituents of 32. Ross R. Atherosclerosis: an inflammatory disease. N Engl
serum: a link between diet and health? J Clin Pathol 1998; J Med 1999; 340:115±26.
51:502±5. 33. Thun MJ. Beyond willow bark: aspirin in the prevention of
16. Blacklock CJ, Lawrence JR, Wiles D, et al. Salicylic acid chronic disease. Epidemiology 2000; 11:371±4.
concentrations in the serum of subjects not taking aspirin. 34. Barnes CJ, Hamby-Mason RL, Hardman WE, Cameron IL,
Comparison of salicylic acid concentrations in the serum of Speeg V, Lee M. Effect of aspirin on prostaglandin
vegetarians, non-vegetarians and patients taking low-dose E2 formation and transforming growth factor a expression
aspirin. J Clin Pathol 2001; 54:553±5. in human rectal mucosa from individuals with a history
17. Balkwill F, Mantovani A. Inflammation and cancer: back to of adenomatous polyps of the colon. Cancer Epidemiol
Biomarkers Prevention 1999; 8:311±15.
Virchow? Lancet 2001; 357:539±45.
35. Cryer B, Feldman M. Effects of very low dose daily,
18. Eberhart CE, Coffey RJ, Radhika A, Giardiella FM,
long-term aspirin therapy on gastric, duodenal and
DuBois RN. Up-regulation of cyclooxygenase 2 gene
rectal prostaglandin levels and on mucosal injury
expression in human colorectal adenomas and adenocarci-
in healthy humans. Gastroenterology 1999;
nomas. Gastroenterology 1994; 107:1183±8.
117:17±25.
19. Lipsky PE, Brooks P, Crofford LJ, et al. Unresolved issues in 36. Frommel TO, Dyavanapalli M, Oldham T, et al. Effect of
the role of cyclooxygenase-2 in normal physiologic aspirin on prostaglandin E2 and leukotriene B4 production in
processes and disease. Arch Intern Med 2000; 160:913±20.
human colonic mucosa from cancer patients. Clin Cancer
20. Williams CS, Tsujii M, Reese J, Dey SK, DuBois RN. Host Res 1997; 3:209±13.
cyclooxygenase-2 modulates carcinoma growth. J Clin 37. Sturmer T, Glynn RJ, Lee I-M, Manson JE, Buring JE,
Invest 2000; 105:1589±94. Hennekens CH. Aspirin use and colorectal cancer: post-trial
21. Janne PA, Mayer RJ. Chemoprevention of colorectal cancer. follow-up data from the physicians' health study. Ann Intern
N Engl J Med 2000; 342:1960±8. Med 1998; 128:713±20.