Q J Med 2002; 95: 267-273
© 2002 Association of Physicians
Review |
Translational medicine: targetting cyclo-oxygenase isozymes to prevent cancer
From the Oncology Department, University of Leicester, Leicester Royal Infirmary, Leicester, UK
 |
Introduction |
|---|
 Top Introduction Prostaglandin biosynthesisTargetting COX-1 for...Sparing COX-1 while inhibiting...COX as a target...Data from clinical trialsConclusionsReferences |
|---|
The formulation of aspirin from salicylate by Felix Hoffman in 1897 represents an early example of translating a clinical need to laboratory pharmacology, and consequently bringing a new derivative of a drug to patients. Although Hippocrates recommended a brew of willow leaves (rich in salicylates) for the relief of the pain of childbirth around 400 BC, it was not until the late nineteenth century that the use of salicylic acid became widespread for the relief of pain and fever.1 Since consumption of salicylic acid caused significant gastric irritation, many chemists sought to formulate more tolerable forms of this drug. Felix Hoffman, working in the laboratory of Friedrich Bayer, formulated a pure and stable form of acetyl salicylic acid, which was given the name ‘a-spirin’.2 Based on its success, the first large-scale pharmaceutical company, Bayer & Co., was established.
Over half a century later, as other non-steroidal anti-inflammatory drugs (NSAIDs) with similar chemical structures entered widespread clinical use, their common biological target was recognized as the cyclo-oxygenase (COX) enzyme. Identification and characterization of the COX-2 isozyme,3 and widespread recognition that each year 0.5–2% of people taking a NSAID have a serious gastrointestinal event such as perforation or bleeding4 (causing an estimated 2000 deaths annually in the UK), has stimulated contemporary chemists to do with NSAIDs what Hoffman did with salicylate: try to formulate a safer derivative. Several comprehensive reviews on the relevance of NSAIDs (and their selective COX-2 inhibiting derivatives) to cardiovascular disease, inflammation and cancer have been published in recent years, to which the reader is referred.5–7 This review does not aim to be comprehensive; its goal is to illustrate how our developing knowledge of molecular pharmacology and the prevention of cardiovascular disease can be translated to the clinical chemoprevention of cancer.
|
Prostaglandin biosynthesis |
|---|
|
TopIntroductionProstaglandin biosynthesisTargetting COX-1 for...Sparing COX-1 while inhibiting...COX as a target...Data from clinical trialsConclusionsReferences |
|---|
Prostaglandins are local hormones that initiate and modulate cellular responses involved in physiological processes such as inflammation and platelet aggregation. Collectively, with the thromboxanes, they form a group of oxygenated fatty acids derivatives called prostanoids or eicosanoids (i.e. derived from a 20-carbon-containing polyunsaturated acid). As shown in Figure 1
, a by-product of prostaglandin biosynthesis is the mutagen and carcinogen, malondialdehyde.8
|
The substrate for both COX and lipo-oxygenase (LOX) enzymes is arachidonic acid, an essential polyunsaturated fatty acid consumed in the diet or derived from elongation and desaturation of dietary linoleic acid.9 COX introduces two molecules of O2 into arachidonic acid, to form prostaglandin (PG) endoperoxides, from which the eicosanoids PGE2, PGD2, PGF2
, PGI2 and thromboxane (Tx) A2 are formed (Figure 1
). As postulated by Masferrer in 1990, COX consists of two isozymes, COX-1 and COX-2. Whereas COX-1 protein can be found in most human tissues, and is required for many normal physiological functions such as regulation of renal vascular flow and maintenance of gastric mucosal integrity, COX-2 is expressed in normal humans at very low levels, and only in the brain, kidney, gonads, uterus, bones, tracheal epithelium and small intestine.6,7 In contrast to COX-1, the expression of COX-2 is highly inducible, upregulated in an autocrine fashion by cytokines and growth factors, oxidative stress and sex hormones. The COX-2 isozyme has been implicated in the pathogenesis of most human inflammatory diseases5 and in cancers of the colorectum, breast, head and neck, lung, pancreas, stomach and prostate.6
