COX-2
Progress in Experimental Tumor Research Vol. 37
Series Editor
Joseph R. Bertino, New Brunswick, N.J.
Basel · Freiburg · Paris · London · New York · Bangalore · Bangkok · Singapore · Tokyo · Sydney
COX-2 A New Target for Cancer Prevention and Treatment
Volume Editors
Andrew J. Dannenberg, New York, N.Y. Raymond N. DuBois, Nashville, Tenn.
32 figures, 6 in color and 20 tables, 2003
Basel · Freiburg · Paris · London · New York · Bangalore · Bangkok · Singapore · Tokyo · Sydney
Progress in Experimental Tumor Research Founded 1960 by F. Homburger, Cambridge, Mass.
Andrew J. Dannenberg
Raymond N. DuBois
Department of Medicine Weill Medical College of Cornell University 525 East 68th Street New York, NY 10021 (USA)
Department of Medicine Vanderbilt University Medical Center 1161 21st Ave. South Nashville, TN 37232-2279 (USA)
Library of Congress Cataloging-in-Publication Data Cox-2 : a new target for cancer prevention and treatment / volume editor, Joseph R. Bertino. p. ; cm. – (Progress in experimental tumor research ; vol. 37) Includes bibliographical references and indexes. ISBN 3–8055–7536–X (hard cover : alk. paper) 1. Cyclooxygenase 2–Inhibitors–Therapeutic use. 2. Cancer–Chemotherapy. I. Bertino, Joseph R. II. Series. [DNLM: 1. Neoplasms–drug therapy. 2. Cyclooxygenase Inhibitors–therapeutic use. QZ 267 C877 2003] RC271.C78C69 2003 616.99⬘4052–dc21 2003044688
Bibliographic Indices. This publication is listed in bibliographic services, including Current Contents® and Index Medicus. Drug Dosage. The authors and the publisher have exerted every effort to ensure that durg selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher. © Copyright 2003 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) www.karger.com Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel ISBN 3–8055–7536–X
Contents
VII Preface 1 Epidemiology of Non-Steroidal Anti-Inflammatory Drugs and Cancer Baron, J.A. (Lebanon, N.H.) 25 Pharmacology of COX-2 Inhibitors Isakson, P.C. (Peapack, N.J.) 52 Regulation of COX-2 Expression in Human Cancers Dixon, D.A. (Nashville, Tenn.) 72 Cyclooxygenase-2 and Skin Carcinogenesis Fürstenberger, G.; Marks, F.; Müller-Decker, K. (Heidelberg) 90 The Role of COX-2 in Breast and Cervical Cancer Dannenberg, A.J.; Howe, L.R. (New York, N.Y.) 107 Cyclooxygenase-2: A Target for the Prevention and Treatment of Cancers of the Upper Digestive Tract Altorki, N.K.; Subbaramaiah, K.; Dannenberg, A.J. (New York, N.Y.) 124 Cyclooxygenase-2 and Colorectal Cancer DuBois, R.N. (Nashville, Tenn.) 138 Cyclooxygenase-2 in Lung Cancer Dubinett, S.M.; Sharma, S.; Huang, M.; Dohadwala, M.; Pold, M.; Mao, J.T. (Los Angeles, Calif.) 163 COX-2 Inhibitors and Other NSAIDs in Bladder and Prostate Cancer Sabichi, A.L.; Lippman, S.M. (Houston, Tex.)
179 Therapeutic Potential of Selective Cyclooxygenase-2 Inhibitors in the Management of Tumor Angiogenesis Gately, S. (Lake Forest, Ill.); Kerbel, R. (Toronto) 193 Potential for Combined Modality Therapy of Cyclooxygenase Inhibitors and Radiation Saha, D.; Choy, H. (Nashville, Tenn.) 210 Non-Steroidal Anti-Inflammatory and Cyclooxygenase-2-Selective Inhibitors in Clinical Cancer Prevention Trials Hawk, E.T.; Viner, J.L.; Umar, A. (Bethesda, Md.) 243 Chemotherapy with Cyclooxygenase-2 Inhibitors in the Treatment of Malignant Disease: Pre-Clinical Rationale and Preliminary Results of Clinical Trials Blanke, C.D. (Portland, Oreg.); Masferrer, J.L. (Chesterfield, Mo.) 261 Role of COX-Independent Targets of NSAIDs and Related Compounds in Cancer Prevention and Treatment Soh, J.-W.; Weinstein, I.B. (New York, N.Y.) 284 Author Index 285 Subject Index
Contents
VI
Preface
Over the last decade, cancer biologists have shifted their emphasis from developing global inhibitors of cell growth (cytotoxic agents) to specifically targeting molecular pathways known to be involved in cell transformation. One of the best examples of the potential success for this kind of strategy is the development of the drug, Gleevec, which inhibits the Bcr-Abl tyrosine kinase enzyme which is known to be involved in the pathogenesis of Philadelphia chromosome-positive (Ph⫹) chronic myeloid leukemia. Hence, a number of investigators have focused their efforts on evaluating the usefulness of targeting specific molecules in key signaling pathways in the hope that this would inhibit tumor growth. The cyclooxygenase-2 (COX-2) enzyme has been utilized as a target for the treatment of patients with chronic inflammatory diseases, such as rheumatoid arthritis. In these diseases, there is a long history of the use of nonsteroidal anti-inflammatory agents (NSAIDs) which typically inhibit both COX-1 and -2. Work over the past 5 years has shown that a significant portion of the gastrointestinal toxicity from NSAIDs use arises from inhibition of COX-1. With the development of new agents (celecoxib, rofecoxib and valdecoxib) which selectively inhibit the COX-2 enzyme, an improved GI safety profile was obtained with equivalent efficacy. Numerous investigators have found increased levels of COX-2 in both pre-malignant and malignant tissues. Recently, two different groups have shown directly that expressing the COX-2 gene in breast or skin tissue leads to a dramatic increase in risk for breast or skin carcinoma. Furthermore, animals engineered to be COX-2 deficient are protected against developing both intestinal and skin tumors.
VII
This book represents an ‘up-to-date’ review of the entire field of COX-2 cancer biology. We have been able to solicit the world’s experts in this area of research and bring all of their expertise together in this volume. Indeed, this is a very fast-paced field and it has been difficult to keep up with the most current developments. We have included a review of the key issues in cancer biology where a link to the COX-2 pathway has been described. We have also provided two chapters reviewing the clinical trials underway and those planned in the near future testing the efficacy of selective COX-2 inhibitors for prevention and/or treatment of cancer. The selective COX-2 inhibitors have shown promise when given in combination with other agents and as a radiation-sensitizer for cancer treatment, however, their ultimate use will not be clear until the clinical trials underway have been completed and carefully analyzed. Andrew J. Dannenberg Raymond N. DuBois
Preface
VIII
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 1–24
Epidemiology of Non-Steroidal Anti-Inflammatory Drugs and Cancer John A. Baron Departments of Medicine, and Community and Family Medicine, Dartmouth-Hitchcock Medical Center, Lebanon, N.H., USA
Introduction
The possibility that non-steroidal anti-inflammatory drugs (NSAIDs) may have anti-carcinogenic effects is derived from observations that many cancers overproduce prostaglandins and the knowledge that NSAIDs inhibit the production of these compounds. These observations led to animal studies in which NSAIDs have consistently exhibited important effects on experimental carcinogenesis. Their activity is present either when given weeks after carcinogen administration, or during the early promotion/late initiation phases of carcinogenesis. The anti-cancer effect is reversible; tumor occurrence increases shortly after discontinuation of the agent, and fairly rapidly equals that in the untreated animals. In murine analogues of human familial polyposis (in which the animals carry a germline APC mutation), NSAIDs (usually sulindac has been studied) inhibit tumor formation and cause regression of existing tumors (see chapter by DuBois, pp 124–137). There has also been considerable human epidemiological data and experimental animal data suggesting that NSAIDs may interfere with carcinogenesis in the large bowel and other organs. However, for several reasons, the observational (non-randomized) investigation of the association between NSAID use and cancer risk is not straightforward. NSAIDs may be taken for a variety of reasons (e.g. inflammatory diseases, pain syndromes). These clinical problems may themselves be directly related to cancer risk, as has been suggested for rheumatoid arthritis, for example [1, 2]. Also, the conditions that prompt individuals to use NSAIDs on their own (or seek medical treatment that results
in prescription NSAIDs) could plausibly lead to enhanced diagnosis of cancer. Indeed, increased testing for cancer and greater numbers of physician visits have been observed in NSAID users in comparison to non-users [3, 4]. Both of these processes would tend to generate a direct association between NSAID use and cancer risk, and cause investigators to underestimate any protective effects of the drugs. On the other hand, individuals who take aspirin for the prevention of atherosclerotic vascular disease may be particularly health conscious, and have an inherently lower risk of some cancers. This would tend to lead investigators to overestimate the preventive potential of NSAIDs. Of course in all observational studies, it is reported aspirin use, not actual use, that is the subject of the analysis. A particular problem is that over-thecounter (OTC) NSAIDs such as aspirin may be taken in irregular patterns; it may be hard to accurately measure their intake and distinguish use from that of other OTC drugs such as acetaminophen. Discordance between actual and reported use has an unpredictable effect on the assessment of the relationship between NSAIDs and cancer. If the errors are ‘random’, associations would tend to be obscured. However, in case-control studies, cases may report use differently than controls and so generate an artifactual association of drug use with cancer occurrence. Moreover, in studies of any design, misreporting could result in the effects of one analgesic being attributed to another. For these reasons, it is particularly desirable for studies to assess use of non-NSAID analgesics as well as aspirin (and other OTC analgesics). A lack of relationship for acetaminophen, for example, would then strengthen the validity of associations observed for aspirin. Another difficulty is that early symptoms of undiagnosed cancer may prompt affected individuals to use NSAIDs for symptomatic relief. This would create an apparent association between NSAID use and cancer risk, and lead investigators to underestimate any protective effect of these drugs. NSAIDs often cause upper gastrointestinal symptoms, and these problems can lead to physician visits and earlier diagnosis of cancer, particularly gastrointestinal cancer. It is also possible that individuals with early cancer symptoms would avoid NSAIDs because of these NSAID side effects. This may be a particular problem for investigations of cancer of the upper gastrointestinal tract. It is virtually impossible for any single observational study to avoid all of these potential biases, and it is often difficult to assess the extent to which the resulting distortions affect a study. However, many of the potential problems can be individually averted through appropriate research strategies, and a series of observational studies of various designs in different populations can elucidate the effect of NSAIDs on cancer risk. Even so, because of the potential limitations of observational research, clinical trials are generally understood to provide the strongest evidence for – or against – a cancer-preventive effect of
Baron
2
NSAIDs. Unfortunately, relatively few trials of NSAIDs can have cancer as the endpoint because of the large sample sizes and long duration of follow-up that is required.
Indirect Epidemiological Evidence of NSAID Effects on Cancer Risk
Indirect evidence that NSAIDs might decrease the risk of cancer may be derived from findings that patients with diseases treated by these drugs have a reduced risk of cancer. This evidence is weak, however, since the disease processes themselves – or other treatments for them – may themselves increase or decrease cancer risk. Nonetheless, such studies are useful for generating hypotheses regarding the associations between NSAID use and cancer risk. This issue has been most intensely studied with regard to rheumatoid arthritis (RA) and colorectal cancer. In several investigations in which RA patients were followed over time, typically through the use of administrative and research databases, a reduced risk of colorectal cancer was found [2, 5–9]. In some small studies there was no association [10; see further references in 6]. Two case-control studies of colorectal cancer reported data consistent with these findings. In one of these, fewer colon cancer cases than controls reported a lower prevalence of diseases treated with NSAIDs (e.g., angina pectoris, atherosclerotic diseases, osteoarthrosis, etc.) [11]. In the other investigation, colorectal cancer cases reported a diagnosis of arthritis less often than controls [12]. There is evidence that this association is due to NSAID treatment: one study reported that RA patients did not have a reduced risk of colorectal adenomas after adjustment for NSAID use [13]. There are limited data available regarding the relationship between RA and cancer at other anatomic sites. Most [2, 5, 6], but not all [7], studies suggest a reduced risk of stomach cancer. In contrast, findings regarding the risk of respiratory (mostly lung) cancer suggest a modestly increased risk in RA patients [2, 5–7, 10], although one study found no association [9]. Risk of female breast cancer may be modestly lower in RA patients [6, 7, 9] although some (smaller) studies reported no association [2, 5, 10]. There appears to be no substantial association between RA diagnosis and prostate cancer risk [2, 6, 7, 9]. RA and other arthritis patients may have an increased risk of hematological malignancies [2, 5–7, 14]. These studies suggest that it would be useful to assess more formally whether NSAID use protects against cancers of the colorectum, perhaps the stomach, and conceivably the breast. More direct, stronger, data regarding these – and other – cancer sites are summarized below.
NSAIDs and Cancer Epidemiology
3
Colorectal Neoplasia: Observational Studies
Epidemiological investigation has repeatedly indicated a beneficial effect of NSAIDs, particularly aspirin, on colorectal cancer risk. In one of the first studies to address this topic, Rosenberg et al. [15] reported approximately a 50% lower risk of large bowel cancer in persons who regularly used NSAIDs, principally aspirin. The reduced risk was apparent for both men and women, for younger and older subjects, and for colon and rectal cancers, but was not maintained after cessation of use. The large American Cancer Society cohort study reported a substantial reduction of colon cancer mortality among subjects taking aspirin as infrequently as 16 days a month [16]; there were similar, though less marked, findings for rectal cancer [17]. Other epidemiological studies have reported similar results. There have been clear protective effects among both men and women [3, 12, 16–22], and similar inverse associations for cancers of the colon and those of the rectum when investigated in the same study (or series of studies) [3, 12, 18–21, 23–25]. Only one major observational study has been negative: a relatively small cohort study carried out in a southern California retirement community found no protection associated with aspirin use; colon cancer incidence was, if anything, increased [26, 27]. The inverse association of colorectal cancer risk has been reported in studies of aspirin specifically [3, 12, 16–18, 20, 21, 24, 26–31], and also in investigations of NSAIDs in general (often largely aspirin) [12, 15, 19, 23, 28, 29, 32–35]. NSAIDs other than aspirin seem to have a similar effect on risk [12, 20, 25, 28–30, 36]. The fact that non-NSAID analgesics (acetaminophen) are not associated with a decreased risk of colorectal cancer [16, 17, 20, 30, 32, 33] provides considerable assurance that the NSAID effect is not due to response biases. Most studies have reported that the inverse association of NSAID use with risk of colorectal cancer dissipates after stopping the drugs for 1–2 years [15, 20, 23–25, 28, 36]. There is relatively little available information on the doses of NSAIDs required for a protective effect. In some studies, as little as one regular aspirin tablet (325 mg) every day or every other day seemed to suffice [3, 20, 21]. For aspirin, there may be a pattern of increasing association with increasing dose, up to a plateau around 325 mg every other day or daily [3, 16, 17, 20] or perhaps at higher doses [21]. Some studies, however, have not seen any dose-response pattern [28]. Patients who take aspirin for cardiovascular protection seem to have a reduced risk of colorectal cancer similar to those who use the drugs for other (higher-dose) indications [24, 28, 33]. There appears to be a relationship between duration of NSAID use and risk of colorectal cancer. Most studies that have addressed the issue have reported
Baron
4
decreasing risks with increasing duration of use [3, 15–17, 20, 24, 29, 30], although a few studies have not seen such trends [33, 36], or have reported apparent inverse associations even with reported short-term use [25]. The most careful data, reported from a large cohort study, suggests that 10–20 years of use is required for a meaningful reduction in the risk of colorectal cancer [3]. A protective effect of aspirin use on the risk of large bowel adenomas has also been repeatedly documented. In an adenoma prevention trial, subjects who used aspirin during the first year of the study had a relative risk of 0.5 relative to non-takers [37]. A case-control study in England found similar results [38], and subsequent investigations have also reported apparent protective effects for aspirin [13, 18, 21, 39, 40] or NSAIDs in general [32, 38–42]. Investigations of non-aspirin NSAIDs have generally also reported inverse associations [38, 43], or suggestions of inverse associations [39, 40]. In a combined analysis of several case-control studies, non-aspirin NSAID use was associated with a reduced risk of advanced adenomas, although aspirin was not [44]. Use of acetaminophen and other non-NSAID analgesic/antipyretics has not been associated with a reduction in risk of colorectal adenomas [32, 38, 40]. Data are relatively limited regarding the details of the association between NSAID use of colorectal adenomas. Several studies have suggested more pronounced effects with increasing doses or frequency of use of NSAIDs [13, 38, 41], although irregular dose-response patterns have also been reported [21]. There seems to be a more pronounced protective effect among subjects who have used NSAIDs for longer periods of time [13, 38, 39], although, again, some investigations have not reported this pattern [32, 41]. The little data available suggest that the inverse association is maintained for a year (or perhaps more) after cessation [13, 40]. Patients with ulcerative colitis have a greatly increased risk of colorectal cancer, and one of the drugs used to treat this condition, sulfasalazine, incorporates a salicylate moiety. A few studies have suggested that ulcerative colitis patients who use sulfasalazine have a decreased risk of colorectal cancer in comparison to ulcerative colitis patients who do not [45, 46]. However, another such analysis was negative [47], and questioned the accuracy of assessment of sulfasalazine in the other investigations. Another study reported that ulcerative colitis patients who also had diagnoses treated with NSAIDs had a non-significantly lower risk of colorectal cancer, and a significantly reduced risk of colorectal cancer mortality [48].
Clinical Trials: Familial Adenomatous Polyposis
Several randomized, double-blind, placebo-controlled clinical trials have investigated the effects of the NSAID sulindac on adenoma risk in patients with
NSAIDs and Cancer Epidemiology
5
familial adenomatous polyposis (FAP). In a cross-over study among 10 patients, 100 mg of sulindac t.i.d. for 4 months was associated with a decrease in the number of polyps and a reduction in the size of the polyps that remained. There was a rebound in the size and number of polyps after the drug was stopped [49]. In another trial that used particularly careful measurement techniques in 22 patients, 150 mg of sulindac b.i.d. for 9 months of treatment resulted in a 44% decrease in the number of polyps and a 35% decrease in the size of the polyps found. In the placebo group, there were increases in both parameters [50]. Similar findings were reported in a third trial [51], and in other less formal investigations [50, 52]. Rectal administration of sulindac and indomethacin has also been reported to be effective in before-after comparisons without placebo controls [53, 54].
Clinical Trials: Sporadic Colorectal Neoplasia
Four intervention studies have focused on the effect of NSAIDs on the regression of sporadic adenomas. In a case series in which intact polyps were marked and left in the bowel, sulindac and piroxicam appeared not to lead to substantial regression of sporadic adenomas [55], although in another uncontrolled study, sulindac was associated with a marked regression of adenomas [56]. A small randomized trial found hints that sulindac may lead to polyp regression, but the findings were not statistically significant [57]. Preliminary findings from a later, placebo-controlled, study suggest more marked efficacy [58]. The Physicians’ Health Study is the only clinical trial of an NSAID that studied sporadic colorectal cancer as an endpoint [59]. The study was designed to assess the effect of 325 mg of aspirin every other day on cardiovascular endpoints in a low-risk population, and assessed colorectal neoplasia in secondary analyses. There was an increased risk of diagnosed colorectal cancer within 3 years of randomization – an effect that would be expected if aspirin use led to the diagnosis of cases present (but unrecognized) at study entry. However, the relative risk fell over time after randomization (p for trend 0.09) to a relative risk of 0.77 for the period 5 or more years after randomization. The risk of diagnosed sporadic adenoma was modestly decreased in the aspirin group, but since no uniform bowel surveillance was applied, the relative incidence of adenomas in general was not estimated [59]. More recent clinical trials have been designed specifically to investigate the effect of aspirin on sporadic colorectal neoplasia. In the Aspirin-Folate Polyp Prevention Study, 3 years’ treatment with 81 mg aspirin daily (but not 325 mg) reduced the risk of new adenomas in the large bowel in comparison to placebo. The risk reductions for more advanced neoplasia suggested more
Baron
6
pronounced effects for both aspirin doses. The results of an aspirin trial (325 mg) conducted in patients with a history of colorectal cancer have been announced: again, a statistically significant reduction was found in the risk of subsequent adenomas [Robert Sandler, pers. commun.]. At the 2002 Digestive Disease meetings, a French adenoma prevention trial using aspirin (160 and 300 mg) reported a borderline-significant reduction in the risk of adenomas at a 1-year follow-up [Robert Benamouzig, pers. commun.].
Cancers of the Esophagus, Stomach and Pancreas
NSAID use has been inversely associated with risk of esophageal cancer in every investigation that has addressed the issue (table 1). In the first National Health and Nutrition Examination Study (NHANES I), there was almost complete suppression, even with ‘occasional’ use in the previous 6 months [60]. However, the study was small: only 14 cases were included in the analysis, and information regarding drug use was not detailed. The large American Cancer Society Cancer Prevention Study II reported a 40% reduction in risk of esophageal cancer mortality among subjects who used aspirin 16 or more times per month in comparison to those who did not use aspirin at all [17]. A population-based case-control study found similarly reduced risks for both squamous carcinoma of the esophagus and adenocarcinoma [61]. The decreased risks did not seem to vary with duration for use for either type of tumor, but the risks did decline with increasing dose for esophageal squamous carcinoma. A hospital-based case-control study also reported modestly (and non-significantly) reduced risks with continuing use of NSAIDs [62]. Most studies that have investigated stomach cancer in relation to NSAID use have also reported an apparent protective effect (table 1). In the large American Cancer Society mortality cohort study, there was more than a 50% reduction of stomach cancer mortality among subjects who used aspirin 16 or more days per month [17]. In large population-based case-control study conducted in the USA, there were non-significant inverse associations of both aspirin use and non-aspirin NSAID use with cardia cancer, and statistically significant inverse associations with risk of other gastric cancers. The protective associations with non-cardia cancers appeared unrelated to dose of drug, but was confined to current users. Even use for less than 5 years conferred a 39% reduction in risk, and the relative risks were lower for longer durations [61]. A Swedish population-based case-control study found 20–30% reductions in risk among aspirin users, with lower risks among those who used higher doses [63]. In the UK General Practice Database case-control study [34], and in
NSAIDs and Cancer Epidemiology
7
Table 1. NSAID use and risk of esophageal and stomach cancer Baron
Design features
Relative risks for aspirin (or predominately aspirin) use
Cohort mortality study; 635,031 subjects followed for approximately 6 years; 176 deaths observed
ASA use per month 0 Occasionally 1–15 16
Funkhouser and Sharp, 1995 [60]
Cohort study; 13,179 subjects, followed for 12–16 years; 14 cases
No ASA use, no pain med. use 1.0 (ref) 0.09 (0.01–0.70) ASA use, no pain med. use 0.00 (0.00–4.69) ASA use, pain med. use
Farrow et al., 1998 [61]
Population-based case-control study; 293 adenocarcinoma, 221 squamous carcinoma (77.1% response of all gastric and esophageal cases) 695 controls (70.2% response)
Adenocarcinoma Never use* Ever use* 10 years of use Squamous carcinoma Never use* Ever use* 10 years of use
Coogan et al., 2000 [62]
Hospital-based case-control study; 215 cases, 5,952 controls (95% response overall)
Never NSAID use Non-regular NSAID use Continuing regular NSAID use
1.0 (ref) 0.9 (0.6–1.2) 0.8 (0.5–1.4)
Langman et al., 2000 [34]
Population (GP practice)-based database case-control study; 550 cases, 1,650 controls
No NSAID prescriptions 7 NSAID prescriptions
1.00 (ref) 0.64 (0.41–0.98)
Esophageal cancer Thun et al., 1993 [73]
Relative risks for non-aspirin NSAID use
1.0 (ref) 0.74 (0.50–1.10) 0.81 (0.53–1.24) 0.59 (0.34–1.03)
1.00 (ref) 0.48 (0.32–0.70) 0.64 (0.37–1.10) 1.00 (ref) 0.52 (0.31–0.87) 0.58 (0.28–1.23)
Adenocarcinoma Never use NANSAIDs Ever 1 tablet/wk, 6 mos 5+ years of use Squamous carcinoma Never use Ever 1 tablet/wk, 6 mos 5 years of use
1.00 (ref) 0.81 (0.51–1.30) 0.88 (0.43–1.78) 1.00 (ref) 0.51 (0.24–1.10) 0.42 (0.12–1.49)
8
Stomach cancer Thun et al., 1993 [73]
NSAIDs and Cancer Epidemiology
Schreinemachers and Everson, 1994 [31]
Farrow et al., 1998 [61]
Zaridze et al., 1999 [64]
Langman et al., 2000 [34]
Coogan et al., 2000 [62]
Akre et al., 2001 [63] 9
Cohort mortality study; 635,031 subjects followed for approx. 6 years; 308 deaths observed
ASA use per month 0 Occasionally 1–15 16
1.0 (ref) 0.83 (0.62–1.10) 0.60 (0.42–0.85) 0.53 (0.34–0.81)
Cohort study; 7,489 women followed for an average of 12.4 years; 147 cases Population-based case-control study; 261 gastric cardia cases 368 non-cardia cases (77.1% response of all gastric and esophageal cases), 695 controls (70.2% response) Hospital-based case-control study; 448 cases (98.3% response), 610 controls (96.5% response) Population (GP practice)-based data-base case-control study; 613 cases, 1,837 controls Hospital-based casecontrol study; 254 cases, 5,952 controls (95% response overall) Population-based casecontrol study; 567 cases (62% response), 1,165 controls (76.0% response)
No aspirin use in 30 days ASA use
1.0 (ref) 0.93 (0.49–1.74)
mos Months; * non cancer control.
Cardia adenocarcinoma Never use* Ever use* 10 years of use Other adenocarcinoma Never use* Ever use* 10 years of use
1.00 (ref) 0.55 (0.39–0.78) 0.48 (0.27–0.82)
No ASA use ASA use Cardia cancer Non-cardia cancer
1.0 (ref) 0.60 (0.41–0.90) 1.14 0.49 (0.31–0.77)
No NSAID prescriptions 7 NSAID prescriptions
1.00 (ref) 0.51 (0.33–0.79)
Never NSAID use Non-regular NSAID use Continuing regular NSAID use Aspirin use No ASA use ASA use, 1 tablets/month 30 tablets/month
1.0 (ref) 0.7 (0.5–0.9) 0.3 (0.1–0.6)
1.00 (ref) 0.88 (0.61–1.27) 0.84(0.48–1.46)
0.4 (0.2–0.8) 1.0 (ref) 0.8 (0.7–1.1) 0.6 (0.3–1.1)
Cardia adenocarcinoma Never use non-ASA NSAIDs Ever 1 tablet/wk, 6 mos 5 years of use Other adenocarcinoma Never use Ever 1 tablet/wk, 6 mos 5 years of use
No use non-ASA NSAIDs Non-ASA NSAID use 1 tablet/wk, 6 mos
1.00 (ref) 0.82 (0.51–1.33) 0.95 (0.45–2.01) 1.00 (ref) 0.55 (0.34–0.89) 0.48 (0.21–1.09)
1.0 (ref) 1.1 (0.6–1.4)
hospital-based studies from the USA [62] and Russia [64] there were also inverse associations between aspirin or general NSAID use and stomach cancer risk. Only the relatively small NHANES I cohort did not report an inverse association, although with wide confidence limits, the study was consistent with a substantial reduction in risk [31]. In two studies, the inverse relationship between NSAID use and stomach cancer was greater among subjects with evidence for infection with Helicobacter pylori [63, 64]. However, there has been discordant data published regarding other details of the effect of NSAIDs on stomach cancer. In the Swedish case-control study, the association was non-significantly more marked for cardia cancer than for non-cardia tumors, and non-cardia cancer of the diffuse type had no association with aspirin use [63]. In contrast, in the Russian case-control study the benefit of NSAIDs was confined to non-cardia cancers. Data regarding pancreatic cancer also suggest a protective effect of NSAIDs. In the Iowa Women’s Cohort Study, there was a strong reduction in the risk of pancreatic cancer among women who used aspirin 6 times a week or more often (RR 0.40 (95% CI 0.20–0.82)). There was no similar reduction associated with non-aspirin NSAIDs, although the estimates of effect were imprecise [65]. A hospital-based case-control study reported an inverse association with use of NSAIDs, particularly long-term use [62]. In the NHANES I cohort, there was a non-significant reduction in risk of pancreatic cancer among subjects who used aspirin [31], but in the case-control study derived from the UK General Practice database, there were indications of increased risks [34].
Breast and Ovarian Cancer
Data regarding the association of NSAID use and risk of breast cancer are conflicting (table 2). In two cohort studies [31, 66], there was a decreased risk, and several case-control studies have reported similar findings [23, 67–70]. In some studies, even short duration of NSAID use has been associated with a reduction in risk [70], with little weakening of the apparent inverse association after cessation [70]. In other studies, continuing aspirin use appears to be required [23], and there has been a trend of decreasing risk with longer duration of use [23]. Some investigations have reported a pattern of decreasing breast cancer risks with increasing doses of aspirin or other NSAIDs [66, 68]. In some analyses, there was an inverse relation between use of ibuprofen and risk of breast cancer [66, 68]. In contrast, several cohort [17, 26, 71] and case-control [34, 35] studies have not found an association between NSAID use and breast cancer risk,
Baron
10
Table 2. NSAID use and risk of breast cancer NSAIDs and Cancer Epidemiology
Design features
Relative risks for aspirin (or predominately aspirin) use
Paganini-Hill et al., 1989 [26]
Cohort study; 8,881 women followed for 6.5 years; 214 cases
No use Daily Daily
1.00 (ref) 0.95 0.96
Thun et al., 1993 [17]
Cohort mortality study; 635,031 subjects followed for approx. 6 years
No use 1–15 times/month 16 times/month
1.00 (ref) 0.98 (0.76–1.26) 0.88 (0.62–1.24)
Schreinemachers and Everson, 1994 [31]
Cohort study; 7,489 women followed for an average of 12.4 years; 147 cases
No aspirin use in 30 days ASA use
1.0 (ref)
Harris et al., 1995 [67] Hospital-based case-control No OTC NSAID use study; 744 cases, 767 1–4 years of use non-cancer controls 5 years of use (response rates not presented) Harris et al., 1996 [68] Case-control study; 511 cases, 1,534 controls (response rates not presented)
Egan et al., 1996 [71]
Relative risks for non-aspirin NSAID use
0.72 (0.52–1.00) 1.00 (ref) 1.12 (0.8–1.6) 0.58 (0.4–0.8)
No continued, regular 1.0 (ref) ASA use Aspirin 3 tablets/wk, 0.69 (0.46–0.99) 1yrs Any NSAID 3/wk, 0.66 (0.52–0.83) 1yrs
No regular NSAID use 1.00 (ref) Ibuprofen 3 tabs/wk, 0.57 (0.36–0.91) 1 yrs
Cohort study; 89,528 women No regular NSAID use 1.00 (ref) No regular NSAID use 1.00 (ref) followed for 12 years; 2,414 2 ASA tablets/wk 1.03 (0.95–1.12) 2 NSAID uses/wk 0.95 (0.78–1.17) cases 1.05 (0.89–1.23) 14 ASA tables/wk
11
Table 2 (continued) Baron
Design features
Relative risks for aspirin (or predominately aspirin) use
Cohort study; 32,505 women attending mammography, followed for median of 4.7 years; 393 cases
No ASA use 1–3 ASA pills/wk 4 ASA pills/wk
Coogan et al., 1999 [23]
Hospital-based case-control study; 6,558 cases, 2,925 non-cancer controls (96% response)
No NSAID use Regular ASA use
Langman et al., 2000 [34]
Population (GP practice)based database case-control study; 3,105 cases, 9,272 controls
No NSAID prescriptions 1.00 (ref) 7 NSAID prescriptions 1.10 (0.92–1.30)
Cotterchio et al., 2001 [70]
Population-based casecontrol study; 3,133 cases (73% response), 3,062 controls (61% response)
Never regular ASA use Ever use 2 months
Harris et al., 1999 [66]
Relative risks for non-aspirin NSAID use
1.00 (ref) No ibuprofen use 0.57 (0.40–0.81) 1–3 ibuprofen pills/wk 0.64 (0.45–0.90) 4 ibuprofen pills/wk
1.00 (ref) 0.7 (0.5–0.8)
1.00 (ref) 0.73 (0.61–087)
1.00 (ref) 0.53 (0.33–0.84) 0.49 (0.30–0.80)
No prescription NSAIDs 7 pills/week
1.00 (ref) 0.38 (0.14–1.02)
No NSAID use Regular non-ASA NSAID use
1.00 (ref) 0.79 (0.66–0.96)
Never NSAID use Ever non-ASA NSAID use
12
although in the American Cancer Society Cohort [17], the relative risk estimates, though not materially reduced, were compatible with a substantial reduction in breast cancer mortality. Studies that have investigated acetaminophen use have reported no association with breast cancer risk [66], a modest inverse association [35], or (apparently) a more substantial inverse association [67]. Although the acetaminophen data have not been presented in detail, these findings – together with those for the NSAIDs leave open the possibility of that the observed NSAID association may be due to response biases (or, conceivably, a confusion with a real effect of acetaminophen). Findings regarding ovarian cancer are also conflicting (table 3). A hospitalbased case-control study reported that long-term use of either aspirin or nonaspirin NSAIDs, was associated with a reduced risk [72]. A population-based case-control study found a non-significant 23% reduction in risk with continuous use of aspirin for 6 months or more, but essentially no reduction in risk with such use of ibuprofen [73]. On the other hand, in a large cohort study, aspirin use was not related to ovarian cancer risk, while non-aspirin NSAIDs were inversely associated with risk, albeit with no dose-response pattern [74]. One hospital-based study reported data consistent with a protective effect of NSAIDs (a reduced risk with frequent use of analgesics) [75], but two other hospital-based case-control studies reported essentially no association between aspirin use and ovarian cancer risk [76, 77]. In contrast to the findings for colorectal cancer, some studies [43, 73, 76], but not all [72, 74], found that acetaminophen use was associated with a reduced risk. These findings further complicate the interpretation of data regarding NSAID use and ovarian cancer.
Cancers of the Urinary Tract
Data regarding the association between NSAID use and bladder cancer are relatively sparse. In a large population-based case-control study from Los Angeles, there was an inverse association of risk with cumulative lifetime dose for each of the NSAID classes investigated (including salicylates) except for pyrazolon derivatives (e.g. phenybutazone). For aspirin, there was a substantially reduced risk (odds ratio 0.63 (95% CI 0.43–0.92)) only in the highest category of use (1,243 g) [78]. In contrast, in the UK General Practice Research Database, there was no association of NSAID use with bladder cancer risk [34, 79], nor was there an association in a case-control study from Germany [80]. There was also no association in two small cohort studies [26, 31], including one from Southern California [26]. Use of phenacetin [81–83], and possibly acetaminophen [83], is a risk factor for renal cell carcinoma, and some studies have found NSAIDs also to be
NSAIDs and Cancer Epidemiology
13
Table 3. NSAID use and risk of ovarian cancer
Baron 14
Design features
Relative risks for aspirin (or predominately aspirin) use
Tzonou et al., 1993 [75]
Hospital-based case-control study; 189 cases (90% response), 200 controls (94% response)
Never/rare analgesic use Infrequent use Frequent use
1.0 (ref) 0.78 (0.43–1.43) 0.51 (0.26–1.02)
Cramer et al., 1998 [73]
Population-based casecontrol study; 563 cases (70% response), 523 controls (response 31–72%)
No OTC ASA use OTC ASA use
1.00 (ref) 0.78 (0.53–1.15)
Tavani et al., 2000 [77]
Hospital-based case-control study; 749 cases, 898 controls (96% response)
Never regular ASA use Ever regular ASA use 5 years regular use
1.00 (ref) 0.93 (0.53–1.62) 1.24 (0.53–2.90)
Rosenberg et al., 2000 [72]
Hospital-based case-control study; 780 cases, 2,570 non-cancer controls (95% response)
No ASA use ASA use ASA use 5 yr
1.0 (ref) 0.8 (0.5–1.2) 0.5 (0.2–1.0)
Moysich et al., 2001 [76]
Hospital-based case-control study; 547 cases, 1,094 controls (~50% response)
Non-ASA users Regular user 11 years of use
1.0 (ref) 1.0 (0.73–1.39) 0.90 (0.61–1.32)
Akhmedkhanov et al., 2001 [97]
Case-control study within cohort; 14,275 women followed for median of 12 years; 68 cases, 680 controls
Never ASA 3/wk for 6 mos Ever ASA 3/wk for 6 mos 5 years of use
1.0 (ref)
0.67 (0.23–1.96)
Fairfield et al., 2002 [74]
Cohort study; 76,821 women followed for 16 years; 333 cases
No ASA use ASA use 15 per week 20 years
1.00 (ref) 1.00 (0.80–1.25) 0.98 (0.63–1.52) 0.99 (0.69–1.43)
Meier et al., 2002 [35]
Population (GP practice)based case-control data-base study; 483 cases, 1877 controls
No NSAID use 30 NSAID prescriptions
1.0 (ref) 1.1 (0.7–1.8)
Relative risks for non-aspirin NSAID use
No OTC ibuprofen use OTC ibuprofen use No prescribed analgesic use Any prescribed analgesic use
1.0 (ref) 1.20 (0.74–1.95) 1.0 (ref) 0.91 (0.53–1.54)
No non-ASA NSAID use Non-ASA NSAID 4 days/wk, 6 wks Non-ASA NSAID use 5 yr
1.0 (ref) 0.5 (0.3–0.9)
No non-ASA NSAID use Non-ASA NSAID use 20 days/month
1.00 (ref) 0.60 (0.38–0.95) 0.59 (0.29–1.19)
0.6 (0.2–2.2)
0.60 (0.26–1.38)
associated with an increased risk (table 4). A population-based case-control study from Los Angeles (associated with the bladder cancer study referred to above) found that patients who used aspirin at doses of 325 mg/day or greater for prolonged periods had more than a doubling of risk. There were similar effects associated with non-aspirin NSAIDs [83]. A cohort study, also from Southern California, also found increased risks in association with aspirin use [26], although only 25 renal cancer cases were observed and there was no consideration in the study of use of other analgesics. In a third study, some aspirin-containing products were associated with increased risks for men (but not women), but few details were reported [84]. Other investigations have not reported an increased risk from NSAIDs for renal carcinoma (table 4). A large, multicenter, international, population-based case-control study carefully assessed all classes of analgesics and found no increased risks, except possibly in association with use of acetaminophen [85]. In particular, relative risks were not materially elevated for use of salicylates or pyrazolones. Other studies from Australia [82, 86], the USA [31, 81] and the UK [79] did not report any substantial association of aspirin or salicylate use with renal carcinoma. Non-aspirin analgesics have also been associated with transitional cell carcinoma of the renal pelvis [81, 86, 87], and some studies have reported indications of increased risks from aspirin or salicylates [87, 88]. However, a case-control study from Australia reported a non-significantly decreased risk of cancer of the renal pelvis with aspirin use [86], and two US case-control studies [81, 89] and one from Germany [80] were essentially negative. Several of these studies are limited by relatively small numbers of cases and consequent lack of statistical precision.
Prostate Cancer
Some investigations have reported an apparent protective effect of NSAID use on risk of prostate cancer (table 5). A population-based case-control study reported borderline-significant 25–30% reductions in the risk of advanced cancer with regular use of aspirin or of non-aspirin NSAIDs [90]. Smaller (nonsignificant) reductions in risk were seen for all prostate cancers. An even more marked inverse association was seen between use of NSAIDs (considered as a group) and prostate cancer risk in a small cohort study [4]. With a median follow-up of 66 months, risk was reduced by more than 50% among daily NSAID users; this effect was particularly pronounced at older ages. A population-based case-control study reported a similarly marked association for use of aspirin/ ibuprofen [91], and a cohort study (using patients attending a multiphasic health
NSAIDs and Cancer Epidemiology
15
Table 4. Aspirin use and cancer of the renal parenchyma Baron
Design features McLaughlin et al., 1985 [81]
McCredie et al., 1988 [82]
McCredie et al., 1993 [86]
Schreinemachers and Everson, 1994 [31] McCredie et al., 1995 [85]
Gago-Dominguez et al., 1999 [83]
Population-based case-control study; 495 cases (98% response, including proxy), 697 controls (98% response) Population-based case-control study; 360 cases (40% response), 985 controls (72% response) Population-based case-control study; 503 cases (68% response), 523 controls (48% response)
16
Cohort study; 12,668 subjects followed for average of 12.4 years; 32 cases Multicentered, international, population-based case-control study; 1,732 cases (72.3% response), 2,309 controls (74.7% response) Population-based case-control study; 1,201 cases (74% response), 1,204 controls (response not described)
Relative risks for aspirin (or predominately aspirin) use
Never aspirin product use Ever aspirin product use Regular aspirin product use 36 months No regular aspirin use Regular aspirin use
Male 1.0 (ref) 0.6 (0.4–0.9) 0.5 (0.2–1.0)
Female 1.0 (ref) 1.6 (0.9–2.8) 1.8 (0.7–4.1)
1.0 (ref) 1.2 (0.7–1.9)
No aspirin use (20 times in lifetime) Aspirin use Phenacetin/aspirin Acetaminophen/aspirin No aspirin use in 30 days ASA use
1.0 (ref)
No salicylate use 0.1–1.0 kg lifetime use 1.1–5.0 kg lifetime use 5.0 kg lifetime use
1.0 (ref) 1.0 (0.8–1.2) 1.1 (0.8–1.5) 1.2 (0.9–1.7)
No regular aspirin Regular ASA use Only ASA use 325 mg, 10 yrs
1.0 (ref) 1.5 (1.2–1.8) 4.3 (1.6–11.3)
1.0 (0.7–1.4) 1.4 (0.9–2.3) 1.4 (0.7–2.9) 1.0 (ref) 0.60 (0.29–1.24)
Table 5. NSAID use and cancer of the prostate NSAIDs and Cancer Epidemiology
Design features
Relative risks for aspirin (or predominately aspirin) use
Norrish et al., 1998 [90]
Population-based casecontrol study; 317 cases (77% response), 480 controls (71% response)
No/infrequent NSAID use Regular ASA use
Nelson and Harris, 2000 [91]
Hospital-based case-control study; 417 cases, 420 controls (response not described)
No regular ASA/ibuprofen 1.0 (ref) ASA/ibuprofen, 1 pill/day 0.52 (0.35–0.78) ASA/ibuprofen, 1 pill/day 0.34 (0.20–0.58)
Relative risks for non-aspirin NSAID use
1.0 (ref) 0.85 (0.61–1.19)
Langman et al., Population (GP practice)2000 [34] based database case-control study; 1,813 cases, 5,354 controls
No NSAID prescriptions 7 NSAID prescriptions
1.00 (ref) 1.33 (1.07–1.64)
Roberts et al., 2002 [4]
Cohort study; 1,362 men followed for median of 66 months; 91 cases
No daily NSAID use Daily NSAID use
1.0 (ref) 0.49 (0.30–0.79)
Habel et al., 2002 [92]
Cohort study; 90,100 men followed for mean 14 years; 2,574 cases
No regular ASA use 6 aspirin, most days
1.0 (ref) 0.76 (0.60–0.98)
No/infrequent NSAID use 1.0 (ref) Regular non-ASA 0.87 (0.49–1.55) NSAID use
17
checkup) reported about a 30% decreased risk among men who took 6 or more aspirin tablets daily [92]. However, in another population-based case-control study that considered prescription NSAIDs 13–36 months before diagnosis, there was a non-significantly increased risk [34], and in a small cohort analysis of elderly men, there was essentially no association between aspirin use and prostate cancer risk [26]. There was also no association of aspirin use with risk of prostate cancer in the NHANES I cohort [31]. In one small study, there was no substantial association between acetaminophen use and prostate cancer risk [91].
Other Cancers
There was a reduced risk of respiratory cancer among aspirin users in one small cohort study in the USA [31], a non-significant reduction among women (but not men) in another cohort analysis [26], and a non-significantly decreased risk of lung cancer mortality in an aspirin clinical trial [93]. However, there was no association of aspirin use with respiratory system cancer mortality in the American Cancer Society cohort [17], nor was there a substantial association in the British general practice research database [34]. One population-based case-control study reported patients who used NSAIDs regularly had a reduced risk of malignant melanoma [94], but in a small cohort analysis, there was no association [31]. A study in China reported no statistically significant association between use of salicylates, indomethacin or phenylbutazone and risk of acute lymphocytic leukemia or acute non-lymphocytic leukemia, but there was an increased risk of chronic myelocytic leukemia and chronic lymphocytic leukemia among subjects who had used salicylates. In the NHANES I cohort there were nonsignificant suggestions of decreased risks of lymphoma and leukemia [31] and in the large American Cancer Society Cohort Study, there was essentially no association between aspirin use and mortality from lymphatic and hematopoietic malignancies [17].
Summary and Conclusion
There is evidence that aspirin – and apparently other NSAIDs – may be protective agents against cancer in the gastrointestinal tract. These effects are particularly well documented in the colon and rectum. Even considered in isolation, the observational data regarding colorectal neoplasia are quite strong, and the reality of a protective effect is buttressed by clinical trial data showing that aspirin prevents sporadic adenomas. Furthermore, the NSAIDs sulindac
Baron
18
celecoxib have actually led to the regression of existing colorectal polyps in patients with FAP. Clearly, NSAIDs have the potential to suppress carcinogenesis in the large bowel. Observational data suggesting inverse associations of NSAIDs with cancers of the stomach and esophagus have emerged from several case-control studies and a few cohort analyses. In some studies the findings display features often associated with causal relationships, for example decreasing risks with increasing doses or duration of use. Nonetheless, the data currently do not support a secure conclusion that NSAIDs protect against these malignancies. The relevant data are not nearly as extensive as those for the colorectum, and case-control investigation of these upper gastrointestinal sites may be particularly delicate. It is conceivable that early symptoms of cancer (or of pre-invasive lesions) may have discouraged NSAID use in the cancer patients, creating the appearance of a protective association of the drugs with the risk of these malignancies. More extensive observational data particularly from cohort studies would be desirable to confirm the existing findings and clarify the doses and durations of use required for an effect. Clinical trial investigation might also be practical for pre-neoplastic endpoints, or – in carefully selected populations – perhaps with cancer as the focus. There are only relatively limited data available regarding the effect of NSAIDs on cancer of the pancreas. However, the studies that have investigated this malignancy have reported indications that NSAIDs may have a protective effect. The effects of NSAIDs on cancers outside the gastrointestinal tract are not clear. Some investigations suggest that NSAID use, particularly aspirin, is inversely associated with risk of cancers of the breast or ovary, but several well-done studies have not seen these associations, and the observations could have been due to bias or confounding. Findings regarding prostate cancer are similarly conflicting. The urinary tract is one organ system in which several studies have reported an increased cancer risk in association with NSAID use. Nonetheless, the effects remain unclear. There is only limited available information regarding carcinoma of the bladder, and no firm conclusions can be drawn at this point. More extensive data have been generated regarding the effect of NSAIDs – largely salicylates – on renal cell carcinoma or cancer or the renal pelvis and ureter. Although some studies have reported increased risks, there are also findings suggesting no association. It is particularly difficult for observational studies to ascertain with confidence the true effects of aspirin because of the suspected relationship of these cancers with use of phenacetin and perhaps acetaminophen. Further data – particularly from careful and large cohort studies – would be important to clarify these issues.
NSAIDs and Cancer Epidemiology
19
As a body of research, the findings discussed here from epidemiological studies and clinical trials have begun to clarify the effect of NSAIDs on carcinogenesis in various organs in humans. There is clear potential for protective effects at several anatomic sites. Even for the colorectum, however, it is probably premature to now begin to use these drugs widely for cancer prevention. To reach that point, a weighing of the risks and benefits of the drugs needs to be made, together with a judgement regarding the benefits of alternative means of prevention. For colorectal cancer, for example, aspirin may provide only limited benefit over regular colonoscopy [95, 96]. Nonetheless, with the increased understanding of the clinical effects of NSAIDs on cancer, the development of effective chemoprevention with these drugs appears to be a real possibility.
References 1 2 3 4
5 6
7 8 9 10 11 12
13
14 15
Macfarlane GJ, Black RJ: Rheumatoid arthritis and lymphatic cancer. Eur J Cancer 1996;32A: 1630–1632. Thomas E, Brewster DH, Black RJ, Macfarlane GJ: Risk of malignancy among patients with rheumatic conditions. Int J Cancer 2000;88:497–502. Giovannucci E, Egan KM, Hunter DJ, Stampfer MJ, Colditz GA, Willett WC, Speizer FE: Aspirin and the risk of colorectal cancer in women. N Engl J Med 1995;333:609–614. Roberts RO, Jacobson DJ, Girman CJ, Rhodes T, Lieber MM, Jacobsen SJ: A population-based study of daily nonsteroidal anti-inflammatory drug use and prostate cancer. Mayo Clin Proc 2002; 77:219–225. Hakulinen T, Isomaki H, Knekt P: Rheumatoid arthritis and cancer studies based on linking nationwide registries in Finland. Am J Med 1985;78:29–32. Gridley G, McLaughlin JK, Ekbom A, Klareskog L, Adami HO, Hacker DG, Hoover R, Fraumeni JF Jr: Incidence of cancer among patients with rheumatoid arthritis. J Natl Cancer Inst 1993;85: 307–311. Mellemkjaer L, Linet MS, Gridley G, Frisch M, Moller H, Olsen JH: Rheumatoid arthritis and cancer risk. Eur J Cancer 1996;32A:1753–1757. Kauppi M, Pukkala E, Isomaki H: Low incidence of colorectal cancer in patients with rheumatoid arthritis. Clin Exp Rheumatol 1996;14:551–553. Cibere J, Sibley J, Haga M: Rheumatoid arthritis and the risk of malignancy. Arthritis Rheum 1997;40:1580–1586. Katusic S, Beard CM, Kurland LT, Weis JW, Bergstralh E: Occurrence of malignant neoplasms in the Rochester, Minnesota, rheumatoid arthritis cohort. Am J Med 1985;78:50–55. Muller AD, Sonnenberg A, Wasserman IH: Diseases preceding colon cancer. A case-control study among veterans. Dig Dis Sci 1994;39:2480–2484. 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: 4399–4404. Garcia Rodriguez LA, Huerta-Alvarez C: Reduced incidence of colorectal adenoma among long-term users of nonsteroidal anti-inflammatory drugs: A pooled analysis of published studies and a new population-based study. Epidemiology 2000;11:376–381. Zheng W, Linet MS, Shu XO, Pan RP, Gao YT, Fraumeni JF Jr: Prior medical conditions and the risk of adult leukemia in Shanghai, PR China. Cancer Causes Control 1993;4:361–368. Rosenberg L, Palmer JR, Zauber AG, Warshauer ME, Stolley PD, Shapiro S: A hypothesis: Nonsteroidal anti-inflammatory drugs reduce the incidence of large-bowel cancer. J Natl Cancer Inst 1991;83:355–358.
Baron
20
16 17 18
19 20
21 22
23
24 25
26 27
28 29
30 31 32
33 34
35
36 37
Thun MJ, Namboodiri MM, Heath CW Jr: Aspirin use and reduced risk of fatal colon cancer. N Engl J Med 1991;325:1593–1596. Thun MJ, Namboodiri MM, Calle EE, Flanders WD, Heath CW Jr: Aspirin use and risk of fatal cancer. Cancer Res 1993;53:1322–1327. 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 Med 1994; 121:241–246. Collet JP, Sharpe C, Belzile E, Boivin JF, Hanley J, Abenhaim L: Colorectal cancer prevention by non-steroidal anti-inflammatory drugs: Effects of dosage and timing. Br J Cancer 1999;81:62–68. Garcia-Rodriguez LA, Huerta-Alvarez C: Reduced risk of colorectal cancer among long-term users of aspirin and non-aspirin nonsteroidal anti-inflammatory drugs. Epidemiology 2001; 12:88–93. Suh O, Mettlin C, Petrelli NJ: Aspirin use, cancer and polyps of the large bowel. Cancer 1993;72: 1171–1177. Rosenberg L, Rao RS, Palmer JR, Strom BL, Zauber A, Warshauer ME, Stolley PD, Shapiro S: Transitional cell cancer of the urinary tract and renal cell cancer in relation to acetaminophen use (United States). Cancer Causes Control 1998;9:83–88. Coogan PF, Rao SR, Rosenberg L, Palmer JR, Strom BL, Zauber AG: The relationship of nonsteroidal anti-inflammatory drug use to the risk of breast cancer. Preventive Medicine 1999; 29:72–76. La Vecchia C, Negri E, Franceschi S, Conti E, Montella M, Giacosa A, Falcini A, Decarli A: Aspirin and colorectal cancer. Br J Cancer 1997;76:675–677. Smalley W, Ray WA, Daugherty J, Griffin MR: Use of nonsteroidal anti-inflammatory drugs and incidence of colorectal cancer: A population-based study. Arch Intern Med 1999;159: 161–166. Paganini-Hill A, Chao A, Ross RK, Henderson BE: Aspirin use and chronic diseases: A cohort study of the elderly. BMJ 1989;299:1247–1250. Paganini-Hill A, Hsu G, Ross RK, Henderson BE: Aspirin use and incidence of large-bowel cancer in a California retirement community. Journal of the National Cancer Institute 1991;83: 1182–1183. Rosenberg L, Louik C, Shapiro S: Nonsteroidal anti-inflammatory drug use and reduced risk of large bowel carcinoma. Cancer 1998;82:2326–2333. Reeves MJ, Newcomb PA, Trentham-Dietz A, Storer BE, Remington PL: Nonsteroidal antiinflammatory drug use and protection against colorectal cancer in women. Cancer Epidemiol Biomarkers Prev 1996;5:955–960. Peleg II, Maibach HT, Brown SH, Wilcox CM: Aspirin and nonsteroidal anti-inflammatory drug use and the risk of subsequent colorectal cancer Arch Intern Med 1994;154:394–399. Schreinemachers DM, Everson RB. Aspirin use and lung, colon and breast cancer incidence in a prospective study. Epidemiology 1994;5:138–146. Peleg I, Lubin M, Cotsonis G, Clark W, Wilcox C: Long-term use of nonsteroidal antiinflammatory drugs and other chemopreventors and risk of subsequent colorectal neoplasia. Dig Dis Sci 1996;41:1319–1326. Muscat JE, Stellman SD, Wynder EL: Nonsteroidal anti-inflammatory drugs and colorectal cancer. Cancer 1994;74:1847–1854. Langman MJ, Cheng KK, Gilman EA, Lancashire RJ: Effect of anti-inflammatory drugs on overall risk of common cancer: Case-control study in general practice research database. BMJ 2000;320:1642–1646. Meier CR, Schmitz S, Jick H: Association between acetaminophen or nonsteroidal antiinflammatory drugs and risk of developing ovarian, breast, or colon cancer. Pharmacotherapy 2002;22:303–309. Friedman G, Coates A, Potter J, Slattery M: Drugs and colon cancer. Pharmacoepidemiol Drug Saf 1998;7:99–106. Greenberg ER, Baron JA, Freeman DH Jr, Mandel JS, Haile R: Reduced risk of large-bowel adenomas among aspirin users. The Polyp Prevention Study Group. J Natl Cancer Inst 1993;85: 912–916.
NSAIDs and Cancer Epidemiology
21
38
39
40 41
42
43 44
45 46
47
48 49
50
51
52 53
54
55 56 57 58
Logan R, Little J, Hawtin P, Hardcastel J: Effect of aspirin and non-steroidal anti-inflammatory drugs on colorectal adenomas: Case-control study of subjects participating in Nottingham faecal occult blood screening programme. BMJ 1993;307:285–289. Breuer-Katschinski B, Nemes K, Rump B, Leiendecker B, Marr A, Breuer N, Goebell H: Longterm use of nonsteroidal anti-inflammatory drugs and the risk of colorectal adenomas. The Colorectal Adenoma Study Group. Digestion 2000;61:129–134. Sandler RS, Galanko JC, Murray SC, Helm JF, Woosley JT: Aspirin and nonsteroidal antiinflammatory agents and risk for colorectal adenomas. Gastroenterology 1998;114:441–447. Martinez ME, McPherson RS, Levin B, Annegers JF: Aspirin and other nonsteroidal antiinflammatory drugs and risk of colorectal adenomatous polyps among endoscoped individuals. Cancer Epidemiol Biomarkers Prev 1995;4:703–707. Rex DK, Cummings OW, Helper DJ, Nowak TV, McGill JM, Chiao GZ, Kwo PY, Gottlieb KT, Ikenberry SO, Gress FG, Lehman GA, Born LJ: Five-year incidence of adenomas after negative colonoscopy in asymptomatic average-risk persons. Gastroenterology 1996;111:1178–1181. Rodriguez C, Henley SJ, Calle EE, Thun MJ: Paracetamol and risk of ovarian cancer mortality in a prospective study of women in the USA. Lancet 1998;352:1354–1355. Terry MB, Neugut AI, Bostick RM, Sandler RS, Haile RW, Jacobson JS, Fenoglio-Preiser CM, Potter JD: Risk factors for advanced colorectal adenomas: A pooled analysis. Cancer Epidemiol Biomarkers Prev 2002;11:622–629. Pinczowski D, Ekbom A, Baron J, Yuen J, Adami HO: Risk factors for colorectal cancer in patients with ulcerative colitis: A case-control study. Gastroenterology 1994;107:117–120. Moody GA, Jayanthi V, Probert CS, MacKay H, Mayberry JF: Long-term therapy with sulphasalazine protects against colorectal cancer in ulcerative colitis: A retrospective study of colorectal cancer risk and compliance with treatment in Leicestershire. Eur J Gastroenterol Hepatol 1996;8:1179–1183. Lindberg BU, Broome U, Persson B: Proximal colorectal dysplasia or cancer in ulcerative colitis. The impact of primary sclerosing cholangitis and sulfasalazine: Results from a 20-year surveillance study. Dis Colon Rectum 2001;44:77–85. Bansal P, Sonnenberg A: Risk factors of colorectal cancer in inflammatory bowel disease. Am J Gastroenterol 1996;91:44–48. Labayle D, Fischer D, Vielh P, Drouhin F, Pariente A, Bories C, Duhamel O, Trousset M, Attali P: Sulindac causes regression of rectal polyps in familial adenomatous polyposis. Gastroenterology 1991;101:635–639. Giardiello FM, Hamilton SR, Krush AJ, Piantadosi S, Hylind LM, Celano P, Booker SV, Robinson CR, Offerhaus GJ: Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med 1993;328:1313–1316. Nugent KP, Farmer KC, Spigelman AD, Williams CB, Phillips RK: Randomized controlled trial of the effect of sulindac on duodenal and rectal polyposis and cell proliferation in patients with familial adenomatous polyposis. Br J Surg 1993;80:1618–1619. Rigau J, Pique JM, Rubio E, Planas R, Tarrech JM, Bordas JM: Effects of long-term sulindac therapy on colonic polyposis. Ann Intern Med 1991;115:952–954. Winde G, Gumbinger HG, Osswald H, Kemper F, Bunte H: The NSAID sulindac reverses rectal adenomas in colectomized patients with familial adenomatous polyposis: Clinical results of a dose-finding study on rectal sulindac administration. Int J Colorectal Dis 1993;8:13–17. Hirota C, Iida M, Aoyagi K, Matsumoto T, Tada S, Yao T, Fujishima M: Effect of indomethacin suppositories on rectal polyposis in patients with familial adenomatous polyposis. Cancer 1996; 78:1660–1665. Hixson LJ, Earnest DL, Fennerty MB, Sampliner RE: NSAID effect on sporadic colon polyps. Am J Gastroenterol 1993;88:1652–1656. Matsuhashi N, Nakajima A, Fukushima Y, Yazaki Y, Oka T: Effects of sulindac on sporadic colorectal adenomatous polyps. Gut 1997;40:344–349. Ladenheim J, Garcia G, Titzer D, Herzenberg H, Lavori P, Edson R, Omary MB: Effect of sulindac on sporadic colonic polyps. Gastroenterology 1995;108:1083–1087. DiSario J, Alberts D, Tietze C, Khullar S, Bohman V, Larsen B, Hixson L, Samowitz W, Reading J, Buys S, Dawson R, Burt R: Sulindac induces regression and prevents progression of sporadic colorectal adenomas. Gastroenterology 1996;112:A555.
Baron
22
59 60 61
62
63 64 65 66 67 68 69
70 71 72
73 74 75
76 77 78 79 80
81 82
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;85:1220–1224. Funkhouser EM, Sharp GB: Aspirin and reduced risk of esophageal carcinoma. Cancer 1995;76: 1116–1119. Farrow DC, Vaughan TL, Hansten PD, Stanford JL, Risch HA, Gammon MD, Chow WH, Dubrow R, Ahsan H, Mayne ST, Schoenberg JB, West AB, Rotterdam H, Fraumeni JF Jr, Blot WJ: Use of aspirin and other nonsteroidal anti-inflammatory drugs and risk of esophageal and gastric cancer. Cancer Epidemiol Biomarkers Prev 1998;7:97–102. Coogan PF, Rosenberg L, Palmer JR, Strom BL, Zauber AG, Stolley PD, Shapiro S: Nonsteroidal anti-inflammatory drugs and risk of digestive cancers at sites other than the large bowel. Cancer Epidemiol Biomarkers Prev 2000;9:119–123. Akre K, Ekstrom AM, Signorello LB, Hansson LE, Nyren O: Aspirin and risk for gastric cancer: A population-based case-control study in Sweden. Br J Cancer 2001;84:965–968. Zaridze D, Borisova E, Maximovitch D, Chkhikvadze V: Aspirin protects against gastric cancer: Results of a case-control study from Moscow, Russia. Int J Cancer 1999;82:473–476. Anderson KE, Johnson TW, Lazovich D, Folsom AR: Association between nonsteroidal antiinflammatory drug use and the incidence of pancreatic cancer. J Natl Cancer Inst 2002;94:1168–1171. Harris RE, Kasbari S, Farrar WB: Prospective study of nonsteroidal anti-inflammatory drugs and breast cancer. Oncol Rep 1999;6:71–73. Harris RE, Namboodiri KK, Stellman SD, Wynder EL: Breast cancer and NSAID use: Heterogeneity of effect in a case-control study. Prev Med 1995;24:119–120. Harris RE, Namboodiri KK, Farrar WB: Nonsteroidal anti-inflammatory drugs and breast cancer. Epidemiology 1996;7:203–205. Sharpe CR, Collet JP, McNutt M, Belzile E, Boivin JF, Hanley JA: Nested case-control study of the effects of non-steroidal anti-inflammatory drugs on breast cancer risk and stage. Br J Cancer 2000;83:112–120. Cotterchio M, Kreiger N, Sloan M, Steingart A: Nonsteroidal anti-inflammatory drug use and breast cancer risk. Cancer Epidemiol Biomarkers Prev 2001;10:1213–1217. Egan KM, Stampfer MJ, Giovannucci E, Rosner BA, Colditz GA: Prospective study of regular aspirin use and the risk of breast cancer. J Natl Cancer Inst 1996;88:988–993. Rosenberg L, Palmer JR, Rao RS, Coogan PF, Strom BL, Zauber AG, Stolley PD, Shapiro S: A case-control study of analgesic use and ovarian cancer. Cancer Epidemiol Biomarkers Prev 2000;9:933–937. Cramer DW, Harlow BL, Titus-Ernstoff L, Bohlke K, Welch WR, Greenberg ER: Over-the-counter analgesics and risk of ovarian cancer. Lancet 1998;351:104–107. Fairfield KM, Hunter DJ, Fuchs CS, Colditz GA, Hankinson SE: Aspirin, other NSAIDs, and ovarian cancer risk (United States). Cancer Causes Control 2002;13:535–542. Tzonou A, Polychronopoulou A, Hsieh CC, Rebelakos A, Karakatsani A, Trichopoulos D: Hair dyes, analgesics, tranquilizers and perineal talc application as risk factors for ovarian cancer. Int J Cancer 1993;55:408–410. Moysich KB, Mettlin C, Piver MS, Natarajan N, Menezes RJ, Swede H: Regular use of analgesic drugs and ovarian cancer risk. Cancer Epidemiol Biomarkers Prev 2001;10:903–906. Tavani A, Gallus S, La Vecchia C, Conti E, Montella M, Franceschi S: Aspirin and ovarian cancer: An Italian case-control study. Ann Oncol 2000;11:1171–1173. Castelao JE, Yuan JM, Gago-Dominguez M, Yu MC, Ross RK: Non-steroidal anti-inflammatory drugs and bladder cancer prevention. Br J Cancer 2000;82:1364–1369. Kaye JA, Myers MW, Jick H: Acetaminophen and the risk of renal and bladder cancer in the general practice research database. Epidemiology 2001;12:690–694. Pommer W, Bronder E, Klimpel A, Helmert U, Greiser E, Molzahn M: Urothelial cancer at different tumour sites: Role of smoking and habitual intake of analgesics and laxatives. Results of the Berlin Urothelial Cancer Study. Nephrol Dial Transplant 1999;14:2892–2897. McLaughlin JK, Blot WJ, Mehl ES, Fraumeni JF Jr: Relation of analgesic use to renal cancer: Population-based findings. Natl Cancer Inst Monogr 1985;69:217–222. McCredie M, Ford JM, Stewart JH: Risk factors for cancer of the renal parenchyma. Int J Cancer 1988;42:13–16.
NSAIDs and Cancer Epidemiology
23
83 84
85
86 87 88 89 90 91 92 93
94 95
96 97
Gago-Dominguez M, Yuan JM, Castelao JE, Ross RK, Yu MC: Regular use of analgesics is a risk factor for renal cell carcinoma. Br J Cancer 1999;81:542–548. Asal NR, Geyer JR, Risser DR, Lee ET, Kadamani S, Cherng N: Risk factors in renal cell carcinoma. II. Medical history, occupation, multivariate analysis and conclusions. Cancer Detect Prev 1988;13:263–279. McCredie M, Pommer W, McLaughlin JK, Stewart JH, Lindblad P, Mandel JS, Mellemgaard A, Schlehofer B, Niwa S: International renal-cell cancer study. II. Analgesics. Int J Cancer 1995;60: 345–349. McCredie M, Stewart JH, Day NE: Different roles for phenacetin and paracetamol in cancer of the kidney and renal pelvis. Int J Cancer 1993;53:245–249. Ross RK, Paganini-Hill A, Landolph J, Gerkins V, Henderson BE: Analgesics, cigarette smoking and other risk factors for cancer of the renal pelvis and ureter. Cancer Res 1989;49:1045–1048. Jensen OM, Knudsen JB, Tomasson H, Sorensen BL: The Copenhagen case-control study of renal pelvis and ureter cancer: Role of analgesics. Int J Cancer 1989;44:965–968. Linet MS, Chow WH, McLaughlin JK, Wacholder S, Yu MC, Schoenberg JB, Lynch C, Fraumeni JF Jr: Analgesics and cancers of the renal pelvis and ureter. Int J Cancer 1995;62:15–18. Norrish AE, Jackson RT, McRae CU: Non-steroidal anti-inflammatory drugs and prostate cancer progression. Int J Cancer 1998;77:511–515. Nelson JE, Harris RE: Inverse association of prostate cancer and non-steroidal anti-inflammatory drugs: Results of a case-control study. Oncol Rep 2000;7:169–170. Habel LA, Zhao W, Stanford JL: Daily aspirin use and prostate cancer risk in a large, multiracial cohort in the US. Cancer Causes Control 2002;13:427–434. Peto R, Gray R, Collins R, Wheatley K, Hennekens C, Jamrozik K, Warlow C, Hafner B, Thompson E, Norton S: Randomised trial of prophylactic daily aspirin in British male doctors. Br Med J Clin Res Ed 1988;296:313–316. Harris RE, Beebe-Donk J, Namboodiri KK: Inverse association of non-steroidal anti-inflammatory drugs and malignant melanoma among women. Oncol Rep 2001;8:655–657. Ladabaum U, Chopra CL, Huang G, Scheiman JM, Chernew ME, Fendrick AM: Aspirin as an adjunct to screening for prevention of sporadic colorectal cancer. A cost-effectiveness analysis. Ann Intern Med 2001;135:769–781. Suleiman S, Rex DK, Sonnenberg A: Chemoprevention of colorectal cancer by aspirin: A costeffectiveness analysis. Gastroenterology 2002;122:78–84. Akhmedkhanov A, Toniolo P, Zeleniuch-Jacquotte A, Kato I, Koenig KL, Shore RE: Aspirin and epithelial ovarian cancer. Prev Med 2001;33:682–768.
John A. Baron, MD Departments of Medicine, and of Community and Family Medicine Dartmouth-Hitchcock Medical Center, Lebanon, NH 03756 (USA) Tel. 1 603 6503456, Fax 1 603 6503473, E-Mail
[email protected]
Baron
24
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 25–51
Pharmacology of COX-2 Inhibitors Peter C. Isakson Pharmacia Research and Development, Peapack, N.J., USA
Introduction
Nonsteroidal anti-inflammatory drugs (NSAIDs) have been used as analgesics in one form or another for centuries. Extracts and preparations from plants such as the willow tree Salix alba have been used for hundreds of years for relief from pain and fever. Derivatives of salicylic acid were characterized as the active components of these plants in the 18th century and were chemically synthesized for the first time in 1860. In 1971, inhibition of the enzyme cyclooxygenase (COX, also known as prostaglandin endoperoxide synthase) was characterized as the mechanism of action of NSAIDs [1]. COX is the ratelimiting enzyme in the production of prostaglandins (PGs), catalyzing the conversion of arachidonic acid to the PG precursor PGH2. PGH2 is then converted to a variety of other PGs including PGE2, PGI2, PGD2 and thromboxane (TxA2). In the late 1980s and early 1990s, a second isoform of COX, COX-2, was identified, cloned and sequenced [2–4]. The two isoforms of COX (COX-1 and COX-2) are distinct with respect to genetic sequence, mechanism of transcriptional regulation, and tissue expression patterns. COX-2 is induced by cytokines in inflammatory cells such as macrophages and monocytes, in tissue at localized sites of injury, and in the spinal cord in response to tissue damage [5–11]. In contrast, COX-1 is constitutively expressed in many tissues, including platelets, gastric mucosa, kidney, etc., and is responsible for maintaining homeostasis in these tissues [12–15]. COX-2 may also be important in other physiological processes, particularly in female reproduction [16, 17]. Although a frequently stated dichotomy is that COX-1 is the constitutive isoform of COX whereas COX-2 is inducible, there are exceptions. Thus, COX-2 is constitutively expressed at low levels in the central nervous system (CNS) and kidney, and basal levels of both COX-1 and COX-2 are observed in the spinal cord [13, 18].
It is widely accepted that COX-2 plays a key role in mediating pain and inflammation in response to tissue damage and in a variety of disease states. COX-2 is induced in response to inflammatory signals, such as the inflammatory cytokines interleukin-1 (IL-1) and tumor necrosis factor-␣ (TNF-␣) [19]. Increased expression of COX-2 mRNA and protein is observed in affected joints of patients suffering from inflammatory diseases such as rheumatoid arthritis (RA) [10, 19] and osteoarthritis (OA) [20]. COX-2 is also induced at peripheral sites of injury in response to tissue damage, as demonstrated by increased PGE2 levels in models of surgical injury and other inflammatory models [21–23], and is upregulated in the CNS in response to peripheral inflammation and humoral factors [9]. Neuroplasticity is characterized by the increased sensitivity of peripheral nociceptors and nociceptive neurons to other painproducing stimuli (peripheral sensitization) [24], which is partly mediated by PGs produced during inflammation by the action of COX-2 [25, 26]. Additionally, the production of excessive PGs both peripherally and centrally leads to modifications in gene expression in the CNS that result in hyperalgesia (heightened and prolonged sensitivity to pain) and allodynia (painful responses to normally nonpainful stimuli) [24–26]. This increased sensitivity to pain induced by phenotypic changes in the CNS is referred to as central sensitization. Unlike COX-2, which is involved in painful and inflammatory responses in a variety of conditions, COX-1 does not appear to have a significant role in the inflammatory process [27] but rather is important in maintaining a variety of homeostatic functions. COX-1 activity, via the synthesis of PGs and TxA2, has been implicated in platelet function and blood clotting, the regulation of blood flow through the kidneys, maintenance of the gastric mucosa, bone metabolism, wound healing and immune responses [12]. Early studies on enzyme inhibition indicated that NSAIDs inhibited both COX isoforms [28]. COX-2 is the therapeutic target of NSAIDs with respect to their analgesic and anti-inflammatory properties. In contrast, inhibition of COX-1 by nonselective NSAIDs is believed to be responsible for many of the adverse side effects of these agents, including those on bleeding and gastrointestinal (GI) mucosa. The introduction of inhibitors that selectively inhibit COX-2 at therapeutic doses has provided a new class of drugs that are as effective as nonselective NSAIDs in treating the pain and inflammation associated with conditions such as RA and OA [29–34], as well as treating other painful conditions such as dysmenorrhea and postoperative pain. The COX-1-sparing nature of COX-2-specific inhibitors accounts for their improved GI and platelet safety profile compared with nonselective NSAIDs [35–39]. The improved long-term upper GI safety profile of the COX-2-specific inhibitors celecoxib and rofecoxib relative to nonselective NSAIDs has been demonstrated in two large-scale trials, each involving over 8,000 arthritis patients [37, 38].
Isakson
26
This chapter will discuss the structural and kinetic features of COX-2specific inhibition, as well as the pharmacology and pharmacokinetic properties of some of the currently available COX-2-specific inhibitors, and will examine the clinical efficacy and safety of this new class of drugs.
Pharmacology of COX-2-Specific Inhibitors
Structural Biology of COX-2-Specific Inhibitors COX-1 and COX-2 are structurally very similar, with two distinct functional active sites (peroxidase and COX) [40]. Both isoforms are 71 kDa and are composed of approximately 600 amino acids [41]. There is a high degree of similarity between the two isoforms, with 63% sequence homology [41] and 77% amino acid similarity [42]. Furthermore, X-ray crystallography studies have demonstrated a high degree of structural similarity between COX-1 and COX-2 [43]; e.g., the structures of both murine and human COX-2 can be superimposed on COX-1 [43–45]. The COX-active site is located at the end of a long and narrow hydrophobic channel [46]. Despite the high degree of structural homology between COX-1 and COX-2, there are important differences that appear to account for inhibitor selectivity. These functional differences appear to be the result of relatively minor structural variations due to single amino acid changes within the catalytic domain (fig. 1). One of the most critical differences is the substitution of the valine residue (Val509) in COX-2 for isoleucine (Ile523) in COX-1 [42, 43, 47]. In the COX-2 isoform, the valine residue, with its smaller side chain, increases accessibility to a side pocket adjacent to the active site, facilitating binding of COX-2-specific inhibitors [43, 44]. In contrast, the COX-1 substrate channel is narrower, due to the presence of the larger isoleucine residue, and is, therefore, unable to accommodate the COX-2-specific inhibitors with their bulky side chains. This has been demonstrated by site-directed mutagenesis of Val509 to isoleucine on the COX-2 isoform, which reduces the ability of COX-2-specific inhibitors to bind and inhibit PGH2 formation [42, 47]. Another valine substitution at position 434 may also play a role in COX-2-specific inhibition by increasing access to the side pocket [43]. Most COX-2-specific inhibitors, i.e., celecoxib, rofecoxib and valdecoxib, have a diarylheterocyclic structure and exhibit marked selectivity for COX-2 (fig. 2). Structure activity studies of COX-2-specific inhibitors have demonstrated the importance of the p-sulfamoylphenyl group in COX-2 inhibition and in vivo efficacy. The phenylsulfonamide moiety binds in a pocket within the channel that leads from the membrane to the COX-active site (fig. 3). As described above, this pocket is more accessible in COX-2 than in
Pharmacology of COX-2 Inhibitors
27
COX-1
COX-2 Catalytic residue lle523
Smaller Val509 opens hydrophilic side pocket
Catalytic residue
Side pocket is inaccessible in COX-1
Val509
Arg120
Arg120 Arachidonic acid
Arachidonic acid PGG2
Phe518, Arg513, Hist90 residues form hydrogen bonds with sulfonamide side chain
PGG2
Fig. 1. Cyclooxygenase active sites of COX-1 and COX-2. The active site of COX isoenzymes is located at the end of a long and narrow hydrophobic channel. COX-1 and COX-2 are characterized by a series of amino acid differences in the hydrophobic channel that are responsible for conformational changes in the protein. A substitution of Ile523 in COX-1 to Val509 residue in COX-2 increases accessibility to a side pocket adjacent to the active site. The COX-1 substrate channel is much narrower, due to the presence of the larger isoleucine residue. Phe518, Arg513 and Hist90 residues in the side pocket of COX-2 are able to form hydrogen bonds with hydrophilic molecules.
O
CH3
N O
O F3C
CH3
N N O
O
S O
Celecoxib
NH2
O
S
O
NH2
S O
CH3
Rofecoxib
Valdecoxib
Fig. 2. Chemical structure of COX-2-specific inhibitors. The COX-2-specific inhibitors celecoxib, rofecoxib and valdecoxib have a diarylheterocyclic structure. To discourage binding to the Arg120 residue in COX-1, there is no carboxylic or enolic acid group, characteristic of nonselective NSAIDs. The bulky side chain of the COX-2-specific inhibitors discourages binding in the narrower hydrophobic channel of the COX-1 isoform. The incorporation of a hydrophilic side chain on COX-2-specific inhibitors encourages binding in the hydrophilic side pocket of the COX-2 isoform.
Isakson
28
Tyr 385 Val 434
Phe 518
Val 523
Ser 530
Arg 513
Arg 120
Arg 516 His 90 SIDE POCKET
Fig. 3. Binding of COX-selective inhibitors to the COX-2 isoform. A COX-2-specific inhibitor (a prototype of celecoxib), depicted in yellow, is shown inside the hydrophobic channel of the catalytic domain of the COX-2 enzyme. The smaller valine residue at position 523 makes the hydrophilic side pocket accessible to the bulky side chain of COX-2-specific inhibitors. Within the side pocket, the Arg513, Phe518 and His90 residues form hydrogen bonds with the sulfonamide group on the side chain of the COX-2-specific inhibitor. Binding of a COX-2-specific inhibitor in this position blocks access of the natural substrate, arachidonic acid, to the catalytic site at Tyr385 [43].
COX-1, as a result of the Ile → Val substitution at position 523 [42, 43, 47]. In COX-2, the smaller valine side chain, together with conformational changes at tyrosine 355 (Tyr355), and potentially a valine substitution at position 434, provides access to the hydrophobic section of the pocket. At position 434, the side chain of the hydrophobic residue rests against Phe518, forming a molecular gate that spans the pocket. This gate is closed in COX-1 as a result of the larger side chain on isoleucine [43]. However, the smaller valine side chain in COX-2 provides enough room for the gate to open, and thus allows the sulfonamide group to enter the pocket. Finally, a third substitution of histidine 513 (His513) in COX-1 to arginine 513 (Arg513) in COX2 may also contribute to the COX-2 selectivity [43, 47]. It appears that one of the consequences of these amino acid differences in the side pocket of COX1 and COX-2 is to create an optimal environment in COX-2 for forming hydrogen-bond interactions with a sulfonamide or methylsulfone-containing compound.
Pharmacology of COX-2 Inhibitors
29
Enzyme Kinetics of COX-2-Specific Inhibitors The enzyme activity of COX-2-specific inhibitors is characterized by a three-step kinetic mechanism that involves reversible competitive binding to both COX isoforms and slow, reversible, time-dependent binding to COX-2 but not COX-1 [48]. The initial binding of COX-2-specific inhibitors to both isoforms is rapid (within seconds) and competitive, typical of most enzyme inhibitors [48–51]. At this stage of the inhibitor/enzyme interaction, there are no differences between the kinetics of interaction of diarylheterocyclic inhibitors with either COX-1 or COX-2. Selectivity for COX-2 is characterized by a subsequent time-dependent step. Time-dependent inhibition of COX was first noted by Rome and Lands [52] with indomethacin, who found that preincubation of COX with this inhibitor was necessary to achieve potent enzyme inhibition; i.e., in a nonselective enzyme assay where inhibitor and enzyme were added to substrate simultaneously, indomethacin showed weak activity, which could be markedly increased by preincubating for several minutes. Preincubation of selective COX-2 inhibitors with the COX enzymes has no effect on potency for COX-1 but shows a slow (minutes), progressive increase in potency against COX-2 as a function of time [48–51]. In other words, interaction of COX-2 inhibitors with COX-2 is uniquely characterized by a tight binding interaction that takes several minutes to reach equilibrium and is noncompetitive kinetically; in contrast, binding of these drugs to COX-1 is rapid, competitive, and low affinity, and thus of no consequence except at very high drug concentrations. Unlike inhibition of COX by aspirin, which entails covalent acetylation of a serine in the active site, COX-2-specific inhibitors do not covalently modify COX-2. However, the observed half-life for inhibitor release from tightly bound enzyme complexes is in the order of hours, suggesting a slow enzyme/inhibitor complex dissociation rate [48–51]. The inhibition kinetics of COX-2-specific inhibitors has been examined using a variety of different in vitro and ex vivo enzyme assays. However, there are distinct limitations in comparing different COX-2-specific inhibitors using these different methods. The relevance of in vitro assays for evaluating inhibition of COX isozymes is limited by a lack of consensus as to the most appropriate assay. Additionally, for certain inhibitory mechanisms such as competitive inhibition and time-dependent reversible inhibition, the results of in vitro enzyme assays are inherently dependent on variables such as enzyme and substrate concentration, presence or absence of membranes, and time and order of addition of substrate and inhibitor. The kinetics of celecoxib, for example, have been measured using the oxygen uptake, peroxidase, and PGE2 ELISA assays [51]. Similar IC50 values were obtained for COX-1, using the three assays; however, IC50 values determined for COX-2 varied considerably
Isakson
30
depending on the assay. Furthermore, lowering the concentration of substrate in an assay will favor the competitive versus time-dependent component of COX-2-specific inhibition, consequently reducing the apparent selectivity of these agents for COX-2 [52]. Thus, while a high degree of selectivity in vitro is an important preliminary criterion, it is only one aspect needed to establish a compound as a COX-2-selective inhibitor. A commonly used method for assessing COX activity is the human whole blood assay [53], wherein whole blood is collected from individuals dosed with a COX inhibitor or control, or an inhibitor is added ex vivo to whole blood. Levels of COX-1 activity are assessed by production of TxA2, generated via platelet COX-1. COX-2 activity is assessed by measuring PGE2 production by monocytes activated by lipopolysaccharide (LPS). The major advantage of this assay is that drugs are evaluated in relevant cell populations, platelets and monocytes, in a physiological milieu. Although widely used, results from different laboratories with a given inhibitor vary considerably, indicating that the assay is not readily transferable, and results from different laboratories are not reliably compared. The whole blood assay typically shows a lesser degree of enzyme selectivity (COX-1 versus COX-2) than the inherent kinetic data obtained with isolated enzyme preparations. Whether the whole blood activity is predictive of results in humans has not been established. The inherent limitations of in vitro and ex vivo assay systems means that it is not possible to predict clinical efficacy or even selectivity based solely on in vitro assays. Therefore, in vivo assays and clinical data must be used to fully assess the pharmacological activity of COX-selective inhibitors. Measurement of PG concentrations in whole animals can be used to monitor the activity of COX-1 or COX-2 in vivo and inhibition of both isozymes can be assessed in a single animal [54]. The relative selectivity of different COX-2-specific inhibitors, celecoxib, valdecoxib and rofecoxib, derived from the use of in vitro recombinant enzyme assays, whole blood assay, and in vivo assay are displayed in table 1. These data demonstrate the extent of variation that arises from the use of these different assays. Therefore, while these different methods are good predictors of the relative COX-2 selectivity of different drugs, it is difficult to make direct comparisons between different agents based on in vitro and animal model data. Nor are these assays necessarily predictive of the clinical efficacy or tolerability of a selective agent. Pharmacokinetics of COX-2-Specific Inhibitors The oral COX-2 inhibitors including celecoxib, rofecoxib and valdecoxib are all highly lipophilic, neutral, nonacidic molecules with limited solubility in aqueous media. All three have been shown to have selectivity for COX-2 versus COX-1 in vitro and in vivo, and to spare COX-1 in humans at therapeutic doses.
Pharmacology of COX-2 Inhibitors
31
Table 1. In vitro, ex vivo and in vivo potency and selectivity
COX-2 specific Celecoxib Rofecoxib Valdecoxib Nonselective Diclofenac Ibuprofen Naproxen
In vitro
Ex vivo
In vivo
enzyme (IC50)
whole blood (IC50)
rat air pouch (ED50)
COX-1 M
COX-2 M
COX-1 M
COX-1 mg/kg
COX-2 mg/kg
15 ⬎1,000 143
0.04 0.5 0.005
8.3 7.0 21.9
0.27 0.22 0.24
⬎200 ⬎200 ⬎200
0.24 1.1 0.06
0.01 117 235
0.16 3.64 9.15
0.02 10.5 ⬎100
0.3 0.2 0.1
0.03 38 32
COX-2 M
0.8 2.5 1.3
Parecoxib sodium is a water-soluble prodrug of valdecoxib that is rapidly converted to the active moiety (valdecoxib) in vivo, and is, therefore, suitable for parenteral IV and IM administration. As with nonselective NSAIDs, COX-2specific inhibitors are highly protein-bound in plasma, primarily to albumin. For example, approximately 97% and 87% of celecoxib and rofecoxib, respectively, are bound to plasma protein. Absorption Oral celecoxib, available in nonselective release capsules, exhibits good absorption, with Cmax reached 2–4 h following a single dose of 100–200 mg [55, 56]. Rofecoxib is also well absorbed in the GI tract with peak plasma concentrations occurring 2–3 h after oral administration [57]. Similarly, valdecoxib is well absorbed with peak plasma concentration occurring within 2 h. Like nonselective NSAIDs, the absorption of COX-2-specific inhibitors is affected by a number of factors, e.g., administration of celecoxib with high-fat meals extends Tmax by 1–2 h [55]. Metabolism and Excretion Celecoxib is converted by hepatic biotransformation to hydroxy, carboxylic acid and glucuronidate derivatives. Oxidative metabolism by the cytochrome P450 (CYP) 2C9 isoenzyme is the primary metabolic pathway, resulting in oxidation of the methyl moiety to a carboxyl group; this is followed by glucuronidation of this carboxyl metabolite to form the major metabolite, SC-62807. Most of the celecoxib metabolites are excreted in the feces, with
Isakson
32
only 2% of celecoxib excreted in the urine unchanged [55]. None of the metabolites of celecoxib have appreciable activity on COX-1 or COX-2. Rofecoxib is metabolized by an uncharacterized cytosolic reductase to cisdihydro and trans-dihydro derivatives, most likely in the liver; these metabolites can undergo further hepatic biotransformation to carboxylic acid and glucuronidate derivatives. Most of a single daily dose is excreted in the urine within 17 h of administration, and less than 1% of excreted drug is unchanged [57]. Valdecoxib also undergoes both CYP-dependent and non-P450-dependent (glucuronidation) metabolism. The CYP-mediated metabolic pathway for valdecoxib primarily involves CYP 3A4 and CYP 2C9. Valdecoxib is metabolized in vivo to a derivative (SC-66905) that is also a COX-2-specific inhibitor, albeit with significantly reduced potency compared with valdecoxib. Drug Interactions of COX-2-Specific Inhibitors COX-selective inhibitors are now used routinely to manage pain and inflammation. Therefore, the likelihood of coadministration with other drugs is high and may have implications for the absorption, pharmacokinetics and elimination of either compound. Unlike nonselective NSAIDs, COX-2-specific inhibitors do not inhibit platelet function, nor do they appear to have any significant interactions with anticoagulant agents [55, 57]. The steady-state concentration of warfarin is not altered by coadministration with celecoxib, and prothrombin times are not significantly affected [55, 56]. Rofecoxib has shown clinically insignificant interactions with warfarin, with coadministration of the two drugs resulting in an 8% increase in prothrombin time in healthy subjects [57]. Coadministration of valdecoxib with warfarin in healthy subjects had no clinically meaningful effect on the pharmacokinetics of warfarin in the majority of healthy adult subjects. Nevertheless, it is recommended that anticoagulant therapy be monitored during the first few weeks after initiating therapy with any of these agents in patients receiving warfarin [58].
Efficacy of Nonselective NSAIDs and COX-2-Specific Inhibitors
Efficacy of COX-2-Specific Inhibitors (table 2) In recent years, two oral COX-2-specific inhibitors, celecoxib and rofecoxib, have been introduced in both US and global markets. Celecoxib was initially approved in the USA in 1999 for the management of pain and inflammation associated with RA and OA, and more recently received an indication for the treatment of primary dysmenorrhea. Rofecoxib was also approved in the USA in 1999 for the symptomatic treatment of OA, and for the treatment of
Pharmacology of COX-2 Inhibitors
33
Table 2. Indications and dosing recommendations for COX-2-specific inhibitors in the USA Celecoxib
Rofecoxib
Valdecoxib
Administration
Oral
Oral
Oral
Osteoarthritis
200 mg/day as 200 mg qd or 100 mg bid
12.5–25 mg qd
10 mg qd
Rheumatoid arthritis
100–200 mg bid
25 mg qd
10 mg qd
Dysmenorrhea
Initial dose: 400 mg (additional 200 mg as needed) Subsequent days: 200 mg bid as needed
Initial dose: 50 mg qd Subsequent doses: 50 mg qd as needed
20 mg bid as needed
Acute pain
Initial dose: 400 mg (additional 200 mg as needed) Subsequent days: 200 mg bid as needed
Initial dose: 50 mg qd Subsequent doses: 50 mg qd as needed
NA
qd ⫽ Once a day; bid ⫽ twice a day.
acute pain including dysmenorrhea, and was approved for use in RA patients more recently. Within the last year, a novel, potent COX-2-specific inhibitor valdecoxib has also been introduced in the USA for the management of pain and inflammation associated with OA and RA, and for the treatment of dysmenorrhea. Celecoxib, the first COX-2-specific inhibitor to be approved in the USA, has demonstrated clinical efficacy that is similar to the nonselective NSAIDs naproxen and diclofenac in treating OA [29, 59, 60]. In randomized, doubleblind, placebo-controlled trials, celecoxib, at a dose of 100 and 200 mg twice daily (bid), was as effective as the nonselective NSAIDs naproxen and diclofenac in treating pain, and in improving physical function in patients suffering from OA of the knee [29, 60]. Rofecoxib (12.5 and 25 mg once daily) also demonstrated efficacy comparable to the nonselective NSAIDs ibuprofen [31, 61] and diclofenac [61] in terms of improving patients’ pain and physical function. Celecoxib and rofecoxib are also recommended for the treatment for RA in adults at doses of 200 mg to 400 mg and 25 mg/day, respectively. At therapeutic doses, celecoxib and rofecoxib effectively reduce the number of tender, painful, or swollen joints, diminish patients’ arthritis pain, and reduce joint stiffness [32, 34]. Thus, COX-2-specific inhibitors are as effective as a number of traditional nonselective NSAIDs in managing pain and inflammation over a
Isakson
34
long-term period in chronic conditions. Furthermore, celecoxib and rofecoxib demonstrate similar efficacy in treating pain and inflammation. In terms of efficacy in treating arthritic conditions, there was no observed difference between celecoxib 200 mg/day and rofecoxib 25 mg/day in treating OA of the knee in a 6-week, randomized, double-blind, placebo-controlled trial [59]. More recently, a third COX-2-specific inhibitor was approved in the USA for the treatment of OA and RA. Valdecoxib, at a therapeutic dose of 10 mg once daily (qd), demonstrated clinical efficacy in treating OA that was similar to the nonselective NSAID naproxen 500 mg bid [33, 62]. Furthermore, valdecoxib 10 mg was also effective in the treatment of RA, demonstrating efficacy similar to naproxen 500 mg bid [30]. Though variability in individual patient response to NSAIDs and to COX-2 specific inhibitors has been observed, in the broader population there appears to be no clinically meaningful differences in efficacy between these agents in treating RA or OA at full therapeutic doses. COX-2-specific inhibitors are also effective in managing other painful conditions. For example, patients suffering from ankylosing spondylitis experienced a significant decrease in pain and functional impairment with celecoxib 200 mg/day and ketoprofen 200 mg/day compared with placebo [63]. Rofecoxib and valdecoxib are both effective and recommended for the treatment of menstrual pain associated with primary dysmenorrhea [64, 65], and celecoxib was recently approved for this condition. Furthermore, celecoxib, valdecoxib and rofecoxib are all effective in managing postoperative pain in a variety of surgical models [66–73], although so far only rofecoxib and celecoxib are approved for the treatment of acute pain in the USA. In addition to demonstrating similar efficacy to nonselective NSAIDs [66, 70–72], celecoxib, rofecoxib and valdecoxib are as effective as mild opioid drug combinations (hydrocodone/ acetaminophen, codeine/acetaminophen and oxycodone/acetaminophen, respectively) in treating postoperative pain [67–69]. Opioid-Sparing Effects of COX-2-Specific Inhibitors Currently available options for the treatment of pain in both acute (e.g., postoperative) and chronic (arthritis, cancer pain, etc.) settings include commonly used agents such as opioids and nonselective NSAIDs. Morphine and other opioids are still commonly used analgesics because of their excellent efficacy [74]. However, opioids are associated with a range of debilitating side effects including respiratory depression, alterations in mental status, ileus, constipation, nausea and vomiting [75–77]. Nonselective NSAIDs have been used for years and are also effective analgesics. In recent years, there has been a growing trend in the use of multimodal treatment strategies, i.e., the use of a combination of analgesic agents to reduce the dose of any one single agent, while still providing adequate or improved
Pharmacology of COX-2 Inhibitors
35
pain relief [75, 78–80]. Reducing opioid intake in patients suffering from acute or chronic pain has the potential to reduce adverse effects and improve quality of life while maintaining appropriate levels of pain relief [78, 80]. Clinical trials have demonstrated that rofecoxib, celecoxib and valdecoxib are all able to reduce opioid requirements when treating postoperative pain in surgical patients [81–83]. The opioid-sparing and analgesic efficacy of COX-2-specific inhibitors, combined with their improved upper GI and platelet safety profile, demonstrate potential for these agents as part of a multimodal treatment strategy, not only in short-term treatment of pain but also in the management of chronic painful conditions.