PGs are not stored, but are synthesized on demand in two kinetically distinct phases.10 In the immediate phase, intracellular mobilization of arachidonic acid by phospholipase A2 results in its supply to the COX-1 isozyme, which is constitutive to most human cells. In the delayed phase, which takes many minutes or many hours depending on a cell's protein synthetic capacity (see below), de novo induction of COX-2 occurs. COX-2 is an isozyme capable of metabolizing lower levels of arachidonic acid to PG endoperoxides than are required for COX-1-mediated catalysis. In many cells, the main PG species formed during the delayed response is PGE2.11 Recent studies have discovered new membrane-associated enzymes required for PGE2 biosynthesis (Figure 2), linked to COX-2 and glutathione metabolism, and therefore classified in the MAPEG (membrane-associated proteins involved in eicosanoid and glutathione metabolism) superfamily.10 Sustained expression of both COX-2 and certain MAPEG enzymes has been shown to result in aberrant growth of cells in vitro.10
|
|
Targetting COX-1 for cardiovascular protection |
|---|
|
TopIntroductionProstaglandin biosynthesisTargetting COX-1 for...Sparing COX-1 while inhibiting...COX as a target...Data from clinical trialsConclusionsReferences |
|---|
Aspirin's effect on ‘platelet stickiness’ was first reported by Morris in 1967.12 Subsequent mechanistic elucidation of this phenomenon provides an example to refute the axiom ‘if a little bit works, a lot works better.’ Aspirin's selectivity (determined by IC50 ratios) for COX-1 over COX-2 is about 5:1, and its tissue specificity can be explained in terms of protein synthesis.9 Non-nucleated cells such as platelets have low protein synthetic capacity. As shown in patients with atherosclerosis, and despite its short half-life in the circulation of approximately 5 min, 40–80 mg aspirin irreversibly acetylates platelet COX-1 in vivo: TxA2 synthesis is thus eliminated until new platelets are produced (usually 3–7 days).13 Normal platelet aggregation requires this eicosanoid. Interestingly, at doses >80 mg, the selectivity of platelet suppression is lost, since COX-1 is eliminated in nucleated cells such as vascular endothelial cells, to degrees which overwhelm these cells’ protein synthetic recovery.13,14
The results of a large number of randomized controlled trials1 have shown that regular use of aspirin reduces the incidence of non-fatal vascular events by 30–40%. As one would expect from the pharmacology described above, there is no clinical evidence to suggest that taking 300 mg daily confers greater protection than a daily dose of 75 mg.15 Prescription of the minimum effective dose is an important concept in medicine, particularly chemoprevention, in order to minimize toxicity. In terms of targetting platelet COX-1, it is unlikely that aspirin will ever be superseded, since it covalently and irreversibly modifies the protein, unlike other NSAIDs, which act by competitive inhibition.9 Unfortunately, the inhibition of COX-1 by aspirin and non-selective NSAIDs is not limited to platelets, and is regarded as the major contributor to the gastrointestinal toxicity of these drugs. In most countries, guidelines exist for the treatment of patients considered at high risk of such toxicity, in which the concurrent administration of gastroprotective agents, such as a proton-pump inhibitors or the synthetic PG analogue misoprostol, is recommended to prevent gastrointestinal complications.16 Proton-pump inhibitors, which reduce NSAID-induced peptic ulcers, are currently less expensive and appear to be better tolerated than misoprostol, but studies using perforation, obstruction and bleeding as endpoints have only been done for misoprostol.17
|
Sparing COX-1 while inhibiting COX-2 |
|---|
|
TopIntroductionProstaglandin biosynthesisTargetting COX-1 for...Sparing COX-1 while inhibiting...COX as a target...Data from clinical trialsConclusionsReferences |
|---|
Felix Hoffman's desire to synthesize a tolerable salicylate was motivated by the suffering of his father, who had severe arthritis. The recent development of ‘designer’ drugs (i.e. designed by understanding of the differences in structures of the COX isozymes) that selectively inhibit COX-2, also called COX-1-sparing drugs, has reduced the risks of gastrointestinal bleeding for thousands of patients with arthritis.7 Like classical NSAIDs, these newer drugs are effective in reducing the inflammation of osteoarthritis and rheumatoid arthritis. Unlike COX-1 inhibitors, however, these drugs do not affect platelet function, and international randomized controlled trials have shown that they reduce gastrointestinal perforation, obstruction and bleeding by 50% or more when compared to older NSAIDs such as ibuprofen or diclofenac.18 However, for patients with multiple pathologies, on aspirin for cardiovascular disease as well as a selective COX-2 inhibitor for arthritis, the incidence of gastrointestinal events may be similar to that for patients on a classical NSAID alone.19