Safety of Nonselective NSAIDs and COX-2-Specific Inhibitors
GI Safety Risk of Upper GI Complications of Nonselective NSAIDs Nonselective NSAIDs have been associated with adverse upper GI effects due to their nonselective inhibition of COX-1. COX-1 is constitutively expressed in the epithelium of the upper GI tract and is thought to play a protective homeostatic role in the gastric mucosa. The range of NSAID-associated adverse effects runs from mild to moderate GI intolerability, through asymptomatic erosions and ulcerations, to serious upper GI adverse effects such as bleeding ulcers and perforation that can lead to hospitalization and death [84–86]. Poor GI tolerability is a common side effect of NSAID use. Approximately 20–30% of chronic NSAID users suffer from persistent GI symptoms such as dyspepsia, epigastric pain, nausea and vomiting [86]. Dyspepsia, for example, has been reported in approximately 10–12% of NSAID users, and can range from 5 to 50% depending on the drug, patient population, and the study design [84]. There is often little correlation between experiencing GI symptoms and developing endoscopic ulcers or serious GI complications. However, low GI tolerability is often enough to cause discontinuation of NSAID therapy [87, 88]. In addition to upper GI intolerability, chronic NSAID use is associated with more serious GI effects. Estimates of the number of patients who are hospitalized every year and the annual number of deaths related to serious GI complications varies. A report in 1991, on NSAID gastropathy, proposed that NSAID GI pathology was responsible for ⬃70,000 hospitalizations and 7,000 deaths per annum in the United States [89]. More recently, reports from the ARAMIS study (Arthritis, Rheumatism and Aging Medical Information System) predicted that there are over 100,000 hospitalizations and 16,500 deaths per annum due to serious
Isakson
36
GI-related complications such as ulcer bleeding and perforation [90]. This is consistent with findings in other countries; in the UK, it is predicted that there are around 12,000 hospitalizations and 2,230 deaths per annum due to serious GI-related complications [91]. NSAID users are 3 times more likely to develop serious GI complications than nonusers [92]. Additionally, OA and RA patients, who are among the largest groups of consumers of NSAIDs, are predicted to be at 2–2.5 times greater risk of hospitalization from GI complications [90]. In total, it is estimated that 1–4% of NSAID users will require hospitalization for ulcer perforation/bleeding during each year of therapy [84]. The FDA has estimated that symptomatic ulcers, GI bleeding and ulcer perforation occurs in ⬃1–2% of patients taking NSAIDs over a 3-month period and in 2–5% of patients taking NSAIDs for over a year [84, 90, 93]. Odds ratios have been calculated at 2.74 for overall serious GI complications [92], 4.5 for NSAID-related peptic ulcer bleeding [94], 4.7 for NSAID-related upper GI ulcer bleeding and perforation [95], and 7.75 for GI-related surgery [84]. A major concern regarding the widespread use of NSAIDs is the fact that the majority of patients who develop serious GI complications have no prior GI symptoms, making it difficult to predict who will develop the most serious GI complications [85, 90]. However, certain factors, such as age, history of peptic ulcer or GI complications and concomitant use of corticosteroids, increase the risk of NSAID-related GI complications even further. Improved Upper GI Safety and Tolerability of COX-2-Specific Inhibitors In contrast to nonselective inhibitors, COX-2-specific inhibitors do not inhibit COX-1 at therapeutic doses. As a result, these newer agents were predicted to demonstrate an improved safety profile compared to NSAIDs in organs where adverse effects are associated with the inhibition of COX-1, e.g., platelets and the gastric mucosa. The impact of COX-2-specific inhibitors on more serious GI complications is potentially of most significance clinically. NSAIDs have been shown by endoscopy to cause frank GI ulceration in 15–40% of patients; most of these ulcers are asymptomatic. In numerous short and longer term studies (up to 24 weeks), celecoxib and rofecoxib have demonstrated a marked reduction in the incidence of endoscopically observed gastroduodenal ulcers and upper GI complications compared with nonselective NSAIDs, with rates similar to those observed with placebo [36, 96]. In clinical trials, valdecoxib also demonstrated an incidence of gastroduodenal ulceration comparable to placebo and significantly lower than the nonselective NSAIDs naproxen, ibuprofen and diclofenac [33, 97]. There is some controversy over whether an increase in endoscopically determined gastroduodenal ulcers translates into a similar increase in the incidence of serious GI complications such as GI bleeding, perforation, etc.
Pharmacology of COX-2 Inhibitors
37
Although COX-2-specific inhibitors such as rofecoxib and celecoxib have a significantly improved GI safety profile compared with nonselective NSAIDs with respect to endoscopically assessed ulceration, it remained to be determined whether they would demonstrate reduced incidence of serious GI complications in a large patient population. Recently, two large studies to determine the rate of GI complications with these agents have been conducted and published. In the VIGOR trial, over 8,000 RA patients over the age of 50 were randomly assigned to receive a supratherapeutic dose of rofecoxib (50 mg/day), or naproxen 500 mg bid (the standard therapeutic dose for this drug). The primary endpoint for this study was confirmed clinical upper GI events, classified as gastroduodenal perforation or obstruction, upper GI bleeding, or symptomatic gastroduodenal ulcers [38]. Similarly in the CLASS trial (Celecoxib Longterm Arthritis Safety Study) over 8,000 patients with OA or RA received celecoxib 400 mg bid (2–4 times the maximum daily dose for RA and OA), ibuprofen 800 mg 4 times a day (tid) or diclofenac 75 mg bid (both therapeutic doses for these drugs) [37]. The main outcome measures were the incidence of prospectively defined ulcer complications (bleeding, perforation and obstruction) and symptomatic GI ulcers. In both studies other adverse events were monitored and renal and cardiovascular events arising from these studies will be discussed in the renal and CV safety sections. Outcomes from these large trials have demonstrated that rofecoxib, used at a dosage 2 times the therapeutic dose, resulted in significantly fewer symptomatic ulcers and upper GI complications compared with naproxen. Furthermore, celecoxib, used for at least 6 months at a dosage 2–4 times that of the recommended therapeutic dose, was also associated with a reduced incidence of upper GI complications compared with the standard therapeutic doses of the nonselective NSAIDs diclofenac and ibuprofen (although this difference was not statistically significant) and a reduced incidence of symptomatic ulcers. One explanation for the lack of significance in the CLASS trial might be that a higher than anticipated percentage of the study population were using low-dose aspirin for CV prophylaxis (almost double compared with other trials). Analysis of non-aspirin users revealed that there was a significant decrease in the incidence of upper GI complications with celecoxib use compared with ibuprofen and diclofenac. Although similar large prospectively designed trials for valdecoxib have not yet been carried out, combined analysis of the incidence of upper GI complications (bleeding, perforation and obstruction) from 8 randomized trials in OA and RA patients, suggest a safety profile for valdecoxib that is similar to the other COX-2-specific inhibitors [98]. Thus, not only is there a significant reduction in incidence of endoscopic ulcers observed with COX-2-specific inhibitors, but there is also a significant reduction in upper GI complications.
Isakson
38
Platelet Safety Another issue of safety that is a regular concern with NSAID use is the ability of these agents to inhibit platelet function and therefore increase the risk of potential bleeding complications. COX-1 is the only COX isoform expressed in platelets and plays an important role in blood clotting, and is responsible for the production of TxA2 [99]. As NSAIDs are nonselective inhibitors of both COX-1 and COX-2, they all can inhibit platelet function to some degree [99–101]. Aspirin irreversibly inhibits COX, due to its ability to covalently acetylate a serine residue in the active site of COX [102, 103]. Platelets in a fully mature state have no nucleus and are therefore unable to generate new protein. Thus, the irreversible inhibition of COX-1 by aspirin results in deactivation of COX-1 for the entire life span of existing platelets [104]. Other NSAIDs have highly variable enzyme kinetics compared with aspirin and do not permanently inhibit platelet COX-1. However, studies have shown that NSAIDs such as naproxen, ibuprofen and indomethacin, when given at full therapeutic doses, are as effective as aspirin in inhibiting platelet aggregation, reducing TxA2 production, and causing prolonged bleeding [100, 101]. In a randomized controlled trial in healthy adults, celecoxib did not inhibit platelet aggregation in response to arachidonate, collagen or ADP, nor did it reduce TxB2 production or cause significant changes in bleeding time compared with placebo [105]. In contrast, naproxen caused significant reduction in serum TxB2 production and almost 100% inhibition of platelet aggregation [105]. In two similar studies in healthy adults and in healthy elderly adults, valdecoxib also demonstrated a complete lack of inhibition of platelet aggregation or serum TxB2 production, compared with significant reductions of both in response to ibuprofen, diclofenac and naproxen [106, 107]. Similarly, rofecoxib does not demonstrate inhibition of platelet function. In a randomized study in healthy adults, rofecoxib did not affect platelet aggregation, serum TxB2 concentrations or bleeding time, compared with low-dose aspirin, which caused almost 100% inhibition of platelet aggregation and reduction of serum TxB2 [108]. The benefits of improved platelet safety of COX-2-specific inhibitors are important in a number of cases. Patients who have a history of GI complications, and who require NSAID therapy, are at high risk of GI bleeding complications. The potential for GI complications with the use of NSAIDs in this patient population is further exacerbated by an even greater risk of bleeding complications due to inhibition of platelet function. This is illustrated by results of the CLASS trial, where diclofenac and ibuprofen caused substantial decreases in hemoglobin and hematocrit compared to celecoxib [37]. The bleeding risk associated with NSAIDs also has implications in analgesia for surgical patients where administration of nonselective NSAIDs is prohibited in
Pharmacology of COX-2 Inhibitors
39
the preoperative or intraoperative period due to the potential risk of increased bleeding during surgery. The absence of platelet effects with COX-2-specific inhibitors could, therefore, have an impact in such therapeutic areas. Cardiovascular Safety Platelets, together with soluble plasma factors and the vascular endothelium, promote hemostasis. If the integrity of the vascular endothelium is compromised, complex hemostatic changes take place in which platelets aggregate in response to collagen, forming a blood clot and produce TxA2, which is a potent vasoconstrictor and causes further platelet aggregation. Aspirin, by virtue of its ability to permanently inactivate platelet COX-1, completely inhibits platelet production of TxA2 even at low doses (80–325 mg/day) and thus reduces platelet aggregation; this is a beneficial effect in patients at risk for thrombotic events and accounts for the utility of low-dose aspirin for prophylaxis of myocardial infarction (MI) in patients at risk [104, 109]. Whether other NSAIDs are effective in this regard has not been established. COX-2-specific inhibitors were not expected to have any cardioprotective effects, due to their COX-1-sparing mechanism of action. It has been suggested, based primarily on in vitro experiments, that platelet function is regulated by prostacylin (PGI2) produced by the vascular endothelium, as this prostanoid inhibits platelet aggregation and is a potent vasodilator [110]. It is well known that endothelial cells express COX-1 and produce PGI2 in response to vasoactive stimuli such as bradykinin; endothelial cells can be induced to express COX-2 by sheer stress in vitro, and by inflammatory stimuli in vivo [110]. Interest in the role of endothelial COX-2 was stimulated by the observation that the COX-2-specific inhibitors celecoxib and rofecoxib reduced excretion of a PGI2 metabolite in the urine of healthy volunteers by ⬃50%, suggesting that COX-2 contributes to systemic production of PGI2 [110–112]. Although vascular endothelium is a rich source of PGI2 synthase, the actual tissue origin of urinary PGI2 is not known. If vascular COX-2 was in fact the source of PGI2 then it is possible that without concomitant inhibition of platelet aggregation, therapeutic doses of COX-2-specific inhibitors might increase the risk of thromboembolic events due to inhibition of PGI2 in the vascular endothelium in patients with a history of CV disease who have a particularly high risk from MI or stroke. This hypothetical risk would be obviated by appropriate prophylactic therapy with low-dose aspirin [110]. In addition, COX-2 inhibitors have no effect on other vascular regulators of platelet function, in particular nitric oxide (NO). Outcomes from the CLASS and VIGOR trials have provided data on whether CV events are more prevalent in celecoxib- or rofecoxib-treated patients [38, 113]. In the CLASS trial, almost 40% of patients had a history of
Isakson
40
CV disease in all treatment groups, the most common being hypertension [113]. The incidence of CV disease was lower in the non-aspirin-treated group, as would be expected. There were no significant differences between celecoxib and NSAIDs in terms of overall serious CV thromboembolic events, nor were there any differences between the two treatment groups in incidence of cardiac, cerebrovascular or peripheral vascular events, or specific serious CV thromboembolic events such as MI [113]. The one exception was that celecoxib was associated with significantly fewer cerebrovascular events than NSAIDs (ibuprofen and diclofenac) and a lower incidence of sudden cardiac death compared with diclofenac [113]. Separate subanalyses of non-aspirin users demonstrated no differences between celecoxib and NSAIDs in overall CV events, subcategories of events, or selective events such as MI, with the exception of a reduced incidence in cardiac death compared with diclofenac [113]. Further analyses were carried out on a cohort of patients not taking aspirin but for whom aspirin was indicated for prophylaxis. Even in this high-risk group, no significant differences were observed between celecoxib and NSAIDs [113]. It should be noted that the CLASS trial was not designed to evaluate CV effects of celecoxib and cannot exclude a small effect by one of the drugs. Nevertheless, these results do not support the hypothesis that inhibition of COX-2-derived PGI2 leads to increased CV risk, at least in the moderate-risk population studied. In the VIGOR trial there was no difference between rofecoxib and naproxen treatment groups in terms of death from CV causes or ischemic cerebrovascular events. However, the incidence of MI in the rofecoxib treatment group (0.4%) was significantly higher when compared with naproxen treatment (0.1%) [38]. Four percent of the study population met the criteria for secondary CV prophylaxis (i.e., history of MI, angina, cerebrovascular accident, transient ischemic attack, angioplasty or coronary artery disease) but were not taking low-dose aspirin. This high-risk population accounted for 38% of the patients who suffered from MI during the study [38]. It is possible that the higher incidence of MI observed in the VIGOR trial was a consequence of a possible cardioprotective effect of naproxen, and the lack of aspirin prophylaxis in a high-risk subset of the study population taking rofecoxib. Alternatively, the increased incidence of MI in the rofecoxib treatment arm could suggest a risk of prothrombotic effects with a supratherapeutic dose of rofecoxib 50 mg/day over a long-term period or could be due to chance alone. Other smaller studies with rofecoxib have shown similar rates of CV adverse events with rofecoxib, ibuprofen, diclofenac, naproxen and nabumetone [114, 115]. Preliminary analysis of CV event rates from the valdecoxib clinical trial program have demonstrated no difference in the rate of thrombotic events and serious thrombotic events in OA and RA patients treated with valdecoxib,
Pharmacology of COX-2 Inhibitors
41
placebo or naproxen [116, 117], but the number of patient years of exposure in clinical trials with valdecoxib is low compared with celecoxib and rofecoxib. Renal Safety COX-1 and COX-2 are each expressed in the kidney but their expression patterns are quite distinct [12]. COX-1 expression has been localized to mesangial cells, arteriolar endothelial cells, parietal epithelial cells of Bowman’s capsule, and cortical and medullary collecting ducts [118]. In contrast, COX-2 has been demonstrated in the macula densa of the rat kidney and in the interstitial cells of the medulla [13, 119]. The macula densa plays an important role in mediating interaction between glomerular filtration, proximal reabsorption and regulation of renin release [120]. The latter is important for sodium balance and fluid volume. PG production plays an important role in kidney function by effects on salt homeostasis, glomerular filtration rate (GFR) and vascular tone. The exact roles of COX-1 and COX-2 in renal function are undefined, but PGs can affect vascular tone, salt and water homeostasis, and cause renin release [121]. Despite these multiple functions, PGs are not primary mediators of basal renal function, but work in conjunction with a number of other factors that maintain homeostasis. However, under stress situations such as decreased renal blood flow or blood volume, PGs are increased to maintain kidney function. It is estimated that approximately 1–5% of NSAID users are likely to suffer from renal function abnormalities [121]. The most common NSAID-related renal abnormalities are an imbalance of fluids and electrolytes, acute deterioration of renal function, nephritic syndrome with interstitial nephritis and papillary necrosis [121–123]. Of these, imbalance of fluids and electrolytes is by far the most common, occurring in less than 5% of the NSAID-using population. The others all occur at a far lower frequency, in less than 1% of the population. Fluid and electrolyte imbalance manifests itself most frequently as retention of NaCl and water, which results in clinically detectable edema in less than 5% of NSAID-treated patients [121]. Most of the adverse renal effects associated with nonselective NSAIDs are reversible with discontinuation of the drug, with the exception of papillary necrosis [121, 122]. In most cases, these effects are minor and do not have an adverse impact. However, in patients with reduced renal blood perfusion there is a danger of developing acute renal failure [121, 124, 125]. Although NSAIDs do not adversely affect GFR in patients with normal renal function, patients with another risk factor such as renal insufficiency may suffer a significant chronic decrease in GFR and renal blood flow [122, 126]. Additionally, adverse blood pressure control in response to NSAID treatment is particularly common in patients who are coprescribed angiotensin-converting enzyme (ACE) inhibitors, -blockers and diuretics [125]. Finally, the risk of renal failure is significantly
Isakson
42
increased in NSAID-treated patients who have CV risk factors and who are receiving diuretic therapy [122, 125]. Acute deterioration of renal function occurs in 0.5–1% of NSAID users [121]. For most patients NSAID-related renal side effects are liable to be minimal, uncomplicated and easily reversed. However, there is the potential for serious renal failure in high-risk patients. As a result, NSAIDs are used with caution in patient populations with known renal abnormalities and a risk of renal failure. Considering the potential for renal side effects associated with nonselective NSAIDs, it is important to examine renal safety issues of COX-2-specific inhibitors. In the CLASS trial, a similar percentage of patients in the celecoxib and NSAID treatment arms experienced peripheral edema (2.8% and 3.5%, respectively), and a lower percentage of patients in the celecoxib group (1.7%) experienced hypertension compared with NSAIDs (2.3%) [37]. The difference in hypertension may be due to a higher rate in ibuprofen-treated patients but not in diclofenac-treated patients [37]. A preliminary study in healthy adults that compared the effects of naproxen 500 mg bid for 10 days versus celecoxib 200 mg bid for 5 days followed by 400 mg bid for a further 10 days, demonstrated similar significant decreases from baseline in urinary PGs (PGE2 and 6-keto-PGF1␣) in both treatment groups and similar small transient decreases in urinary sodium excretion [127]. However, treatment with naproxen resulted in a greater decrease from baseline in GFR compared with celecoxib and this difference became significant after 6 days of treatment [127]. Retrospective analyses of adverse event data from the celecoxib clinical trial program, involving 50 studies and over 13,000 patients, were carried out to assess the renal safety of celecoxib versus placebo and nonselective NSAIDs [128]. These analyses revealed an overall increase in the incidence of renal adverse events in celecoxib-treated patients compared with placebo that was similar to nonselective NSAIDs tested [128]. The most common events were peripheral edema (2.1%), hypertension (0.8%) and exacerbation of existing hypertension (0.6%); none of these events appeared to be dose- or time-related [128]. Furthermore, there were no adverse drug interactions between celecoxib and ACE inhibitors, -blockers, calcium ion channel blockers or diuretics [128]. Similar post-hoc analyses were carried out using data from the rofecoxib clinical trials program, involving over 5,000 patients [129]. As with celecoxib and NSAIDs, the incidence of lower extremity edema (or peripheral edema) and hypertension were increased among patients receiving rofecoxib compared with patients treated with placebo [129]. In general, the incidence of lower extremity edema and hypertension in the rofecoxib and NSAID treatment arms were similar, although there were some exceptions [129]. The incidence of hypertension was similar between the rofecoxib (12.5 and 25 mg) and ibuprofen treatment groups, but these treatment groups demonstrated an increased rate of hypertension
Pharmacology of COX-2 Inhibitors
43
compared with diclofenac, nabumetone and placebo. Furthermore, rofecoxib 25 mg/day demonstrated an increase in mean systolic blood pressure over baseline that was similar to ibuprofen but greater than that observed with diclofenac, nabumetone and placebo [129]. The similarity between rofecoxib and NSAIDs, with respect to renal-related adverse events, was confirmed by outcomes from the VIGOR trial, where the overall incidence in renal adverse events was similar between the rofecoxib (1.2%) and naproxen (0.9%) treatment groups [38]. A recent study compared the effects of celecoxib, rofecoxib and placebo on the incidence of renal adverse events in elderly OA patients suffering from hypertension. Results from this study indicated that both drugs caused increases in the incidence of peripheral edema and hypertension; however, rofecoxib caused significantly more hypertension and edema compared to celecoxib [130]; this suggests that there may be differences within the COX-2 inhibitor class with respect to effects on the kidney. Although celecoxib demonstrated a lower incidence of renal-related adverse events compared with rofecoxib, both agents as well as NSAIDs can cause increases in renal adverse events, suggesting that both COX-1 and COX-2 play a role in renal physiology. Preliminary data from the valdecoxib clinical trials program demonstrate similar outcomes with an incidence of edema and hypertension that is similar to, but no worse than, that observed with nonselective NSAIDs [131].
Conclusions
The discovery of a novel COX isoform in the late 1980s and characterization of its structure, expression patterns and function has revolutionized our understanding of the role of COX isoforms in pain and inflammation. A major breakthrough in anti-inflammatory and analgesic therapy has been the ability to develop inhibitors that selectively inhibit COX-2 at therapeutic doses. COX-2specific inhibitors have proven to have efficacy that is similar to traditional nonselective NSAIDs in treating pain and inflammation in a number of acute and chronic painful conditions. Furthermore, the selective binding of COX-2 by these agents provides improved GI and platelet safety and tolerability profiles versus nonselective NSAIDs. There are now a number of large, multicentered trials that demonstrate the improved GI safety of these agents, with respect to GI tolerability, development of endoscopic ulcers, and the development of serious, life-threatening GI complications. Since the release of COX-2-specific inhibitors on to the market, there has been some concern about their relative cardiovascular and renal safety. There is a degree of variability among these agents with respect to CV and renal safety. Generally, COX-2-specific inhibitors are associated with a low risk of CV and renal events in arthritis patients and
Isakson
44
are not readily differentiated from nonselective NSAIDs in this regard. Further studies are necessary to understand the physiological role of COX-2, particularly in organs such as the kidney where both COX isoforms appear to play a role.
References 1 2
3
4
5 6
7
8 9
10
11 12 13
14
15 16 17
Vane JR: Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nat New Biol 1971;231:232–235. Kujubu DA, Fletcher BS, Varnum BC, Lim RW, Herschman HR: TIS10, a phorbol ester tumor promoter-inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/ cyclooxygenase homologue. J Biol Chem 1991;266:12866–12872. Merlie JP, Fagan D, Mudd J, Needleman P: Isolation and characterization of the complementary DNA for sheep seminal vesicle prostaglandin endoperoxide synthase (cyclooxygenase). J Biol Chem 1988;263:3550–3553. Xie WL, Chipman JG, Robertson DL, Erikson RL, Simmons DL: Expression of a mitogenresponsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Natl Acad Sci USA 1991;88:2692–2696. Hay C, de Belleroche J: Carrageenan-induced hyperalgesia is associated with increased cyclooxygenase-2 expression in spinal cord. Neuroreport 1997;8:1249–1251. Huang Z, Massey J: Differential regulation of cyclooxygenase-2 mRNA stability by interleukin-1 and tumor necrosis factor-␣ in human in vitro differentiated macrophages. Biochem Pharmacol 2000;59:187–194. Porreca E, Reale M, Di Febbo CD, Di Gioacchino M, Barbacane R, Castellani M, Baccante G, Conti P, Cuccurullo F: Down-regulation of cyclooxygenase-2 by interleukin-receptor antagonist in human monocytes. Immunology 1996;89:424–429. Kang RY, Freire-Moar, Sigal E, Chu CQ: Expression of cyclooxygenase-2 in human and an animal model of rheumatoid arthritis. Br J Rheumatol 1996;35:711–718. Samad TA, Moore KA, Sapirstein A, Billet S, Allchorne A, Poole S, Bonventre JV, Woolf CJ: Interleukin-1-mediated induction of COX-2 in the CNS contributes to inflammatory pain hypersensitivity. Nature 2001;410:471–475. Seigle I, Klein T, Backman JT, Saal JG, Nusing RM, Fritz P: Expression of cyclooxygenase-1 and cyclooxygenase-2 in human synovial tissue: Differential elevation of cyclooxygenase in inflammatory joint disease. Arthritis Rheum 1998;41:122–129. Vanegas H, Schaible HG: Prostaglandins and cyclooxygenases in the spinal cord. Prog Neurobiol 2001;64:327–363. DuBois RN, Abramson SB, Crofford L, Gupta RA, Simon LS, Van De Putte LB, Lipsky PE: Cyclooxygenase in biology and disease. Faseb J 1998;12:1063–1073. Harris R, McKanna J, Akai Y, Jacobson H, DuBois R, Breyer M: Cyclooxygense-2 is associated with the macula densa of rat kidney and increases with salt restriction. J Clin Invest 1994;94: 2504–2510. Seibert K, Zhang Y, Leahy K, Hauser S, Masferrer J, Perkins W, Lee L, Isakson P: Pharmacological and biochemical demonstration of the role of cyclooxygenase-2 in inflammation and pain. Proc Natl Acad Sci USA 1994;91:12013–12017. Smith WL, Garavito RM, DeWitt DL: Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem 1996;271:33157–33160. Matsumoto H, Ma W, Smalley W, Trzaskos J, Breyer R, Dey S: Diversification of cyclooxygenase2-derived prostaglandins in ovulation and implantation. Biol Reprod 2001;64:1557–1565. Pall M, Frieden B, Brännström M: Induction of delayed follicular rupture in the human by the selective COX-2 inhibitor rofecoxib: A randomised double-blind study. Hum Reprod 2001;16: 1323–1328.
Pharmacology of COX-2 Inhibitors
45
18 19
20
21
22
23
24 25 26 27
28
29
30
31
32
33
34
35
36
Lipsky LP, Abramson SB, Crofford L, DuBois RN, Simon LS, van de Putte LB: The classification of cyclooxygenase inhibitors. J Rheumatol 1998;25:2298–2303. Crofford LJ, Wilder RL, Ristimaki AP, Sano H, Remmers EF, Epps HR, Hla T: Cyclooxygenase1 and -2 expression in rheumatoid synovial tissues. Effects of interleukin-1, phorbol ester and corticosteroids. J Clin Invest 1994;93:1095–1101. Amin AR, Attur M, Patel RN, Thakker GD, Marshall PJ, Rediske J, Stuchin SA, Patel IR, Abramson SB: Superinduction of cyclooxygenase-2 activity in human osteoarthritis-affected cartilage. Influence of nitric oxide. J Clin Invest 1997;99:1231–1237. O’Brien TP, Roszkowski MT, Wolff LF, Hinrichs JE, Hargreaves KM: Effect of a non-steroidal anti-inflammatory drug on tissue levels of immunoreactive prostaglandin E2, immunoreactive leukotriene and pain after periodontal surgery. J Periodontol 1996;67:1307–1316. Cominelli F, Nast CC, Duchini A, Lee M: Recombinant interleukin-1 receptor antagonist blocks the proinflammatory activity of endogenous interleukin-1 in rabbit immune colitis. Gastroenterology 1992;103:65–71. Roszkowski MT, Swift JQ, Hargreaves KM: Effect of NSAID administration on tissue levels of immunoreactive prostaglandin E2, leukotriene B4, and (S)-flurbiprofen following extraction of impacted third molars. Pain 1997;73:339–345. Ji RR, Woolf CJ: Neuronal plasticity and signal transduction in nociceptive neurons: Implications for the initiation and maintenance of pathological pain. Neurobiol Dis 2001;8:1–10. Woolf CJ, Costigan M: Transcriptional and posttranslational plasticity and the generation of inflammatory pain. Proc Natl Acad Sci USA 1999;96:7723–7730. Woolf CJ, Salter MW: Neuronal plasticity: Increasing the gain in pain. Science 2000;288:1765–1768. Smith CJ, Zhang Y, Koboldt CM, Muhammad J, Zweifel BS, Shaffer A, Talley JJ, Masferrer JL, Seibert K, Isakson PC: Pharmacological analysis of cyclooxygenase-1 in inflammation. Proc Natl Acad Sci USA 1998;95:13313–13318. Mitchell JA, Akarasereenont P, Thiemermann C, Flower RJ, Vane JR: Selectivity of nonsteroidal anti-inflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA 1993;90:11693–11697. Bensen WG, Fiechtner JJ, McMillen JI, Zhao WW, Yu SS, Woods EM, Hubbard RC, Isakson PC, Verburg KM, Geis GS: Treatment of osteoarthritis with celecoxib, a cyclooxygenase-2 inhibitor: A randomized controlled trial. Mayo Clin Proc 1999;74:1095–1105. Bensen WG, Weaver A, Espinoza L, Zhao WW, Riley W, Paperiello B, Recker D: Efficacy and safety of valdecoxib in treating the signs and symptoms of rheumatoid arthritis: A randomised, controlled comparison with placebo and naproxen. Rheumatology 2002 (in press). Day R, Morrison B, Luza A, Castaneda O, Strusberg A, Nahir M, Helgetveit KB, Kress B, Daniels B, Bolognese J, Krupa D, Seidenberg B, Ehrich E: A randomized trial of the efficacy and tolerability of the COX-2 inhibitor rofecoxib vs. ibuprofen in patients with osteoarthritis. Rofecoxib/Ibuprofen Comparator Study Group. Arch Intern Med 2000;160:1781–1787. Emery P, Zeidler H, Kvien TK, Guslandi M, Naudin R, Stead H, Verburg KM, Isakson PC, Hubbard RC, Geis GS: Celecoxib versus diclofenac in long-term management of rheumatoid arthritis: A randomised double-blind comparison. Lancet 1999;354:2106–2111. Kivitz AJ, Eisen G, Zhao WW, Bevirt T, Recker D: The COX-2-specific inhibitor valdecoxib is as effective as naproxen in treating symptomatic osteoarthritis of the knee and demonstrates reduced gastrointestinal ulceration. J Fam Pract 2002;51:530–537. Schnitzer TJ, Truitt K, Fleischmann RM, Dalgin P, Block J, Zeng Q, Bolognese JA, Seidenberg B, Ehrich E: The safety profile, tolerability and effective dose range of rofecoxib in the treatment of rheumatoid arthritis. Phase II Rofecoxib Rheumatoid Arthritis Study Group. Clin Ther 1999;21: 1688–1702. Bensen WG, Zhao SZ, Burke TA, Zabinski RA, Makuch RW, Maurath CJ, Agrawal NM, Geis GS: Upper gastrointestinal tolerability of celecoxib, a COX-2-specific inhibitor, compared to naproxen and placebo. J Rheumatol 2000;27:1876–1883. Simon LS, Lanza FL, Lipsky PE, Hubbard RC, Talwalker S, Schwartz BD, Isakson PC, Geis GS: Preliminary study of the safety and efficacy of SC-58635, a novel cyclooxygenase-2 inhibitor: Efficacy and safety in two placebo-controlled trials in osteoarthritis and rheumatoid arthritis, and studies of gastrointestinal and platelet effects. Arthritis Rheum 1998;41:1591–1602.
Isakson
46
37
38
39
40 41 42
43
44 45 46
47
48
49
50
51 52 53
54
55
Silverstein FE, Faich G, Goldstein JL, Simon LS, Pincus T, Whelton A, Makuch R, Eisen G, Agrawal NM, Stenson WF, Burr AM, Zhao WW, Kent JD, Lefkowith JB, Verburg KM, Geis GS: Gastrointestinal toxicity with celecoxib versus nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: A randomized controlled trial. Celecoxib Long-Term Arthritis Safety Study. JAMA 2000;284:1247–1255. Bombardier C, Laine L, Reicin AS, Shapiro A, Burgos-Vargas R, Davis B, Ferraz M, Hawkey C, Hochberg M, Kvien TK, Schnitzer TJ: Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:1520–1528. Watson DJ, Harper SE, Zhao PL, Quan H, Bolognese JA, Simon TJ: Gastrointestinal tolerability of the selective COX-2 inhibitor rofecoxib compared with nonselective COX-1 and COX-2 inhibitors in osteoarthritis. Arch Intern Med 2000;160:2998–3003. Hawkey C: COX-2 inhibitors. Lancet 1999;353:307–314. Vane JR, Bakhle Y, Botting R: Cyclooxygenases 1 and 2. Annu Rev Pharmacol Toxicol 1998;39: 97–120. Gierse JK, McDonald JJ, Hauser SD, Rangwala SH, Koboldt CM, Seibert K: A single amino acid difference between COX-1 and COX-2 reverses the selectivity of COX-2-specific inhibitors. J Biol Chem 1996;271:15810–15814. Kurumbail RG, Stevens AM, Gierse JK, McDonald JJ, Stegeman RA, Pak JY, Gildehaus D, Miyashiro JM, Penning TD, Seibert K, Isakson PC, Stallings WC: Structural basis for selective inhibition of cyclooxygenase-2 by anti-inflammatory agents. Nature 1996;384:644–648. Luong C, Miller A, Barnett J, Chow J, Ramesha CS, Browner MF: Flexibility of the NSAID binding site in the structure of human cyclooxygenase-2. Nat Struct Biol 1996;3:927–933. Picot D, Loll PJ, Garavito RM: The X-ray crystal structure of the membrane protein prostaglandin H2 synthase. Nature 1994;367:243–249. Garavito RM: The three-dimensional structure of cyclooxygenases; in Vane JR, Botting J, Botting R (eds): Improved Non-Steroidal Anti-Inflammatory Drugs: COX-2 Enzyme Inhibitors. Dordrecht, Kluwer Academic, 1996, pp 29–43. Wong E, Bayly C, Waterman H, Riendeau D, Mancini J: Conversion of prostaglandin G/H synthase-1 into an enzyme sensitive to PGHS-2-selective inhibitors by a double His513-Arg and Ile523-Val mutation. J Biol Chem 1997;272:9280–9286. Walker M, Kurumbail RG, Kiefer J, Moreland K, Koboldt C, Isakson PC, Seibert K, Gierse JK: A three-step kinetic mechanism for selective inhibition of cyclo-oxygenase-2 by diarylheterocyclic inhibitors. Biochem J 2001;357:709–718. Chan CC, Boyce S, Brideau C, Charleson S, Cromlish E, Ethier D, Evans J, Ford-Hutchison AW, Forrest M, Gauthier J, Gordon R, Gresser M, Guay J, Kargman S, Kennedy B, Leblanc Y, Leger S, Mancini J, O’Niell G, Ouellet M, Patrick D, Percival MD, Perrier H, Prasit P, Rodger IW, Tagari P, Therien M, Vickers PJ, Visco D, Wang Z, Webb JK, Wong E, Xu LJ, Young R, Zamboni R, Riendeau D: Rofecoxib (Vioxx, MK-0966; 4-(4⬘-methylsulfonylphenyl)-3-phenyl-2(5H)furanone): a potent and orally active cyclooxygenase-2 inhibitor. Pharmacological and biochemical profiles. J Pharmacol Exp Ther 1999;290:551–560. Copeland RA, Williams JM, Giannaras J, Nurnberg S, Covington M, Pinto D, Pick S, Trzaskos JM: Mechanism of selective inhibition of the inducible isoform of prostaglandin G/H synthase. Proc Natl Acad Sci USA 1994;91:11202–11206. Gierse JK, Koboldt CM, Walker MC, Seibert K, Isakson PC: Kinetic basis for selective inhibition of cyclo-oxygenases. Biochem J 1999;339:607–614. Rome LH, Lands WE: Structural requirements for time-dependent inhibition of prostaglandin biosynthesis by anti-inflammatory drugs. Proc Natl Acad Sci USA 1975;72:4863–4865. Patrignani P, Panara MR, Greco A, Fusco O, Natoli C, Iacobelli S, Cipollone F, Ganci A, Creminon C, Maclouf J: Biochemical and pharmacological characterization of the cyclooxygenase activity of human blood prostaglandin endoperoxide synthases. J Pharmacol Exp Ther 1994;271:1705–1712. Masferrer JL, Zweifel BS, Manning PT, Hauser SD, Leahy KM, Smith WG, Isakson PC, Seibert K: Selective inhibition of inducible cyclooxygenase 2 in vivo is anti-inflammatory and nonulcerogenic. Proc Natl Acad Sci USA 1994;91:3228–32. Davies NM, McLachlan AJ, Day RO, Williams KM: Clinical pharmacokinetics and pharmacodynamics of celecoxib: A selective cyclooxygenase-2 inhibitor. Clin Pharmacokinet 2000;38:225–242.
Pharmacology of COX-2 Inhibitors
47
56
57 58
59
60 61
62
63
64
65
66
67
68
69
70 71
72 73
74 75 76
Karim A, Tolbert D, Piergies A, Hubbard RC, Harper K, Wallemark CB, Slater M, Geis GS: Celecoxib does not significantly alter the pharmacokinetics or hypoprothrombinemic effect of warfarin in healthy subjects. J Clin Pharmacol 2000;40:655–663. Scott LJ, Lamb HM: Rofecoxib. Drugs 1999;58:499–505. Karim A, Schwartz L, Boynton J, Slater M, Snabes MC, Piergies A: Multiple dose pharmacokinetics and tolerability of parecoxib sodium, a prodrug of the COX-2-specific inhibitor, following intravenous and intramuscular administration. American Society of Health-System Pharmacists Annual Meeting. Am J Health-System Pharmacy 2001;36:abstr P-456E. McKenna F, Weaver A, Fiechtner JJ, Bello AE, Fort JG: COX-2-specific inhibitors in the management of osteoarthritis of the knee: A placebo-controlled, randomized, double-blind study. J Clin Rheumatol 2001;7:151–159. McKenna F, Borenstein D, Wendt H, Wallemark C, Lefkowith JB, Geis GS: Celecoxib versus diclofenac in the management of osteoarthritis of the knee. Scand J Rheumatol 2001;30:11–18. Saag K, Van der Heijde D, Fisher C, Samara A, DeTora L, Bolognese J, Sperling R, Daniels B: Rofecoxib, a new cyclooxygenase-2 inhibitor, shows sustained efficacy, comparable with other nonsteroidal anti-inflammatory drugs. Arch Fam Med 2000;9:1124–1134. Makarowski W, Zhao W, Bevirt T, Recker D: Efficacy and safety of the COX-2-specific inhibitor valdecoxib in the management of osteoarthritis of the hip: A randomized, double-blind, placebocontrolled comparison with naproxen. Osteoarthritis Cartilage 2002;10:290–296. Dougados M, Behier JM, Jolchine I, Calin A, Van der Heijde D, Olivieri I, Zeidler H, Herman H: Efficacy of celecoxib, a cyclooxygenase 2-specific inhibitor, in the treatment of ankylosing spondylitis. Arthritis Rheum 2001;44:180–185. Morrison BW, Daniels S, Kotey P, Cantu N, Seidenberg B: Rofecoxib, a specific cyclooxygenase-2 inhibitor, in primary dysmenorrhea: A randomized controlled trial. Obstet Gynecol 1999;94: 504–508. Daniels S, Talwalker S, Torri S, Kuss M, Snabes MC, Recker D, Verburg KM: Valdecoxib, a COX2-specific inhibitor, is effective and well tolerated in treating women with primary dysmenorrhea. Obstet Gynecol 2001 (in press). Chang D, Desjardins PJ, Chen E, Polis A, McAvoy M, Mockoviack S, Geba G: Comparison of the analgesic effects of rofecoxib and enteric-coated diclofenac sodium in the treatment of postoperative dental pain: A randomized, placebo-controlled clinical trial. Clin Ther 2002;24:490–503. Chang D, Fricke J, Bird S, Bohidar N, Dobbins T, Geba G: Rofecoxib versus codeine/acetaminophen in postoperative dental pain: A double-blind, randomized, placebo- and active comparator-controlled clinical trial. Clin Ther 2001;23:1446–1455. Daniels SE, Desjardins PJ, Talwalker S, Kuss M, Recker D, Verburg KM: Single oral doses of valdecoxib are as effective as oxycodone/acetaminophen in reducing pain following oral surgery: Results of two randomized, double-blind trials. J Am Dent Assoc 2002;133:611–621. Gimbel JS, Brugger A, Zhao W, Verburg KM, Geis GS: Efficacy and tolerability of celecoxib versus hydrocodone/acetaminophen in the treatment of pain after ambulatory orthopedic surgery in adults. Clin Ther 2001;23:228–241. Fricke J, Varkalis J, Zwillich S, Adler R, Forester E, Recker D, Verburg KM: Valdecoxib is more efficacious than rofecoxib in relieving pain associated with oral surgery. Am J Ther 2002;9:89–97. Malmstrom K, Daniels S, Kotey P, Seidenberg BC, Desjardins PJ: Comparison of rofecoxib and celecoxib, two cyclooxygenase-2 inhibitors, in postoperative dental pain: A randomized, placebo and active comparator-controlled clinical trial. Clin Ther 1999;21:1653–1663. Reuben SS, Fingeroth R, Krushell R, Maciolek H: Evaluation of the safety and efficacy of the perioperative administration of rofecoxib for total knee arthroplasty. J Arthroplast 2002;17:26–31. Desjardins P, Talwalker S, Recker D, Verburg K, Woolf C: A single pre-operative oral dose of valdecoxib, a long-acting COX-2-specific inhibitor, relieves post-oral surgery or bunionectomy pain. Anesthesiology 2002 (in press). Hamilton GR, Baskett TF: In the arms of Morpheus, the development of morphine for postoperative pain relief. Can J Anaesth 2000;47:367–374. Kehlet H, Rung GW, Callesen T: Postoperative opioid analgesia: Time for a reconsideration? J Clin Anesth 1996;8:441–445. Mulroy MF: Monitoring opioids. Reg Anesth 1996;21:89–93.
Isakson
48
77 78 79 80 81 82 83
84
85
86 87
88 89 90 91
92 93
94
95 96
97
98
99
Wheeler M, Oderda G, Ashburn M, Lipman A: Adverse events associated with postoperative opioid analgesia: A systematic review. J Pain 2002;3:159–180. Kehlet H, Dahl JB: The value of ‘multimodal’ or ‘balanced analgesia’ in postoperative pain treatment. Anesth Analg 1993;77:1048–1056. Rawal N: Management of Acute and Chronic Pain. London, BMJ Books, 1998. Rawal N: Ten years of acute pain services – Achievements and challenges. Reg Anesth Pain Med 1999;24:68–73. Camu F, Beecher T, Recker D, Verburg KM: Valdecoxib, a COX-2-specific inhibitor, is an efficacious, opioid-sparing analgesic in patients undergoing hip arthroplasty. Am J Ther 2002;9:43–51. Reuben SS, Connelly NR: Postoperative analgesic effects of celecoxib or rofecoxib after spinal fusion surgery. Anesth Analg 2000;91:1221–1225. Reynolds L, Brill R, Recker D, Sainati S, Verburg KM: The COX-2-specific inhibitor, valdecoxib, is an effective, opioid-sparing analgesic in patients undergoing total knee arthroplasty. J Pain Sympt Manag 2002 (in press). Singh G, Ramey DR: NSAID-induced gastrointestinal complications: The ARAMIS perspective 1997. Arthritis, Rheumatism and Aging Medical Information System. J Rheumatol 1998; 51(suppl):8–16. Singh G, Ramey DR, Morfeld D, Shi H, Hatoum HT, Fries JF: Gastrointestinal tract complications of nonsteroidal anti-inflammatory drug treatment in rheumatoid arthritis. A prospective observational cohort study. Arch Intern Med 1996;156:1530–1536. Verburg KM, Maziasz TJ, Weiner E, Loose L, Geis GS, Isakson PC: COX-2-specific inhibitors: Definition of a new therapeutic concept. Am J Ther 2001;8:49–64. Barrier C, Hirschowitz B: Controversies in the detection and management of nonsteroidal anti-inflammatory drug-induced side effects in the upper gastrointestinal tract. Arthitis Rheum 1989;32:926–932. Scholes D, Stergachis A, Penna P, Normand E, Hansten P: Nonsteroidal anti-inflammatory drug discontinuation in patients with osteoarthritis. J Rheumatol 1995;22:708–712. Fries JF: NSAID gastropathy: Epidemiology. J Musculoskel Med 1991;8:21–28. Singh G: Recent considerations in nonsteroidal anti-inflammatory drug gastropathy. Am J Med 1998;105:31S–38S. Blower AL, Brooks A, Fenn GC, Hill A, Pearce MY, Morant S, Bardhan KD: Emergency admissions for upper gastrointestinal disease and their relation to NSAID use. Aliment Pharmacol Ther 1997;11:283–291. Gabriel SE, Jaakkimainen L, Bombardier C: Risk for serious gastrointestinal complications related to use of nonsteroidal anti-inflammatory drugs. A meta-analysis. Ann Intern Med 1991;115:787–796. Paulus H: FDA Arthritis Advisory Committee Meeting: Risks of agranulocytosis/aplastic anemia, flank pain and adverse gastrointestinal effects with nonsteroidal anti-inflammatory drugs. Arthitis Rheum 1987;30:593–595. Langman MJ, Weil J, Wainwright P, Lawson DH, Rawlins MD, Logan RF, Murphy M, Vessey MP, Colin-Jones DG: Risks of bleeding peptic ulcer associated with individual non-steroidal anti-inflammatory drugs. Lancet 1994;343:1075–1078. Garcia Rodriguez LA, Jick H: Risk of upper gastrointestinal bleeding and perforation associated with individual nonsteroidal anti-inflammatory drugs. Lancet 1994;343:769–772. Langman MJ, Jensen DM, Watson DJ, Harper SE, Zhao PL, Quan H, Bolognese JA, Simon TJ: Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs. JAMA 1999;282: 1929–1933. Goldstein JL, Kent JD, Zhao W, Recker D, Verburg KM: Reduced incidence of gastroduodenal ulcers with valdecoxib compared to ibuprofen and diclofenac in patients with rheumatoid arthritis: A multicenter trial (abstract). Am J Gastroenterol 2001;96(suppl):9. Goldstein JL, Stenson WF, Agrawal N, Eisen G, Omachi J, Kent JD, Recker D, Verburg KM: Valdecoxib is associated with a significantly lower incidence of ulcer complications and symptomatic ulcers in arthritis patients as compared to NSAIDs (Abstract – Digestive Disease Week). Gastroenterology 2002;122(suppl 1):5. Schror K: Aspirin and platelets: The antiplatelet action of aspirin and thrombosis treatment and prophylaxis. Semin Thromb Hemost 1997;23:349–356.