Despite the elegance of this rational approach to drug development, the change from classical NSAIDs to selective COX-2 inhibitors such as celecoxib and rofecoxib has proved somewhat disappointing in terms of clinical tolerability. Although endoscopically-observed lesions in the stomach are fewer for the new agents, the incidences of dyspepsia, abdominal pain and nausea appear to be similar to those induced by conventional NSAIDs.20 The reasons for this gap between theoretical biology and clinical experience are unknown, but may be related to pre-existing gastrointestinal inflammation or the involvement of COX-2 in ulcer healing. Co-administration with other drugs may thus be necessary, severely restricting cost-effectiveness. New drugs derived from indomethacin are in preclinical development, including those that release nitric oxide to protect the gastric mucosa,21 or sulphone and amide derivatives22 which appear to have greater COX-2 selectivity than celecoxib. Greater understanding of enantiomers (mirror images) of drugs such as ibuprofen, which is now available in Austria and Switzerland as the pure (S)-isomer, may also contribute to the refinement of therapeutic effects including isozyme inhibition. The (S)-isomer is more water soluble than the (R)-isomer and consequently has a faster onset of action, but the (R)-isomer may have greater COX-2-inhibiting potential.23 In addition to preclinical assays, such as purified recombinant enzyme systems and transfected cell culture, the clinical selectivity of novel agents should be measured ex vivo using human blood.24
|
COX as a target for preventing cancer |
|---|
|
TopIntroductionProstaglandin biosynthesisTargetting COX-1 for...Sparing COX-1 while inhibiting...COX as a target...Data from clinical trialsConclusionsReferences |
|---|
Retrospective epidemiological studies suggest a decreased incidence of cancers of the oesophagus, stomach, colon and rectum in regular users of NSAIDs.25 The most convincing evidence exists for regular usage of aspirin, which may reduce the incidence of colorectal cancer by up to 50%, although there may be a delay of a decade or more before the benefits of daily usage are seen.26
The role of COX in colorectal cancer provides a contemporary example of translational chemoprevention. Since the first report of elevated levels of PGs in colon cancer tissue,27 significant increases in COX-2 mRNA or protein levels have been observed in up to 90% of sporadic colorectal cancers as well as in 40% of adenomas, generally regarded as premalignant lesions.28 A third of the population of developed countries have detectable adenomas by the age of 50 years, and half the population at 70 years.25 Preclinical evidence for the role of COX-2 in colon carcinogenesis is particularly robust; for example, in mouse models of defects in the adenomatous polyposis coli (APC) tumour suppressor gene (Figure 3).29 In 1983, Waddell and Loughry first reported that the NSAID sulindac reduced the size and number of rectal adenomas in patients with familial adenomatous polyposis (FAP).30 Sulindac (inactive) is reduced to sulindac sulphide (potent COX-inhibitor) by colonic bacteria, and may therefore be more effective than other NSAIDs in this target organ.9 Overexpression of COX-2 and receptors of a related pathway, peroxisome proliferator-activated receptor 
(PPAR), appears greatest in the distal bowel in a chemical rodent model of intestinal carcinogenesis.31 Although COX-1 is thought to be important in the regulation of new blood vessel formation by normal endothelial cells, COX-2 appears more important to the malignant behaviour of carcinoma cells.32 Interestingly, another metabolite of sulindac formed by colonic bacteria, sulindac sulphone, has little or no ability to inhibit COX, but effectively prevents tumour formation in a chemical rodent model of colorectal cancer.33 NSAIDs, particularly aspirin, may therefore possess cancer preventive potential, independent of their ability to inhibit COX.34 Such activity may relate to inhibition of angiogenesis, induction of apoptosis or other pathways involved in inflammation such as nitric oxide synthase.35
|
|
Data from clinical trials |
|---|
|
TopIntroductionProstaglandin biosynthesisTargetting COX-1 for...Sparing COX-1 while inhibiting...COX as a target...Data from clinical trialsConclusionsReferences |
|---|
More recently, interest has focused on the ability of the selective COX-2 inhibitors to prevent polyp formation in patients with FAP, in whom APC gene defects result in a 100% chance of developing colorectal cancer.25 In 2000, celecoxib was approved by the US Food and Drug Administration as an adjunct to usual care, based on the results of a clinical trial in 83 patients with FAP, which demonstrated that 400 mg twice daily for 6 months resulted in 28% fewer polyps than placebo.36 Although piroxicam and sulindac are more potent inhibitors of adenoma formation than celecoxib in mouse models of this disease (R. Jacoby, unpublished data), the selective COX-2 inhibitor was approved because of its potential for less toxicity, particularly important in considering any drug for chemoprevention. As was observed for sulindac in patients with FAP, indefinite treatment may be required to suppress adenoma development.37 In this regard, diet-derived agents likely to cause less toxicity but capable of inhibiting PG biosynthesis, such as green tea or curcumin (from the spice turmeric), may offer clinical promise. Rather than competitive inhibition, curcumin inhibits the transcription of COX-2, in a manner also discovered as a mechanism of action for aspirin and salicylate.38 Greater understanding of the effects of COX-inhibiting agents on signal transduction pathways involving nuclear factor
B (NFB), PPAR, and the epidermal growth factor receptor family,31,39,40 all of which are currently being targetted in clinical trials of novel agents, will enrich the translational approach to drug development. NFB, in particular, represents a molecular target that is responsible for cross-talk between many intracellular signalling pathways, and is an attractive target for altering the chemosensitivity of cells.39
Like the contemporary development of aspirin as a colorectal cancer chemopreventive agent, a structured Phase I–III approach, including measurement of surrogate biomarker levels, is currently being adopted for other COX-inhibiting agents, including those derived from the diet.38,40,41 It is perhaps not surprising that such an approach is also being adopted in patients with established malignancies known to overexpress COX-2, in combination with standard chemotherapy, since overexpression of this isozyme appears to be a prognostic factor for at least one form of human cancer.42 Large-scale clinical trials of selective COX-2 inhibitors currently recruiting patients with cancer or with premalignant conditions are shown in Table 1.
|
|
Conclusions |
|---|
|
TopIntroductionProstaglandin biosynthesisTargetting COX-1 for...Sparing COX-1 while inhibiting...COX as a target...Data from clinical trialsConclusionsReferences |
|---|
The advancement of aspirin over the past century from the treatment of arthritis to the prevention of cardiovascular disease has been accompanied by greater understanding of its pharmacology and the role of COX isozymes. This knowledge has resulted in rationally designed drugs for the treatment of arthritis, which have also been approved for the chemoprevention of colorectal cancer. Because of the high overexpression of COX-2 found in a wide variety of human malignancies, large-scale clinical trials of selective COX-2 inhibitors are currently ongoing in patients with premalignant or malignant lesions of the colon and rectum, oesophagus, bladder and skin. New approaches to the selective inhibition of COX-2 in the laboratory, involving sulphone and amide derivatives of NSAIDs, nitric oxide-releasing NSAIDs and pure enantiomer formulations may offer avenues for clinical cancer chemoprevention with less potential for toxicity. Greater understanding of the effects of NSAIDs and their contemporary derivatives on signal transduction pathways, particularly those involving NF
B, PPARs, and the epidermal growth factor receptor family, will enrich the translational approach to drug development. Targetting COX isozymes demonstrates the two-way dynamic interaction that exists between clinical practice and molecular pharmacology.
|
Acknowledgments |
|---|
I am grateful to Sarah Perkins for help with Figure 3
, and to Sue Spriggs for checking the accuracy of the references. My research is supported by the University Hospitals of Leicester, the Medical Research Council and the Prostate Research Campaign, UK. |
Notes |
|---|
Address correspondence to Dr R.A. Sharma, Oncology Department, University of Leicester, Leicester Royal Infirmary, Leicester LE1 5WW. e-mail: ras20{at}le.ac.uk
|
References |
|---|
|
TopIntroductionProstaglandin biosynthesisTargetting COX-1 for...Sparing COX-1 while inhibiting...COX as a target...Data from clinical trialsConclusionsReferences |