Pharmacology of COX-2 Inhibitors
49
100 Cheng JC, Siegel L, Katari B, Traynoff S, Ro J: Nonsteroidal anti-inflammatory drugs and aspirin: A comparison of the antiplatelet effect. Am J Ther 1997;4:62–65. 101 Nadell J, Bruno J, Varaday J, Segre E: Effect of naproxen and of aspirin on bleeding time and platelet aggregation. J Clin Pharmacol 1974;14:176–182. 102 Roth G, Majerus PW: The mechanism of the effect of aspirin on human platelets. J Clin Invest 1975;56:624–632. 103 Roth G, Stanford N, Majerus PW: Acetylation of prostaglandin synthase by aspirin. Biochemistry 1975;72:3073–3076. 104 Patrono C, Coller B, Dalen J, Fitzgerald G, Fuster V, Gent M, Hirsh J, Roth G: Platelet-active drugs: The relationship among dose, effectiveness and side effects. Chest 2001;119:39S–63S. 105 Leese PT, Hubbard RC, Karim A, Isakson PC, Yu SS, Geis GS: Effects of celecoxib, a novel cyclooxygenase-2 inhibitor, on platelet function in healthy adults: A randomized, controlled trial. J Clin Pharmacol 2000;40:124–132. 106 Leese P, Recker D, Kuss M: A double-blind, placebo-controlled study to evaluate the effects of valdecoxib, a novel COX-2-specific inhibitor, on platelet function in the elderly. Ann Rheum Dis 2001;60(suppl 10):abstr SAT0057. 107 Leese P, Talwalker S, Recker D, Kent J: Valdecoxib, a novel COX-2-specific inhibitor, does not impair platelet function in healthy adults: results of a randomized controlled trial. Am J Emerg Med 2002 (in press). 108 Greenberg H, Gottesdiener K, Huntington M, Wong P, Larson P, Wildonger L, Gillen L, Dorval E, Waldman S: A new cyclooxygenase-2 inhibitor, rofecoxib (VIOXX), did not alter the antiplatelet effect of low-dose aspirin in healthy volunteers. J Clin Pharmacol 2000;40:1509–1515. 109 Antiplatelet Trialists’ Collaboration: Collaborative overview of randomised trials of antiplatelet therapy-I: Prevention of death, myocardial infarction and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ 1994;308:81–106. 110 Fitzgerald G, Patrono C: The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med 2001;345:433–442. 111 Catella-Lawson F, McAdam B, Morrison BW, Kapoor SC, Kujubu DA, Antes L, Lasseter K, Quan H, Gertz BJ, Fitzgerald GA: Effects of specific inhibition of cyclooxygenase-2 on sodium balance, hemodynamics and vasoactive eicosanoids. Pharmacology 1999;289:735–741. 112 McAdam B, Catella-Lawson F, Mardini I, Kapoor SC, Lawson J, Fitzgerald G: Systemic biosynthesis of prostacyclin by COX-2: The human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA 1999;96:2722–277. 113 White W, Faich G, Whelton A, Maurath C, Ridge N, Verburg K, Geis G, Lefkowith J: Comparison of thromboembolic events in patients treated with celecoxib, a cyclooxygenase-2 specific inhibitor, versus ibuprofen and diclofenac. Am J Cardiol 2002;89:425–430. 114 Konstam M, Weir M, Reicin AS, Shapiro D, Sperling R, Barr E, Gertz BJ: Cardiovascular thrombotic events in controlled clinical trials of rofecoxib. Circulation 2001;104:2280–2288. 115 Reicin A, Shapiro D, Sperline R, Barr E, Yu Q: Comparison of cardiovascular thrombotic events in patients with osteoarthritis treated with rofecoxib versus nonselective nonsteroidal antiinflammatory drugs (ibuprofen, diclofenac and nambutone). Am J Cardiol 2002;89:204–209. 116 White W, Recker D, Kent JD, Verburg KM, Whelton A: The COX-2-specific inhibitor, valdecoxib, is not associated with an increased risk of serious thrombotic events in arthritis patients (abstract). Am Coll Cardiol 2002 (submitted). 117 Whelton A, Kent JD, Recker D: No difference in thrombotic events in rheumatoid arthritis patients: Valdecoxib vs. naproxen and placebo (abstract). Am Coll Rheumatol. Arthritis Rheum 2002 (submitted). 118 Smith WL, Bell T: Immunohistochemical localization of the prostaglandin-forming cyclooxygenase in renal cortex. Am J Physiol 1978;235:F451–F457. 119 Guan Y, Chang M, Cho W, Zhang Y, Redha R, Davis L, Chang S, DuBois R, Hao C, Breyer M: Cloning, expression, and regulation of rabbit cyclooxygenase-2 in renal medullary interstitial cells. Am J Physiol 1997;273:F18–F26. 120 Harris R: The macula densa: Recent developments. J Hypertens 1996;14:815–822. 121 Whelton A, Hamilton C: Nonsteroidal anti-inflammatory drugs: Effects on kidney function. J Clin Pharmacol 1991;31:588–598.
Isakson
50
122 Clive D, Stoff J: Renal syndromes associated with nonsteroidal anti-inflammatory drugs. N Engl J Med 1984;310:563–572. 123 Murray M, Lazaridis E, Brizendine E, Haag K, Becker P, Brater D: The effect of nonsteroidal anti-inflammatory drugs on electrolyte homeostasis and blood pressure in young and elderly persons with and without renal insufficiency. Am J Med Sci 1997;314:80–88. 124 Perneger T, Whelton P, Klag M: Risk of kidney failure associated with the use of acetaminophen, aspirin and nonsteroidal anti-inflammatory drugs. N Engl J Med 1994;331:1675–1679. 125 Whelton A: Nephrotoxicity of nonsteroidal anti-inflammatory drugs: Physiologic foundations and clinical implications. Am J Med 1999;106:13S–24S. 126 Murray M, Black P, Kuzmik D, Haag K, Manatunga A, Mullin M, Hall S, Brater D: Acute and chronic effects of nonsteroidal anti-inflammatory drugs on glomerular filtration rate in elderly patients. Am J Med Sci 1995;310:188–197. 127 Whelton A, Schulman G, Wallemark C, Drower EJ, Isakson PC, Verburg KM, Geis GS: Effects of celecoxib and naproxen on renal function in the elderly. Arch Intern Med 2000;160:1465–1470. 128 Whelton A, Maurath CJ, Verburg KM, Geis GS: Renal safety and tolerability of celecoxib, a novel cyclooxygenase-2 inhibitor. Am J Ther 2000;7:159–175. 129 Gertz B, Krupa D, Bolognese JA, Sperling R, Reicin A: A comparison of adverse renovascular experiences among osteoarthritis patients treated with rofecoxib and comparator nonselective nonsteroidal anti-inflammatory agents. Curr Med Res Opin 2002;18:82–91. 130 Whelton A, Fort JG, Puma JA, Normandin D, Bello AE, Verburg KM: Cyclooxygenase-2-specific inhibitors and cardiorenal function: A randomized, controlled trial of celecoxib and rofecoxib in older hypertensive osteoarthritis patients. Am J Ther 2001;8:85–95. 131 Whelton A, White W, Recker D, Kent JD, Verburg KM: Valdecoxib is associated with an incidence of renal effects similar to nonselective NSAIDs (abstract). Am Coll Rheumatol. Arthritis Rheum 2001 (submitted).
Peter C. Isakson, PhD Pharmacia Research and Development, Pharmacia Corporation 100 Route 206 North, Peapack, NJ 07977 (USA) Tel. ⫹1 908 9018025, Fax ⫹1 908 9011755, E-Mail
[email protected]
Pharmacology of COX-2 Inhibitors
51
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 52–71
Regulation of COX-2 Expression in Human Cancers Dan A. Dixon Surgical Oncology Research Laboratory, Departments of Surgery and Cancer Biology, Vanderbilt University Medical Center, Nashville, Tenn., USA
Introduction
Metabolites of arachadonic acid participate in normal growth responses and in aberrant cellular growth and proliferation, including carcinogenesis [1, 2]. The key step in the conversion of free arachidonic acid to prostaglandins is catalyzed by the cyclooxygenase enzyme (COX). Until recently, only one isoform of the COX enzyme (COX-1) had been purified and cloned. However, it is now known that a second inducible isoform of the COX enzyme (COX-2) exists. The COX-1 enzyme is constitutively expressed at low levels in a majority of tissues and presumably makes prostaglandins for normal physiological functions. By contrast, COX-2 is not normally present in most cells, but tight regulation allows it to be rapidly expressed in response to growth-related signals resulting in increased prostaglandin synthesis associated with inflammation and carcinogenesis. Insight into the molecular regulation of COX-2 expression preceded its discovery. Work in the late 1980s identified an inducible COX activity that was temporally regulated on both transcriptional and post-transcriptional levels [3]. Near the same time, at least two independent groups had identified and cloned COX-2 as an immediate-early response gene whose expression was highly induced in response to cellular transformation by v-src [4] or treatment of cells with phorbol ester [5]. Subsequent work led to the cloning of the human COX-2 cDNA from vascular endothelial cells [6, 7] and isolation of the COX-2 gene [8]. Examination of the COX-2 5⬘ promoter region has identified several transcription factor regulatory elements. A number of mRNA instability elements are contained within the 3⬘ untranslated region (3⬘UTR) of COX-2
mRNA. The fact that COX-2 expression is regulated at both transcriptional and post-transcriptional levels implies that fine control of expression is important. The connection between COX-2 expression and carcinogenesis was first suggested by studies that demonstrated the efficacy of aspirin and NSAIDs to reduce the relative risk of colon cancer and also promote tumor regression in both humans and experimental animal models of colon cancer [2]. Investigation of the molecular basis of these observations showed that high levels of COX-2 protein were present in both human and animal colorectal tumors, whereas the normal intestinal mucosa has low to undetectable COX-2 expression [9–12]. Similar to the situation with colorectal cancer, other solid malignancies such as breast, lung, prostate, pancreas, bladder, stomach, esophagus, and head and neck exhibit elevated COX-2 levels resulting from changes in the regulation of COX-2 expression. This association of COX-2 overexpression and carcinogenesis was further established in genetic studies demonstrating a significant reduction in intestinal polyposis in mice deficient for COX-2 gene [13] along with demonstration of COX-2 overexpression is sufficient to induce breast tumorigenesis [14] and pre-malignant lesions in the skin of mice [15]. Taken together, these findings clearly indicate that chronic elevation of COX-2 is pathological and suggest that inhibition of COX-2 via pharmacological means or regulation of its expression can limit the development or progression of human cancers. Several lines of evidence gathered from epidemiological, experimental models, and cellular studies indicate that unregulated COX-2 expression is an important step in tumorigenesis and indicate that dysregulation occurs early in carcinogenesis, particularly colorectal cancer. The control of COX-2 expression is a complex regulatory process that requires input from multiple signal transduction pathways. Cellular defects in these signaling pathways can differentially promote the expression of COX-2 through the loss of transcriptional and post-transcriptional regulatory mechanisms. The sum of these effects results in a dramatic increase in COX-2 protein levels and associated prostaglandin production. This review is intended to summarize the major advances in the field of molecular regulation of COX-2 expression in cancer, largely focusing on the recent findings pertaining to colorectal cancer. Particular attention is paid to novel mechanisms controlling COX-2 expression at the post-transcriptional level.
Transcriptional Regulation of COX-2
A variety of studies support the hypothesis that an early event in tumorigenesis, such as the mutation of the ‘gate-keeper’ adenomatous polyposis coli
Regulation of COX-2 Expression in Cancer
53
NF-B
NF-IL-6
⫺223/⫺214
⫺132/⫺124
PEA3
CRE
TATA
⫺75/⫺72 ⫺59/⫺53 ⫺31/⫺25
⫹1
Fig. 1. Proximal region of the human COX-2 promoter. The transcription start site (⫹1) is indicated by an arrow. The TATA box located at –31/–25 is shown as a box and the well-characterized transcription factor-binding sites are depicted as ovals.
(APC) gene, initiates the unregulated expression of COX-2 [16, 17]. However, the molecular events leading to the constitutive transcription of COX-2 in human cancer are not totally understood. Evidence of constitutive activation of the COX-2 promoter occurring in colon cancer cells [12] suggested that the increased levels of COX-2 mRNA detected in colorectal adenomas, adenocarcinomas [9, 12] and colon cancer cell lines [12] occurs through increased transcription. Similarly, increased COX-2 transcription has been observed in transformed mammary epithelial cells [18]. Although the underlying mechanism is not completely clear, several key cis-acting elements localized within 250 bases of the transcription start site of the COX-2 gene promoter region have been shown to play a decisive role in the regulation of COX-2 expression in human carcinoma cells (fig. 1). The Proximal Region of the COX-2 Promoter Regulates Transcription TATA Box: A characteristic element present in many immediate-early response genes, such as COX-2, is the TATA box. Recognition of the TATA box is conferred by the TATA-binding protein (TBP) which promotes assembly of a functional transcription initiation complex. The p53 tumor suppressor protein has been demonstrated to inhibit expression of genes containing a TATA box [19] and recent findings have shown that wild-type p53 protein causes a marked decrease in COX-2 gene transcription as a result of competition for TATA-box binding with TBP [20]. This notion is supported by observations demonstrating correlation between p53 mutation and COX-2 overexpression in several human cancers [21]. Thus, the occurrence of p53 mutational inactivation, a common genetic event in human cancer [22], may prove to be an important link in understanding why constitutive expression of COX-2 occurs in many cancers. CRE: Original observations showing the involvement of the cAMPresponse element site (CRE) in COX-2 transcriptional regulation were demonstrated in cells responding to v-src activation or growth factor treatment [23, 24]. As a consequence of JNK kinase cascade activation, enhanced phosphorylation
Dixon
54
of c-Jun occurs allowing it to activate transcription as either a homodimer or heterodimer with c-Fos or ATF-2, producing the AP-1 transcription factor. Due to altered MAPK signaling detected in colon cancer cells and transformed mammary epithelial cells, increased activation of AP-1 contributes to enhanced COX-2 gene transcription through the CRE site [25–27]. Furthermore, AP-1-mediated COX-2 transcription is an effective target of inhibition by chemopreventive agents. Studies of cancer cells treated with peroxisome proliferator-activated receptor (PPAR) ligands or retinoids display decreased COX-2 transcription as a result of their ability to squelch CBP/p300, a co-activator of AP-1 mediated gene expression which links AP-1 to the basal transcription machinery [28, 29]. PEA3: The transcriptional activation of COX-2 has been observed under conditions where the transcriptional activator -catenin is stabilized as a consequence of APC mutation in intestinal tumors and Wnt signaling in breast tumors [18, 30, 31]. However, since COX-2 has not been shown to be a direct target of -catenin-mediated transcription, this suggests that -catenin may possibly activate COX-2 transcription through an intermediary transcription factor [32]. Recent findings have shown the PEA3 subfamily of Ets transcription factors promote COX-2 transcription in breast and colorectal cancer lines through a PEA3/Ets site present in the COX-2 promoter [27, 32]. Accordingly, high PEA3 transcription factor levels have been detected in both intestinal and breast tumors [33]. Although the connection between -catenin and PEA3 is not known, the correlation between PEA3 and COX-2 expression in both breast and intestinal tumors suggests that PEA3 contributes to COX-2 transcriptional upregulation in response to both Wnt1 signaling and mutations in APC. NF-IL-6: The trans-acting factors that promote transcription through the NF-IL-6 site include members of the CCAAT/enhancer-binding protein (C/EBP) family of basic leucine zipper transcription factors [34]. Binding of C/EBP to the NF-IL-6 site has been reported to be crucial for COX-2 promoter activation by stimuli such as mitogens and growth factors in a number of normal and carcinoma cell lines [35–39]. The NF-IL-6 site is also important for COX-2 transcription in a number of colorectal cancer cell lines [26]. Phosphorylation of C/EBP protein through a Ras-dependent MAPK cascade is essential for C/EBP activation and increases C/EBP binding to the COX-2 promoter [40, 41], suggesting that constitutive COX-2 transcription detected in colon cancer cells is influenced through enhanced C/EBP phosphorylation as a result of the presence of oncogenic Ras and increased MAPK activity [26, 42, 43]. Thus, the C/EBP proteins may regulate COX-2 gene expression in cancer cells through direct interaction with the NF-IL-6-binding site and also modulate the binding of other factors known to participate in COX-2 transcriptional activation [32, 36, 39].
Regulation of COX-2 Expression in Cancer
55
NF-kB: A growing body of evidence indicates that NF-B activation may play a important role in promoting COX-2 transcription in cancer associated with inflammation [44]. When exposed to inflammatory agonists, NF-B has been shown to be a positive regulator of COX-2 transcription in human colon cancer cells [45]. These findings suggest a role for NF-B-mediated activation of COX-2 expression in inflammatory bowel disease [46] and are of significance since chronic inflammation is a known risk factor for several types of human cancers [1].
Post-Transcriptional Regulation of COX-2
Messenger RNA turnover is a highly regulated process that plays a central role in the regulation of mammalian gene expression. In normal cells, the expression of growth-related gene products encoding cytokines, growth factors, and proto-oncogenes are tightly regulated via rapid decay of their respective mRNA transcripts. These mRNAs are inherently unstable due to the presence of a common cis-acting element known as the AU-rich element (ARE) which targets the mRNA for rapid decay [47, 48]. This element, most often present within the 3⬘ untranslated region (3⬘UTR) of the mRNA, can also regulate protein translation by acting as a translation inhibitory element [49–51]. The ability of the ARE to regulate gene expression can be modulated by endogenous and extracellular stimuli allowing for rapid changes in the abundance of these mRNAs when needed [47]. The COX-2 mRNA Contains an AU-Rich RNA Element That Targets It for Rapid Decay and Translational Inhibition The first evidence suggesting that COX-2 might be regulated at a posttranscriptional level was the identification of multiple copies of the AUUUA sequence motif within its 3⬘UTR [6, 7]. In the human COX-2 gene, the 3⬘UTRcontaining exon 10 contains 22 copies of the AUUUA sequence [8] and can be processed to yield COX-2 mRNA transcripts of multiple lengths due to alternative polyadenylation sites [52, 53]. Of particular significance was a cluster of 6 AUUUA sequence elements localized in the proximal part of the 3⬘UTR (fig. 2). The context of these AUUUA motifs within the 3⬘UTR strongly suggests the involvement of this region in regulating mRNA stability. This AU-rich region is highly conserved in both sequence and location among human, mouse, rat, chicken, pig, cow and sheep COX-2 mRNA transcripts, implying that the function of the ARE had been evolutionary conserved. Earlier investigations suggested that COX-2 expression could be regulated through its 3⬘UTR [52, 54]. More recent experiments have substantiated these findings by directly
Dixon
56
COX-2
5´-uagAAGUCUAAUGAUCAUAUUUAUUUAUUUAUAUGAACCA UGUCUAUUAAUUUAAUUAUUUAAUAAUAUUUAUAUUAAAC UCCUUAUGUUACUUAACAUCUUCUGUAACAGAAGUCAGU-3´
Fig. 2. The human COX-2 AU-rich element. The representation of the human COX-2 mRNA is not to scale. The filled circles represent AU-rich sequences, AUUUA, contained within the 3⬘UTR (open bar); circles adjacent to one another indicate multiple repeat elements. The 116 nucleotide sequence of the COX-2 ARE in uppercase letters contains six AU-rich sequence motifs (AUUUA). The COX-2 termination codon is shown in lowercase letters. Solid triangles represent canonical AAUAAA polyadenylation sites; open triangles indicate non-canonical AUUAAA sites.
showing that the COX-2 3⬘UTR can confer post-transcriptional regulation through rapid mRNA turnover and translational inhibition [55–60]. Posttranscriptional regulation was shown to be dependent on the COX-2 3⬘UTR or the conserved AU-rich region since their presence conferred rapid decay on a normally stable chimera reporter mRNA [55–57]. It should be noted that other regions within the COX-2 3⬘UTR have also been implicated to play a role in the regulation of COX-2 expression [60]. A number of observations suggest that defects in the ability of AREs to regulate gene expression on a post-transcriptional level play a role in neoplastic transformation of cells [61–63]. When AU-rich elements are removed from the proto-oncogenes c-fos and c-myc, there is a correlation with increased oncogenicity [64, 65] and cells show enhanced tumorigenicity when expressing IL-3 lacking the normal ARE-containing 3⬘UTR [66]. Additionally, a variety of human tumor cells show enhanced mRNA stability of ARE-containing cytokine genes [67] and a reporter gene containing the 3⬘UTR of GM-CSF is stable in monocytic tumor cells [68]. Similar findings have also been observed in human colon carcinoma cells with regard to regulation of COX-2 expression [26, 69]. Interestingly, colon cancer cells maintained the ability to rapidly degrade c-myc mRNA, suggesting that defects in rapid mRNA decay are specific to the transcript [69]. The c-myc mRNA contains a class I ARE, which is characterized by dispersed AUUUA motifs in association with stretches of U-rich regions. In contrast, COX-2 mRNA contains a class II ARE, which has multiple copies of AUUUA motifs clustered together [47]. The class II type ARE is common feature of many growth factor mRNAs associated with angiogenesis such as VEGF and IL-8. These and other ARE-containing growth factors have been shown to be upregulated in cancer cells that have altered post-transcriptional regulation of COX-2 [26, 69].
Regulation of COX-2 Expression in Cancer
57
COX-2 Expression Is Regulated by AU-Rich RNA Element-Binding Proteins AREs appear to mediate their regulatory function through association of multiple trans-acting RNA-binding proteins that have an affinity for the AUUUA sequence motif [47, 48]. The ability of AREs to regulate mRNA stability occurs through recruitment of a complex of RNA-degrading exonucleases termed the exosome [70, 71]. It appears that ARE-binding proteins associated with rapid decay provide a functional link between the exosome and ARE-containing mRNAs; this occurs either through its direct association within the exosome [71] or targeting of the exosome to mRNA through recruitment [70]. The COX-2 AU-rich element is also a target of cellular ARE-binding factors. Cytoplasmic proteins have been shown to form stable complexes with the COX-2 3⬘UTR [54, 56, 57]; the AREs from COX-2 and GM-CSF are recognized by similar RNA/protein complexes [56]. This complex, when crosslinked to the COX-2 ARE, is comprised of several distinct proteins with sizes ranging from 90 to 35 kDa [56]. It appears that these factors may play distinct roles in regulating both COX-2 mRNA stability and translation through ARE binding, since the proteins of lower mass appear to be preferentially polysome associated [56]. A number of ARE-binding factors have been identified [47, 62] and the particular factor bound to the RNA can promote rapid mRNA decay, increase mRNA stability, or regulate translational efficiency. Thus, the findings demonstrating enhanced stability of COX-2 mRNA and reporter constructs containing the COX-2 3⬘UTR in both colon tumor cells [26, 69] and transformed intestinal epithelial cells [42, 43, 58, 59], suggested that the observed post-transcriptional defects were a result of altered ARE recognition by trans-acting regulatory factors. Significant differences in ARE binding have been observed in the HT29 colon cancer cell line which explains the observed lack of COX-2 posttranscriptional regulation in this cell system [69]. These defects in ARE binding have also been found in other human colon cancer cells displaying increased COX-2 expression [unpublished observations]; this suggests that loss of proper ARE-recognition is a common cellular defect occurring during colon tumorigenesis. Currently, 4 factors have been identified to bind the COX-2 ARE-containing 3⬘UTR and influence COX-2 expression. These factors, their role in regulating COX-2 expression, and the potential importance in pathogenic states are discussed below: HuR (HuA): The HuR protein is a ubiquitously expressed member of the ELAV (Embryonic-Lethal Abnormal Vision in Drosophila) family of RNA-binding proteins [72]. HuR protein has a high affinity and specificity for AREs and overexpression of HuR stabilizes ARE-containing mRNA transcripts
Dixon
58
and promotes their translation [73–75]. HuR is also a nuclear-cytoplasmic shuttling protein [76] and it is thought that the ability of HuR to promote mRNA stabilization requires its translocation to the cytoplasm [73, 74, 76]. Based on its ability to bind the COX-2 ARE, HuR has been identified as a trans-acting factor involved in regulating COX-2 expression [69, 77]. The enhanced stabilization of COX-2 mRNA observed in colon cancer cells is, in part, due to elevated levels of HuR, resulting in enhanced binding to the COX-2 ARE [69]. Increased levels of cytoplasmic HuR protein were detected in neoplastic epithelial cells of colon tumors [69], which may reflect its ability to promote epithelial COX-2 expression during the later stages of carcinogenesis. Accordingly, the ectopic overexpression of HuR promoted the expression of endogenous COX-2 and the ARE-containing angiogenic factors VEGF and IL-8 [69]. Through its enhanced ARE binding, HuR could indirectly inhibit COX-2 mRNA deadenylation by promoting binding of the poly(A)-binding protein to the poly(A) tail [47]. Alternatively, HuR binding may impede the formation or recruitment of the exosome to the COX-2 transcript [70, 71]. Based on the ability of HuR to shuttle between the nucleus and cytoplasm, potential alterations in the HuR-accompanied transport of COX-2 mRNA from the nucleus to the cytoplasm in tumor cells may have direct implications on COX-2 mRNA turnover and association with the translational apparatus [75]. Taken together, these findings define a role for aberrant HuR expression in colorectal neoplasia by promoting the unregulated expression of COX-2 and associated angiogenic factors similar to other pathogenic states [77–80]. TTP (Tristetraprolin): TTP is a member of a small family of zinc finger-containing proteins of the CCCH class [81]. Originally identified as an immediate-early response gene whose expression was transiently induced by extracellular stimuli, TTP was proposed to be a transcription factor. However, growth factor or mitogen treatment of cells resulted in the enhanced cytoplasmic localization of TTP, presumably occurring through increased nucleocytoplasmic shuttling of the protein [81, 82], suggesting an alternative role for TTP. More recently, it has been demonstrated that TTP can promote the rapid decay of AREcontaining mRNAs by directly binding to the ARE [82, 83]. The binding of TTP to AREs from inflammatory mediators, such as TNF-␣ and GM-CSF, targets them for rapid deadenylation and decay through exosome recruitment [70, 84]. TTP is also rapidly phosphorylated through activation of the ERK and p38 MAPK pathways by extracellular stimuli [85, 86]. It is hypothesized that the post-translational modification of TTP inhibits ARE binding, thus allowing for mRNA stabilization [85, 86]. Recent investigations in a human colon carcinoma cell line (HCA-7) have postulated that TTP is involved in the regulation of COX-2 mRNA
Regulation of COX-2 Expression in Cancer
59
turnover [87, 88]. Binding of TTP to the COX-2 3⬘UTR has been demonstrated in conjunction with an inverse relationship between TTP mRNA and COX-2 mRNA with changes in cellular growth status. Clearly, determining the role of TTP in regulating COX-2 expression in cancer will be of great interest, since TTP deficiency in mice results in syndromes of inflammatory arthritis and bowel disease owing to defective turnover of inflammatory mediators [82, 84]. AUF1(hnRNP D): The AUF1 protein was identified as part of a complex that accelerates the rapid decay of the ARE-containing c-myc proto-oncogene mRNA [89]. Through its ability to bind AREs with high affinity, AUF1 expression has been correlated with rapid mRNA decay [90]. The AUF1 gene yields 4 protein isoforms as a result of alternative splicing and differences exist among their ability to promote rapid decay [91]. Interestingly, AUF1 can enhance mRNA stabilization under cellular responses invoked from stress and differentiation signals, suggesting that specific signals cause AUF1 to be sequestered in an inactive cytoplasmic or nuclear complex inhibiting the cytoplasmic mRNA decay function of AUF1 [92]. These types of altered cellular signaling are characteristic of cancer and thus may promote the increased cytoplasmic expression of AUF1, as detected in lung cancers [79]. Notably, the overexpression of AUF1 has been shown to promote sarcoma development in transgenic mice [93]. Presumably this increase in AUF1 expression could occur during tumor progression thereby leading to the stabilization of ARE-containing mRNAs, including COX-2, since the AUF1 protein has been demonstrated to bind the COX-2 ARE [57]. TIA-1/TIAR: TIA-1 protein was identified as an apoptosis-promoting factor in thymocytes and is 80% identical to that of the TIA-1-related protein, TIAR. Both factors are RNA-binding proteins and have been demonstrated to bind regions of RNA containing short stretches of uridylates [94]. Under normal conditions, both TIA-1 and TIAR are primarily localized in the nucleus [95]. Yet, both proteins are translocated to the cytoplasm in response to cellular stress signals where they co-localize with untranslated mRNAs in discrete cytoplasmic foci, implicating a role in translational regulation [95]. Consistent with this observation, both factors have been shown to specifically bind the ARE of TNF-␣ and can regulate its expression through translational inhibition [50, 51]. Deletion of TIA-1 in mice leads to elevated levels of TNF-␣ and increased sensitivity to LPS [51]. TIA-1 and TIAR interact with the COX-2 ARE [69], although the precise role of these factors in COX-2 regulation is unknown. Presumably, binding of these factors acts to repress COX-2 translation as part of a repressor complex, similar to TNF-␣, since they do not appear to influence rapid COX-2 mRNA decay [unpublished observation].
Dixon
60
Oncogenic Ras
Wnt APC
GSK
Translation
Rho
Rac1
Raf1
-Catenin
MEKK1
MEK
p38 MAPK
PDK
PEA3
JNK
ERK
MAPKAP2
Akt/PKB
Transcription
MKK6
PI3-K
mRNA stability
Fig. 3. Signaling pathways regulating COX-2 expression. Schematic diagram outlining the major signal transduction pathways participating in regulation of COX-2 gene expression. Relevant signaling molecules are discussed in the text and their role in COX-2 regulation is indicated.
Regulation of COX-2 Expression by Signal Transduction Pathways
Multiple Signaling Pathways Are Involved in Regulating COX-2 Expression The regulation of COX-2 expression is a complex process involving multiple signal transduction pathways. This level of complexity underscores the requirement for tight regulation of the enzymatic action of COX-2, which has pathogenic effects if its expression is left unchecked. As a result of genetic mutations commonly detected in a majority of human tumors, cellular defects in these signaling pathways promote the expression of COX-2 through the loss of transcriptional regulation, rapid mRNA decay, and translational inhibition (fig. 3). These primary signaling pathways and their involvement in regulating COX-2 expression in human cancer are discussed below. Wnt/APC: Wnt1 is a mammary oncogene encoding a secreted signaling factor and expression of Wnt1 promotes mammary carcinoma formation [96]. A downstream modulator of Wnt signaling is the tumor suppressor gene, APC. Mutations in the APC gene occur frequently in both familial and sporadic colorectal cancers [97] and these defects in APC predispose mice to intestinal and mammary tumors [13, 98]. A consequence of Wnt misexpression or APC mutations results in cytoplasmic stabilization and nuclear accumulation of the transcriptional activator -catenin [96]. COX-2 appears to be a common downstream target of Wnt expression and APC mutations. In Wnt-1 transformed murine mammary epithelial cells, elevated COX-2 levels are detected as a result of transcriptional activation [18].
Regulation of COX-2 Expression in Cancer
61
An association between increased nuclear -catenin levels and COX-2 expression is seen in human and murine colon cancer cells and is related to defects in APC [30, 31], suggesting a possible role for -catenin in stimulating COX-2 transcription [32]. Interestingly, the role of APC in COX-2 regulation appears to be multifaceted. In colon cancer cells containing mutations in APC, the expression of full-length APC downregulates COX-2 protein expression, while mRNA levels are unchanged [99], suggesting a link between APC and the regulation of COX-2 translation. Ras Signaling: Ras proteins are membrane-bound GDP/GTP-regulated switch molecules that convert signals from the cell membrane to the nucleus. Ras mutations are found to occur in a wide variety of human malignancies. K-ras mutations are found in colon, lung and pancreatic cancers, whereas H-ras is frequently mutated in cancers of the urinary tract and bladder [100]. Approximately 50% of large colorectal adenomas contain mutant Ras, so its activation is believed to be a relatively early event in neoplastic progression [22]. Oncogenic mutations in the Ras gene lead to constitutive activation of this small GTPase resulting in activation of multiple downstream signal transduction MAPK cascades (fig. 3). With regard to regulating COX-2 expression, both oncogenic K-ras and H-ras stimulate COX-2 transcription through activation of the Rho, JNK, ERK, and p38 MAPK pathways, whereas COX-2 mRNA stability is modulated through activation of signaling pathways involving ERK, p38 MAPK, and Akt/PKB [25–27, 42, 43, 101]. Oncogenic Ras may also modulate the translation of COX-2 since it has been shown to promote phosphorylation of the translation initiation factor eIF-4e and promote enhanced protein synthesis [102]. Ras signaling has also been shown to participate with growth factors to enhance the expression of COX-2. There is mounting evidence that TGF- can enhance malignant transformation and tumor progression for several different epithelial tumors [103] and TGF- is abnormally expressed in over 90% of human colon cancers [104]. Accordingly, treatment of intestinal epithelial cells with TGF- can induce or augment COX-2 expression [105–108], and recent work has demonstrated that TGF- synergistically enhances the expression of COX-2 in Ras-transformed intestinal epithelial cells through mechanisms involving COX-2 mRNA stabilization [58]. Rho Pathway: The Rho family of proteins are key components of oncogenic cellular transformation of cells by Ras [109]. Rho-dependent activation of COX-2 transcription has been demonstrated in a variety of cell types, including colon and breast cancer cells, through cis-acting elements present in the proximal region of the COX-2 promoter (fig. 1) [26, 110, 111]. JNK Pathway: It is well documented that activation of the Ras/Rac1/ MEKK1/JNK signal transduction pathway activates the transcription of COX-2.
Dixon
62
Ras activation of the JNK kinase cascade leads to phosphorylation of c-Jun, resulting in transcriptional activation of COX-2 via the CRE element [23, 24, 36]. Increased COX-2 promoter activity, owing in part to JNK activation, has been observed in cancer cells [26, 27]. Tumor-promoting agents such as bile acids and ceramide also activate the JNK pathway leading to increased COX-2 transcription [25, 112], whereas COX-2 mRNA stability does not appear to be regulated through this pathway [59]. ERK Pathway: Activation of the Ras/Raf1/MEK/ERK signal transduction pathway was originally shown to modulate COX-2 transcriptional regulation in fibroblasts responding to v-src, PDGF, or serum [23, 24]. This MAPK pathway is equally important in constitutive activation of COX-2 expression in cancer. ERK activation parallels tumor progression in murine intestinal neoplasia and ERK inhibition suppresses colonic tumor growth [113]. In colorectal cancer cell lines displaying elevated levels of COX-2, constitutive activation of the ERK pathway has been observed [69] and inhibition of this pathway reduced both the COX-2 promoter activity and stabilization of its mRNA [26]. Similarly, in Ras-transformed intestinal epithelial cells, the ERK pathway is essential for Ras-mediated induction of COX-2 promoter and mRNA stabilization [42, 43, 59]. Furthermore, in HER2/neu-transformed mammary epithelial cells, HER2/neu induced COX-2 promoter activity via ERK activation [27]. p38 MAPK Pathway: Activation of the p38 MAPK pathway is primarily associated with cellular stress and pro-inflammatory stimuli. A number of studies have demonstrated the ability of proinflammatory signals to induce COX-2 expression in a variety of cell types through p38 MAPK activation [101, 114, 115]. Considerable evidence suggests that activation p38 MAPK promotes stabilization of COX-2 and other ARE-containing mRNAs through activation of MAPKAP-2, a kinase downstream of p38 MAPK [57, 114]. Interestingly, the ability of the anti-inflammatory glucocorticoid dexamethasone to inhibit COX-2 expression [52] has recently been shown to promote rapid destabilization of COX-2 mRNA by inhibition of the p38 MAPK pathway [116]. The ability of p38 MAPK to promote COX-2 mRNA stabilization has been detected in intestinal and breast carcinoma cells [59, 117], this may be relevant because activated p38 MAPK has been observed in neoplastic tissue [118]. p38 MAPK has also been shown to promote COX-2 transcription. In transformed mammary epithelial and cervical carcinoma cells, induction of COX-2 promoter activity was attributed, in part, to p38 MAPK activation [25, 27, 111], presumably through its ability to promote AP-1 activity via phosphorylation of ATF-2. Akt/PKB: The kinase Akt, or protein kinase B (Akt/PKB), is a downstream effector of phosphophatidylinositol 3-kinase (PI3-K) [119]. Sequential activation of the PI3K/PDK/Akt/PKB kinase cascade by various growth and
Regulation of COX-2 Expression in Cancer
63
survival factors promotes cell survival through inactivation of several apoptosisassociated factors. Activation of this pathway is required for Ras-mediated transformation of a variety of cell types, including intestinal epithelial cells, and Akt/PKB activation facilitates a number of cellular events associated with cellular transformation (i.e. cytoskeletal reorganization) [120, 121]. Constitutive activation of Akt/PKB has been observed in colon cancer cell lines [69] and the Akt/PKB pathway predominantly regulates COX-2 expression by modulating the stability of COX-2 mRNA [26]. In intestinal epithelial cells, oncogenic Rasmediated post-transcriptional stabilization of COX-2 mRNA occurs through activation of the Akt/PKB pathway [43]. These findings indicate that COX-2 may represent a downstream mediator of this oncogenic pathway.
Conclusions
Although the molecular events leading to overexpression of COX-2 protein in human cancer have not been definitively characterized, substantial progress has been made toward understanding the mechanisms regulating COX-2 expression. Under normal cellular growth conditions, the expression of COX-2 is tightly controlled with rapid induction of COX-2 transcription occurring through a variety of stimuli. In colorectal neoplasia, a loss of transcriptional regulation causes the increased levels of COX-2 mRNA detected in a majority of colorectal adenomas, adenocarcinomas and colon cancer cell lines [9, 12, 26, 69, 122]. Notably, concomitant increase in the amounts of COX-2 protein and prostaglandins has not been observed in all adenomas or adenocarcinomas [10, 123, 124]. Moreover, increased levels of COX-2 protein were not detected all colon cancer cell lines that overexpress COX-2 mRNA [26, 69]. To reconcile the observed apparent discrepancies between dysregulated expression of COX-2 message and protein, a provisional model is presented (fig. 4). Colon cancer is used as an example, but this model could be applied to other human cancers as well. In this model, transcriptional activation of COX-2 is an early event in the initiation of a colon tumor [12, 26, 69]. However, enhanced expression of COX-2 protein also requires aberrant post-transcriptional regulation. This occurs during the transition from adenoma to carcinoma and suggests that loss of post-transcriptional regulation of COX-2 may be a crucial step in colon carcinogenesis. This notion is supported by the increase in COX-2 mRNA stability that is readily observed in colon cancer cells that constitutively express COX-2 [26, 69]. Furthermore, the observation of a correlation between increased levels of COX-2 and increased tumor size and invasiveness [13, 125, 126] suggests a link between tumor progression and defects in regulation of COX-2 gene transcription and subsequent mRNA decay. Taken together, these
Dixon
64
APC mutation
RAS mutation
TGF IIR MMR
p53 mutation
Normal epithelium
Invasive carcinoma Early adenoma
⫺ COX-2
⫹ COX-2 Loss of transcriptional regulation
Late adenoma ⫹⫹ COX-2
⫹⫹⫹ COX-2
Loss of post-transcriptional control
Fig. 4. Molecular regulation of COX-2 expression in colorectal cancer. Multiple genetic mutations have been identified to promote colorectal tumorigenesis (TGF IIR: TGF type II receptor; MMR: DNA mismatch repair) [16]. As a result of these genetic changes, alterations in cellular signaling promote overexpression of COX-2 (and associated prostaglandin synthesis) through the loss of COX-2 transcriptional regulation and posttranscriptional control of rapid mRNA decay and translational inhibition.
control defects result in unregulated expression of COX-2 protein and presumably other immediate-early response genes that are detected in the later stages of adenoma development.
Acknowledgements I gratefully thank D. Beauchamp, A. Dannenberg, N. Deene, R. DuBois, R. Gupta, J. Oates, S. Prescott, D. Simmons, and members of the Surgical Oncology Research Laboratory for helpful discussions on the manuscript. This work was supported by research grants from the American Heart Association (9930102N) and National Institute of Health (CA77839-04).
References 1 2 3
4
5
Prescott SM, Fitzpatrick FA: Cyclooxygenase-2 and carcinogenesis. Biochim Biophys Acta 2000;1470:M69–M78. Gupta RA, DuBois RN: Colorectal cancer prevention and treatment by inhibition of cyclooxygenase-2. Nat Rev Cancer 2001;1:11–21. Raz A, Wyche A, Needleman P: Temporal and pharmacological division of fibroblast cyclogenase expression into transcriptional and translation phases. Proc Natl Acad Sci USA 1989;86: 1657–1661. Xie W, Chipman JG, Robertson DL, Erikson RL, Simmons DL: Expression of a mitogenresponsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Natl Acad Sci USA 1991;88:2692–2696. Kajubu DA, Fletcher BS, Varnum BC, Lim RW, Herschman HR: TIS10, a phorbol ester tumor promoter-inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/ cyclooxygenase homologue. J Biol Chem 1991;226:12866–12872.
Regulation of COX-2 Expression in Cancer
65
6 7
8 9
10 11
12
13
14
15
16 17 18
19 20 21
22
23
24 25
26 27
Hla T, Neilson K: Human cyclooxygenase-2 cDNA. Proc Natl Acad Sci USA 1992;89:7384–7388. Jones DJ, Carlton DP, McIntyre TM, Zimmerman GA, Prescott SM: Molecular cloning of human prostaglandin endoperoxide synthase type II and demonstration of expression in response to cytokines. J Biol Chem 1993;268:9049–9054. Appleby SB, Ristimaki A, Neilson K, Narko K, Hla T: Structure of the human cyclo-oxygenase-2 gene. Biochem J 1994;302:723–727. Eberhart CE, Coffey RJ, Radhika A, Giardiello FM, Ferrenbach S, DuBois RN: Up-regulation of cycloxygenase-2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 1994;107:1183–1188. Kargman S, O’Neill G, Vickers P, Evans J, Mancini J, Jothy S: Expression of prostaglandin G/H synthase-1 and -2 protein in human colon cancer. Cancer Res 1995;55:2556–2559. Sano H, Kawahito Y, Wilder RL, Hashiramoto A, Mukai S, Asai K, Kumura S, Kato H, Kondo M, Hla T: Expression of cyclooxygenase-1 and -2 in human colorectal cancer. Cancer Res 1995;55: 3785–3789. Kutchera W, Jones DA, Matsunami N, Groden J, McIntyre TM, Zimmerman GA, White RL, Prescott SM: Prostaglandin synthase 2 is abnormally expressed in human colon cancer: Evidence for a transcriptional effect. Proc Natl Acad Sci USA 1996;93:4816–4820. Oshima M, Dinchuk JE, Kargman SL, Oshima H, Hancock B, Kwong E, Trzaskos JM, Evans JF, Taketo MM: Suppression of intestinal polyposis in APC⌬716 knockout mice by inhibition of cyclooxygenase-2. Cell 1996;87:803–809. Liu CH, Chang SH, Narko K, Trifan OC, Wu MT, Smith E, Haudenschild C, Lane TF, Hla T: Overexpression of cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem 2001;276:18563–18569. Neufang G, Furstenberger G, Heidt M, Marks F, Muller-Decker K: Abnormal differentiation of epidermis in transgenic mice constitutively expressing cyclooxygenase-2 in skin. Proc Natl Acad Sci USA 2001;98:7629–7634. Kinzler KW, Vogelstein B: Lessons from hereditary colorectal cancer. Cell 1996;87:159–170. Prescott SM, White RL: Self-promotion? Intimate connections between APC and prostaglandin H synthase-2. Cell 1996;87:783–786. Howe LR, Subbaramaiah K, Chung WJ, Dannenberg AJ, Brown AM: Transcriptional activation of cyclooxygenase-2 in Wnt-1-transformed mouse mammary epithelial cells. Cancer Res 1999;59: 1572–1577. Mack DH, Vartikar J, Pipas JM, Laimins LA: Specific repression of TATA-mediated but not initiator-mediated transcription by wild-type p53. Nature 1993;363:281–283. Subbaramaiah K, Altorki N, Chung WJ, Mestre JR, Sampat A, Dannenberg AJ: Inhibition of cyclooxygenase-2 gene expression by p53. J Biol Chem 1999;274:10911–10915. Leung WK, To KF, Ng YP, Lee TL, Lau JY, Chan FK, Ng EK, Chung SC, Sung JJ: Association between cyclo-oxygenase-2 overexpression and missense p53 mutations in gastric cancer. Br J Cancer 2001;84:335–339. Fearon E, Gruber S: Molecular abnormalities in colon and rectal cancer; in Mendelsohn J, Howley P, Israel M, Liotta L (eds): The Molecular Basis of Cancer. Philadelphia, Saunders, 2001, pp 289–312. Xie W, Herschman HR: v-src induces prostaglandin synthase 2 gene expression by activation of the c-Jun N-terminal kinase and the c-Jun transcription factor. J Biol Chem 1995;270: 27622–27628. Xie W, Herschman HR: Transcriptional regulation of prostaglandin synthase 2 gene expression by platelet-derived growth factor and serum. J Biol Chem 1996;271:31742–31748. Subbaramaiah K, Chung WJ, Dannenberg AJ: Ceramide regulates the transcription of cyclooxygenase-2. Evidence for involvement of extracellular signal-regulated kinase/c-Jun N-terminal kinase and p38 mitogen-activated protein kinase pathways. J Biol Chem 1998;273:32943–32949. Shao J, Sheng H, Inoue H, Morrow JD, DuBois RN: Regulation of constitutive cyclooxygenase-2 expression in colon carcinoma cells. J Biol Chem 2000;43:33951–33956. Subbaramaiah K, Norton L, Gerald W, Dannenberg AJ: Cyclooxygenase-2 is overexpressed in HER-2/neu-positive breast cancer. Evidence for involvement of AP-1 and PEA3. J Biol Chem 2002;277:18649–18657.
Dixon
66
28
29
30
31
32
33
34 35
36 37
38
39
40
41
42
43 44
45
46
Subbaramaiah K, Lin DT, Hart JC, Dannenberg AJ: Peroxisome proliferator-activated receptor ␥ ligands suppress the transcriptional activation of cyclooxygenase-2. Evidence for involvement of activator protein-1 and CREB-binding protein/p300. J Biol Chem 2001;276:12440–12448. Subbaramaiah K, Cole PA, Dannenberg AJ: Retinoids and carnosol suppress cyclooxygenase-2 transcription by CREB-binding protein/p300-dependent and -independent mechanisms. Cancer Res 2002;62:2522–2530. Mei JM, Hord NG, Winterstein DF, Donald SP, Phang JM: Differential expression of prostaglandin endoperoxide H synthase-2 and formation of activated -catenin-LEF-1 transcription complex in mouse colonic epithelial cells contrasting in APC. Carcinogenesis 1999;20: 737–740. Dimberg J, Hugander A, Sirsjo A, Soderkvist P: Enhanced expression of cyclooxygenase-2 and nuclear -catenin are related to mutations in the APC gene in human colorectal cancer. Anticancer Res 2001;21:911–915. Howe LR, Crawford HC, Subbaramaiah K, Hassell JA, Dannenberg AJ, Brown AM: PEA3 is up-regulated in response to Wnt1 and activates the expression of cyclooxygenase-2. J Biol Chem 2001;276:20108–20115. Crawford HC, Fingleton B, Gustavson MD, Kurpios N, Wagenaar RA, Hassell JA, Matrisian LM: The PEA3 subfamily of Ets transcription factors synergizes with -catenin-LEF-1 to activate matrilysin transcription in intestinal tumors. Mol Cell Biol 2001;21:1370–1383. Akira S, Kishimoto T: NF-IL-6 and NF-B in cytokine gene regulation. Adv Immunol 1997; 65:1–46. Kim Y, Fischer SM: Transcriptional regulation of cyclooxygenase-2 in mouse skin carcinoma cells. Regulatory role of CCAAT/enhancer-binding proteins in the differential expression of cyclooxygenase-2 in normal and neoplastic tissues. J Biol Chem 1998;273:27686–27694. Reddy ST, Wadleigh DJ, Herschman HR: Transcriptional regulation of the cyclooxygenase-2 gene in activated mast cells. J Biol Chem 2000;275:3107–3113. Thomas B, Berenbaum F, Humbert L, Bian H, Bereziat G, Crofford L, Olivier JL: Critical role of C/EBP␦ and C/EBP factors in the stimulation of the cyclooxygenase-2 gene transcription by interleukin-1 in articular chondrocytes. Eur J Biochem 2000;267:6798–6809. Wadleigh DJ, Reddy ST, Kopp E, Ghosh S, Herschman HR: Transcriptional activation of the cyclooxygenase-2 gene in endotoxin-treated RAW 264.7 macrophages. J Biol Chem 2000; 275:6259–6266. Zhu Y, Saunders MA, Yeh H, Deng WG, Wu KK: Dynamic regulation of cyclooxygenase-2 promoter activity by isoforms of CCAAT/enhancer-binding proteins. J Biol Chem 2002; 277:6923–6928. Nakajima T, Kinoshita S, Sasagawa T, Sasaki K, Naruto M, Kishimoto T, Akira S: Phosphorylation at threonine-235 by a ras-dependent mitogen-activated protein kinase cascade is essential for transcription factor NF-IL-6. Proc Natl Acad Sci USA 1993;90:2207–2211. Saunders MA, Sansores-Garcia L, Gilroy DW, Wu KK: Selective suppression of CCAAT/ enhancer-binding protein  binding and cyclooxygenase-2 promoter activity by sodium salicylate in quiescent human fibroblasts. J Biol Chem 2001;276:18897–18904. Sheng H, Williams CS, Shao J, Liang P, DuBois RN, Beauchamp RD: Induction of cyclooxygenase-2 by activated Ha-ras oncogene in Rat-1 fibroblasts and the role of mitogenactivated protein kinase pathway. J Biol Chem 1998;273:22120–22127. Sheng H, Shao J, DuBois RN: K-Ras-mediated increase in cyclooxygenase-2 mRNA stability involves activation of the protein kinase B. Cancer Res 2001;61:2670–2675. Surh YJ, Chun KS, Cha HH, Han SS, Keum YS, Park KK, Lee SS: Molecular mechanisms underlying chemopreventive activities of anti-inflammatory phytochemicals: Down-regulation of COX-2 and iNOS through suppression of NF-B activation. Mutat Res 2001;480–481:243–268. Kojima M, Morisaki T, Izuhara K, Uchiyama A, Matsunari Y, Katano M, Tanaka M: Lipopolysaccharide increases cyclo-oxygenase-2 expression in a colon carcinoma cell line through nuclear factor-B activation. Oncogene 2000;19:1225–1231. Singer II, Kawka DW, Schloemann S, Tessner T, Riehl T, Stenson WF: Cyclooxygenase-2 is induced in colonic epithelial cells in inflammatory bowel disease. Gastroenterology 1998;115: 297–306.
Regulation of COX-2 Expression in Cancer
67
47 48 49 50
51
52
53
54 55
56
57
58
59 60
61 62 63 64 65
66
67 68
Wilusz CJ, Wormington M, Peltz SW: The cap-to-tail guide to mRNA turnover. Nat Rev Mol Cell Biol 2001;2:237–246. Van Hoof A, Parker R: Messenger RNA degradation: Beginning at the end. Curr Biol 2002; 12:R285–R287. Kruys V, Marinx O, Shaw G, Deschamps J, Huez G: Translational blockade imposed by cytokinederived UA-rich sequences. Science 1989;245:852–855. Gueydan C, Droogmans L, Chalon P, Caput D, Kruys V: Identification of TIAR as a protein binding to the translational regulatory AU-rich element of tumor necrosis factor ␣ mRNA. J Biol Chem 1999;274:2322–2326. Piecyk M, Wax S, Beck AR, Kedersha N, Gupta M, Maritim B, Chen S, Gueydan C, Kruys V, Streuli M, Anderson P: TIA-1 is a translational silencer that selectively regulates the expression of TNF-␣. EMBO J 2000;19:4154–4163. Ristimaki A, Narko K, Hla T: Down-regulation of cytokine-induced cyclo-oxygenase-2 transcript isoforms by dexamethasone: Evidence for post-transcriptional regulation. Biochem J 1996;318: 325–331. Newton R, Seybold J, Liu SF, Barnes PJ: Alternate COX-2 transcripts are differentially regulated: Implications for post-transcriptional control. Biochem Biophys Res Commun 1997; 234:85–89. Srivastava SK, Tetsuka T, Daphna-Iken D, Morrison AR: IL-1 stabilizes COX II mRNA in renal mesangial cells: Role of the 3⬘-untranslated region. Am Physiol Soc 1994;94:504–508. Gou Q, Liu CH, Ben-Av P, Hla T: Dissociation of basal turnover and cytokine-induced transcript stabilization of the human cyclooxygenase-2 mRNA by mutagenesis of the 3⬘-untranslated region. Biochem Biophys Res Commun 1998;242:508–512. Dixon DA, Kaplan CD, McIntyre TM, Zimmerman GA, Prescott SM: Post-transcriptional control of cyclooxygenase-2 gene expression. The role of the 3⬘-untranslated region. J Biol Chem 2000; 275:11750–11757. Lasa M, Mahtani KR, Finch A, Brewer G, Saklatvala J, Clark AR: Regulation of cyclooxygenase-2 mRNA stability by the mitogen-activated protein kinase p38 signaling cascade. Mol Cell Biol 2000;20:4265–4274. Sheng H, Shao J, Dixon DA, Williams CS, Prescott SM, DuBois RN, Beauchamp RD: Transforming growth factor-1 enhances Ha-ras-induced expression of cyclooxygenase-2 in intestinal epithelial cells via stabilization of mRNA. J Biol Chem 2000;275:6628–6635. Zhang Z, Sheng H, Shao J, Beauchamp RD, DuBois RN: Posttranscriptional regulation of cyclooxygenase-2 in rat intestinal epithelial cells. Neoplasia 2000;2:523–530. Cok SJ, Morrison AR: The 3⬘-untranslated region of murine cyclooxygenase-2 contains multiple regulatory elements that alter message stability and translational efficiency. J Biol Chem 2001; 276:23179–23185. Conne B, Stutz A, Vassalli JD: The 3⬘-untranslated region of messenger RNA: A molecular ‘hotspot’ for pathology? Nat Med 2000;6:637–641. Brewer G: Misregulated posttranscriptional checkpoints: Inflammation and tumorigenesis. J Exp Med 2001;193:F1–F4. Mendell JT, Dietz HC: When the message goes awry: Disease-producing mutations that influence mRNA content and performance. Cell 2001;107:411–414. Raymond V, Atwater JA, Verma IM: Removal of an mRNA destabilizing element correlates with the increased oncogenicity of proto-oncogene fos. Oncogene Res 1989;5:1–12. Aghib DF, Bishop JM, Ottolenghi S, Guerrasio A, Serra A, Saglio G: A 3⬘ truncation of MYC caused by chromosomal translocation in a human T-cell leukemia increases mRNA stability. Oncogene 1990;5:707–711. Wang XY, McCubrey JA: Malignant transformation induced by cytokine genes: A comparison of the abilities of germline and mutated interleukin-3 genes to transform hematopoietic cells by transcriptional and posttranscriptional mechanisms. Cell Growth Differ 1996;7:487–500. Ross HJ, Santo N, Ueyama Y, Koeffler HP: Cytokine messenger RNA stability is enhanced in tumor cells. Blood 1991;77:1787–1795. Schuler GD, Cole MD: GM-CSF and oncogene mRNA stabilities are independently regulated in trans in a mouse monocytic tumor. Cell 1988;55:1115–1122.
Dixon
68
69
70
71
72 73 74 75 76 77
78 79
80
81
82 83
84
85
86
87
88
89
Dixon DA, Tolley ND, King PH, Nabors LB, McIntyre TM, Zimmerman GA, Prescott SM: Altered expression of the mRNA stability factor HuR promotes cyclooxygenase-2 expression in colon cancer cells. J Clin Invest 2001;108:1657–1665. Chen CY, Gherzi R, Ong SE, Chan EL, Raijmakers R, Pruijn GJ, Stoecklin G, Moroni C, Mann M, Karin M: AU binding proteins recruit the exosome to degrade ARE-containing mRNAs. Cell 2001;107:451–464. Mukherjee D, Gao M, O’Connor JP, Raijmakers R, Pruijn G, Lutz CS, Wilusz J: The mammalian exosome mediates the efficient degradation of mRNAs that contain AU-rich elements. EMBO J 2002;21:165–174. Brennan CM, Steitz JA: HuR and mRNA stability. Cell Mol Life Sci 2001;58:266–277. Fan XC, Steitz JA: Overexpression of HuR, a nuclear-cytoplasmic shuttling protein, increases the in vivo stability of ARE-containing mRNAs. EMBO J 1998;17:3448–3460. Peng SSY, Chen CYA, Xu N, Shyu AB: RNA stabilization by the AU-rich element binding protein, HuR, an ELAV protein. EMBO J 1998;17:3461–3470. Keene J: Why is Hu where? Shuttling of early-response messenger RNA subsets. Proc Natl Acad Sci USA 1999;96:5–7. Fan XC, Steitz JA: HNS, a nuclear-cytoplasmic shuttling sequence in HuR. Proc Natl Acad Sci USA. 1998;95:15293–15298. Nabors LB, Gillespie GY, Harkins L, King PH: HuR, an RNA stability factor, is expressed in malignant brain tumors and binds to adenine- and uridine-rich elements within the 3⬘-untranslated regions of cytokine and angiogenic factor mRNAs. Cancer Res 2001;61:2154–2161. Levy NS, Chung S, Furneaux H, Levy AP: Hypoxic stabilization of vascular endothelial growth factor mRNA by the RNA-binding protein HuR. J Biol Chem 1998;273:6417–6423. Blaxall BC, Dwyer-Nield LD, Bauer AK, Bohlmeyer TJ, Malkinson AM, Port JD: Differential expression and localization of the mRNA binding proteins, AU-rich element mRNA binding protein (AUF1) and Hu antigen R (HuR), in neoplastic lung tissue. Mol Carcinog 2000;28:76–83. Wang W, Fan J, Furer-Galban S, Lopez de Silanes I, von Kobbe C, Guo J, Foufelle F, Hardie DG, Carling D, Gorospe M: AMP-activated kinase regulates cytoplasmic HuR. Mol Cell Biol 2002;22:3425–3436. Phillips RS, Ramos SB, Blackshear PJ: Members of the tristetraprolin family of tandem CCCH zinc finger proteins exhibit CRM1-dependent nucleocytoplasmic shuttling. J Biol Chem 2002; 277:11606–11613. Carballo E, Lai WS, Blackshear PJ: Feedback inhibition of macrophage tumor necrosis factor-␣ production by tristetraprolin. Science 1998;281:1001–1005. Lai WS, Carballo E, Thorn JM, Kennington EA, Blackshear PJ: Interactions of CCCH zinc finger proteins with mRNA. Binding of tristetraprolin-related zinc finger proteins to Au-rich elements and destabilization of mRNA. J Biol Chem 2000;275:17827–17837. Carballo E, Lai WS, Blackshear PJ: Evidence that tristetraprolin is a physiological regulator of granulocyte-macrophage colony-stimulating factor messenger RNA deadenylation and stability. Blood 2000;95:1891–1899. Carballo E, Cao H, Lai WS, Kennington EA, Campbell D, Blackshear PJ: Decreased sensitivity of tristetraprolin-deficient cells to p38 inhibitors suggests the involvement of tristetraprolin in the p38 signaling pathway. J Biol Chem 2001;276:42580–42587. Mahtani KR, Brook M, Dean JL, Sully G, Saklatvala J, Clark AR: Mitogen-activated protein kinase p38 controls the expression and posttranslational modification of tristetraprolin, a regulator of tumor necrosis factor ␣ mRNA stability. Mol Cell Biol 2001;21:6461–6469. Sawaoka H, Oates JA, Boutaud O: A 2577 base-pair long mRNA is the main transcript of cyclooxygenase-2 in the human colon cancer cell line, HCA-7: Eicosanoids and Other Bioactive Lipids in Cancer, Inflammation and Related Diseases, Nashville 2001, vol 7. Sawaoka H, Dixon DA, Oates JA, Boutaud O: Tristetrapolin binds to the COX-2 mRNA 3⬘-untranslated region in cancer cells. International Conference on Advances in Prostaglandins, Leukotrienes and Other Bioactive Lipid Research, Istanbul 2002, vol 12. Zhang W, Wagner BJ, Ehrenman K, Schaefer AW, DeMaria CT, Crater D, DeHaven K, Long L, Brewer G: Purification, characterization and cDNA cloning of an AU-rich element RNA-binding protein, AUF1. Mol Cell Biol 1993;13:7652–7665.
Regulation of COX-2 Expression in Cancer
69
90 DeMaria CT, Brewer G: AUF1 binding affinity to A+U-rich elements correlates with rapid mRNA degradation. J Biol Chem 1996;271:12179–12184. 91 Loflin P, Chen CY, Shyu AB: Unraveling a cytoplasmic role for hnRNP D in the in vivo mRNA destabilization directed by the AU-rich element. Genes Dev 1999;13:1884–1897. 92 Shyu AB, Wilkinson MF: The double lives of shuttling mRNA binding proteins. Cell 2000;102: 135–138. 93 Gouble A, Grazide S, Meggetto F, Mercier P, Delsol G, Morello D: A new player in oncogenesis: AUF1/hnRNPD overexpression leads to tumorigenesis in transgenic mice. Cancer Res 2002;62: 1489–1495. 94 Dember LM, Kim ND, Liu KQ, Anderson P: Individual RNA recognition motifs of TIA-1 and TIAR have different RNA binding specificities. J Biol Chem 1996;271:2783–2788. 95 Kedersha N, Gupta M, Li W, Miller I, Anderson P: RNA-binding proteins TIA-1 and TIAR link the phosphorylation of eIF-2a to the assembly of mammalian stress granules. J Cell Biol 1999; 147:1431–1441. 96 Polakis P: Wnt signaling and cancer. Genes Dev 2000;14:1837–1851. 97 Cottrell S, Bicknell D, Kaklamanis L, Bodmer WF: Molecular analysis of APC mutations in familial adenomatous polyposis and sporadic colon carcinomas. Lancet 1992;340:626–630. 98 Moser AR, Mattes EM, Dove WF, Lindstrom MJ, Haag JD, Gould MN: ApcMin, a mutation in the murine Apc gene, predisposes to mammary carcinomas and focal alveolar hyperplasias. Proc Natl Acad Sci USA 1993;90:8977–8981. 99 Hsi LC, Angerman-Stewart J, Eling TE: Introduction of full-length APC modulates cyclooxygenase-2 expression in HT-29 human colorectal carcinoma cells at the translational level. Carcinogenesis 1999;20:2045–2049. 100 Adjei AA: Ras signaling pathway proteins as therapeutic targets. Cur Pharm Design 2001;7: 1581–1594. 101 Guan Z, Buckman SY, Pentland AP, Templeton DJ, Morrison AR: Induction of cyclooxygenase-2 by the activated MEKK1 → SEK1/MKK4 → p38 mitogen-activated protein kinase pathway. J Biol Chem 1998;273:12901–12908. 102 Flynn A, Proud CG: The role of eIF4 in cell proliferation. Cancer Surv 1996;27:293–310. 103 Wakefield LM, Roberts AB: TGF- signaling: Positive and negative effects on tumorigenesis. Curr Opin Gen Dev 2002;12:22–29. 104 Avery A, Paraskeva C, Hall P, Flanders KC, Sporn M, Moorghen M: TGF- expression in the human colon: Differential immunostaining along crypt epithelium. Br J Cancer 1993;68:137–139. 105 Gilbert RS, Reddy ST, Targan S, Herschman HR: TGF-1 augments expression of the TIS10/ prostaglandin synthase-2 gene in intestinal epithelial cells. Cell Mol Biol Res 1994;40:653–660. 106 Sheng H, Shao J, Hooton EB, Tsujii M, DuBois RD, Beauchamp RD: Cyclooxygenase-2 induction and transforming growth factor- growth inhibition in rat intestinal epithelial cells. Cell Growth Differ 1997;8:463–470. 107 Shao J, Sheng H, Aramandla R, Pereira MA, Lubet RA, Hawk E, Grogan L, Kirsch IR, Washington MK, Beauchamp RD, DuBois RN: Coordinate regulation of cyclooxygenase-2 and TGF-1 in replication error-positive colon cancer and azoxymethane-induced rat colonic tumors. Carcinogenesis 1999;20:185–191. 108 Sheng H, Shao JO, Mahony CA, Lamps L, Albo D, Isakson PC, Berger DH, DuBois RN, Beauchamp RD: Transformation of intestinal epithelial cells by chronic TGF-1 treatment results in downregulation of the type II TGF- receptor and induction of cyclooxygenase-2. Oncogene 1999;18:855–867. 109 Zohn IM, Campbell SL, Khosravi-Far R, Rossman KL, Der CJ: Rho family proteins and Ras transformation: The RHOad less traveled gets congested. Oncogene 1998;17:1415–1438. 110 Slice LW, Bui L, Mak C, Walsh JH: Differential regulation of COX-2 transcription by Ras- and Rho-family of GTPases. Biochem Biophys Res Commun 2000;276:406–410. 111 Subbaramaiah K, Hart JC, Norton L, Dannenberg AJ: Microtubule-interfering agents stimulate the transcription of cyclooxygenase-2. Evidence for involvement of ERK1/2 and p38 mitogenactivated protein kinase pathways. J Biol Chem 2000;275:14838–14845. 112 Zhang F, Subbaramaiah K, Altorki N, Dannenberg AJ: Dihydroxy bile acids activate the transcription of cyclooxygenase-2. J Biol Chem 1998;273:2424–2428.
Dixon
70
113 Sebolt-Leopold JS, Dudley DT, Herrera R, Van Becelaere K, Wiland A, Gowan RC, Tecle H, Barrett SD, Bridges A, Przybranowski S, Leopold WR, Saltiel AR: Blockade of the MAP kinase pathway suppresses growth of colon tumors in vivo. Nat Med 1999;5:810–816. 114 Ridley SH, Dean JL, Sarsfield SJ, Brook M, Clark AR, Saklatvala J: A p38 MAP kinase inhibitor regulates stability of interleukin-1-induced cyclooxygenase-2 mRNA. FEBS Lett 1998; 439:75–80. 115 Dean JL, Brook M, Clark AR, Saklatvala J: p38 mitogen-activated protein kinase regulates cyclooxygenase-2 mRNA stability and transcription in lipopolysaccharide-treated human monocytes. J Biol Chem 1999;274:264–269. 116 Lasa M, Brook M, Saklatvala J, Clark AR: Dexamethasone destabilizes cyclooxygenase-2 mRNA by inhibiting mitogen-activated protein kinase p38. Mol Cell Biol 2001;21:771–780. 117 Jang BC, Sanchez T, Schaefers HJ, Trifan OC, Liu CH, Creminon C, Huang CK, Hla T: Serum withdrawal-induced post-transcriptional stabilization of cyclooxygenase-2 mRNA in MDA-MB231 mammary carcinoma cells requires the activity of the p38 stress-activated protein kinase. J Biol Chem 2000;275:39507–39515. 118 Hardwick JC, van den Brink GR, Offerhaus GJ, van Deventer SJ, Peppelenbosch MP: NF-B, p38 MAPK and JNK are highly expressed and active in the stroma of human colonic adenomatous polyps. Oncogene 2001;20:819–827. 119 Franke TF, Kaplan DR, Cantley LC: PI3K: downstream AKTion blocks apoptosis. Cell 1997;88: 435–437. 120 Rodriguez-Viciana P, Warne PH, Khwaja A, Marte BM, Pappin D, Das P, Waterfield MD, Ridley A, Downward J: Role of phosphoinositide 3-OH kinase in cell transformation and control of the actin cytoskeleton by Ras. Cell 1997;89:457–467. 121 Sheng H, Shao J, DuBois RN: Akt/PKB activity is required for Ha-Ras-mediated transformation of intestinal epithelial cells. J Biol Chem 2001;276:14498–14504. 122 Dimberg J, Samuelsson A, Hugander A, Soderkvist P: Gene expression of cyclooxygenase-2, group II and cytosolic phospholipase A2 in human colorectal cancer. Anticancer Res 1998;18: 3283–3287. 123 Rigas B, Goldman IS, Levine L: Altered eicosanoid levels in human colon cancer. J Lab Clin Med 1993;122:518–523. 124 Dimberg J, Samuelsson A, Hugander A, Soderkvist P: Differential expression of cyclooxygenase-2 in human colorectal cancer. Gut 1999;45:730–732. 125 Fujita T, Matsui M, Takaku K, Uetake H, Ichikawa W, Taketo MM, Sugihara K: Size- and invasion-dependent increase in cyclooxygenase-2 levels in human colorectal carcinomas. Cancer Res 1998;58:4823–4826. 126 Zhang H, Sun XF: Overexpression of cyclooxygenase-2 correlates with advanced stages of colorectal cancer. Am J Gastroenterol 2002;97:1037–1041.
Dan A. Dixon Vanderbilt University Medical Center, Department of Surgery, 1161 21st Ave South, D-2300 MCN, Nashville, TN 37232-2733 (USA) Tel. +1 615 3225244, Fax +1 615 3226174, E-Mail
[email protected]
Regulation of COX-2 Expression in Cancer
71
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 72–89
Cyclooxygenase-2 and Skin Carcinogenesis Gerhard Fürstenberger, Friedrich Marks, Karin Müller-Decker Department of Eicosanoids and Epithelial Carcinogenesis, Deutsches Krebsforschungszentrum, Heidelberg, Germany
Introduction
Cyclooxygenase (COX)-derived prostaglandins (PGs) exhibit manifold functions in skin inflammation and cancer. Squamous cell cancer of skin is an example of epithelial neoplasias that overexpress COX-2 in the parenchyme and the mesenchyme of premalignant and malignant lesions. Using the multistage model of mouse skin carcinogenesis it can be demonstrated that aberrant COX-2 expression and activity are causally involved in tumor promotion and progression. In addition, the genetic ablation of COX-2 and COX-1 leads to a reduced tumor burden in the skin of mice. Vice versa, transgenic overexpression of COX-2 causes an auto-promoted skin phenotype, i.e. it dramatically sensitizes the tissue for carcinogenesis. Disturbance of terminal keratinocyte differentiation seems to be a major mechanism by which COX-2 contributes to epidermal tumor formation. Based on this data, COX-2-selective inhibitors may rank among the most promising agents for skin cancer prevention and therapy.
Cyclooxygenases and Cyclooxygenase Inhibitors
COX enzymes catalyze the oxygenation of arachidonic acid to the 15-hydroperoxy-PG endoperoxide (PGG2). The hydroperoxide group at C15 is immediately reduced to the corresponding alcohol PGH2 by a peroxidase activity that is a component of the COX enzyme. PGH2 is then converted by a series of specific isomerases into PGs, thromboxane and prostacyclin (PGI2).
In vertebrates, two isoforms of COX have been identified. COX-1 is expressed constitutively and ubiquitously and is thought to generate PGs that control normal physiological function such as maintenance of the gastric mucosa and platelet function. On the other hand, COX-2 is undetectable in most tissues except brain, pancreatic islets, ovary, uterus and kidney, but becomes strongly induced upon hormonal stimulation or environmental stress, for example in the course of inflammatory processes and tissue repair. Despite a similar overall protein structure, the catalytic sites of the COX isozymes differ in that the COX-2-active site has an additional side pocket. Access to this side pocket is controlled by the valine residue at position 523 in the human COX-2 instead of the bulkier isoleucine residue as present in COX-1. This diversity has been used to develop COX isozyme-selective inhibitors [1–3]. COX represent the major cellular targets of non-steroidal anti-inflammatory drugs (NSAIDs) including the non-selective COX inhibitors and the COX-2selective ‘coxibs’ [4]. Other less sensitive NSAID targets are IB kinase, stressactivated cJun-N-terminal kinase, the transcription factor AP1, and peroxisome proliferator-activated receptor delta (PPAR-␦) which are inhibited as well as PPAR-␣ and -␥ and the stress-activated protein kinase p38 which become activated by distinct NSAIDs and salicylate [5]. As a rule, the doses required to affect these targets are up to two orders of magnitude higher than those sufficient for COX inhibition [6].
Skin Architecture
As the body’s most exposed tissue, the skin fullfils a variety of functions including thermoregulation, sensory perception, non-specific immune stimulation, and protection against external physical, chemical, or microbial attacks. This functional multiplicity is provided by a complex tissue organization at the cellular level and by unique metabolic and endocrinologic activities of the different cell types. The skin consists of the mesodermal dermis and the overlaying ectodermal epidermis. Although anatomically distinct, the two tissue compartments function in an interdependent and cooperative manner. The epidermis is a cornifying stratified epithelium with the keratinocyte which make up the major cellular constituent. The undermost cell layer(s) located just above the dermis is (are) formed by proliferatively active basal keratinocytes. These divide periodically and the daughter cells move upwards to the skin surface. As they ascend they lose their ability to proliferate and pass through different stages of terminal differentiation corresponding to distinct suprabasal cell layers such as the spinous, granular and cornified cell layers. In the course of terminal differentiation, keratins
COX-2 and Skin
73
and other epidermis-specific proteins are cross-linked to form the protein envelope of cornified cells. Finally, a complex lipid envelope is deposited between the cornified cells to control transepidermal water loss and permeability for exogenous material. Thus, the cornified cell envelope represents a structural and functional interface between the body and the external environment. Scattered throughout the epidermis are non-keratinocyte cells which have immigrated into the epithelium during embryonic development. These include neural crestderived melanocytes, bone-marrow-derived Langerhans’ cells, Merkel cells and distinct epidermal T-cell populations. The coordinated interaction of the different cell types is a prerequisite for the structural and functional integrity of the epidermis [7]. The skin is a major site of PG synthesis [8]. The major COX products of keratinocytes are PGE2 and PGF2␣ in addition to minor amounts of PGD2 while fibroblasts and endothelial cells mainly produce PGI2 [9]. The epidermis constitutively expresses COX-1 in individual keratinocytes of the interfollicular epidermis and the distal part of the hair follicle while COX-2 is found predominantly in outer root sheath cells of the growing hair follicle [10–12]. Growth factors such as TGF-␣, the cytokine IL-1␣, and 12-O-tetradecanoylphorbol 13-acetate (TPA) are known to induce COX-2 expression (fig. 1). There is substantial evidence attributing PGs important functions for the induction of skin inflammation and the development of benign and malignant skin diseases.
Cyclooxygenases and Acute Skin Inflammation
The general response of skin to a wide variety of environmental stimuli is acute inflammation. Erythema, heat, edema and pain followed by tanning and a regenerative epidermal hyperplasia and hyperkeratosis are the most prominent symptoms [13]. Dilatation of local arterioles leads to increased blood flow and subsequent erythema and heat while an increased vascular permeability causes edema and infiltration of the dermis by neutrophils. Upon exposure of skin to noxious agents such as UV light and chemicals like the phorbol ester TPA or arachidonic acid, keratinocytes are able to initiate acute inflammation by expressing an array of critical proteins such as growth factors, cytokines and cognate receptors, cell adhesion molecules, and proteins involved in lipid-derived proinflammatory mediator synthesis [14]. A modulatory role of PGs in this process is indicated by several observations. Irritants such as UV-B [15] and TPA [16, 17] enhance PG synthesis in keratinocytes as well as mesenchymal cells. The time course of PGE2 and PGI2 synthesis correlates well with the development of erythema and edema [18]. The effect of UV-B is partially mediated by histamine-induced PG synthesis [19]. Individual
Fürstenberger/Marks/Müller-Decker
74
IL-1␣
Co
Kda 97
TPA
N
TGF␣
MM
SCC
SCC
SCC SCC
N
BCC
N
BCC BCC
b PC
a
COX-2 69
c
f
d
N
Pap
e
SCC
COX-2
Fig. 1. Constitutive COX-2 overexpression in benign and malignant epidermal tumors of man and mice. a Immunoblot analysis of COX-2 protein expression in normal human skin (N), basal cell carcinomas (BCC), squamous cell carcinomas (SCC), and malignant melanoma (MM; superficial spreading, pT3). PC ⫽ Positive control. b Immunoblot analysis of COX-2 protein immunoprecipitated from untreated human keratinocytes HaCaT (Co), and from HaCaT treated with TGF-␣ (100 ng/ml), TPA (10⫺6 M), or IL-1␣ (500 U/ml) for 4 h. c–e Immunohistochemical localization of COX-2 in sections of normal human skin (c), an actinic keratosis (d), and a carcinoma in situ (e). Nuclei were counterstained with hematoxylin. f Immunoblot analysis of COX-2 protein immunoprecipitated from normal mouse epidermis (N), from benign papillomas (Pap), or from SCC.
PGs act synergistically with other eicosanoids and with structurally unrelated proinflammatory mediators [20, 21]. PGE2 and PGI2, for instance, potentiate the histamine-, bradykinin-, PAF- or LTB4-induced edema [20, 22–24] or pain receptor sensitization towards chemical and mechanical stimuli [24, 25]. Intradermal injection of PGE2, PGD2 and PGI2 induces different aspects of the inflammatory response including erythema and edema [20, 22]. Furthermore, the induction of epidermal hyperproliferation by TPA depends on immediate COX-1-mediated PGE2 synthesis [26, 27], while more delayed PG synthesis is mainly COX-2-dependent [26]. Among NSAIDs the non-selective COX
COX-2 and Skin
75
inhibitor indomethacin exhibits potent anti-inflammatory activity in human skin when applied immediately after UV-B exposure [28]. Moreover, in mouse skin a selective COX-2 inhibitor effectively suppresses UV-B-induced edema, infiltration and activation of dermal neutrophils as well as the formation of DNA-damaged sunburn cells [29]. Arachidonic acid, the predominant precursor of PGs, is esterified to cellular phospholipids at the sn-2 position. Therefore, arachidonic acid release is a major control point for PG synthesis [30]. Cleavage by the cytoplasmic 105-kDa phospholipase A2 (cPLA2) was found to be critical in UV-B- or TPAinduced PGE2 synthesis and erythema formation in keratinocytes in vitro [31] and human skin in vivo [32, 33]. Early PGE2 synthesis as catalyzed by COX-1 depends on both the activation of cPLA2 by phosphorylation and increased cPLA2 synthesis [32]. Upon progression of the inflammatory response, COX-2 protein and activity are shown to be transiently upregulated in epidermis, further augmenting PGE2 synthesis [11, 34, 35]. Depending on the type of stimulus, the time point of stimulation, and the species, COX-2 is localized to various compartments of the epidermis [11, 34, 36]. UV-B-induced generation of free radicals is probably involved in the synthesis of cPLA2 and COX-2 induced by this stimulus [35, 37]. The relative contributions of the two COX isoforms to skin inflammation were also evaluated in COX knockout mice using arachidonic acid or TPA to induce mouse ear edema [38, 39]. These studies revealed COX-1 rather than COX-2 to be relevant for the acute phase of the arachidonic acid-induced ear edema. This view is in line with the observation that heterozygous transgenic overexpression of COX-2 in basal cells of the follicular and interfollicular epidermis resulting in highly elevated levels of PGE2, PGF2␣ and 15-deoxy⌬12,14PGJ2 but not of PGI2 are not sufficient to evoke inflammation in dorsal and tail skin [40]. Thus, the extent to which each COX isoform contributes to PG formation may depend on the inflammatory stimulus, the time after insult, and the relative expression level in the targeted body site. In addition to the proinflammatory effects, chronic exposure to UV-B is a major cause of chronic cutaneous damage and a predominant etiologic factor for the induction of skin cancer [41].
COX-2 Expression in Human Skin Cancer
Malignant melanoma and non-melanoma skin cancers (NMSC) such as basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) belong to the most frequent neoplasias in the Caucasian population worldwide and their incidence has reached an epidemic dimension. NMSC are about 20 times more
Fürstenberger/Marks/Müller-Decker
76
frequent than malignant melanomas. The standardized ratio of BCC and SCC is about 4:1 [42]. Several etiologic factors have been implicated in NMSC development including UV irradiation as the most important factor although the association is less clear for BCC [43], chemical carcinogens such as arsenic and coal tar [42], immune suppression [42, 44], and chronic inflammatory and degenerative skin alterations [42]. Infection with human papilloma virus [45] as well as germline mutations of genes involved in nucleotide excision repair are additional risk factors for the development of SCC [42]. In contrast to BCC [46], SCC is preceded by precancerous lesions. These comprise Bowen’s disease, radiation, coal tar, arsenic and scar keratoses, and most commonly actinic keratoses (AK) being attributable to UV exposure [43, 47]. The morbidity and mortality of BCC and SCC are different. While BCC display a potential for local invasion and destruction, they possess no general tendency to metastasize [46]. On the other hand, SCC represent invasive tumors and, on average, about 10% will metastasize [47]. For desmoplastic SCC the rate of local and regional metastasis is up to 23% [48]. Although NMSC are easily detected and surgically removed, in patients with multiple lesions or in cases with tumors in critical locations, disfigurement and disease recurrence may generate serious problems. BCC and SCC differ with respect to COX expression. BCC biopsies show a regular expression of COX-1 mRNA and protein resembling that of normal epidermis. COX-2 is only weakly albeit consistently expressed [12, 34, 49]. In contrast, COX-2 is strongly overexpressed in the majority of SCC while being undetectable in normal skin of the tumor patients (fig. 1) [12, 34, 49, 50]. Immunohistochemical analysis revealed both COX isozymes to be expressed throughout the tumor parenchyme in the cytoplasm, the perinuclear membrane and the nucleus of keratinocytes (fig. 1). In addition, cells of the inflammatory infiltrate and dendritic cells are positive for both COX isozymes [12]. COX-2 expression is also strongly upregulated in SCC of the head and neck while undetectable in matched control samples [51]. COX-2 protein is localized to the cytoplasm of the tumor cells. A weaker immune signal of COX-2 was found in the apparently normal epithelium adjacent to the tumors. Aberrant COX-2 expression has also been reported for spontaneous SCC of skin as well as of head and neck in dogs. The protein is localized to the perinuclear membranes of neoplastic keratinocytes and stromal cells whereas normal skin and mucosa cells adjacent to or within the tumors are negative. A weak COX-1 staining is found in normal and neoplastic canine skin resembling the situation in humans [52]. Most importantly, a constitutive COX-2 expression is observed in precursor lesions of SCC such as leukoplakias and AK (fig. 1) [12, 34, 53]. Thus, COX-2 upregulation appears to be an early pre-malignant event in the development
COX-2 and Skin
77
of SCC. Since UV-B irradiation strongly induces COX-2 expression and AK are primarily found on chronically sun-exposed body sites, UV-B-induced COX-2 may be causally related to the development of AK [34, 41]. Accordingly, inhibition of COX-2 may be a strategy to halt or reverse skin carcinogenesis (see below). The fact that AK has reached epidemic dimensions ranging between 5 and 14% of the population in the USA and up to 60% in the adult population of Australia urgently calls for preventive measures against skin cancer [54]. Cutaneous malignant melanoma is the most rapidly increasing cancer in Caucasian populations with an estimated annual increase between 3 and 7% suggesting a doubling of rates every 10–20 years [55]. The increased risk is thought to be related to the frequency of severe sunburns in childhood and altered patterns of sun exposure in fair-skinned people [43]. A high cure rate can be achieved upon treatment at an early stage whereas advanced stages of melanomas are mostly incurable. In fact, malignant melanoma is the most frequent cause of death due to a skin disease. Malignant melanomas express COX-1 to a similar degree as the adjacent normal keratinocytes [12, 56]. Contradictory results have been reported with respect to COX-2 expression. In one study, a moderate to strong expression of COX-2 is found in malignant melanomas but not in benign nevi [56] while two other groups failed to verify COX-2 expression in melanomas [12, 49].
Regulation of COX-2 Expression in Keratinocytes
A transient COX-2 expression in response to various stimuli and a permanent upregulation of COX-2 in tumors (fig. 1) have been reported to be predominantly due to transcriptional activation [1, 57–60] but also increased mRNA stability [58, 61, 62]. Signaling kinases linked to COX-2 gene activation in epidermis include tyrosine kinases [63], protein kinase A [64], protein kinase C␣ [65, 66], the extracellular signal-regulated kinase (ERK) and p38 mitogen-activated protein kinase (MAPK) [58]. The human [57, 67] and murine [68] COX-2 gene promoters have been cloned. Being considerably different from the COX-1 promoter [69], the COX-2 promoter contains only one transcription start site, a TATA box and, besides multiple copies of the SP-1 element, several other putative enhancer elements, including NF-B, NF-IL-6 (C/EBP), ATF/CRE, E-box, AP-2, PEA-3, STAT3, GATA sites, as well as a glucocorticoid and an insulin-responsive element [1, 57, 67, 68]. This complex structure of the COX-2 promoter explains the enormous variability of stimulatory effects. The trans-acting factors and cis-acting DNA elements that mediate trans-activation appear to depend on the cell type, on the stimulus, and on species-specific promoter elements.
Fürstenberger/Marks/Müller-Decker
78
Using reporter constructs under the control of distinct COX-2 promoter sequences, the transcriptional induction in normal human keratinocytes by IFN-␥ has been found to be mediated at least partially through the EGF receptor-controlled Ras-Raf-MAP kinase cascade. This signaling pathway leads to the activation of transcription factors that bind to the proximal CRE/ATF/E-box element located at nucleotide ⫺124 [58]. In a murine SCC-derived cell line that constitutively overexpresses COX-2, the E-box (located at ⫺66/⫺38) and the NF-IL-6 site (located at ⫺138/⫺131) are identified as positive regulatory sequences of the mouse COX-2 promoter [60]. Upstream stimulatory factors and CCAAT/enhancer-binding proteins (C/EBPs) bind to and activate these sites. A comparison of normal epidermis with carcinoma cells in electrophoretic mobility and supershift assays using NF-IL-6 and ATF/CRE-E-box probes reveals different shift patterns only for NF-IL-6. Especially C/EBP isoforms are expressed differentially during mouse skin carcinogenesis. The constitutive upregulation of COX-2 expression in early papillomas correlates with the downregulation of C/EBP-␣ and simultaneous upregulation of C/EBP- and -␦ expression, suggesting that they contribute to the overexpression of COX-2 in epidermal tumors [60].
COX Inhibition and Skin Carcinogenesis
There is substantial epidemiological and clinical evidence that NSAIDs and coxibs decrease the risk for colorectal and other cancers of the gastrointestinal tract [70, 71]. Studies with various animal models in which cancer was induced by gene ablation or chemical carcinogens have corroborated this data by showing a strong reduction of tumor multiplicity following NSAID treatment [72, 73]. Allowing a clear-cut experimental separation of tumor development into defined stages, the multistage approach of mouse skin carcinogenesis [74, 75] reflects stepwise cancer development in man. Neoplastic development is thought to be initiated by an activation of a proto-oncogene or an inactivation of a tumor suppressor gene in a single cell. This mutated cell is clonally expanded by a long-lasting disturbance of tissue homeostasis which is achieved for instance by long-term treatment with a tumor promoter. This promotion step gives the initiated cells a selective advantage provided that they respond more sensitively to mitogenic stimuli and/or less sensitively to differentiationinducing or apoptotic signals than their non-initiated counterparts. The clonal expansion of initiated cells raises the possibility of additional genetic alterations which are required for progression to malignancy. In addition, endogenous mechanisms leading to genetic instability may become involved during
COX-2 and Skin
79
tumor promotion. Such mechanisms include an impairment of gene repair and the production of genotoxic metabolites. Experimental skin carcinogenesis in mice can be initiated by a single application of a low dose of a genotoxic carcinogen such as dimethylbenz[a]anthracene (DMBA) and promoted by continuous treatment with the phorbol ester TPA. The majority of the primary tumors are papillomas, some of which convert spontaneously into SCC [74, 75]. This malignant progression can be augmented by application of genotoxic carcinogens. In papillomas and SCC generated according to this protocol, COX-2 is constitutively overexpressed whereas COX-1 expression is similar in both normal and neoplastic skin [17] (fig. 1). COX-1 protein is found in a few keratinocytes scattered throughout the normal and neoplastic epithelium. In contrast, COX-2 expression occurs in essentially all basal cells of papillomas and in the tumor parenchyme of SCC as well as in the tumor vasculature [76, 77] (fig. 3). In tumor tissue the levels of PGs are up to 50-fold higher than in normal epidermis indicating aberrant COX-2 activity to be important for skin tumor development [17]. This view is supported by the observation that COX-2-selective inhibitors such as SC-58125 and celebrex suppress papilloma formation by inhibiting the promotion but not the initiation stage [76, 77]. The association between COX activity and tumor promotion can be drawn back to a single COX product in that the inhibitory activity of indomethacin, a nonselective COX inhibitor, is specifically reversed when PGF2␣ is administered together with TPA [78]. Thus, PGF2␣ appears to fullfil the function of an endogenous skin tumor promoter. On the other hand, the cognate PGF2␣ receptor is downregulated at the mRNA level in papillomas but not in the surrounding hyperplastic tissue [79]. This receptor downregulation may represent an adaptation of the cells to the high PGF2␣ level. Overexpression of COX-2 has been reported for other experimental models of skin carcinogenesis. UV-B-induced SCC in mouse skin consistently expresses COX-2 and the development of SCC is sensitive to inhibition by the COX-2selective inhibitor celebrex and the non-selective COX inhibitor indomethacin [80–82]. In addition, non-selective COX inhibitors such as piroxicam and nabumetone decrease tumor growth and weight in mice subcutaneously injected with cells from SCC of the head and neck [83]. Finally, a COX-2-selective inhibitor suppresses the growth of tumors in nude mice which were inoculated with human head and neck squamous carcinoma cells [84]. The finding that therapeutic doses of COX inhibitors strongly reduce the number of UV-B-induced skin tumors initiated studies to determine whether or not such a chemopreventive effect also occurs in humans. According to a multicenter, double-blind, placebo-controlled study, topical treatment with the non-selective COX inhibitor diclofenac is effective in the treatment of
Fürstenberger/Marks/Müller-Decker
80
AK [85]. Moreover, several multicenter trials are in progress investigating the chemopreventive effect of COX-2-selective inhibitors using AK and BCC as endpoints [86].