|---|
1. Elwood PC. Aspirin: past, present and future. Clin Med2001; 1:132–7.[Web of Science][Medline]
2. Wilthauer J, Wohlgemut J. Uber aspirine (acetylsalicylsaure). Ther Mh (Halbmh)1899; 13:276.
3. Luong C, Miller A, Barnett J, Chow J, Ramasha C, Browner MF. Flexibility of the NSAID binding site in the structure of human cyclooxygenase-2. Nat Struct Biol1996; 3:927–33.[Web of Science][Medline]
4. Singh G, Rosen Ramey D. NSAID induced gastrointestinal complications: the ARAMIS perspective, 1997. J Rheumatol1998; 25(Suppl. 51):8–16.
5. Hawkey CJ. COX-2 inhibitors. Lancet1999; 353:307–14.[Web of Science][Medline]
6. Prescott SM, Fitzpatrick FA. Cyclooxygenase-2 and carcinogenesis. Biochim Biophys Acta2000; 1470:M68–78.
7. Isakson P, Zweifel B, Masferrer J, et al. Specific COX-2 inhibitors, from bench to bedside. In: Vane JR, Botting JH, eds. Selective COX-2 inhibitors: Pharmacology, Clinical Effects and Therapeutic Potential. London, William Harvey Press, 1997:127–33.
8. Sharma RA, Gescher A, Plastaras JP, et al. Cyclooxygenase-2, malondialdehyde and pyrimidopurinone adducts of deoxyguanosine in human colon cells. Carcinogenesis2001; 9:1557–60.
9.
Marnett LJ. Aspirin and the potential role of prostaglandins in colon cancer. Cancer Res1992; 52:5575–89.
10.
Murakami M, Naraba H, Tanioka T, et al. Regulation of prostaglandin E2 biosynthesis by inducible membrane-associated prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. J Biol Chem2000; 275:32783–92.
11.
Brock TG, McNish RW, Peters-Golden M. Arachidonic acid is preferentially metabolized by cyclooxygenase-2 to prostacyclin and prostaglandin E2. J Biol Chem1999; 274:11660–6.
12. Morris CDW. Acetylsalicylic acid and platelet stickiness. Lancet1967; i:279–80.
13. Weksler BB, Pett SB, Alonso D, Richter RC, Stelzer P. Differential inhibition by aspirin of vascular and platelet prostaglandin synthesis in atherosclerotic patients. N Engl J Med1983; 308:800–5.[Abstract]
14. Vane J, Botting R. Inflammation and the mechanism of action of anti-inflammatory drugs. FASEB J1987; 1:89–96.[Abstract]
15. De Caterina R, Giannessi D, Boem A, et al. Equal antiplatelet effects of aspirin 50 or 324 mg/day in patients after acute myocardial infarction. Thromb Haemost1985; 54:528–32.[Web of Science][Medline]
16. Boers M. NSAIDs and selective COX-2 inhibitors: competition between gastroprotection and cardioprotection. Lancet2001; 357:1222–3.[Web of Science][Medline]
17.
Silverstein FE, Graham DY, Senior JR, et al. Misoprostol reduces serious gastrointestinal complications in patients with rheumatoid arthritis receiving nonsteroidal anti-inflammatory drugs: a randomised, double-blind, placebo-controlled trial. Ann Intern Med1995; 123:241–9.
18.
Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med2000; 343:1520–8.
19.
Silverstein FE, Faich G, Goldstein JL, et al. Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomised controlled trial. JAMA2000; 284:1247–55.
20. Anon. Are rofecoxib and celecoxib safer NSAIDs? Drug Ther Bulletin2000; 38:81–6.
21. Takeuchi K, Mizoguch H, Araki H, Komoike Y, Suzuki K. Lack of gastric toxicity of nitric oxide-releasing indomethacin, NCX-530, in experimental animals. Dig Dis Sci2001; 46:1805–18.[Web of Science][Medline]
22. Timofeevski SL, Marnett LJ. Isoform-selective interaction of cyclooxygenase-2 with indomethacin amides studied by real-time fluorescence and inhibition kinetics. Biochem2001; 40:254.
23. Tucker GT. Chiral switches. Lancet2000; 355:1085–7.[Web of Science][Medline]
24.
Patrignani P, Panara MR, Greco A, et al. Biochemical and pharmacological characterization of the cyclooxygenase activity of human blood prostaglandin endoperoxide synthases. J Pharmacol Exp Ther1994; 271:1705–12.
25.
Langman MJS, Cheng KK, Gilman EA, Lancashire RJ. Effect of anti-inflammatory drugs on overall risk of common cancer: a case-control study in a general practice research database. Br Med J2000; 320:1642–6.
26. Collet J-P, Sharpe C, Belzile E, Boivin J-F, Hanley J, Abenhaim L. Colorectal cancer prevention by non-steroidal anti-inflammatory drugs: effects of dosage and timing. Br J Cancer1999; 81:62–8.[Web of Science][Medline]
27. Bennett A, Tacca MD, Stamford IF, Zebro T. Prostaglandins from tumours of human large bowel. Br J Cancer1977; 35:881–4.[Web of Science][Medline]
28.
Giovannucci E, Rimm EB, Stampfer MJ, Colditz GA, Ascherio A, Willett WC. Aspirin use and the risk for colorectal cancer and adenoma in male health professionals. Ann Intern Med1994; 121:241–6.
29. Oshima M, Dinchuk JE, Kargman SL, et al. Suppression of intestinal polyposis in APCD716 knockout mice by inhibition of cyclooxygenase 2 (COX-2). Cell1996; 87:803–9.[Web of Science][Medline]
30. Waddell WR, Loughry RW. Sulindac for polyposis of the colon. J Surg Oncol1983; 24:83–7.[Web of Science][Medline]
31. Roy HK, Karolski WJ, Ratashak A. Distal bowel selectivity in the chemoprevention of experimental colon carcinogenesis by the non-steroidal anti-inflammatory drug nabumetone. Int J Cancer2001; 92:609–15.[Web of Science][Medline]
32. Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN. Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell1998; 93:705–16.[Web of Science][Medline]
33. Alberts DS, Hixson L, Ahnen D, et al. Do NSAIDs exert their colon cancer chemoprevention activities through the inhibition of mucosal prostaglandin synthetase? J Cell Biol Suppl1995; 22:18–23.
34. Patten EJ, DeLong MJ. Effects of sulindac, sulindac metabolites, and aspirin on the activity of detoxification enzymes in HT-29 human colon adenocarcinoma cells. Cancer Lett1999; 147:95–100.[Web of Science][Medline]
35. Sharma RA, Harris AL, Dalgleish AG, Steward WP, O'Byrne KJ. Angiogenesis as biomarker and target in cancer chemoprevention. Lancet Oncol2001; 2:726–32.[Web of Science][Medline]
36.
Steinbach G, Lynch PM, Phillips RKS, et al. The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med2000; 342:1946–52.
37.
Giardiello FM, Hamilton SR, Krush AJ, Piantadosi S, Hylind LM, Celano P. Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med1993; 328:1313–16.
38. Gescher AJ, Sharma RA, Steward WP. Cancer chemoprevention by dietary constituents: A salutary tale of failure and promise. Lancet Oncol2001; 2:371–9.[Medline]
39.
Waddick KG, Uckun FM. Innovative treatment programs against cancer II: nuclear factor-B as a molecular target. Biochem Pharmacol1999; 57:9–17.[Web of Science][Medline]
40.
Ruffin MT, Krishnan K, Rock CL, et al. Suppression of human colorectal mucosal prostaglandins: determining the lowest effective aspirin dose. J Natl Cancer Inst1997; 89:1152–60.
41.
Plummer SM, Hill KA, Festing MFW, Steward WP, Gescher A, Sharma RA. Curcumin inhibits prostaglandin biosynthesis via cyclooxygenase-2 in human blood. Cancer Epidemiol Biomarkers Prev2001; 10:1295–9.
42.
Khuri FR, Wu H, Lee JJ, et al. Cyclooxygenase-2 overexpression is a marker of poor prognosis in stage I non-small cell lung cancer. Clin Cancer Res2001; 7:861–7.
43. Sharma RA, Manson MM, Gescher A, Steward WP. Colorectal cancer chemoprevention: biochemical targets and clinical development of promising agents. Eur J Cancer2001; 37:12–22.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||