Functional Analysis of COX-2 in vivo by Transgenic and Knockout Approaches
COX-2 deficiency protects against the formation of intestinal tumors in Apc⫹/⫺ knockout [87] and in Min⫹/⫺ mice [88], i.e. two animal models of human familial adenomatous polyposis coli. In line with this data, recent transgenic and knockout studies demonstrate a causal relationship between COX-2 and the development of epidermal tumors and give insight into the mechanisms by which COX-2 contributes to epidermal homeostasis and squamous cell cancer development. The skin of COX-2-deficient mice shows a reduced hyperplastic response to TPA treatment being the result of an enhanced keratinocyte transit rate rather than a decreased proliferation rate or an increased level of apoptosis [89]. In a multistage skin carcinogenesis experiment with DMBA as an initiator and TPA as a promoter, COX-2 deficiency led to a gene dosage-dependent reduction of the papilloma yield and incidence as compared to wild-type mice. The reduced tumor burden is not the result of a decreased initiation rate, suggesting that COX-2 is not essential for metabolic activation of DMBA [89, 90]. This observation is keeping with the pharmacological data showing an inhibition of promotion but not initiation by the COX-2-selective celebrex [77]. The cellular event leading to the reduced rate of tumor formation in COX-2-deficient mice seems to be a premature onset of terminal differentiation in initiated keratinocytes. Keratins 1 and 10, both indicators of keratinocytes committed to differentiate [91], are more frequently expressed in basal keratinocytes and papillomas of COX-2-deficient than of wild-type mice, while expression of loricrin, a marker of late terminal differentiation, is only weakly impaired. Thus, early rather than late stages of epidermal differentiation are affected in the COX-2 knockout mice. Apoptotis does not account for the smaller size of the papillomas, which developed despite COX-2 deficiency [89]. Similar effects have been reported for COX-1-deficient mice indicating that this isoform might also contribute to skin tumor development [89]. On the other hand, the expression levels of COX-1 are not altered in hyperplastic, dysplastic and neoplastic epidermal lesions of mice [11, 17, 26, 50] and humans as compared to normal tissue [8, 12]. The results obtained with COX-2-deficient mice can be fully reconciled with a previous study. Transgenic mice overexpressing COX-2 under the
COX-2 and Skin
81
control of a keratin 5 promoter in the basal cells of interfollicular and follicular epidermis show a delayed terminal differentiation but no increased proliferation rate in the epidermis [40] (fig. 2). In such animals, keratin 10, a marker of immediate early, involucrin, a marker of early, and loricrin, a marker of late differentiation [91] are found to be incompletely expressed. In addition, Ki67 protein, a marker of proliferation, is not restricted to the basal cell compartment as in wild-type epidermis but is also found in the suprabasal spinous layers of the transgenic epidermis. This disturbance results in a basal cell hyperplasia and hyperkeratosis in interfollicular epidermis. Moreover, the epidermis is dysplastic showing a complete loss of cell polarity, a formation of horn pearls, and an endophytic papillary growth into the underlying dermis. In addition, sebaceous glands are hyperplastic [40] (fig. 2) and sebum production is strongly enhanced leading to strong epicutanous sebum deposition and a greasy fur. Furthermore, the blood vessel density is increased in COX-2 transgenic skin as compared to wild-type skin indicating a proangiogenic effect by COX-2 [40]. The phenotypic changes depend on COX-2 expression and activity, since feeding of transgenic mice with a COX-2-selective inhibitor led to normal PG levels, a regular expression of the differentiation markers in transgenic epidermis, and a complete suppression of the transgenic phenotype [77]. The dysplastic lesions resemble the lesions induced by DMBA/TPA. Nevertheless, the transgenics do not develop skin tumors spontaneously, but only upon a single application of an initiating dose of DMBA. However, longterm treatment with the tumor promoter TPA, as required for tumorigenesis in wild-type mice, is not necessary for transgenics [77]. Pathological analysis of tumors revealed papillomas and SCC and, in addition, sebaceous gland adenomas, indicating that COX-2 also plays a role in the development of this tumor type (fig. 3). It was concluded that COX-2 overexpression leading to high tissue levels of epidermal PGE2, PGF2␣ and 15-deoxy-⌬12,14-PGJ2 is insufficient for skin tumor induction but transforms epidermis into an ‘auto-promoted’ state, i.e. dramatically sensitizes the tissue for genotoxic carcinogens. Tumorigenesis in COX-2 overexpressing transgenic mice is not restricted to the skin. In fact, in mammary glands, focal hyperplasia and dysplasia were observed. In contrast to skin, transgenic mammary glands spontaneously develop metastatic carcinomas [92]. In this case, decreased apoptosis of mammary epithelial cells seems to contribute to tumorigenesis. However, there is no clear distinction between tumor initiation and tumor promotion in this model. Thus, depending on the epithelium and cell type, COX-2 may contribute to tumorigenesis by various mechanisms including impairment of differentiation [40, 89], inhibition of apoptosis [93], stimulation of angiogenesis [40, 94, 95], but probably also by stimulation of invasiveness [96] and by immune suppression [97]. In skin, delay of terminal differentiation appears to be the predominant
Fürstenberger/Marks/Müller-Decker
82
a
b
c
d
e
f
g
h
i
j
Fig. 2. COX-2-induced delay of terminal keratinocyte differentiation and sebaceous gland hyperplasia. Immunoperoxidase (a, b, g, h) or immunofluorescence (c–f, i, j) staining of cryosections from wild-type (a, c, e, g, i) and transgenic tail skin (b, d, f, h, j) with anti-COX-2 (a, b), anti-keratin 5 (c, d, i, j), anti-involucrin (e, f), and anti-Ki67 (g, h) antisera. Note COX-2 expression in follicular outer root sheath but not interfollicular keratinocytes of wild-type and in interfollicular basal keratinocytes of transgenic tail epidermis; the local extension of keratin 5 (d, j), the restriction of involucrin to the uppermost spinous layer (f), and the suprabasal Ki67-positive keratinocytes (h) in the hyperplastic transgenic epidermis. Nuclei were counterstained with hematoxylin (a, b, g, h). Sebaceous gland hyperplasia in transgenics (j) as compared to wild-type litters (i). Lobules, enclosed by a single layer of keratin 5-positive basal cells and grouped around an enlarged gland duct (j), were increased in number and were built up by well-differentiated sebocytes.
COX-2 and Skin
83
a
b
c
d
e
f
g
h
i
j
k
l
Fig. 3. Characterization of tumors in the skin of COX-2 transgenic mice treated with DMBA alone or according to the DMBA/TPA protocol. a–c Representative HE staining of sections from sebaceous gland adenoma (a), papilloma (b) and SCC (c) obtained by DMBA application to the skin of COX-2 transgenic mice. d–l COX-2 protein in the proliferative compartment of the tumor parenchyme and tumor vasculature. A double immunofluorescence analysis was performed for a sebaceous gland adenoma (d, g, j) and papillomas from DMBA-initiated (e, h, k) and from DMBA/TPA-treated transgenics ( f, i, l) using anti-COX2 and anti-CD31 antisera revealing that COX-2 (d–f, green) and CD31 (g–i, red) colocalize in tumor vessels ( j–l, yellow).
Fürstenberger/Marks/Müller-Decker
84
mechanism by which COX-2 supports the clonal outgrowth of initiated epidermal cells. Whether this mechanism of COX-2 action may play a role in other epithelia in vivo needs to be investigated.
References 1 2 3 4 5 6
7 8 9 10
11
12
13 14 15 16
17
18
19
20
Smith WL, Dewitt DL, Garavito RM: Cyclooxygenases: Structural, cellular and molecular biology. Annu Rev Biochem 2000;69:145–182. Herschman HR: Prostaglandin-2 synthase. Biochim Biophys Acta 1996;1299:125–140. Turini ME, DuBois RN: Cyclooxygenase-2: A therapeutic target. Annu Rev Med 2002;53:35–57. Marnett LJ, Dubois RN: COX-2: A target for colon cancer prevention. Annu Rev Pharmacol Toxicol 2002;42:55–80. Tegeder I, Pfeilschifter J, Geisslinger G: Cyclooxygenase-independent actions of cyclooxygenase inhibitors. FASEB J 2001;15:2057–2072. Raz A: Is inhibition of cyclooxygenase required for the anti-tumorigenic effects of non-steroidal, anti-inflammatory drugs (NSAIDs)? In vitro versus in vivo results and the relevance for the prevention and treatment of cancer. Biochem Pharmacol 2002;63:343–347. Fuchs E, Raghavan S: Getting under the skin of epidermal morphogenesis. Nat Rev Genet 2002; 3:199–209. Goldyne ME: Cyclooxygenase isoforms in human skin. Prostaglandins Other Lipid Mediat 2000; 63:15–23. Ziboh VA: The significance of polyunsaturated fatty acids in cutaneous biology. Lipids 1996;31: 249–253. Neufang G, Füstenberger G, Leder C, Marks F, Müller-Decker K: Cyclooxygenase-2 expression in the outer root sheath of anagen but not telogen hair follicles of mouse skin. J Invest Dermatol 2000;115:151. Müller-Decker K, Scholz K, Neufang G, Marks F, Fürstenberger G: Localization of prostaglandinH synthase-1 and -2 in mouse skin: Implications for cutaneous function. Exp Cell Res 1998; 242:84–91. Müller-Decker K, Reinerth G, Krieg P, Zimmermann R, Heise H, Bayerl C, Marks F, Fürstenberger G: Prostaglandin-H-synthase isozyme expression in normal and neoplastic human skin. Int J Cancer 1999;82:648–656. Hruza LL, Pentland AP: Mechanisms of UV-induced inflammation. J Invest Dermatol 1993; 100:35S–41S. Williams IR, Kupper TS: Immunity at the surface: Homeostatic mechanisms of the skin immune system. Life Sci 1996;58:1485–1507. Pentland AP, Jacobs SC: Bradykinin-induced prostaglandin synthesis is enhanced in keratinocytes and fibroblasts by UV injury. Am J Physiol 1991;261:R543–R547. Goldyne ME, Evans CB: 12-O-tetradecanoylphorbol-13-acetate and the induction of prostaglandin E2 generation by human keratinocytes: A re-evaluation. Carcinogenesis 1994;15: 141–143. Müller-Decker K, Scholz K, Marks F, Fürstenberger G: Differential expression of prostaglandin H synthase isozymes during multistage carcinogenesis in mouse epidermis. Mol Carcinog 1995; 12:31–41. Black AK, Greaves MW, Hensby CN, Pummer NA: Increased prostaglandins E2 and F2␣ in human skin at 6 and 24 h after ultraviolet B irradiation (290–320 nm). Br J Clin Pharmacol 1978;5: 431–436. Pentland AP, Mahoney M, Jacobs S, Holtzmann MJ: Enhanced prostaglandin synthesis after ultraviolet injury is mediated by endogenous histamine stimulation. J Clin Invest 1990;86: 566–574. Williams TJ: Prostaglandin E2, prostaglandin I2 and the vascular changes of inflammation. Br J Pharmacol 1979;65:517–524.
COX-2 and Skin
85
21
22 23 24
25 26
27
28 29
30 31
32
33
34
35 36
37 38
39
40
41
Rhodes LE, Belgi G, Parslew R, McLoughlin L, Clough GF, Friedman PS: Ultraviolet-B-induced erythema is mediated by nitric oxide and prostaglandin E2 in combination. J Invest Dermatol 2001;117:880–885. Williams TJ, Peck MJ: Role of prostaglandin-mediated vasodilatation in inflammation. Nature 1977;270:530–532. Davis P, Bailey PJ, Goldenberg MM, Ford-Hutchinson AW: The role of arachidonic acid oxygenation products in pain and inflammation. Annu Rev Immunol 1984;2:335–357. Murata T, Ushikubi F, Matsuoka T, Hirata M, Yamasaki A, Sugimoto Y, Ishikawa A, Aze Y, Tanaka T, Yoshida N, Uedo A, Ohishi S, Narumiya S: Altered pain perception and inflammatory response in mice lacking prostacyclin receptor. Nature 1997;388:678–682. Ferreira SH: Inflammatory pain, prostaglandin hyperalgesia and the development of peripheral analgesia. Trends Pharmacol Sci 1981;12:183–186. Scholz K, Fürstenberger G, Müller-Decker K, Marks F: Differential expression of prostaglandin H-synthase isoenzymes in normal and activated keratinocytes in vivo and in vitro. Biochem J 1995;306:263–269. Fürstenberger G, Marks F: Early prostaglandin E synthesis is an obligatory event in the induction of cell proliferation in mouse epidermis in vivo by the phorbol ester TPA. Biochem Biophys Res Commun 1980;92:749–756. Snyder DS, Eaglestein WH: Intradermal anti-prostaglandin agents and sunburn. J Invest Dermatol 1974;62:47–50. Wilgus TA, Ross MS, Parrett ML, Oberyszyn TM: Topical application of a selective cyclooxygenase inhibitor suppresses UVB-mediated cutaneous inflammation. Prostaglandins Other Lipid Mediat 2000;62:367–384. Balsinde J, Balboa MA, Insel PA, Dennis EA: Regulation and inhibition of phopholipase A2. Annu Rev Pharmacol Toxicol 1999;39:175–189. Kast R, Fürstenberger G, Marks F: Phorbol-ester TPA- and bradykinin-induced arachidonic acid release from keratinocytes is catalyzed by a cytosolic phospholipase A2. J Invest Dermatol 1993;101:567–572. Gresham A, Masferrer J, Chen X, Leal-Khouri S, Pentland AP: Increased synthesis of highmolecular-weight cPLA2 mediates early UV-induced PGE2 in human skin. Am J Physiol 1996; 270:C1037–C1050. Kang-Rotondo CH, Miller CC, Morrison AR, Pentland AP: Enhanced keratinocyte prostaglandin synthesis after UV injury is due to increased phospholipase activity. Am J Physiol 1993;264: C396–C401. Buckman SY, Gresham A, Hale P, Hruza G, Anast J, Masferrer J, Pentland AP: COX-2 expression is induced by UVB exposure in human skin: Implications for the development of skin cancer. Carcinogenesis 1998;19:723–729. Soriani M, Luscher P, Tyrrell RM: Direct and indirect modulation of ornithine decarboxylase and cyclooxygenase by UVB radiation in human skin cells. Carcinogenesis 1999;20:727–732. Athar M, An KP, Morel KD, Kim AL, Aszterbaum M, Longley J, Epstein EH, Bickers DR: Ultraviolet B-induced COX-2 expression in murine skin: An immunohistochemical study. Biochem Biophys Res Commun 2001;280:1042–1047. Chen X, Gresham A, Morrison A, Pentland AP: Oxidative stress mediates synthesis of cytosolic phospholipase A2 after UVB injury. Biochim Biophys Acta 1996;1299:23–33. Morham SG, Langenbach R, Loftin CD, Tiano HF, Vouloumanos N, Jennette JC, Mahler JF, Kluckmann KD, Ledford A, Lee CA, Smithies O: Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell 1995;83:473–482. Langenbach R, Morham SG, Tiano HF, Loftin CD, Ghanayem BI, Chulada PC, Mahler JF, Lee CA, Goulding EH, Kluckman KD, Kim HS, Smithies O: Prostaglandin synthase-1 gene disruption in mice reduced arachidonic acid-induced inflammation and indomethacin-induced gastric ulceration. Cell 1995;83:483–492. Neufang G, Fürstenberger G, Heidt M, Marks F, Müller-Decker K: Abnormal differentiation of epidermis in transgenic mice constitutively expressing cyclooxygenase-2 in skin. Proc Natl Acad Sci USA 2001;98:7629–7634. Brash DE: Sunlight and the onset of skin cancer. Trends Genet 1997;13:410–414.
Fürstenberger/Marks/Müller-Decker
86
42 43 44 45 46 47 48
49
50
51
52 53 54 55 56 57
58
59
60 61
62
63 64
Diepgen TL, Mahler V: The epidemiology of skin cancer. Br J Dermatol 2002;146(suppl 61):1–6. Armstrong BK, Kricker A: The epidemiology of UV induced skin cancer. J Photochem Photobiol B 2001;63:8–18. DiGiovanna JJ: Post-transplantation skin cancer: Scope of the problem, management, and role of systemic retinoid chemoprevention. Transplant Proc 1998;30:2771–2778. Zur Hausen H: Papilloma virus causing cancer: Evasion from host-cell control in early events in carcinogenesis. J Natl Cancer Inst 2000;92:690–698. Randle HW: Basal cell carcinoma. Identification and treatment of the high-risk patients. Dermatol Surg 1996;22:255–261. Stratton SP, Dorr RT, Alberts DS: The state-of-the-art in chemoprevention of skin cancer. Eur J Cancer 2000;36:1292–1297. Breuninger H, Schaumburg-Lever G, Holzschuh J, Horny HP: Desmoplastic squamous cell carcinoma of skin and vermilion surface: A highly malignant subtype of skin cancer. Cancer 1997;79: 915–919. Vogt T, McClelland M, Jung B, Popova S, Bogenrieder T, Becker B, Rumpler G, Landthaler M, Stolz W: Progression and NSAID-induced apoptosis in malignant melanomas are independent of cyclooxygenase II. Melanoma Res 2001;11:587–599. Leong J, Hughes-Fulford M, Rakhlin N, Habib A, Maclouf J, Goldyne ME: Cyclo-oxygenase in human and mouse skin and cultured human keratinocytes: Association of COX-2 expression with human keratinocyte differentiation. Exp Cell Res 1996;224:79–87. Chan G, Boyle JO, Yang EK, Zhang F, Sacks PG, Shah JP, Edelstein D, Soslow RA, Koki AT, Woerner BM, Masferrer JL, Dannenberg AJ: Cyclooxygenase-2 expression is up-regulated in squamous cell carcinoma of the head and neck. Cancer Res 1999;59:991–994. Pestili de Almeida AM, Piché C, Sirois J, Doré M: Expression of cyclo-oxygenase-2 in naturally occurring squamous cell carcinomas in dogs. J Histochem Cytochem 2001;49:867–875. Renkonen J, Wolff H, Paavonen T: Expression of cyclo-oxygenase-2 in human tongue carcinoma and its precursor lesions. Virchows Arch 2002;440:594–597. Stratton SP: Prevention of non-melanoma skin cancer. Curr Oncol Rep 2001;3:295–300. Gilchrest BA, Eller MS, Geller AC, Yaar M: The pathogenesis of melanoma induced by ultraviolet irradiation. N Engl J Med 1999;340:1341–1348. Denkert C, Köbel M, Berger S, Siegert A, Leclere A, Trefzer, U, Hauptmann S: Expression of cyclooxygenase 2 in human malignant melanoma. Cancer Res 2001;61:303–308. Lukiw WJ, Pelaez RP, Martinez J, Bazan NG: Budesonide epimer R or dexamethasone selectively inhibit platelet-activating factors and cyclooxygenase-2 gene expression in human epidermal keratinocytes. Proc Natl Acad Sci USA 1998;95:3914–3919. Matsuura H, Sakaue M, Subbaramaiah K, Kamitani H, Eling TE, Dannenberg AJ, Tanabe T, Inoue H, Arata J, Jetten AM: Regulation of cyclooxygenase-2 by interferon-␥ and transforming growth factor-␣ in normal human epidermal keratinocytes and squamous carcinoma cells. J Biol Chem 1999;274:29138–29148. Kutchera W, Jones DA, Matsunami N, Groden J, MyIntyre TM, Zimmermann GA, White RL, Prescott SM: Prostaglandin H synthase-2 is expressed abnormally in human colon cancer: Evidence for a transcriptional effect. Proc Natl Acad Sci USA 1996;93:4816–4820. Kim Y, Fischer SM: Transcriptional regulation of cyclooxygenase-2 in mouse skin carcinoma cells. J Biol Chem 1998;273:27686–27694. Ristimäki A, Garfinkel S, Wessendorf J, Maciag T, Hla T: Induction of cyclo-oxygenase-2 by interleukin-1␣. Evidence for posttranscriptional regulation. J Biol Chem 1994;269: 11769–11775. Dixon DA, Tolley ND, King PH, Nabors LB, McIntyre TM, Zimmermann GA, Prescott SM: Altered expression of the mRNA stability factor HuR promotes cyclooxygenase-2 expression in colon cancer cells. J Clin Invest 2001;108:1657–1665. Miller CC, Hale P, Pentland AP: Ultraviolet B injury increases prostaglandin synthesis through a tyrosine kinase-dependent pathway. J Biol Chem 1994;264:3529–3533. Maldve RE, Kim Y, Muga SJ, Fischer SM: Prostaglandin E2 regulation of cyclooxygenase expression in keratinocytes is mediated via cyclic nucleotide-linked prostaglandin receptors. J Lipid Res 2000;41:873–881.
COX-2 and Skin
87
65
66
67
68
69
70 71
72 73
74 75 76
77
78
79 80
81 82
83
84
85
Wang HQ, Smart RC: Overexpression of protein kinase C-␣ in the epidermis of transgenic mice results in striking alterations in phorbol ester-induced inflammation and COX-2, MIP-2, and TNF-␣ expression but not tumor promotion. J Cell Sci 1999;112:3497–3506. Wang HQ, Kim MP, Tiano HF, Langenbach R, Smart RC: Protein kinase C-␣ coordinately regulates cytosolic phospholipase A2 activity and expression of cyclooxygenase-2 through different mechanisms in mouse keratinocytes. Mol Pharmacol 2001;59:860–866. Kosaka T, Miyata A, Ihara H, Hara S, Sugimoto T, Takeda O, Takahashi E, Tanabe T: Characterization of the human gene (PTGS2) encoding prostaglandin-endoperoxide synthase Eur J Biochem 1994;221:889–897. Fletscher BS, Kujubu DA, Perrin DM, Herschman HR: Structure of the mitogen-inducible TIS 10 gene and demonstration that the TIS-10-encoded protein is a functional prostaglandin G/H synthase. J Biol Chem 1992;267:4338–4344. Kraemer SA, Meade EA, DeWitt DL: Prostaglandin endoperoxide synthase gene structure: Identification of the transcriptional start site and 5⬘ flanking regulatory sequences. Arch Biochem Biophys 1992;293:391–400. Thun MJ, Henley SJ, Patrono C: Nonsteroidal anti-inflammatory drugs as anticancer agents: Mechanistic, pharmacologic and clinical issues. J Natl Cancer Inst 2002;94:252–266. Steinbach G, Lynch PM, Phillips RK, Wallace MH, Hawk E, Gordon GB, Wakabayashi N, Saunders B, Shen Y, Fujimura T, Su LK, Levin B: The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–1952. Smalley WE, Dubois RN: Colorectal cancer and nonsteroidal anti-inflammatory drugs. Adv Pharmacol 1997;39:1–20. Marks F, Furstenberger G: Cancer prevention through interruption of multistage carcinogenesis: The lessons learnt by comparing mouse skin carcinogenesis and human large bowel cancer. Eur J Cancer 2000;36:314–329. DiGiovanni J: Multistage carcinogenesis in mouse skin. Pharmacol Ther 1992;54:63–128. Marks F, Fürstenberger G: Tumor promotion in skin; in Arcos CE, Arcos MF, Woo YT (eds): Chemical Induction of Cancer. Boston, Birkhäuser, 1995, pp 125–160. Müller-Decker K, Kopp-Schneider A, Marks F, Seibert K, Fürstenberger G: Localization of prostaglandin H synthase isozymes in murine epidermal tumors: Suppression of skin tumor promotion by inhibition of prostaglandin H synthase-2. Mol Carcinog 1998;23:36–44. Müller-Decker K, Neufang G, Berger I, Neumann M, Marks F, Fürstenberger G: Transgenic cyclooxygenase-2 overexpression sensitizes mouse skin for carcinogenesis. Proc Natl Acad Sci USA 2002;99:12483–12488. Fürstenberger G, Gross M, Marks F: Eicosanoids and multistage carcinogenesis in NMRI mouse skin: Role of prostaglandin E and F in conversion (first stage of tumor promotion) and promotion (second stage of tumor promotion). Carcinogenesis 1989;10:91–96. Müller K, Krieg P, Marks F, Fürstenberger G: Expression of PGF2␣ mRNA in normal, hyperplastic and neoplastic skin. Carcinogenesis 2000;21:1063–1066. Fischer SM, Lo HH, Gordon GB, Seibert K, Kelloff G, Lubet RA, Conti CJ: Chemopreventive activity of celecoxib, a specific cyclooxygenase-2 inhibitor, and indomethacin against ultravioletinduced skin carcinogenesis. Mol Carcinog 1999;25:231–240. Pentland AP, Schoggins JW, Scott GA, Khan KN. Han R: Reduction of UV-induced skin tumors in hairless mice by selective COX-2 inhibition. Carcinogenesis 1999;20:1939–1944. Orengo IF, Gerguis J, Phillips R, Guevara A, Lewis AT, Black HS: Celecoxib, a cyclooxygenase2 inhibitor as potential chemopreventive to UV-induced skin cancer: A study in the hairless mouse model. Arch Dermatol 2002;138:751–755. Scioscia KA, Snyderman CH, Rueger R Reddy J, D’Amico F, Comsa S Collins B: Role of arachidonic acid metabolites in tumor growth inhibition by nonsteroidal anti-inflammatory drugs. Am J Otolaryngol 1997;18:1–8. Nishimura G, Yanoma S, Mizuno H, Kawakami K, Tsukuda M: A selective cyclooxygenase-2 inhibitor suppresses tumor growth in nude mice xenografted with human head and neck squamous cell carcinoma cells. Jpn J Cancer Res 1999;90:1152–1162. Rivers JK, Arlette J, Shear N, Guenther L, Carey W, Poulin Y: Topical treatment of actinic keratoses with 3.0% diclofenac in 2.5% hyaluronan gel. Br J Dermatol 2002;146:1–7.
Fürstenberger/Marks/Müller-Decker
88
86 87
88
89
90 91 92
93 94 95 96 97
Pentland AJ: Editorial: Cyclooxygenase inhibitors for skin cancer prevention. Are they beneficial enough? Arch Dermatol 2002;138:751–755. Oshima M, Dinchuk JE, Kargman SL, Oshima H, Hancock B, Kwong E, Trazaskos JM, Evans JF, Taketo MM: Suppression of intestinal polyposis in APC716 knockout mice by inhibition of cyclooxygenase-2 (COX-2). Cell 1996;87:803–809. Chulada PC, Thompson MB, Mahler JF, Doyle CM, Gaul BW, Lee C, Tiano HF, Morham SG, Smithies O, Langenbach R: Genetic disruption of Ptgs-1, as well as of Ptgs-2, reduces intestinal tumorigenesis in min mice. Cancer Res 2000;60:4705–4708. Tiano HF, Loftin CD, Akunda J, Lee CA, Spalding J, Sessoms A, Dunson DB, Rogan EG, Morham SG, Smart RC, Langenbach R: Deficiency of either cyclooxygenase (COX)-1 or COX-2 alters epidermal differentiation and reduces mouse skin tumorigenesis. Cancer Res 2002;62:3395–3401. Wiese FW, Thompson PA, Kadlubar FF: Carcinogen substrate specificity of human COX-1 and COX-2. Carcinogenesis 2001;21:5–10. Fuchs E: Keratins and the skin. Annu Rev Cell Dev Biol 1995;11:123–153. Liu CH, Chang SH, Narko K, Trifan OC, Wu MT, Smith E, Haudenschild C, Lane TF, Hla T: Overexpression of cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem 2001;276:18563–18569. Tsujii M, DuBois RN: Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase-2. Cell 1995;83:493–501. Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN: Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705–716. Prescott SM: Is cyclooxygenase-2 the alpha and the omega in cancer? J Clin Invest 2000;105: 1511–1513. Tsujii M, Kawano S, DuBois RN: Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA 1997;94:3336–3340. Shreedhaar V, Giese T, Sung VW, Ullrich SE: A cytokine cascade including prostaglandin E2, IL-4 and IL-10 is responsible for UV-induced systemic immune suppression. J Immunol 1998; 160:3783–3789.
Karin Müller-Decker Deutsches Krebsforschungszentrum Heidelberg B0502, Eicosanoide und epitheliale Tumorentwicklung INF 280, D–69120 Heidelberg (Germany) Tel. ⫹49 6221 424506, Fax ⫹49 6221 424406, E-Mail
[email protected]
COX-2 and Skin
89
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 90–106
The Role of COX-2 in Breast and Cervical Cancer Andrew J. Dannenberg b,c, Louise R. Howe a,c Departments of aCell and Developmental Biology and bMedicine, Weill Medical College of Cornell University, and cStrang Cancer Research Laboratory, New York, N.Y., USA
The inducible prostaglandin (PG) synthase COX-2 is strongly implicated in colorectal cancer by a wealth of evidence including epidemiological analyses and expression data. More recently, it has become apparent that COX-2 is also overexpressed in numerous other human malignancies and premalignant conditions, and may be a broadly relevant molecular target for chemopreventive and therapeutic intervention. Normally, cyclooxygenase-derived prostanoids contribute to many physiological processes including hemostasis, platelet aggregation, kidney and gastric function, reproduction, pain and fever. In particular, COX-2 is required for post-natal renal development and several female reproductive processes. Additionally, COX-2 is upregulated in response to growth factors, oncogenes and inflammatory stimuli, and thus contributes to increased PG synthesis in inflamed and neoplastic tissues. Cyclooxygenase activity is inhibited by nonsteroidal anti-inflammatory drugs (NSAIDs), such as aspirin and sulindac, which are commonly used to relieve pain and inflammation. Several epidemiological studies have reported inverse correlations between colon cancer incidence and NSAID use [1–5], suggesting that aberrant COX activity contributes to colorectal neoplasia. Consistent with this, NSAID administration reduces tumor incidence and multiplicity in rodent models of intestinal tumorigenesis, and selective COX-2 inhibitors are similarly effective in preventing intestinal tumors [6]. Two complementary genetic approaches have provided definitive evidence that COX-2 is important for tumorigenesis. Firstly, tumor formation in a mouse model of familial adenomatous polyposis is markedly reduced by genetic ablation of COX-2 [7]. Intestinal tumor incidence was reduced to 34% in COX-2 heterozygotes and to 14% in COX-2-null mice, relative to COX-2 wild-type
mice. Conversely, COX-2 overexpression is sufficient to cause breast cancer in multiparous mice [8]. Targeted expression of human COX-2 in murine mammary gland caused breast tumor formation in greater than 85% of multiparous mice. Together, these data demonstrate an important role for COX-2 in tumorigenesis, and suggest that inhibition of COX-2 may prove to be a useful strategy for preventing and possibly treating cancer. Notably, selective COX-2 inhibitors do not inhibit platelet function and have reduced gastrointestinal side effects relative to traditional NSAIDs [9, 10]. Thus, selective COX-2 inhibitors may be sufficiently safe to allow administration on a chronic basis, and thus be potentially useful agents for cancer chemoprevention. Here we review the role of COX-2 in breast and cervical cancer. COX-2 is expressed in about 40% of human breast cancers, and in mammary tumors from rodent breast cancer models. Aberrant COX-2 expression has also been detected in human cervical cancers. Strikingly, COX-2 positivity in both breast and cervical neoplasia is associated with poor patient prognosis. The mechanisms by which COX-2 contributes to cancer are considered. Additionally, we review pharmacological studies which demonstrate that NSAIDs and selective COX-2 inhibitors are effective for both prevention and treatment of breast cancer. Less is known about the potential use of COX inhibitors to prevent or treat experimental cervical cancer.
COX-2 and Breast Cancer
Expression and Epidemiology The issue of COX-2 expression in breast cancer has been somewhat controversial. Enhanced COX expression in breast cancer was first suggested by reports of elevated PG levels in breast tumors [11–13], particularly in those from patients with metastatic disease [11, 12]. PG production and COX-2 expression in breast cancer-derived cell lines was also reported [14–16]. However, preliminary analyses of human breast cancers revealed conflicting frequencies of COX-2 expression ranging from 100% (n ⫽ 13) to 5% (n ⫽ 44) [17, 18]. Clarification was provided by a recent large-scale analysis of COX-2 expression, comprised of 1,576 invasive breast cancers, of which 37.4% had moderate to strong expression of COX-2 protein detected by antibody staining [19]. COX-2 immunoreactivity localized exclusively to the cytoplasm of tumor cells, and was not elevated in stromal cells, consistent with observations in other studies [17, 20, 21]. A similar frequency of COX-2-positive breast cancers was also reported in a separate study: COX-2 protein was detected at intermediate/high levels in 43% of invasive breast cancers (n ⫽ 57) and 63% of
Breast and Cervical Cancer
91
Distant disease-free survival
100
90 Score 0 80
Score 1
70
Score 2 Score 3
60
0
1
2
3 4 5 6 Years of follow-up
7
8
Fig. 1. Disease-free survival of breast cancer patients as a function of COX-2 distribution scores. Distant disease-free survival was plotted as a function of COX-2 expression: Score 0, no COX-2 expression (n ⫽ 133); score 1, weak COX-2 expression (n ⫽ 854); score 2, moderate COX-2 expression (n ⫽ 511); score 3, strong COX-2 expression (n ⫽ 78). Elevated expression of COX-2 protein correlated with reduced survival (p ⬍ 0.0001; log-rank test). Reproduced from Ristimaki et al. [19] with permission from the American Association for Cancer Research.
ductal carcinomas in situ (n ⫽ 16) [21]. In light of these data, it seems likely that the variable frequency of COX-2 positivity in earlier studies reflects the small sample size and differences in technique used to assess COX-2 expression in these studies. In the study of Ristimaki et al. [19], COX-2 positivity correlated with decreased distant disease-free survival (fig. 1). Elevated COX-2 protein levels correlated with several parameters characteristic of aggressive breast cancer including; large tumor size, high proliferation rate, axillary node metastases, ductal histology and HER2 gene amplification (table 1). Interestingly, a correlation between COX-2 and HER2 positivity was also observed in another recent study [22]. High levels of COX-2 protein were detected in 14 of 15 HER2overexpressing breast cancers. In contrast, COX-2 was detected in only 4 of 14 HER2-negative breast cancers, and was expressed at significantly lower levels than in the HER2-positive samples. Together, these data sets suggest that COX-2 expression occurs in a significant proportion of human breast cancers, including HER2-positive cancers, and might thus represent a useful chemopreventive and therapeutic target. The fact that COX-2 is not uniformly expressed in breast cancer may have confounded epidemiological analyses of the relationship between NSAID use and breast cancer incidence. Such studies revealed an inverse correlation
Dannenberg/Howe
92
Table 1. Association of elevated COX-2 immunopositivity with clinicopathological parameters COX-2 staining positive/total
COX-2-positive %
P 2 test
Age at diagnosis, years ⬍50 ⱖ50
147/418 442/1,158
35.2 38.2
NS, 0.167
Tumor size, cm ⱕ2 cm ⬎2 cm
288/894 276/601
32.2 45.9
⬍0.0001
Axillary node status Negative Positive
331/983 231/530
33.7 43.6
⫽0.0001
Histologic grade I II III
47/183 181/492 148/284
25.7 36.8 52.1
⬍0.0001
Histologic type Ductal Lobular Special
461/1,155 74/251 48/156
39.9 29.5 30.8
⫽0.0017
ER Negative Positive
234/447 330/988
52.3 33.4
⬍0.0001
PgR Negative Positive
314/643 249/786
48.8 31.7
⬍0.0001
Tumor proliferation Ki-67 ⬍ 20% Ki-67 ⱖ 20%
277/840 264/489
33.0 54.0
⬍0.0001
p53 expression Negative/low High
391/1,070 143/242
36.5 59.1
⬍0.0001
HER2 oncogene amplification Negative Positive
425/1,157 132/262
36.7 50.4
⬍0.0001
Clinicopathological parameter
Reproduced from Ristimaki et al. [19] with permission from the American Association for Cancer Research.
Breast and Cervical Cancer
93
between colon cancer incidence and regular use of NSAIDs such as aspirin [1–5], leading to the hypothesis that aberrant PG biosynthesis could contribute to colorectal neoplasia. However, approximately 85% of human colorectal adenocarcinomas overexpress COX-2 [23–26]. Similar studies examining breast cancer statistics have reported conflicting findings. Several studies failed to find a significant relationship between aspirin use and breast cancer risk [2, 27, 28]. In contrast, other analyses revealed an association between NSAID consumption and decreased breast cancer incidence. Friedman and Ury [29] found significantly reduced breast cancer incidence in 4,867 women who used indomethacin, compared with age-matched controls. Harris et al. [30] compared NSAID use in 511 women with newly diagnosed breast cancer with 1,534 population control subjects, and found that the relative risk of breast cancer was reduced to 66% in women using NSAIDs at least 3 times per week for at least 1 year. Two additional studies also found that NSAIDs protected against breast cancer [31, 32]. In hindsight, this lack of consistency among different studies could reflect the fact that COX-2 is not ubiquitously expressed in human breast cancers. COX-2 Expression in Rodent Breast Cancers The association between COX-2 expression and breast cancer has also been evaluated using rodent models. In the rat, COX-1 is ubiquitously expressed in virgin, pregnant, lactating and post-lactational mammary glands, but COX-2 expression is only detectable during lactation [33]. In contrast, COX-2 expression is readily apparent in rat mammary tumors induced by various carcinogens [34–36]. As in human breast cancers, COX-2 protein is localized in the epithelial tumor cells [35–38]. Additionally, COX-2 expression has been analyzed in mouse breast cancer models in which mammary tumor development is induced by mammary-targeted oncogene expression. COX-2 protein is present in murine breast carcinomas induced by HER2/neu overexpression [39], consistent with the findings in HER2-positive human breast cancers [19, 22], and is also detectable in mammary tumors from Wnt1 transgenic mice [40]. Since COX-2 is expressed in rodent breast cancers induced both by carcinogens and mammary oncogenes, such systems provide useful models for translational studies of COX inhibitors. Effect of COX Inhibitors on Mammary Tumorigenesis Carcinogen-induced rat mammary tumors have been used as a model system to test various NSAIDs, including selective COX-2 inhibitors, for their chemopreventive potential (table 2). In general, indomethacin has been found to reduce the incidence and multiplicity of tumors induced by dimethylbenz[a]anthracene (DMBA) [41–44], although one study did not
Dannenberg/Howe
94
Table 2. Chemoprevention of rodent mammary tumorigenesis using COX inhibitors Reference (first author)
Tumor induction
Drug
Effects
Carter, 1983 [44]
DMBA/18% corn oil
Indomethacin
54% inhibition of tumor multiplicity Reduction in tumor incidence
McCormick, 1983 [47]
NMU
Flurbiprofen
McCormick, 1985 [42]
DMBA
Indomethacin
Abou-el-Ela, 1989 [45] Carter, 1989 [43]
DMBA DMBA/20% fat
Indomethacin Indomethacin
Noguchi, 1991 [41]
DMBA/20% corn oil
Indomethacin
Kitagawa, 1994 [85]
Piroxicam Aspirin
Mori, 1999 [48]
DMBA/20% soybean oil PhIP/high corn oil PhIP/high fat
Reduction in tumor incidence and multiplicity at low NMU dose Reduction in benign or malignant tumors according to drug administration regimen No inhibition Inhibition of tumorigenesis in animals fed 4 or 12% linoleate 61% inhibition of tumor multiplicity Reduction in tumor incidence No inhibition
Harris, 2000 [49]
DMBA
Celecoxib
86% inhibition of tumor multiplicity 68% reduction in tumor incidence
Nakatsugi, 2000 [35]
PhIP/24% corn oil
Nimesulide
54% inhibition of tumor multiplicity 28% reduction in tumor incidence
Abou-Issa, 2001 [37]
DMBA
Celecoxib
Dose-dependent reduction in tumor multiplicity, incidence and volume
Howe, 2002 [39]
HER-2/neu
Celecoxib
Reduction in tumor incidence
Suzui, 1997 [46]
Aspirin
44% inhibition of tumor multiplicity Inhibited tumor multiplicity
DMBA ⫽ Dimethylbenz[a]anthracene; NMU ⫽ N-nitrosomethyl urea; PhIP ⫽ 2-amino1-methyl-6-phenylimidazol[4,5-b]pyridine.
Breast and Cervical Cancer
95
Incidence (% rats with tumors)
Control Ibuprofen Celecoxib
100 80 60 40 20 0
30
40
50
60
70
80
90 100 110
Days post DMBA intubation
Fig. 2. Celecoxib reduces mammary tumor incidence in a rat breast cancer model. Rats were assigned to a control diet (䊏), or a diet containing 1,500 ppm ibuprofen (䊉) or 1,500 ppm celecoxib ( ) 7 days prior to a single intragastric dose of DMBA. Tumor incidence was measured for 16 weeks. Reproduced from Harris et al. [49] with permission from the American Association for Cancer Research.
detect a protective effect [45]. Two additional NSAIDs, flurbiprofen and aspirin, are also capable of reducing carcinogen-induced mammary tumorigenesis [46–48]. Recent studies have evaluated the effects of selective COX-2 inhibitors on mammary tumorigenesis. Tumor onset in DMBA-treated rats is significantly delayed by celecoxib (fig. 2). Dietary administration of 1,500 ppm celecoxib reduced incidence, multiplicity and volume of malignant breast tumors by 68, 86 and 81%, respectively, relative to the control group [49]. A follow-up study demonstrated that the chemopreventive effects of celecoxib were dosedependent over the range 250–1,500 ppm [37]. The chemopreventive properties of another COX-2 inhibitor, nimesulide, was tested in rats in which COX-2 expression and mammary tumorigenesis were induced using the environmental carcinogen 2-amino-1-methyl-6-phenylimidazol[4,5-b]pyridine in conjunction with a 24% corn oil diet [35]. A 28% reduction in tumor incidence was achieved by administration of 400 ppm nimesulide. In addition, both size and multiplicity of tumors were significantly reduced in the nimesulide-treated animals (73 and 54%, respectively). Together, these studies represent the first direct evidence that selective COX-2 inhibitors can protect against experimental breast cancer. The observed correlation between HER2 positivity and COX-2 expression [19, 22] suggests that COX-2 inhibitors might be particularly relevant for HER2positive breast cancer. To evaluate this possibility, we tested celecoxib in mice with mammary overexpression of HER2/neu. We found that HER2/neu-induced tumor onset was significantly delayed by dietary administration of 500 ppm
Dannenberg/Howe
96
Mice with tumors (%)
100
Control Celecoxib
80 60 40 20 0 0
5
10
15
20 25 30 Weeks of age
35
40
45
50
Fig. 3. Celecoxib delays the onset of HER2/neu-induced breast cancer. Female HER2/neu transgenic mice were fed control diet (䊏) or diet containing 500 ppm celecoxib ( ) from weaning onwards. Mammary tumor incidence was significantly delayed in the celecoxib-fed cohort (p ⫽ 0.003, log-rank test). Reproduced with permission from Howe et al. [39].
celecoxib (fig. 3) [39]. Our data suggest a potential application of selective COX-2 inhibitors for the prevention of HER2-positive breast cancer. Additional studies suggest that COX inhibition may also be a useful strategy for treating breast cancer. Robertson et al. [36] measured tumor size in rats that were maintained for 100 days post DMBA treatment, then fed control or ibuprofen-containing diet for an additional 5 weeks prior to sacrifice. Tumors from the control animals increased in volume by approximately 180%. In contrast, those from the ibuprofen-treated cohort decreased in volume by almost 40%. More recently, a similar study was conducted in which the effects of the selective COX-2 inhibitor celecoxib were investigated [50]. After 6 weeks of treatment, tumor volume had increased by 518% in control animals, but decreased by 32% in the group fed celecoxib. This report of regression of mammary tumors in vivo by a selective COX-2 inhibitor is consistent with earlier studies showing that various NSAIDs reduced the growth of mammary tumor xenografts [51, 52]. More recently, Kundu and Fulton [53] have shown that xenograft growth can be slowed by either a selective COX-1 or COX-2 inhibitor. These data suggest that both COX isoforms can impact on tumorigenesis, an idea which is also supported by genetic evidence in the Min mouse model of intestinal tumorigenesis [54]. Taken together, this body of work suggests that COX inhibition may represent a viable strategy for both prevention and treatment of human breast cancer. In this respect, since selective COX-2 inhibitors appear to cause fewer
Breast and Cervical Cancer
97
serious side effects than nonselective NSAIDs, selective COX-2 inhibitors offer great promise. How Does COX-2 Contribute to Cancer? Potential mechanisms by which COX-2 may contribute to tumorigenesis have recently been reviewed in detail [6], and are only discussed briefly here, with a focus on breast-related mechanisms. One important aspect of tumorigenesis in which cyclooxygenases are strongly implicated is angiogenesis. This is a crucial facet of tumorigenesis since neovascularization is required for tumors to grow beyond 2–3 mm in size. COX-2 inhibition reduces angiogenesis both in vivo and in vitro [55–59], and neovascularization of xenografted tumors is diminished in COX-2-null mice [60]. Experimental data also implicate COX-1 in angiogenesis, suggesting that both COX isoforms contribute to tumor vascularization. COX-2 overexpression in cultured cell lines has suggested several additional functions for COX-2 during tumorigenesis. COX-2-overexpressing cells demonstrate increased invasiveness, and diminished apoptosis [61, 62], both of which are considered likely to favor tumor formation. Consistent with these in vitro data, COX-2 expression also appears to suppress apoptosis in vivo. Hla and colleagues [8] have generated a transgenic mouse in which mammarytargeted COX-2 expression is driven by the mouse mammary tumor virus promoter, leading to development of breast cancers in over 85% of multiparous mice. Interestingly, mammary gland involution after weaning is delayed in this strain. Significant decreases in apoptotic cells were detected relative to those observed in wild-type mice 2 days after weaning, demonstrating that COX-2 expression in vivo can suppress apoptosis, and suggesting an additional mechanism whereby COX-2 contributes to tumorigenesis. Other likely consequences of COX-2 overexpression include PG-mediated reduction of immune surveillance and enhanced epithelial cell proliferation. PGs are known to depress proliferation of some cell types, particularly those of the immune system, and these antiproliferative effects may cause immune suppression, facilitating tumor avoidance of immune surveillance. Conversely, PGs can stimulate mitogenesis of other cell types, including mammary epithelial cells, potentially favoring tumor development. In breast tissue, PGs may also stimulate proliferation indirectly via increased estrogen biosynthesis [63]. Recently, PGE2 has been demonstrated to increase aromatase activity, and may also induce aromatase expression [64–66]. Thus, PG overproduction may lead to increased estrogen synthesis via aromatase induction. Consistent with this, a positive correlation has been observed between aromatase and COX expression in human breast cancer specimens [67, 68]. It is possible, therefore, that PG-mediated estrogen overproduction may be an
Dannenberg/Howe
98
important organ-site-specific consequence of COX-2 upregulation in breast cancer. Thus, COX-2 overexpression apparently disturbs the equilibrium between proliferation and apoptosis required for normal maintenance of organ homeostasis, thereby promoting tumorigenesis. Additionally, the involvement of COX-2 in angiogenesis and invasiveness, suggests that COX-2 also contributes to later stages of tumorigenesis. This would certainly be consistent with the observation that COX-2 positivity is associated with aggressive breast cancers [19] and with metastastic involvement in cervical cancer patients [69] (see below).
COX-2 and Cervical Cancer
Currently, there is considerable interest in the role of COX-2 in cervical neoplasia. Analysis of cervical COX-2 expression in rodents and primates suggests that normally COX-2 is absent or weakly expressed in cervix, but increases during labor and parturition [70, 71]. Increased PG production is required for cervical ripening during labor, and is most likely generated as a consequence of upregulated COX-2 expression in multiple tissues including the cervix, amnion, chorion and decidua. Human cervical cancers have been examined for COX-2 expression by several groups using immunohistochemistry, RT-PCR and Western blotting approaches. COX-2 expression has been detected in squamous cell carcinomas, adenocarcinomas, adenosquamous carcinomas and in a single case of adenoid cystic carcinoma [69, 72–78], whereas no COX-2 expression was detected in histologically normal cervix tissue [72, 75, 79]. Additionally, a single observation of COX-2 in CIN III suggests that COX-2 upregulation may occur in precancerous tissues [72]. The proportion of COX-2-positive cancers reported in the different studies ranges from 100% in 36 premenopausal stage IB-IIA cancers [73] to 43% of 84 patients with stage IB-IVA cancer [78]. Ryu et al. [69] used immunohistochemistry to demonstrate that COX-2 expression was especially strong in areas of tumor invasion, and observed an inverse correlation between COX-2 intensity and apoptosis, consistent with the ability of COX-2 to suppress apoptosis. Strikingly, COX-2 expression at the tumor invasion site was significantly elevated in tumors from patients with lymph node or parametrial involvement relative to expression in patients without metastases. Consistent with this observation, several studies have found high COX-2 expression to be associated with poor overall and disease-free survival [76–78], as has also been reported for breast cancer [19]. Elevated COX-2 expression correlated with diminished
Breast and Cervical Cancer
99
Fraction surviving
1.0
(92%, n⫽12)
0.8 p⫽0.0152
0.6 0.4
(28%, n⫽12)
0.2 0.0 0
50
100
150 Months
200
250
Fig. 4. Disease-free survival of cervical cancer patients as a function of COX-2 distribution score. Distant disease-free survival in cervical cancer patients was plotted as a function of COX-2 distribution staining. Patients were divided into those whose tumors displayed ⬍10% (solid line) and ⱖ10% (dashed line) COX-2-positive cells, respectively. The numbers in parentheses indicate 5-year survival values. Elevated expression of COX-2 protein correlated with reduced survival (p ⫽ 0.015; log-rank test). Reproduced from Gaffney et al. [76] with permission from Elsevier Science.
survival in patients treated with either radiotherapy [76] or platinum-based chemotherapy [77, 78]. For example, in the study of Gaffney et al. [76], the 5-year disease-free survival rates for tumors with low vs. high COX-2 distribution values were 92 and 28%, respectively (fig. 4). Thus, high COX-2 expression in cervical cancers is associated with the presence of metastases and reduced patient survival. Intriguingly, one study compared COX-2 expression in cervical cancers resected at different stages of the patients’ menstrual cycle, and observed that COX-2 was 3- to 3.5-fold higher in cancers removed during the proliferative phase than in those removed during the secretory phase [73]. The rationale for this approach derives from several reports linking post-surgical survival rates of breast cancer patients to the stage of menstrual cycle at time of surgery. These studies suggest that women undergoing breast cancer resection during the proliferative phase tend to have poorer survival than those operated on during the secretory phase. Further investigation will be required to determine whether surgical removal of cervical cancers at different stages of the menstrual cycle is associated with differing patient outcomes, and to evaluate the contribution of COX-2. To date, we are aware of only one study that has evaluated COX inhibitors for chemoprevention of cervical neoplasia [80]. The incidence of methylcholanthrene-induced squamous cell carcinomas in mice was reduced by
Dannenberg/Howe
100
dietary administration of the NSAID indomethacin in a dose-dependent manner. The maximal dose tested (40 mg/kg) caused a striking reduction in tumor incidence from 91 to 26%, suggesting that COX inhibition may be useful for the prevention of cervical neoplasia.
Clinical Implications
COX-2 expression has been detected in approximately 40% of human breast cancers, and also in a substantial proportion of cervical carcinomas. Strikingly, for both breast and cervical cancer, COX-2 positivity is associated with a poor patient prognosis. Consistent with this, COX-2 expression correlates with several parameters characteristic of aggressive breast cancer, and in cervical cancer patients there is an association between COX-2 and metastatic involvement. Evaluation of COX inhibitors in rodent breast cancer models suggests that NSAIDs and selective COX-2 inhibitors may be effective for the prevention of breast cancer. In particular, the enhanced safety profile of selective COX-2 inhibitors relative to traditional NSAIDs may render them appropriate for prevention in high-risk patients. This strategy has already been successfully tested in familial adenomatous polyposis patients [81]. Intriguingly, the detection of COX-2 in cervical neoplasia (CIN III) suggests COX-2 may also be a useful target for chemoprevention of cervical cancer, although analysis of a larger cohort will be required to confirm this observation. Translational studies in rodents suggest that selective COX-2 inhibitors may also be useful in a therapeutic setting. For example, COX-2 is commonly overexpressed in ductal carcinoma in situ, raising the possibility that selective COX-2 inhibitors may be useful in treating this condition. Additionally, the observed correlation between COX-2 and HER2 positivity in breast cancers raises the possibility of using COX-2 inhibitors in conjunction with HER2neutralizing antibodies such as Herceptin (Trastuzumab). This approach has previously been validated in a rodent colon cancer model, where therapy combining celecoxib and Herceptin proved more effective than either agent alone for suppressing tumor growth [82]. Interestingly, some therapeutic strategies are associated with COX-2 induction, which may decrease therapeutic efficacy. For example, both radiation and taxol can induce COX-2 [83, 84]. Therefore, simultaneous administration of a selective COX-2 inhibitor might enhance the efficacy of radiotherapy or taxol-based chemotherapy resulting in clinical benefit. Additionally, it has been hypothesized that the myalgias and arthralgias associated with taxol may be due, in part, to upregulation of COX-2 [84]. Thus, COX-2 inhibitors might also be useful to minimize the side effects associated with taxol therapy. Overall there appear to be several appropriate avenues for
Breast and Cervical Cancer
101
investigating the use of selective COX-2 inhibitors in combination with standard anti-cancer therapies.
Acknowledgements This work was funded in part by NIH grant CA089578 (to A.J.D.). The assistance of Kelly Tolle and Leslie Castelo-Soccio during manuscript preparation is gratefully acknowledged.
References 1 2 3
4
5
6 7
8
9
10
11
12 13 14 15
Suh O, Mettlin C, Petrelli NJ: Aspirin use, cancer and polyps of the large bowel. Cancer 1993;72: 1171–1177. Thun MJ, Namboodiri MM, Heath CW Jr: Aspirin use and reduced risk of fatal colon cancer. N Engl J Med 1991;325:1593–1596. Logan RF, Little J, Hawtin PG, Hardcastle JD: Effect of aspirin and non-steroidal anti-inflammatory drugs on colorectal adenomas: Case-control study of subjects participating in the Nottingham faecal occult blood screening programme. Br Med J 1993;307:285–289. Greenberg ER, Baron JA, Freeman DH Jr, Mandel JS, Haile R: Reduced risk of large-bowel adenomas among aspirin users. The Polyp Prevention Study Group. J Natl Cancer Inst 1993;85: 912–916. Reeves MJ, Newcomb PA, Trentham-Dietz A, Storer BE, Remington PL: Nonsteroidal antiinflammatory drug use and protection against colorectal cancer in women. Cancer Epidemiol Biomarkers Prev 1996;5:955–960. Howe LR, Subbaramaiah K, Brown AMC, Dannenberg AJ: Cyclooxygenase-2: A target for the prevention and treatment of breast cancer. Endocr Relat Cancer 2001;8:97–114. Oshima M, Dinchuk JE, Kargman SL, Oshima H, Hancock B, Kwong E, Trzaskos JM, Evans JF, Taketo MM: Suppression of intestinal polyposis in Apc⌬716 knockout mice by inhibition of cyclooxygenase-2. Cell 1996;87:803–809. Liu CH, Chang SH, Narko K, Trifan OC, Wu MT, Smith E, Haudenschild C, Lane TF, Hla T: Overexpression of cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem 2001;276:18563–18569. Simon LS, Weaver AL, Graham DY, Kivitz AJ, Lipsky PE, Hubbard RC, Isakson PC, Verburg KM, Yu SS, Zhao WW, Geis GS: Anti-inflammatory and upper gastrointestinal effects of celecoxib in rheumatoid arthritis: A randomized controlled trial. JAMA 1999;282:1921–1928. Langman MJ, Jensen DM, Watson DJ, Harper SE, Zhao PL, Quan H, Bolognese JA, Simon TJ: Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs. JAMA 1999;282: 1929–1933. Rolland PH, Martin PM, Jacquemier J, Rolland AM, Toga M: Prostaglandin in human breast cancer: Evidence suggesting that an elevated prostaglandin production is a marker of high metastatic potential for neoplastic cells. J Natl Cancer Inst 1980;64:1061–1070. Bennett A, Charlier EM, McDonald AM, Simpson JS, Stamford IF, Zebro T: Prostaglandins and breast cancer. Lancet 1977;ii:624–626. Tan WC, Privett OS, Goldyne ME: Studies of prostaglandins in rat mammary tumors induced by 7,12-dimethylbenz[a]anthracene. Cancer Res 1974;34:3229–3231. Gilhooly EM, Rose DP: The association between a mutated ras gene and cyclooxygenase-2 expression in human breast cancer cell lines. Int J Oncol 1999;15:267–270. Liu XH, Rose DP: Differential expression and regulation of cyclooxygenase-1 and -2 in two human breast cancer cell lines. Cancer Res 1996;56:5125–5127.
Dannenberg/Howe
102
16
17 18 19
20 21
22
23
24
25 26
27 28 29 30 31 32
33
34
35
36
Subbaramaiah K, Telang N, Ramonetti JT, Araki R, DeVito B, Weksler BB, Dannenberg AJ: Transcription of cyclooxygenase-2 is enhanced in transformed mammary epithelial cells. Cancer Res 1996;56:4424–4429. Parrett ML, Harris RE, Joarder FS, Ross MS, Clausen KP, Robertson FM: Cyclooxygenase-2 expression in human breast cancer. Int J Oncol 1997;10:503–507. Hwang D, Scollard D, Byrne J, Levine E: Expression of cyclooxygenase-1 and cyclooxygenase-2 in human breast cancer. J Natl Cancer Inst 1998;90:455–460. Ristimaki A, Sivula A, Lundin J, Lundin M, Salminen T, Haglund C, Joensuu H, Isola J: Prognostic significance of elevated cyclooxygenase-2 expression in breast cancer. Cancer Res 2002;62:632–635. Soslow RA, Dannenberg AJ, Rush D, Woerner BM, Khan KN, Masferrer J, Koki AT: COX-2 is expressed in human pulmonary, colonic and mammary tumors. Cancer 2000;89:2637–2645. Half E, Tang XM, Gwyn K, Sahin A, Wathen K, Sinicrope FA: Cyclooxygenase-2 expression in human breast cancers and adjacent ductal carcinoma in situ. Cancer Res 2002;62: 1676–1681. Subbaramaiah K, Norton L, Gerald W, Dannenberg AJ: Cyclooxygenase-2 is overexpressed in HER-2/neu-positive breast cancer. Evidence for involvement of AP-1 and PEA3. J Biol Chem 2002;277:18649–18657. Sano H, Kawahito Y, Wilder RL, Hashiramoto A, Mukai S, Asai K, Kimura S, Kato H, Kondo M, Hla T: Expression of cyclooxygenase-1 and -2 in human colorectal cancer. Cancer Res 1995;55: 3785–3789. Kutchera W, Jones DA, Matsunami N, Groden J, McIntyre TM, Zimmerman GA, White RL, Prescott SM: Prostaglandin H synthase 2 is expressed abnormally in human colon cancer: Evidence for a transcriptional effect. Proc Natl Acad Sci USA 1996;93:4816–4820. Kargman SL, O’Neill GP, Vickers PJ, Evans JF, Mancini JA, Jothy S: Expression of prostaglandin G/H synthase-1 and -2 protein in human colon cancer. Cancer Res 1995;55:2556–2559. Eberhart CE, Coffey RJ, Radhika A, Giardiello FM, Ferrenbach S, DuBois RN: Up-regulation of cyclooxygenase 2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 1994;107:1183–1188. Egan KM, Stampfer MJ, Giovannucci E, Rosner BA, Colditz GA: Prospective study of regular aspirin use and the risk of breast cancer. J Natl Cancer Inst 1996;88:988–993. Paganini-Hill A, Chao A, Ross RK, Henderson BE: Aspirin use and chronic diseases: A cohort study of the elderly. BMJ 1989;299:1247–1250. Friedman GD, Ury HK: Initial screening for carcinogenicity of commonly used drugs. J Natl Cancer Inst 1980;65:723–733. Harris RE, Namboodiri KK, Farrar WB: Nonsteroidal anti-inflammatory drugs and breast cancer. Epidemiology 1996;7:203–205. Schreinemachers DM, Everson RB: Aspirin use and lung, colon and breast cancer incidence in a prospective study. Epidemiology 1994;5:138–146. Sharpe CR, Collet JP, McNutt M, Belzile E, Boivin JF, Hanley JA: Nested case-control study of the effects of non-steroidal anti-inflammatory drugs on breast cancer risk and stage. Br J Cancer 2000;83:112–120. Badawi AF, el-Sohemy A, Stephen LL, Ghoshal AK, Archer MC: Modulation of the expression of the cyclooxygenase-1 and -2 genes in rat mammary glands: Role of hormonal status and dietary fat. Adv Exp Med Biol 1999;469:119–124. Hamid R, Singh J, Reddy BS, Cohen LA: Inhibition by dietary menhaden oil of cyclooxygenase-1 and -2 in N-nitrosomethylurea-induced rat mammary tumors. Int J Oncol 1999;14: 523–528. Nakatsugi S, Ohta T, Kawamori T, Mutoh M, Tanigawa T, Watanabe K, Sugie S, Sugimura T, Wakabayashi K: Chemoprevention by nimesulide, a selective cyclooxygenase-2 inhibitor, of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP)-induced mammary gland carcinogenesis in rats. Jpn J Cancer Res 2000;91:886–892. Robertson FM, Parrett ML, Joarder FS, Ross M, Abou-Issa HM, Alshafie G, Harris RE: Ibuprofen-induced inhibition of cyclooxygenase isoform gene expression and regression of rat mammary carcinomas. Cancer Lett 1998;122:165–175.
Breast and Cervical Cancer
103
37
38 39
40
41
42 43
44
45
46
47 48 49 50 51 52
53 54
55
56 57
Abou-Issa HM, Alshafie GA, Seibert K, Koki AT, Masferrer JL, Harris RE: Dose-response effects of the COX-2 inhibitor, celecoxib, on the chemoprevention of mammary carcinogenesis. Anticancer Res 2001;21:3425–3432. Kundu N, Yang Q, Dorsey R, Fulton AM: Increased cyclooxygenase-2 expression and activity in a murine model of metastatic breast cancer. Int J Cancer 2001;93:681–686. Howe LR, Subbaramaiah K, Patel J, Masferrer JL, Deora A, Hudis C, Thaler HT, Muller WJ, Du B, Brown AMC, Dannenberg AJ: Celecoxib, a selective cyclooxygenase-2 inhibitor, protects against human epidermal growth factor receptor 2 (HER-2)/neu-induced breast cancer. Cancer Res 2002; 62:5405–5407. Howe LR, Crawford HC, Subbaramaiah K, Hassell JA, Dannenberg AJ, Brown AMC: PEA3 is upregulated in response to Wnt1 and activates the expression of cyclooxygenase-2. J Biol Chem 2001;276:20108–20115. Noguchi M, Taniya T, Koyasaki N, Kumaki T, Miyazaki I, Mizukami Y: Effects of the prostaglandin synthetase inhibitor indomethacin on tumorigenesis, tumor proliferation, cell kinetics, and receptor contents of 7,12-dimethylbenz[a]anthracene-induced mammary carcinoma in Sprague-Dawley rats fed a high- or low-fat diet. Cancer Res 1991;51:2683–2689. McCormick DL, Madigan MJ, Moon RC: Modulation of rat mammary carcinogenesis by indomethacin. Cancer Res 1985;45:1803–1808. Carter CA, Ip MM, Ip C: A comparison of the effects of the prostaglandin synthesis inhibitors indomethacin and carprofen on 7,12-dimethylbenz[a]anthracene-induced mammary tumorigenesis in rats fed different amounts of essential fatty acid. Carcinogenesis 1989;10:1369–1374. Carter CA, Milholland RJ, Shea W, Ip MM: Effect of the prostaglandin synthetase inhibitor indomethacin on 7,12-dimethylbenz[a]anthracene-induced mammary tumorigenesis in rats fed different levels of fat. Cancer Res 1983;43:3559–3562. Abou-el-Ela SH, Prasse KW, Farrell RL, Carroll RW, Wade AE, Bunce OR: Effects of D,L2-difluoromethylornithine and indomethacin on mammary tumor promotion in rats fed high n-3 and/or n-6 fat diets. Cancer Res 1989;49:1434–1440. Suzui N, Sugie S, Rahman KM, Ohnishi M, Yoshimi N, Wakabayashi K, Mori H: Inhibitory effects of diallyl disulfide or aspirin on 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridineinduced mammary carcinogenesis in rats. Jpn J Cancer Res 1997;88:705–711. McCormick DL, Moon RC: Inhibition of mammary carcinogenesis by flurbiprofen, a non-steroidal anti-inflammatory agent. Br J Cancer 1983;48:859–861. Mori H, Sugie S, Rahman W, Suzui N: Chemoprevention of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine-induced mammary carcinogenesis in rats. Cancer Lett 1999;143:195–198. Harris RE, Alshafie GA, Abou-Issa H, Seibert K: Chemoprevention of breast cancer in rats by celecoxib, a cyclooxygenase 2 inhibitor. Cancer Res 2000;60:2101–2103. Alshafie GA, Abou-Issa HM, Seibert K, Harris RE: Chemotherapeutic evaluation of celecoxib, a cyclooxygenase-2 inhibitor, in a rat mammary tumor model. Oncol Rep 2000;7:1377–1381. Fulton AM: In vivo effects of indomethacin on the growth of murine mammary tumors. Cancer Res 1984;44:2416–2420. Karmali RA, Marsh J: Antitumor activity in a rat mammary adenocarcinoma: The effect of cyclooxygenase inhibitors and immunization against prostaglandin E2. Prostaglandins Leukot Med 1986;23:11–14. Kundu N, Fulton AM: Selective cyclooxygenase (COX)-1 or COX-2 inhibitors control metastatic disease in a murine model of breast cancer. Cancer Res 2002;62:2343–2346. Chulada PC, Thompson MB, Mahler JF, Doyle CM, Gaul BW, Lee C, Tiano HF, Morham SG, Smithies O, Langenbach R: Genetic disruption of Ptgs-1, as well as Ptgs-2, reduces intestinal tumorigenesis in Min mice. Cancer Res 2000;60:4705–4708. Jones MK, Wang H, Peskar BM, Levin E, Itani RM, Sarfeh IJ, Tarnawski AS: Inhibition of angiogenesis by nonsteroidal anti-inflammatory drugs: Insight into mechanisms and implications for cancer growth and ulcer healing. Nat Med 1999;5:1418–1423. Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN: Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705–716. Daniel TO, Liu H, Morrow JD, Crews BC, Marnett LJ: Thromboxane A2 is a mediator of cyclooxygenase-2-dependent endothelial migration and angiogenesis. Cancer Res 1999;59:4574–4577.
Dannenberg/Howe
104
58 59
60 61 62 63
64
65
66
67
68
69
70
71
72 73
74 75
76
77
Yamada M, Kawai M, Kawai Y, Mashima Y: The effect of selective cyclooxygenase-2 inhibitor on corneal angiogenesis in the rat. Curr Eye Res 1999;19:300–304. Masferrer JL, Leahy KM, Koki AT, Zweifel BS, Settle SL, Woerner BM, Edwards DA, Flickinger AG, Moore RJ, Seibert K: Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 2000;60:1306–1311. Williams CS, Tsujii M, Reese J, Dey SK, DuBois RN: Host cyclooxygenase-2 modulates carcinoma growth. J Clin Invest 2000;105:1589–1594. Tsujii M, Kawano S, DuBois RN: Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA 1997;94:3336–3340. Tsujii M, DuBois RN: Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell 1995;83:493–501. Harris RE, Robertson FM, Abou-Issa HM, Farrar WB, Brueggemeier R: Genetic induction and upregulation of cyclooxygenase (COX) and aromatase (CYP19): An extension of the dietary fat hypothesis of breast cancer. Med Hypoth 1999;52:291–292. Purohit A, Singh A, Ghilchik MW, Reed MJ: Inhibition of tumor necrosis factor ␣-stimulated aromatase activity by microtubule-stabilizing agents, paclitaxel and 2-methoxyestradiol. Biochem Biophys Res Commun 1999;261:214–217. Zhao Y, Agarwal VR, Mendelson CR, Simpson ER: Estrogen biosynthesis proximal to a breast tumor is stimulated by PGE2 via cyclic AMP, leading to activation of promoter II of the CYP19 (aromatase) gene. Endocrinology 1996;137:5739–5742. Brueggemeier RW, Richards JA, Joomprabutra S, Bhat AS, Whetstone JL: Molecular pharmacology of aromatase and its regulation by endogenous and exogenous agents. J Steroid Biochem Mol Biol 2001;79:75–84. Brueggemeier RW, Quinn AL, Parrett ML, Joarder FS, Harris RE, Robertson FM: Correlation of aromatase and cyclooxygenase gene expression in human breast cancer specimens. Cancer Lett 1999;140:27–35. Brodie AM, Lu Q, Long BJ, Fulton A, Chen T, Macpherson N, DeJong PC, Blankenstein MA, Nortier JW, Slee PH, van de Ven J, van Gorp JM, Elbers JR, Schipper ME, Blijham GH, Thijssen JH: Aromatase and COX-2 expression in human breast cancers. J Steroid Biochem Mol Biol 2001;79:41–47. Ryu HS, Chang KH, Yang HW, Kim MS, Kwon HC, Oh KS: High cyclooxygenase-2 expression in stage IB cervical cancer with lymph node metastasis or parametrial invasion. Gynecol Oncol 2000;76:320–325. Dong YL, Gangula PR, Fang L, Yallampalli C: Differential expression of cyclooxygenase-1 and -2 proteins in rat uterus and cervix during the estrous cycle, pregnancy, labor and in myometrial cells. Prostaglandins 1996;52:13–34. Wu WX, Ma XH, Smith GC, Koenen SV, Nathanielsz PW: A new concept of the significance of regional distribution of prostaglandin H synthase 2 throughout the uterus during late pregnancy: Investigations in a baboon model. Am J Obstet Gynecol 2000;183:1287–1295. Kulkarni S, Rader JS, Zhang F, Liapis H, Koki AT, Masferrer JL, Subbaramaiah K, Dannenberg AJ: Cyclooxygenase-2 is overexpressed in human cervical cancer. Clin Cancer Res 2001;7:429–434. Formenti S, Felix J, Salonga D, Danenberg K, Pike MC, Danenberg P: Expression of metastasesassociated genes in cervical cancers resected in the proliferative and secretory phases of the menstrual cycle. Clin Cancer Res 2000;6:4653–4657. Ishiko O, Sumi T, Yoshida H, Tokuyama O, Wakasa K, Haba T, Ogita S: Cyclooxygenase-2 expression in an adenoid cystic carcinoma of the uterine cervix. Oncol Rep 2001;8:1023–1025. Sales KJ, Katz AA, Davis M, Hinz S, Soeters RP, Hofmeyr MD, Millar RP, Jabbour HN: Cyclooxygenase-2 expression and prostaglandin E2 synthesis are up-regulated in carcinomas of the cervix: A possible autocrine/paracrine regulation of neoplastic cell function via EP2/EP4 receptors. J Clin Endocrinol Metab 2001;86:2243–2249. Gaffney DK, Holden J, Davis M, Zempolich K, Murphy KJ, Dodson M: Elevated cyclooxygenase-2 expression correlates with diminished survival in carcinoma of the cervix treated with radiotherapy. Int J Radiat Oncol Biol Phys 2001;49:1213–1217. Ishiko O, Sumi T, Yoshida H, Matsumoto Y, Honda K, Deguchi M, Yamada R, Ogita S: Association between overexpression of cyclooxygenase-2 and suppression of apoptosis in
Breast and Cervical Cancer
105
78
79
80 81
82
83 84
85
advanced cancer of the uterine cervix after cyclic balloon-occluded arterial infusion. Oncol Rep 2001;8:1259–1263. Ferrandina G, Lauriola L, Distefano MG, Zannoni GF, Gessi M, Legge F, Maggiano N, Mancuso S, Capelli A, Scambia G, Ranelletti FO: Increased cyclooxygenase-2 expression is associated with chemotherapy resistance and poor survival in cervical cancer patients. J Clin Oncol 2002;20: 973–981. Sales KJ, Katz AA, Howard B, Soeters RP, Millar RP, Jabbour HN: Cyclooxygenase-1 is up-regulated in cervical carcinomas: autocrine/paracrine regulation of cyclooxygenase-2, prostaglandin E receptors, and angiogenic factors by cyclooxygenase-1. Cancer Res 2002;62: 424–432. Rao AR, Hussain SP: Modulation of methylcholanthrene-induced carcinogenesis in the uterine cervix of mouse by indomethacin. Cancer Lett 1988;43:15–19. Steinbach G, Lynch PM, Phillips RK, Wallace MH, Hawk E, Gordon GB, Wakabayashi N, Saunders B, Shen Y, Fujimura T, Su LK, Levin B: The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–1952. Mann M, Sheng H, Shao J, Williams CS, Pisacane PI, Sliwkowski MX, DuBois RN: Targeting cyclooxygenase 2 and HER-2/neu pathways inhibits colorectal carcinoma growth. Gastroenterology 2001;120:1713–1719. Milas L: Cyclooxygenase-2 enzyme inhibitors as potential enhancers of tumor radioresponse. Semin Radiat Oncol 2001;11:290–299. Subbaramaiah K, Hart JC, Norton L, Dannenberg AJ: Microtubule-interfering agents stimulate the transcription of cyclooxygenase-2. Evidence for involvement of ERK1/2 AND p38 mitogenactivated protein kinase pathways. J Biol Chem 2000;275:14838–14845. Kitagawa H, Noguchi M: Comparative effects of piroxicam and esculetin on incidence, proliferation, and cell kinetics of mammary carcinomas induced by 7,12-dimethylbenz[a]anthracene in rats on high- and low-fat diets. Oncology 1994;51:401–410.
Louise R. Howe, PhD Strang Cancer Research Laboratory Rockefeller University, Box 231 1230 York Avenue, New York, NY 10021 (USA) Tel. ⫹1 212 7340567/ext 221, Fax ⫹1 212 4729471, E-Mail
[email protected]
Dannenberg/Howe
106
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 107–123
Cyclooxygenase-2: A Target for the Prevention and Treatment of Cancers of the Upper Digestive Tract Nasser K. Altorki a, Kotha Subbaramaiah b, Andrew J. Dannenberg b Departments of aCardiothoracic Surgery and bMedicine, Weill Medical College of Cornell University and Strang Cancer Prevention Center, New York, N.Y., USA
Cancers of the upper digestive tract including the head and neck, esophagus and stomach are among the most lethal malignancies worldwide. Approximately 40,000 new cases of head and neck cancer were reported in the USA in 2001 [1]. Despite advances in radiotherapy and chemotherapy, the survival of patients with head and neck squamous cell cancer (HNSCC) has not improved significantly over the past three decades. Esophageal cancer is the third most common gastrointestinal malignancy and ranks among the ten most common cancers worldwide [2]. Approximately 12,000 new cases are diagnosed annually in the USA and the overall survival rate is in the range of 5–10% [1]. The majority of patients are incurable by current treatment modalities including surgical resection, chemotherapy, radiotherapy or any combination thereof. Gastric cancer is the second leading cause of cancer deaths in the world. Nearly 750,000 cases are detected annually worldwide and overall 5-year survival is less than 25% [3]. Given the poor prognosis associated with cancers of the upper digestive tract, new preventive and/or treatment strategies are needed. In this chapter, evidence is presented that COX-2, an inducible enzyme that catalyzes the synthesis of prostaglandins (PGs), represents a potential therapeutic target [4]. PGs appear to be important in the pathogenesis of cancer because they modulate cell proliferation, apoptosis, angiogenesis and immune surveillance [5–10]. Moreover, a variety of cancers including those of the upper digestive tract form more PGs than the normal tissues from which they arise [11–14]. The notion that PGs contribute to carcinogenesis is supported by epidemiological and
experimental evidence that nonsteroidal anti-inflammatory drugs (NSAIDs), prototypic inhibitors of PG biosynthesis, reduce the risk of tumorigenesis [15–25].
Evidence that NSAIDs Reduce Tumorigenesis
Numerous epidemiological studies suggest that NSAIDs including aspirin reduce cancer risk [15–20]. The association between NSAID use and risk of esophageal cancer has been evaluated [17, 18, 20]. Thun et al. [20] observed a reduced risk of esophageal cancer among individuals using aspirin at least 16 times per month. More recently, Farrow et al. [17] found that the use of aspirin or other NSAIDs was associated with a reduced risk of esophageal squamous cell carcinoma, esophageal adenocarcinoma and noncardia gastric adenocarcinoma. Similarly, a study by the American Cancer Society [20] reported a relative risk of 0.5 for gastric cancer among those using aspirin regularly. To the best of our knowledge, epidemiological studies have not been carried out to assess whether use of NSAIDs alters the risk of developing HNSCC. In addition to the epidemiological evidence, experimental studies in animals have consistently shown that NSAIDs protect against cancers of the digestive tract [21–25]. For example, indomethacin protected against chemicallyinduced esophageal cancer [24]. NSAIDs have also been observed to decrease the incidence of experimental HNSCC [22].
Synthesis of Prostaglandins
The best known effect of aspirin and other NSAIDs is the inhibition of PG synthesis. The precursor of PGs is arachidonic acid, a 20-carbon polyunsaturated fatty acid. The first step in the synthesis of PGs is the hydrolysis of phospholipids to produce free arachidonate, a reaction catalyzed by phospholipase A2. Molecular oxygen is then added to arachidonic acid in a reaction catalyzed by the cyclooxygenase activity of COX. This reaction produces an unstable product, PGG2. PGG2 is converted rapidly to PGH2 by the peroxidase activity of COX. PGH2 is the common precursor of all other prostanoids in reactions catalyzed by distinct, specific synthases. Each of the products derived from PGH2 has its own range of biologic activities [26]. The types and amounts of PGs and thromboxanes formed by COX-catalyzed oxidation of arachidonate are highly variable in different cell types at least, in part, due to the differences in the composition of the distal synthases.
Altorki/Subbaramaiah/Dannenberg
108
Distinct genes encode COX-1 and COX-2, the two known isoforms of COX. COX-1 is constitutively expressed in most tissues where it mediates PG synthesis necessary for normal physiologic functions including platelet aggregation, regulation of glomerular blood flow and cytoprotection in the gastrointestinal mucosa. The inducible isoform, COX-2, is regulated by a variety of factors [27–33]. Growth factors, oncogenes and tumor suppressor genes (P53) can modulate the expression of COX-2. For example, stimulation of epidermal growth factor receptor (EGFR), a receptor that is commonly overexpressed in cancers of the digestive tract, induces COX-2 expression by a MAP kinasedependent mechanism [34]. Activation of protein kinase C signaling also induces COX-2 in epithelial cells [35]. Conversely, it has been shown that wildtype but not mutant P53 inhibits COX-2 transcription [36]. P53 mutations are commonly detected in cancers of the head and neck, esophagus and stomach suggesting that P53 status could be an important determinant of COX-2 expression in these tumors [37]. In support of this notion, higher levels of COX-2 have been observed in human tumors containing P53 mutations including cancers of the stomach and esophagus [38, 39]. The COX-2 gene is also induced by exogenous carcinogens contained in tobacco smoke as well as by endogenous tumor promoters such as bile acids [31, 40]. Cigarette smoking is one of the major risk factors for HNSCC as well as carcinoma of the esophagus. Benzo[a]pyrene (B[a]P), a constituent of tobacco smoke and charbroiled food, induces COX-2 in cultured cells [31]. Reflux of bile acids into the esophagus has been implicated in the pathogenesis of esophageal adenocarcinoma [41, 42]. Interestingly, treatment of esophageal cells with either conjugated or unconjugated bile acids induces COX-2 (fig. 1) [43]. Bile acids induce COX-2 via both EGFR and protein kinase C-mediated signaling pathways [40, 43, 44].
Preclinical Evidence That COX Contributes to Carcinogenesis
Several lines of evidence suggest that COX-2 contributes to carcinogenesis. Increased amounts of COX-2 are commonly found in cancers of the head and neck, esophagus and stomach [45–47]. In addition, overexpression of COX-2 has been observed in premalignant conditions including oral leukoplakia, squamous dysplasia of the esophagus, Barrett’s esophagus and gastric metaplasia [45, 47–50]. Helicobacter pylori infection, a risk factor for gastric cancer, has been linked to increased COX-2 expression [48]. In fact, COX-2 expression was reduced after successful eradication of H. pylori [48]. There is extensive evidence, beyond the finding that COX-2 is commonly overexpressed in neoplastic tissue, to suggest that COX-2 is mechanistically linked to
COX-2 and Cancers of the Upper Digestive Tract
109
CD 200 M
CD 100 M
CD 50 M
CD 25 M
Control
GCD 1,250 M
GCD 1,000 M
GCD 750 M
GCD 500 M
Control COX-2 mRNA
18S rRNA
Fig. 1. Conjugated and unconjugated bile acids induce COX-2 mRNA in esophageal cells. A cell line derived from squamous carcinoma of the esophagus was treated with vehicle, glycochenodeoxycholate (GCD, 500–1250 M) or chenodeoxycholate (CD, 25–200 M) for 3 h. Total cellular RNA was isolated and 10 g of RNA was added to each lane. The Northern blot was hybridized sequentially with probes that recognized COX-2 mRNA and 18S ribosomal RNA, respectively. Reprinted from Zhang et al. [43] with permission from Elsevier Science.
the development of cancer. The most specific data supporting a cause-and-effect connection between overexpression of COX-2 and carcinogenesis comes from genetic studies. In one study, knocking out the COX-2 gene led to a marked reduction in the number and size of intestinal polyps in APC⌬716 mice, an animal model for familial adenomatous polyposis (FAP) [51]. In another study, mammary cancer developed in multiparous female transgenic mice that overexpressed the COX-2 gene in mammary glands [52]. Similar genetic studies have not been reported in animal models of upper digestive tract cancers.
Mechanisms by Which COX-2 Contributes to Carcinogenesis
COX-2 can affect multiple mechanisms that are important in carcinogenesis, which makes it an attractive therapeutic target. These mechanisms are described below. Xenobiotic Metabolism COX-2 is a bifunctional enzyme that has both peroxidase and cyclooxygenase activities. The peroxidase activity catalyzes the conversion of procarcinogens to carcinogens [53, 54]. In the liver, these kinds of oxidative reactions are catalyzed principally by cytochrome P450s. Extrahepatic tissues such as
Altorki/Subbaramaiah/Dannenberg
110
COX P450
Epoxide hydrolase
P450
HO
O
HO OH
Benzo[a]pyrene (B[a]P)
B[a]P-7,8-epoxide
O
B[a]P-7,8-diol
OH B[a]P-7,8-diol-9,10-epoxide
Fig. 2. Schematic illustration of the metabolism of benzo[a]pyrene. COX can convert benzo[a]pyrene-7,8-diol to the ultimate carcinogen benzo[a]pyrene-7,8-diol-9,10-epoxide. Reprinted from Dannenberg et al. [4] with permission from the Lancet Publishing group.
those of the upper digestive tract, however, frequently have low concentrations of P450s [55] and therefore significant amounts of xenobiotics may be co-oxidized to mutagens by the peroxidase activity of COX. In addition to catalyzing the synthesis of mutagens, COX-2 can be induced by procarcinogens. As noted above, B[a]P, a procarcinogen in tobacco smoke, stimulates COX-2 transcription. In turn, COX-2 catalyzes the oxidation of B[a]P-7,8-diol to B[a]P-7,8-diol-9,10-epoxide, which is a highly reactive and strongly mutagenic compound (fig. 2) [56]. This raises the possibility that B[a]P-mediated induction of COX-2 facilitates its own conversion to B[a]P-7, 8-diol-9,10-epoxide, thereby amplifying the effect of a given dose of B[a]P on tumor initiation. The ability of COX to convert procarcinogens in tobacco smoke to carcinogens suggests a potential role for selective COX-2 inhibitors in preventing tobacco smoke-related DNA damage. Apoptosis Apoptosis, or programmed cell death, has been observed to decrease during carcinogenesis. Forced overexpression of COX-2 in epithelial cells inhibited apoptosis [9]. This effect was attributed, at least in part, to increased levels of the antiapoptotic protein Bcl-2. In a related study, treatment with PGE2, a COX-2 derived eicosanoid, induced Bcl-2 in cultured cells [10]. Recently, increased levels of Bcl-2 and reduced amounts of Bax, a proapoptotic protein, were detected in tumors derived from COX-2 overexpressing transgenic mice [52]. Consistent with these findings, NSAIDs and selective COX-2 inhibitors have been shown to induce apoptosis in a variety of experimental test systems including esophageal cancer cell lines [9, 57–60]. Taken together, these data show a clear causal link between expression of COX-2 and inhibition of programmed cell death. Possibly, up-regulation of COX-2 prolongs the survival of abnormal cells and thereby favors the accumulation of sequential genetic changes that increase the risk of tumorigenesis.
COX-2 and Cancers of the Upper Digestive Tract
111
Inflammation and Immunosuppression Chronic inflammation is a recognized risk factor for epithelial carcinogenesis [61]. Cytokine-mediated up-regulation of COX-2 contributes to increased synthesis of PGs in inflamed tissues. The findings discussed above, therefore, provide a basis for a cause-and-effect link between chronic inflammation and carcinogenesis via overexpression of COX-2. Chronic inflammation may be particularly important in the pathogenesis of esophageal and gastric cancer. Endemic atrophic esophagitis, presumably resulting from dietary deficiencies or ingestion of mutagens, is a known risk factor for squamous cell carcinoma of the esophagus in the Middle and Far East [62, 63]. In the western hemisphere, reflux esophagitis and Barrett’s esophagus are well-recognized risk factors for the development of esophageal adenocarcinoma [64, 65]. There is considerable evidence that the concentration of bile acids in the esophageal refluxate correlates with the degree of mucosal injury in patients with gastroesophageal reflux disease [66, 67]. In experimental animals, the surgical creation of duodeno-esophageal reflux led to severe esophagitis, marked thickening of the esophageal mucosa, and enhanced expression of COX-2 [43]. The latter effect is likely to reflect, in part, the ability of bile acids to induce COX-2 (fig. 1). H. pylori plays a central role in the etiology of chronic superficial gastritis and peptic ulcer disease. H. pylori infection is associated with an increased risk of gastric cancer but the underlying mechanism is incompletely understood [48]. The fact that H. pylori induces COX-2 may be important for understanding the link between chronic gastritis and gastric cancer. PGs are believed to contribute to tumor-related immune suppression [68, 69]. Colony-stimulating factors released by tumor cells activate monocytes and macrophages to synthesize PGE2, which inhibits the production of immune regulatory lymphokines, T- and B-cell proliferation, and the cytotoxic activity of natural killer cells, thus favoring tumor growth [70, 71]. PGE2 also inhibits the production of tumor necrosis factor-␣ and induces the production of IL-10, an immunosuppressive cytokine [72, 73]. Dubinett et al. [74] have demonstrated that abrogation of COX-2 expression promotes antitumor reactivity by restoring the balance of IL-10 and IL-12 in vivo. Importantly, inhibitors of COX activity including selective COX-2 inhibitors have been shown to attenuate tumor-mediated immune suppression [74, 75]. Angiogenesis Any significant increase in tumor mass must be preceded by an increase in vascular supply to deliver nutrients and oxygen to the tumor [76]. Recently, levels of COX-2 were found to correlate with both levels of VEGF and tumor vascularization in HNSCC [77]. In gastric cancer, overexpression of COX-2
Altorki/Subbaramaiah/Dannenberg
112
was also reported to be associated with enhanced angiogenesis [13]. These findings in human tissues are consistent with prior evidence that overexpression of COX-2 in epithelial cells led to enhanced production of vascular growth factors and the formation of capillary-like networks [78, 79]. Several other studies also link COX-2 and angiogenesis. In one study, tumor growth was attenuated in COX-2 knockout mice compared with wildtype or COX-1 knockout mice [79]. Knocking out or pharmacologically inhibiting COX-2 led to decreased production of VEGF, a result that could explain the reduction in microvessel density and tumor growth in COX-2 knockout mice. In another study, celecoxib, a selective COX-2 inhibitor, blocked basic fibroblast growth factor induced angiogenesis in the rat cornea [80]. Finally, Nishimura et al. [81] showed that a selective COX-2 inhibitor suppressed angiogenesis in association with reduced growth of HNSCC xenografts. Invasion and Metastasis Human cancer cells engineered to overexpress COX-2 produce increased amounts of PGs and become more invasive [82]. Biochemical changes associated with enhanced invasiveness include increased amounts of mRNA for membrane-type metalloproteinase-1 and activation of metalloproteinase-2. These enzymes digest the collagen matrix of the basement membrane stimulating the motile and invasive phenotype of cancer cells. In another study, forced overexpression of COX-2 was associated with increased amounts of CD44, the cell surface receptor for hyaluronate, and specific blockade of CD44 significantly decreased the invasiveness of tumor cells [83]. Consistent with these in vitro findings, selective COX-2 inhibitors have been observed to suppress metastases in experimental animals [80, 84].
COX-2 as a Target for Prevention and Treatment of Cancer
There is excellent experimental evidence that targeting the COX-2 enzyme is an effective strategy for the prevention or treatment of tumors in animal models [85–93]. Selective COX-2 inhibitors reduced the formation, growth and metastases of a variety of experimental tumors including those of the head and neck, esophagus and stomach. Shiotani et al. [85] showed that treatment with nimesulide, a selective inhibitor of COX-2, decreased the incidence and multiplicity of chemically-induced tongue carcinoma in rats (table 1). JTE-522, a selective COX-2 inhibitor, significantly reduced the incidence and multiplicity of NMBA-induced esophageal cancer [92]. Recently, the selective COX-2 inhibitor MF-tricyclic was found to reduce the relative risk of esophageal cancer by 55% in an animal model of Barrett’s esophagus (fig. 3) [93].
COX-2 and Cancers of the Upper Digestive Tract
113
Table 1. Nimesulide inhibits the development of tongue SCC initiated by 4-NQO Treatment
Rats n
Incidence (%)
Multiplicity (rats, n)
4-NQO alone 4-NQO ⫹ 150 ppm NIM 4-NQO ⫹ 300 ppm NIM
11 12 13
9 (81.8) 4 (33.3)a 1 (7.7)b
1.00 ⫾ 0.77 0.33 ⫾ 0.49a 0.08 ⫾ 0.28b
p ⬍ 0.05, bp ⫽ 0.01. Adapted from Shiotani et al. [85]. a
60 18/35
Percent animal
50 40 30 8/35* 20 4/35** 10 51%
23%
11%
Control
MF-tricyclic
Sulindac
0
Fig. 3. Treatment with a selective COX-2 inhibitor or a nonselective NSAID reduces the incidence of esophageal adenocarcinoma. Rats were subjected to esophagojejunostomy to induce Barrett’s esophagus and esophageal cancer. The ability of MF-tricyclic, a selective COX-2 inhibitor, or sulindac, a dual inhibitor of COX-1/COX-2, to inhibit tumorigenesis was assessed. The incidence of esophageal cancer (%) is represented by gray bars. The risk of esophageal cancer is significantly lower in the MF-tricyclic-treated group (p ⫽ 0.013,*) and the sulindac-treated group (p ⬍ 0.001,**) than in the control group. Reprinted from Buttar et al. [93] with permission from Elsevier Science.
In addition to possessing chemopreventive properties, selective COX-2 inhibitors are potentially useful for treating established tumors. Treatment with either JTE-522 or celecoxib reduced the growth of HNSCC xenografts in nude mice [81]. Similarly, the use of another selective COX-2 inhibitor, NS-398, suppressed the growth of gastric cancer xenografts, at least in part, by inducing apoptosis [94]. Importantly, there is also evidence that selective COX-2
Altorki/Subbaramaiah/Dannenberg
114
inhibitors can augment the antitumor effects of chemotherapy or radiotherapy [95, 96].
Use of Selective COX-2 Inhibitors to Prevent or Treat Human Cancers
As detailed above, there is excellent evidence that selective COX-2 inhibitors suppress tumorigenesis in experimental animals. To be useful in humans, a chemoprentive agent must have an acceptable safety profile in addition to being effective. Because peptic ulcer disease is a major side effect of traditional NSAIDs, it is noteworthy that selective COX-2 inhibitors appear to cause less injury to the mucosa of the upper gastrointestinal tract than classical NSAIDs [97, 98]. The first human trial to evaluate the anticancer properties of a selective COX-2 inhibitor has been completed [99]. This study was conducted in FAP patients because of the strength of the preclinical data and prior evidence that sulindac, an NSAID, reduced the number of colorectal polyps in this patient population [100]. Treatment with celecoxib 400 mg bid for 6 months caused a 28% reduction in the number of colorectal polyps compared to a 4.5% reduction for placebo (p ⫽ 0.003). Similarly, the total polyp burden was significantly reduced in subjects receiving celecoxib 400 mg bid. Based on the results of this study, celecoxib was approved by the FDA as adjunctive therapy for patients with FAP. Recently, a pilot study was carried out in 12 patients with Barrett’s esophagus [101]. Ten days of therapy with rofecoxib, a selective COX-2 inhibitor, led to a significant reduction in PGE2 release and cell proliferation in Barrett’s esophagus. Additional studies are in progress to determine whether selective COX-2 inhibitors are effective in the treatment of oral premalignant lesions and dysplastic Barrett’s esophagus [19]. A major objective of these clinical trials is to determine whether a selective COX-2 inhibitor induces regression of either oral premalignant lesions or dysplasia of the esophagus. Another trial is underway to determine whether celecoxib can prevent the recurrence of Barrett’s esophagus after thermal ablation. Given the frequent need for surgical intervention in these conditions, identification of a pharmacological approach to cause either regression or stabilization of disease would represent a significant clinical advance. Although the potential chemopreventive properties of selective COX-2 inhibitors are being actively investigated, less is known about the utility of these agents in the treatment of cancer. Importantly, there is evidence that the extent of COX-2 expression in tumors correlates with various clinicopathologic parameters, including patient prognosis. A correlation has been observed between the level of COX-2 expression and disease stage and/or patient survival in
COX-2 and Cancers of the Upper Digestive Tract
115
Probability of overall survival
1.0
0.8
COX-2 low
0.6
COX-2 high 0.4
0.2
0 0
20
40
60
80
100
Survival (months)
Fig. 4. Elevated COX-2 expression in esophageal adenocarcinoma is associated with reduced overall survival. Kaplan-Meier curves of 145 patients with esophageal adenocarcinoma arising in Barrett’s esophagus. There were 30 patients with low COX-2 expression and 115 with high COX-2 expression. A statistically significant difference was observed between the two groups (p ⫽ 0.002; log rank test). Reprinted from Buskens et al. [106] with permission from Elsevier Science.
a variety of tumors including cancers of the esophagus and stomach [102–106]. In gastric adenocarcinoma, elevated expression of COX-2 was found to be associated with enhanced lymphatic invasion and metastasis [104, 105]. Increased expression of COX-2 protein in esophageal adenocarcinomas was significantly associated with nodal metastases and reduced survival (fig. 4) [106]. The results of these correlative studies support the goal of initiating clinical trials to determine whether there is a role for selective COX-2 inhibitors as adjuvant therapy for cancer.
Future Directions
Although major progress has been made in understanding the link between COX-2 and carcinogenesis, many unanswered questions remain. To begin with, it will be important to establish whether selective COX-2 inhibitors are effective in the treatment of premalignant lesions. The results of trials already underway in patients with oral leukoplakia and Barrett’s esophagus should provide
Altorki/Subbaramaiah/Dannenberg
116
important insights. Similarly, it would be worthwhile to determine whether selective COX-2 inhibitors are useful in the management of premalignant gastric lesions. Because chemopreventive agents may require prolonged use, the safety profile of selective COX-2 inhibitors needs to be established in each of these patient populations. Patients with a history of HNSCC are at increased risk for second primary tumors. Increased amounts of COX-2 have been detected in premalignant lesions throughout the upper aerodigestive tract. Hence, the potential role of selective COX-2 inhibitors in decreasing the incidence of second primary tumors should be investigated in this patient population. Whether selective COX-2 inhibitors can delay the recurrence of tumors after surgical resection also remains a significant unanswered question. In addition to COX-2, other potential pharmacological targets have been identified and more will continue to be discovered in the years ahead. Combination therapy, a common strategy in cancer treatment, may be equally applicable in chemoprevention. Low doses of combinations of agents may be more effective than either agent alone, and with less toxicity. This notion is supported by the recent findings of Torrance et al. [107]. These investigators showed that combining an NSAID with a novel inhibitor of EGFR kinase was more effective than either agent alone in reducing intestinal tumor formation in experimental animals. To the best of our knowledge, comparable studies utilizing combinations of agents have not been done in experimental models of upper digestive tract cancers. Retinoids are active in patients with oral leukoplakia although toxicity has been problematic [108]. The transcriptional activation of COX-2 is blocked by retinoic acid in oral epithelial cells [34, 35]. If a selective COX-2 inhibitor proves to be effective in the ongoing oral leukoplakia trial, a future study could investigate whether a selective COX-2 inhibitor combined with a retinoid is more effective than either agent alone. Importantly, this type of approach might permit a lower dose of retinoid to be used to avoid toxic effects. Clearly, this type of strategy can be tested in an experimental model of HNSCC before being evaluated in humans. Thus far, major emphasis has been placed on evaluating the role of selective COX-2 inhibitors in preventing cancer. It is important to emphasize, therefore, that there is growing interest in determining whether these agents are also useful in treating cancer. In most preclinical studies, selective COX-2 inhibitors reduced the growth rate of established tumors rather than causing tumor regression [80, 81, 94]. This strongly suggests that selective COX-2 inhibitors will need to be given in combination with standard cytotoxic therapy. Several studies in experimental systems have demonstrated that co-treatment with a selective COX-2 inhibitor can augment the efficacy of chemotherapy or radiotherapy [95, 96]. Whether the same will prove true in the treatment of cancers of the head and neck, esophagus and stomach in humans awaits investigation.
COX-2 and Cancers of the Upper Digestive Tract
117
Finally, additional mechanistic studies are needed. Overexpression of COX-2 can inhibit apoptosis or stimulate angiogenesis, but the relative importance of these effects compared with the immunosuppressive effects of PGs is uncertain. The fact that COX-2 converts tobacco procarcinogens to carcinogens could be extremely important for developing strategies to prevent tobacco smoke-induced carcinogenesis. More work is needed to relate these preclinical findings to altering the risk of cancers of the upper digestive tract in humans.
Acknowledgment Support from National Institutes of Health grant CA82578 is acknowledged.
References 1 2
3
4
5 6 7 8
9 10 11 12 13 14 15
Greenlee RT, Hill-Harmon MB, Murray T, Thun M: Cancer Statistics. CA Cancer J Clin 2001;51:15–36. Schrump DS, Altorki NK, Forastiere A, Minsky BD: Cancer of the esophagus; in Devita VT, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology. Philadelphia, Lippincott Williams & Wilkins, 2001, pp 1051–1091. Karpeh MS, Kelsen DP, Tepper JE: Cancer of the stomach; in Devita VT, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology. Philadelphia, Lippincott Williams & Wilkins 2001, pp 1092–1126. Dannenberg AJ, Altorki NK, Boyle JO, Dang C, Howe LR, Weksler BB, Subbaramaiah K: Cyclo-oxygenase-2: A pharmacological target for the prevention of cancer. Lancet Oncol 2001; 2:544–551. Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori H, DuBois RN: Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705–716. Balch CM, Dougherty PA, Cloud GA, Tilden AB: Prostaglandin E2-mediated suppression of cellular immunity in colon cancer patients. Surgery 1984;95:71–77. Goodwin JS, Ceuppens JL: Regulation of immune response by prostaglandins. J Clin Immunol 1983;3:295–315. Goodwin JS, Bankhurst AD, Messner RP: Suppression of human T-cell mitogenesis by prostaglandin: Existence of a prostaglandin-producing suppressor cell. J Exp Med 1977;146: 1719–1734. Tsujii M, DuBois RN: Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase-2. Cell 1995;83:493–501. Sheng H, Shao J, Morrow JD, Beauchamp RD, DuBois RN: Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells. Cancer Res 1998;58:362–366. Bennett A, Carter RL, Stamford IF, Tanner NS: Prostaglandin-like material extracted from squamous cell carcinomas of the head and neck. Br J Cancer 1980;41:204–208. Jung TTK, Berlinger NT, Juhn SK: Prostaglandins in squamous cell carcinoma of the head and neck: A preliminary study. Laryngoscope 1985;95:307–312. Uefuji K, Ichikura T, Mochizuki H: Cyclooxygenase-2 expression is related to prostaglandin biosynthesis and angiogenesis in human gastric cancer. Clin Cancer Res 2000;6:135–138. Rigas B, Goldman IS, Levine L: Altered eicosanoid levels in human colon cancer. J Lab Clin Med 1993;122:518–523. Thun MJ, Namboodiri MM, Heath CW Jr: Aspirin use and reduced risk of fatal colon cancer. N Engl J Med 1991;325:1593–1596.
Altorki/Subbaramaiah/Dannenberg
118
16 17
18
19 20 21
22 23 24 25 26 27
28
29
30
31
32
33
34
35
36
Schreinemachers DM, Everson RB: Aspirin use and lung, colon, and breast cancer incidence in a prospective study. Epidemiology 1994;5:138–146. Farrow DC, Vaughan TL, Hansten PD, Stanford JL, Risch HA, Gammon MD, Chow WH, Dubrow R, Ahsan H, Mayne ST, Schoenberg JB, West AB, Rotterdam H, Fraumeni JF, Blot WJ: Use of aspirin and other nonsteroidal anti-inflammatory drugs and risk of esophageal and gastric cancer. Cancer Epidemiol Biomarkers Prev 1998;7:97–102. Langman MJS, Cheng KK, Gilman EA, Lancashire RJ: Effect of anti-inflammatory drugs on overall risk of common cancer: Case-control study in general practice research database. BMJ 2000;320:1642–1646. Thun MJ, Henley SJ, Patrono C: Nonsteroidal anti-inflammatory drugs as anticancer agents: Mechanistic, pharmacologic and clinical issues. J Natl Cancer Inst 2002;94:252–266. Thun MJ, Namboodiri MM, Calle EE, Flanders WD, Heath CW: Aspirin use and risk of fatal cancer. Cancer Res 1993;53:1322–1327. Reddy BS, Maruyama H, Kelloff G: Dose-related inhibition of colon carcinogenesis by dietary piroxicam, a nonsteroidal anti-inflammatory drug, during different stages of rat colon tumor development. Cancer Res 1987;47:5340–5346. Cornwall H, Odukoya O, Shklar G: Oral mucosal tumor inhibition by ibuprofen. J Oral Maxillofac Surg 1983;41:795–800. Castonguay A, Rioux N: Inhibition of lung tumorigenesis by sulindac: Comparison of two experimental protocols. Carcinogenesis 1997;18:491–496. Rubio CA: Antitumoral activity of indomethacin on experimental esophageal tumors. J Natl Cancer Inst 1984;72:705–707. Rubio CA: Further studies on the therapeutic effect of indomethacin on esophageal tumors. Cancer 1986;58:1029–1031. Kerins DM, Murray R, Fitzgerald GA: Prostacyclin and prostaglandin E1: Molecular mechanisms and therapeutic utility. Prog Hemost Thromb 1991;10:307–337. Kujubu DA, Fletcher BS, Varnum BC, Lim RW, Herschman HR: TIS10, a phorbol ester tumor promoter-inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/ cyclooxygenase homologue. J Biol Chem 1991;266:12866–12872. DuBois RN, Awad J, Morrow J, Roberts LJ, Bishop PR: Regulation of eicosanoid production and mitogenesis in rat intestinal epithelial cells by transforming growth factor-␣ and phorbol ester. J Clin Invest 1994;93:493–498. Jones DA, Carlton DP, McIntyre TM, Zimmerman GA, Prescott SM: Molecular cloning of human prostaglandin endoperoxide synthase type II and demonstration of expression in response to cytokines. J Biol Chem 1993;268:9049–9054. Rimarachin JA, Jacobson JA, Szabo P, Maclouf J, Creminon C, Weksler BB: Regulation of cyclooxygenase-2 expression in aortic smooth muscle cells. Arterioscler Thromb 1994;14: 1021–1031. Kelley DJ, Mestre JR, Subbaramaiah K, Sacks PG, Schantz SP, Tanabe T, Inoue H, Ramonetti JT, Dannenberg AJ: Benzo[a]pyrene up-regulates cyclooxygenase-2 gene expression in oral epithelial cells. Carcinogenesis 1997;18:795–799. Xie W, Herschman HR: v-src induces prostaglandin synthase-2 gene expression by activation of the c-Jun N-terminal kinase and the c-Jun transcription factor. J Biol Chem 1995;270: 27622–27628. Subbaramaiah K, Telang N, Ramonetti JT, Araki R, DeVito B, Weksler BB, Dannenberg AJ: Transcription of cyclooxygenase-2 is enhanced in transformed mammary epithelial cells. Cancer Res 1996;56:4424–4429. Mestre JR, Subbaramaiah K, Sacks PG, Schantz SP, Tanabe T, Inoue H, Dannenberg AJ: Retinoids suppress epidermal growth factor-induced transcription of cyclooxygenase-2 in human oral squamous cell carcinoma cells. Cancer Res 1997;57:2890–2895. Mestre JR, Subbaramaiah K, Sacks PG, Schantz SP, Tanabe T, Inoue H, Dannenberg AJ: Retinoids suppress phorbol ester-mediated induction of cyclooxygenase-2. Cancer Res 1997;57: 1081–1085. Subbaramaiah K, Altorki N, Chung WJ, Mestre JR, Sampat A, Dannenberg AJ: Inhibition of cyclooxygenase-2 gene expression by p53. J Biol Chem 1999;274:10911–10915.
COX-2 and Cancers of the Upper Digestive Tract
119
37
38
39
40 41
42 43
44
45
46 47 48
49
50
51
52
53
54 55
56
Boyle JO, Hakim J, Koch W, van der Riet P, Hruban RH, Roa RA, Correo R, Eby YJ, Ruppert JM, Sidransky D: The incidence of p53 mutations increases with progression of head and neck cancer. Cancer Res 1993;53:4477–4480. Leung WK, To KF, Ng YP, Lee TL, Lau JYW, Chan FKL, Ng EKW, Chung SCS, Sung JJY: Association between cyclo-oxygenase-2 overexpression and missense p53 mutations in gastric cancer. Br J Cancer 2001;84:335–339. Biramijamal F, Allameh A, Mirbod P, Groene HJ, Koomagi R, Hollstein M: Unusual profile and high prevalence of p53 mutations in esophageal squamous cell carcinomas from northern Iran. Cancer Res 2001;61:3119–3123. Zhang F, Subbaramaiah K, Altorki N, Dannenberg AJ: Dihydroxy bile acids activate the transcription of cyclooxygenase-2. J Biol Chem 1998;273:2424–2428. Attwood SEA, Ball CS, Barlow AP, Jenkinson L, Norris TL, Watson A: Role of intragastric and intraoesophageal alkalinisation in the genesis of complications in Barrett’s columnar lined lower oesophagus. Gut 1993;34:11–15. Vaezi MF, Richter JE: Role of acid and duodenogastroesophageal reflux in gastroesophageal reflux disease. Gastroenterology 1996;111:1192–1199. Zhang F, Altorki NK, Wu YC, Soslow RA, Subbaramaiah K, Dannenberg AJ: Duodenal reflux induces cyclooxygenase-2 in the esophageal mucosa of rats: Evidence for involvement of bile acids. Gastroenterology 2001;121:1391–1399. Yoon JH, Higuchi H, Werneburg NW, Kaufmann SH, Gores GJ: Bile acids induce cyclooxygenase-2 expression via the epidermal growth factor receptor in a human cholangiocarcinoma cell line. Gastroenterology 2002;122:985–993. Chan G, Boyle JO, Yang EK, Zhang F, Sacks PG, Shah JP, Edelstein D, Soslow RA, Koki AT, Woerner BM, Masferrer JL, Dannenberg AJ: Cyclooxygenase-2 expression is up-regulated in squamous cell carcinoma of the head and neck. Cancer Res 1999;59:991–994. Zimmermann KC, Sarbia M, Weber AA, Borchard F, Gabbert HE, Schror K: Cyclooxygenase-2 expression in human esophageal carcinoma. Cancer Res 1999;59:198–204. Ristimaki A, Honkanen N, Jankala H, Sipponen P, Harkonen M: Expression of cyclooxygenase-2 in human gastric carcinoma. Cancer Res 1997;57:1276–1280. Sung JJY, Leung WK, Go MYY, To KF, Cheng ASL, Ng EKW, Chan FKL: Cyclooxygenase-2 expression in Helicobacter pylori-associated premalignant and malignant gastric lesions. Am J Pathol 2000;157:729–735. Wilson KT, Fu S, Ramanujam KS, Meltzer SH: Increased expression of inducible nitric oxide synthase and cyclooxygenase-2 in Barrett’s esophagus and associated adenocarcinomas. Cancer Res 1998;58:2929–2934. Shamma A, Yamamoto H, Doki Y, Okami J, Kondo M, Fujiwara Y, Yano M, Inoue M, Matsuura N, Shiozaki H, Monden M: Up-regulation of cyclooxygenase-2 in squamous carcinogenesis of the esophagus. Clin Cancer Res 2001;6:1229–1238. Oshima M, Dinchuk JE, Kargman SL, Oshima H, Hancock B, Kwong E, Trzaskos JM, Evans JF, Taketo MM: Suppression of intestinal polyposis in APC⌬716 knockout mice by inhibition of cyclooxygenase-2. Cell 1996;87:803–809. Liu CH, Chang SH, Narko K, Trifan OC, Wu MT, Smith E, Haudenschild C, Lane TF, Hla T: Overexpression of cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem 2001;276:18563–18569. Marnett LJ, Reed GA, Dennison DJ: Prostaglandin synthetase-dependent activation of 7,8dihydro-7,8-dihydroxy-benzo[a]pyrene to mutagenic derivatives. Biochem Biophys Res Commun 1978;82:210–216. Eling TE, Thompson DC, Foureman GL, Curtis JF, Hughes MF: Prostaglandin H synthase and xenobiotic oxidation. Annu Rev Pharmacol Toxicol 1990;30:1–45. Janot F, Massaad L, Ribrag V, de Waziers I, Beaune PH, Luboinski B, Parise O, Gouyette A, Chabot GG: Principal xenobiotic-metabolizing enzyme systems in human head and neck squamous cell carcinoma. Carcinogenesis 1993;14:1279–1283. Wiese FW, Thompson PA, Kadlubar FF: Carcinogen substrate specificity of human COX-1 and COX-2. Carcinogenesis 2001;21:5–10.
Altorki/Subbaramaiah/Dannenberg
120
57
58 59
60
61 62
63
64 65 66 67
68 69 70 71 72 73
74
75 76 77
78
Liu XH, Yao S, Kirschenbaum A, Levine AC: NS398, a selective cyclooxygenase-2 inhibitor, induces apoptosis and down-regulates Bcl-2 expression in LNCaP cells. Cancer Res 1998;58: 4245–4249. Li M, Wu X, Xu XC: Induction of apoptosis by cyclooxygenase-2 inhibitor NS398 through a cytochrome c-dependent pathway in esophageal cancer cells. Int J Cancer 2001;93:218–223. Souza RF, Shewmake K, Beer DG, Cryer B, Spechler SJ: Selective inhibition of cyclooxygenase-2 suppresses growth and induces apoptosis in human esophageal adenocarcinoma cells. Cancer Res 2000;60:5767–5772. Li M, Lotan R, Levin B, Tahara E, Lippman SM, Xu XC: Aspirin induction of apoptosis in esophageal cancer: A potential for chemoprevention. Cancer Epidemiol Biomarkers Prev 2000; 9:545–549. Weitzman SA, Gordon LI: Inflammation and cancer: Role of phagocyte-generated oxidants in carcinogenesis. Blood 1990;76:655–663. Chang-Claude JC, Wahrendorf J, Liang QS, Rei YG, Muñoz N, Crespi M, Raedsch R, Thurnham DI, Correa P: An epidemiological study of precursor lesions of esophageal cancer among young persons in a high-risk population in Huixian, China. Cancer Res 1990;50:2268–2274. Crespi M, Munoz N, Grassi A, Qiong S, Jing WK, Jien LJ: Precursor lesions of oesophageal cancer in a low-risk population in China: Comparison with high-risk populations. Int J Cancer 1984;34:599–602. Lagugren J, Bergstrom R, Lindgren A, Nyren O: Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med 1999;340:825–831. Hameeteman W, Tytgat GNJ, Houthoff HJ, van den Tweel JG: Barrett’s esophagus: Development of dysplasia and adenocarcinoma. Gastroenterology 1989;96:1249–1256. Oberg S, Peters JH, DeMeester TR, Lord RV, Johansson J, DeMeester SR, Hagen JA: Determinants of intestinal metaplasia within the columnar-lined esophagus. Arch Surg 2000;135:651–656. Oberg S, Ritter MP, Crookes PF, Fein M, Mason RJ, Gadensytatter M, Brenner CG, Peters JH, DeMeester TR: Gastroesophageal reflux disease and mucosal injury with emphasis on shortsegment Barrett’s esophagus and duodenogastroesophageal reflux. J Gastrointest Surg 1998; 2:547–553. Plescia OJ, Smith AH, Grinwich K: Subversion of immune system by tumor cells and role of prostaglandins. Proc Natl Acad Sci USA 1975;72:1848–1851. Brunda MJ, Herberman RB, Holden HT: Inhibition of murine natural killer cell activity by prostaglandins. J Immunol 1980;124:2682–2687. Taffet SM, Russell SW: Macrophage-mediated tumor cell killing: Regulation of expression of cytolytic activity by prostaglandin E. J Immunol 1981;126:424–427. Baker PE, Fahey JV, Munck A: Prostaglandin inhibition of T-cell proliferation is mediated at two levels. Cell Immunol 1981;61:52–61. Kunkel SL, Wiggins RC, Chensue SW, Larrick J: Regulation of macrophage tumor necrosis factor production by prostaglandin E2. Biochem Biophys Res Commun 1986;137:404–410. Kambayashi T, Alexander HR, Fong M, Strassmann G: Potential involvement of IL-10 in suppressing tumor-associated macrophages. Colon-26-derived prostaglandin E2 inhibits TNF-␣ release via a mechanism involving IL-10. J Immunol 1995;154:3383–3390. Stolina M, Sharma S, Lin Y, Dohadwala M, Gardner B, Luo J, Zhu L, Kronenberg M, Miller PW, Portanova J, Lee JC, Dubinett SM: Specific inhibition of cyclooxygenase-2 restores antitumor reactivity by altering the balance of IL-10 and IL-12 synthesis. J Immunol 2000;164:361–370. Grinwich KD, Plescia OJ: Tumor-mediated immunosuppression: Prevention by inhibitors of prostaglandin synthesis. Prostaglandins 1977;14:1175–1182. Folkman J: What is the evidence that tumors are angiogenesis-dependent? J Natl Cancer Inst. 1990;82:4–6. Gallo O, Franchi A, Magnelli L, Sardi I, Vannacci A, Boddi V, Chiarugi V, Masini E: Cyclooxygenase-2 pathway correlates with VEGF expression in head and neck cancer. Implications for tumor angiogenesis and metastasis. Neoplasia 2001;3:53–61. Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN: Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705–716.
COX-2 and Cancers of the Upper Digestive Tract
121
79 80
81
82 83
84
85
86 87
88 89
90 91
92
93
94
95
96 97
98
Williams CS, Tsujii M, Reese J, Dey SK, DuBois RN: Host cyclooxygenase-2 modulates carcinoma growth. J Clin Invest 2000;105:1589–1594. Masferrer JL, Leahy KM, Koki AT, Zweifel BS, Settle SL, Woerner BM, Edwards DA, Flickinger AG, Moore RJ, Seibert K: Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 2000;16:1306–1311. Nishimura G, Yanoma S, Mizuno H, Kawakami K, Tsukuda M: A selective cyclooxygenase-2 inhibitor suppresses tumor growth in nude mouse xenografted with human head and neck squamous carcinoma cells. Jpn J Cancer Res 1999;90:1152–1162. Tsujii M, Kawano S, DuBois RN: Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA 1997;94:3336–3340. Dohadwala M, Luo J, Zhu L, Lin Y, Dougherty GJ, Sharma S, Huang M, Pold M, Batra RK, Dubinett SM: Non-small cell lung cancer cyclooxygenase-2-dependent invasion is mediated by CD44. J Biol Chem 2001;276:20809–20812. Tomozawa S, Nagawa H, Tsuno N, Hatano K, Osada T, Kitayama J, Sunami E, Nita ME, Ishihara S, Yano H, Tsuruo T, Shibata Y, Muto T: Inhibition of haematogenous metastasis of colon cancer in mice by a selective COX-2 inhibitor, JTE-522. Br J Cancer 1999;81:1274–1279. Shiotani H, Denda A, Yamamoto K, Kitayama W, Endoh T, Sasaki Y, Tsutsumi M, Sugimura M, Konishi Y: Increased expression of cyclooxygenase-2 protein and 4-nitroquinoline-1-oxideinduced rat tongue carcinomas and chemopreventive efficacy of a specific inhibitor, nimesulide. Cancer Res 2001;61:1451–1456. Kawamori T, Rao CV, Seibert K, Reddy BS: Chemopreventive activity of celecoxib, a specific cyclo-oxygenase-2 inhibitor, against colon carcinogenesis. Cancer Res 1998;58:409–412. Oshima M, Murai N, Kargman S, Arguello M, Luk P, Kwong E, Taketo MM, Evans JF: Chemoprevention of intestinal polyposis in APC⌬716 mouse by rofecoxib, a specific cyclooxygenase-2 inhibitor. Cancer Res 2001;16:1733–1740. Harris RE, Alshafie GA, Abou-Issa H, Seibert K: Chemoprevention of breast cancer in rats by celecoxib, a cyclooxygenase-2 inhibitor. Cancer Res 2000;60:2101–2103. Fischer SM, Lo HH, Gordon GB, Seibert K, Kelloff G, Lubet RA, Conti CJ: Chemopreventive activity of celecoxib, a specific cyclooxygenase-2 inhibitor, and indomethacin against ultraviolet light-induced skin carcinogenesis. Mol Carcinog 1999;25:231–240. Rioux N, Castonguay A: Prevention of NNK-induced lung tumorigenesis in A/J mice by acetylsalicylic acid and NS-398. Cancer Res 1998;58:5354–5360. Grubbs CJ, Lubet RA, Koki AT, Leahy KM, Masferrer JL, Steele VE, Kelloff GJ, Hill DL, Seibert K: Celecoxib inhibits N-butyl-N-(4-hydroxybutyl)nitrosamine-induced urinary bladder cancers in male B6D2F1 mice and female Fischer-344 rats. Cancer Res 2000;60:5599–5602. Li Z, Shimada Y, Kawabe A, Sato F, Maeda M, Komoto I, Hong T, Ding Y, Kaganoi J, Imamura M: Suppression of N-nitrosomethylbenzylamine (NMBA)-induced esophageal tumorigenesis in F344 rats by JTE-522, a selective COX-2 inhibitor. Carcinogenesis 2001;22:547–551. Buttar NS, Wang KK, Leontovich O, Westcott JY, Pacifco RJ, Anderson MA, Krishnadath KK, Lutzke LS Burgart LJ: Chemoprevention of esophageal adenocarcinoma by COX-2 inhibitors in an animal model of Barrett’s esophagus. Gastroenterology 2002;122:1101–1112. Sawaoka H, Kawano S, Tsuji S, Tsuji M, Gunawan ES, Takei Y, Nagano K, Hori M: Cyclooxygenase-2 inhibitors suppress the growth of gastric cancer xenografts via induction of apoptosis in nude mice. Am J Physiol 1998;274:G1061–G1067. Moore RJ, Zweifel BS, Heuvelman DM, Leahy KM, Edwards DA, Woerner BM, Ornberg RL, Seibert K, Koki AT, Masferrer JL: Enhanced antitumor activity by co-administration of celecoxib and the chemotherapeutic agents cyclophosphamide and 5-FU. Proc Am Assoc Cancer Res 2000; 41:409. Milas L, Kishi K, Hunter N, Mason K, Masferrer JL, Tofilon PJ: Enhancement of tumor response to ␥-radiation by an inhibitor of cyclooxygenase-2 enzyme. J Natl Cancer Inst 1999;91:1501–1504. Langman MJ, Jensen DM, Watson DJ, Harper SE, Zhao PL, Quan H, Bolognese JA, Simon TJ: Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs. JAMA 1999;282: 1929–1933. Silverstein FE, Faich G, Goldstein JL, Simon LS, Pincus T, Whelton A, Makuch R, Eisen G, Agrawal NM, Stenson WF, Burr AM, Zhao WW, Kent JD, Lefkowith JB, Verburg KM,
Altorki/Subbaramaiah/Dannenberg
122
99
100
101
102
103
104
105
106
107
108
Geis GS: Gastrointestinal toxicity with celecoxib vs. nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: The CLASS study. JAMA 2000;284:1247–1255. Steinbach G, Lynch PM, Phillips RKS, Wallace MH, Hawk E, Gordon GB, Wakabayashi N, Saunders B, Shen Y, Fujimura T, Su LK, Levin B: The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–1952. Giardiello FM, Hamilton SR, Krush AJ, Piantadosi S, Hylind LM, Celano P, Booker SV, Robinson CR, Offerhaus GJ: Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med 1993;328:1313–1316. Kaur BS, Khamnehei N, Iravani M, Namburu SS, Lin O, Triadafilopoulos G: Rofecoxib inhibits cyclooxygenase-2 expression and activity and reduces cell proliferation in Barrett’s esophagus. Gastroenterology 2002;123:60–67. Khuri FR, Wu H, Lee JJ, Kemp BL, Lotan R, Lippman SM, Feng L, Hong WK, Xu XC: Cyclooxygenase-2 overexpression is a marker of poor prognosis in stage I non-small cell lung cancer. Clin Cancer Res 2001;7:861–867. Sheehan KM, Sheahan K, O’Donoghue DP, MacSweeney F, Conroy RM, Fitzgerald DJ, Murray FE: The relationship between cyclooxygenase-2 expression and colorectal cancer. JAMA 1999;282:1254–1257. Murata H, Kawano S, Tsuji S, Tsuji M, Sawaoka H, Kimura Y, Shiozaki H, Hori M: Cyclooxygenase-2 overexpression enhances lymphatic invasion and metastasis in human gastric carcinoma. Am J Gastroenterol 1999;94:451–455. Lim HY, Joo HJ, Choi JH, Yi JW, Yang MS, Cho DY, Kim HS, Nam DK, Lee KB, Kim HC: Increased expression of cyclooxygenase-2 protein in human gastric carcinoma. Clin Cancer Res 2000;6:519–525. Buskens CJ, Van Rees BP, Sivula A, Reitsma JB, Haglund C, Bosma PJ, Offerhaus GJA, Van Lanschot J, Ristimäki A: Prognostic significance of elevated cyclooxygenase-2 expression in patients with adenocarcinoma of the esophagus. Gastroenterology 2002;122:1800–1807. Torrance CJ, Jackson PE, Montgomery E, Kinzler KW, Vogelstein B, Wissner A, Nunes M, Frost P, Discafani CM: Combinatorial chemoprevention of intestinal neoplasia. Nat Med 2000;6: 1024–1028. Hong WK, Endicott J, Itri LM, Doos W, Batsakis JG, Bell R, Fofonoff S, Byers R, Atkinson EN, Vaughan C, Toth BB, Kramer A, Dimery IW, Skipper P, Strong S: 13-cis-Retinoic acid in the treatment of oral leukoplakia. N Engl J Med 1986;315:1501–1505.
Nasser K. Altorki, MD Division of General Thoracic Surgery, Weill Medical College of Cornell University 525 East 68th Street, Box 110, New York, NY 10021 (USA) Tel. ⫹1 212 7465156, Fax ⫹1 212 7468426, E-Mail
[email protected]
COX-2 and Cancers of the Upper Digestive Tract
123
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 124–137
Cyclooxygenase-2 and Colorectal Cancer Raymond N. DuBois Departments of Medicine and Cell Biology, Vanderbilt University Medical Center and Department of Veterans Affairs Medical Center, Nashville, Tenn., USA
Introduction
Colorectal cancer (CRC) leads to approximately 550,000 annual deaths worldwide and is thus a major public health concern [1]. Once an individual is diagnosed with CRC, the disease is often advanced and current treatment regimens are often ineffective. Recently, cyclooxygenase-2 (COX-2) inhibitor therapy has emerged as a promising novel therapy for CRC prevention and/or treatment. Since currently available treatment regimens for CRC are not effective, there is renewed interest in the area of cancer prevention. This chapter will review some of the key findings from both clinical and laboratory settings that has led to the hypothesis that COX-2 inhibitors may be useful in reducing morbidity and mortality from CRC.
Basic Science Studies on NSAIDs and CRC
The Function of the COX Enzymes The COX enzymes catalyze the enzymatic conversion of the endoperoxide intermediate PGH2 from arachidonic acid [2]. PGH2 is then converted to one of several structurally related prostaglandins (PGs), including PGE2, PGD2, PGF2␣, PGI2, and thromboxane A2 (TxA2), by the enzymatic activity of specific PG synthases. PGs play roles in a broad number of physiological processes including blood clotting, ovulation, initiation of labor, bone metabolism, nerve growth and development, wound healing, kidney function, blood vessel tone,
and immune responses [3]. The importance of the cyclooxygenase pathway in human disease became obvious with the discovery that non-steroidal antiinflammatory drugs (NSAIDs), a group of compounds that have been used for more than a century in the treatment of various inflammatory disorders, were shown in the 1970s to inhibit the activity of COX. Discovery of COX-2 and the Biochemistry of COX-2-Selective Inhibitors Until the early 1990’s, only one isoform of COX (now known as COX-1) had been cloned and identified. However, approximately 10 years ago, at least two independent groups identified and cloned a second COX isoform, COX-2. Subsequent research has suggested a paradigm in which COX-1 is responsible for ‘housekeeping’ PG biosynthesis and is constitutively expressed in many tissues in the body [4]. COX-2, on the other hand, is not normally expressed in most tissues but can be induced by a wide spectrum of growth factors and pro-inflammatory cytokines in specific pathophysiological conditions. It was thus hypothesized that the anti-inflammatory and analgesic properties of traditional NSAIDs, which inhibit both COX-1 and COX-2, are most likely due to their ability to inhibit COX-2 [5]. In contrast, the erosion and ulceration of the gastric mucosa seen with chronic NSAID therapy may be due, in large measure, to inhibition of PG production mediated by COX-1. To test the above hypothesis, compounds that selectively inhibit the COX-2 isoform were developed. The most well-known non-selective COX inhibitor, aspirin, inactivates COX enzymatic activity by covalent modification via acetylation of serine residue 530 in COX-1 and serine residue 516 in COX-2. In contrast, most other non-selective NSAIDs are competitive inhibitors that compete with the substrate arachidonic acid for active site binding in both COX-1 and COX-2. Based on the availability of the three-dimensional structure of both COX-1 and COX-2, compounds were found that preferentially bind to the active site of COX-2 [6]. Currently available clinical evidence suggests that these COX-2-specific inhibitors (‘coxib’ drugs) offer the therapeutic benefits of traditional NSAIDs with reduced toxicity to the gastrointestinal mucosa [7]. Finally, a new aspirin like COX-2 inhibitor, APHS (o-(acetoxypheny)hept2-ynyl sulfide), has been developed which is 21 times more selective for COX-2 than COX-1 [8]. While this level of selectivity is not as significant as that of some other COX-2 inhibitors, APHS offers a unique advantage in that it covalently modifies the enzyme, thus insuring permanent inactivation of COX-2. Effects of NSAIDs on Tumor Growth in Animals Early observational studies in the clinical setting suggested NSAIDs might restrict the growth of CRC (see below: Clinical Studies on NSAIDs and CRC). These observations prompted a number of follow-up studies that examined the
Cyclooxygenase-2 and Colorectal Cancer
125
Table 1. Effects of NSAID treatment in animal models of CRCa Animal model
NSAID treatment
Outcome
Min/⫹ mouse
Sulindacb Piroxicamb Celecoxibc
↓ Polyp multiplicity ↓ Polyp multiplicity ↓ Polyp multiplicity
Azoxymethane (AOM) – rat
Aspirinb Celecoxibc NS-398c Nimesulidec
↓ Tumor incidence and multiplicity ↓ Tumor incidence and multiplicity ↓ Tumor incidence and multiplicity ↓ Tumor incidence and multiplicity
a
Reviewed in reference 9. Non-selective NSAID. c COX-2-selective NSAID. b
chemopreventive properties of these drugs in animal models of intestinal neoplasia (summarized in table 1 and reviewed in Williams et al. [9]). One wellknown murine model of colon cancer is the Min/⫹ mouse, which represents the murine homolog of the polyposis syndrome Familial adenomatous polyposis (FAP) in humans. Multiple investigators have examined the ability of both nonselective and COX-2-selective NSAIDs to inhibit polyposis using this model. Without exception, either non-selective or COX-2-selective inhibitors have proven to be potent suppressors of polyp formation in these mice. Another widely used animal model for colon cancer is the azoxymethane (AOM)-treated rat, in which the chemical carcinogen AOM induces pre-neoplastic colonic lesions termed aberrant crypt foci (ACF) which later progress to carcinomas. Both non-selective and COX-2-selective NSAIDs reduce the incidence, multiplicity and size of colonic carcinomas in the AOM rat model. Role of COX-2 in CRC Growth The findings in animal models served as a stimulus for basic science investigations into the mechanism(s) by which NSAIDs inhibit CRC growth. Because the COX enzymes are the most well-defined pharmacological targets of NSAIDs, one of the first ideas to emerge was the hypothesis that these drugs were inhibiting the presumed pro-tumorigenic activity of either COX-1 or COX-2. This reasoning was strengthened by studies that demonstrated increased levels of PGs in human colorectal tumors compared to normal adjacent colon [10] as well as in vitro experiments showing that some PGs can increase the proliferation rate of human CRC cells [11]. In early attempts to examine the role of COX in CRC, several groups determined the levels of
DuBois
126
COX-1 and COX-2 in human adenomas and adenocarcinomas [12, 13]. These results point to essentially the same conclusion: comparison of neoplastic and adjacent normal tissue finds little change in COX-1 expression but increased levels of COX-2 in adenoma and adenocarcinoma tissue compared to the normal colonic mucosa. This expression data was suggestive that NSAIDs could be suppressing tumor growth via inhibition of COX-2. However, it could still be argued that elevated levels of COX-2 are simply a consequence of the carcinogenic process and that the enzyme has no direct role in promoting CRC growth. One way to determine the role of COX-2 in intestinal neoplasia is to use a genetic approach. Oshima et al. [14] did this by assessing the development of intestinal polyposis in Apc⌬716 mice (a model similar to the Min/⫹ mouse) in a wild-type and homozygous null COX-2 genetic background. The number and size of polyps was reduced dramatically in the COX-2 null mice compared to COX-2 wild-type mice. In addition, treatment of the Apc⌬716 COX-2 wild-type mice with a novel COX-2-selective inhibitor, MF tricyclic, reduced polyp number more significantly than sulindac. This experiment was one of the first to offer rigorous evidence in support of the hypothesis that NSAIDs inhibit tumor growth via inhibition of COX-2. More recent studies have confirmed a pro-oncogenic role for COX-2. For example, Liu et al. [15] have demonstrated that overexpression of COX-2 alone is sufficient to induce cellular transformation. This group developed transgenic mice in which the murine mammary tumor virus promoter/enhancer directs COX-2 expression. Although virgin mice overexpressing COX-2 did not develop mammary tumors, multiparous mice showed significant increases in mammary gland carcinomas compared with age-matched controls. A related study by Neufang et al. [16] has reported that transgenic expression of COX-2 in basal keratinocytes results in epidermal hyperplasia and dysplasia, suggesting a causal association between COX-2 expression and the development of pre-neoplastic lesions in the skin. COX-Independent Effects of NSAIDs Multiple groups have now documented that high doses of NSAIDs can modify the biology of cultured cells independently of their ability to bind and inhibit COX-1 or COX-2. For example, transformed fibroblast cell lines derived from wild-type, COX-1⫺/⫺, COX-2⫺/⫺, or COX-1⫺/⫺/COX-2⫺/⫺ mice all show comparable sensitivity to NSAID-induced cell death [17]. Such results are not surprising, since any xenobiotic agent is likely to have multiple targets depending on the dose of the drug used. Non-COX cellular targets that have been proposed include IB kinase  [18], the peroxisome proliferator-activated receptor (PPAR) family of nuclear hormone receptors [19, 20], and the pro-apoptotic gene BAX [21].
Cyclooxygenase-2 and Colorectal Cancer
127
Paracrine Other cells
Autocrine PG
Stromal effects Fibroblast ↑COX-2
Angiogenic growth factor
↑COX-2 Blood vessel
Epithelial cell ↓Apoptosis ↑Cell migration ↑Proliferation
Angiogenesis ↑COX-2 Endothelial cell
Fig. 1. Mechanisms by which elevated COX-2 promotes CRC growth. Forced expression of COX-2 in CRC (CRC) cell lines leads to changes in key regulatory genes that promote blood vessel growth (angiogenesis). In other model systems, cultured skin fibroblasts from COX-2⫺/⫺ mice exhibit defects in the basal secretion of several angiogenic growth factors arguing that COX-2 may modify tumor growth by limiting the ability of fibroblasts to support neovascularization within the microenvironment of a tumor. COX-2 has also been localized to the tumor endothelial cells and several groups have demonstrated that COX-2 inhibitors can block both the migration of endothelial cells.
However, in all instances, ‘COX-independent’ effects are seen at drug concentration(s) in the 50–1,000 M range. This is 10- to 200-fold higher than the serum concentration of celecoxib (approx. 2–5 M) required to inhibit tumor growth in animal models of CRC [22]. At these low concentrations, the best characterized biochemical target of NSAIDs remains the COX enzymes, although it is likely that other unknown ‘high affinity’ targets could be affected as well. Mechanisms Involved in COX-2-Mediated Promotion of CRC COX-2 has been localized to both tumor epithelial cells and adjacent stromal cells. Thus, COX-2-derived PGs may be acting on the malignant epithelial cells (a cell autonomous effect) or on the surrounding stroma (cell nonautonomous or ‘landscaping’ effect) to promote tumor development (fig. 1). There is evidence to support both theories. If COX-2 inhibitors reduce tumor growth only though blockade of COX-2derived PGs within the stromal compartment, then one would predict that most CRC cell lines, irrespective of their level of COX-2 expression, would be sensitive to therapy with COX-2 inhibitors. This does not appear to be the case: in the nude mouse xenograft model, tumor cells that express COX-2 are more sensitive to treatment with selective COX-2 inhibitors [23]. Consistent
DuBois
128
with this, several studies suggest that forced expression of COX-2 in intestinal epithelial cell lines leads to changes in cellular pathways linked to carcinogenesis. For example, rat intestinal epithelial cells engineered to overexpress COX-2 have elevated levels of the anti-apoptotic protein Bcl-2 and exhibit increased resistance to apoptosis induced by sodium butyrate [24]. In the Caco-2 cancer cell line, overexpression of COX-2 leads to an increase in cell migration and invasion that is associated with elevated levels of several members of the matrix metalloprotease (MMP) family [25]. This same cell line also secretes higher levels of several angiogenic factors compared with vector-transfected control cells, and promotes the formation of endothelial cell tubes when co-cultured with human umbilical vein endothelial cells [26]. The in vivo relevance of these findings has not yet been determined. Several reports have suggested that COX-2 can also promote tumorigenesis through direct actions on the stromal compartment. These studies have largely centered on the ability of COX-2-derived PGs to stimulate tumorassociated angiogenesis. Using a model in which angiogenesis is assessed in sponge implants injected with various growth factors, Majima et al. [27, 28] were one of the first groups to demonstrate that COX-2 inhibitors could block neovascularization. COX-2 inhibitors also blocked the migration of human microvascular endothelial cells and growth factor-induced corneal angiogenesis, effects that could be reconstituted with a TXA2 agonist [29]. A related study reported strong COX-2 immunoreactivity in tumor neovasculature in human colon, breast, prostate and lung cancer biopsy tissue [30]. In addition, corneal blood vessel formation in rats was potently suppressed by selective COX-2, but not COX-1, inhibitors. Jones et al. [31] reported that both nonselective and COX-2-selective NSAIDs inhibit angiogenesis through direct actions on endothelial cells through both COX and non-COX mechanisms. Finally, a recent study provides genetic evidence that stroma-derived COX-2 can promote tumor growth by a landscaping mechanism [32]. In this study, the growth of a lung cancer cell line was attenuated if engrafted on to COX-2⫺/⫺ versus wild-type control mice. COX-2 expression within the stroma surrounding the tumor was localized primarily to fibroblasts, and cultured skin fibroblasts from the COX-2⫺/⫺ mice exhibited defects in the basal secretion of several angiogenic growth factors. Collectively, this last set of experiments argues that COX-2 may modify tumor growth by limiting the ability of fibroblasts to support neovascularization within the microenvironment of a tumor. Recent experiments by Dormond et al. [33] have helped clarify the mechanism by which COX-2-derived PGs promote angiogenesis. Their studies demonstrated that COX-2-derived PGE2 and PGI2 play essential roles in ␣V3 integrin-mediated endothelial cell spreading and migration.
Cyclooxygenase-2 and Colorectal Cancer
129
Specifically, both PGs appear to be important in the activation of the small GTPases Cdc42/Rac that occurs following engagement of ␣V3 integrin with its substrate. PG Receptors in CRC The different classes of PGs exert their effects by binding to a G-proteincoupled cell-surface receptor that then leads to changes in the cellular levels of cAMP and Ca2⫹ [for review, see 34]. In addition, increasing evidence points to the possibility that PGs can also modulate cellular pathways by directly acting within the nucleus. For example, COX-2 has been localized to the perinuclear envelope [35] and the PGE2 receptor EP1 has been localized to the nuclear envelope in certain cell types, and activation of the receptor led to changes in nuclear levels of Ca2⫹ [36]. Studies on the role of specific PGs and PG receptors that act downstream of COX-2 during the progression of CRC have been limited. Several groups have reported elevated levels of PGE2 in CRC biopsies compared with normal colonic mucosa [10, 37]. However, because many PGs are highly unstable, the physiological relevance of these studies is not known. Furthermore, there have been no systematic surveys documenting the expression of the known PG synthases and PG receptors in CRC. Several studies have reported a pro-carcinogenic effect of PGE2 in cultured CRC cells. For example, Sheng et al. [38] reported a reduction in the basal apoptotic rate and increased levels of Bcl-2 after treatment of a human CRC cell line with PGE2. Exposure of a different colorectal cell line to PGE2 led to an increase in cell proliferation and motility associated with activation of the phosphatidylinositol 3-kinase (PI3K)/Akt pathway, an effect likely to be due to activation of the PGE2 receptor subtype EP4 [39]. A related study implicated PGE2 in the ability of heregulin-1 to induce colon cancer cell migration and invasion [40]. Genetic studies in mice suggest that PGE2 promotes tumorigenesis at least in part by activating the EP1 receptor subtype. Mice with homozygous deletions in the EP1, but not the EP3, gene were partially resistant to AOM-mediated induction of ACF [41, 42]. Moreover, in AOM-treated wildtype mice, an EP1 receptor antagonist also decreased the incidence of ACF. Finally, ApcMin mice treated with the same EP1 receptor antagonist had 57% fewer intestinal polyps than untreated mice. In contrast, Sonoshita et al. [43] identified an important role for the EP2 receptor in CRC development. They examined the genetic role of all four PGE2 receptors in the development of intestinal polyposis in Apc⌬716 mice. The number and size of intestinal polyps was significantly reduced only in mice also harboring a homozygous deletion of the EP2 receptor.
DuBois
130
Clinical Studies on NSAIDs and CRC
Observational Studies of CRC and NSAIDs There are over 13 published observational studies of the relationship between NSAID use and CRC incidence or mortality [reviewed in 44]. This group of studies is remarkable for its consistency in demonstrating a protective effect of NSAIDs against the development of CRC. Only one of the studies failed to demonstrate a protective effect and this study did not offer any methodological advantage over other studies in the group [45]. The studies were carried out in three countries in a variety of settings (hospital-based, population-based, national surveys) utilizing various measures of aspirin and nonaspirin NSAID use (questionnaire and computerized pharmacy records). Past history suggests that it is unusual for a putative chemopreventive factor to have such a consistent effect across several different studies. Observational studies have demonstrated a protective effect for both aspirin and non-aspirin NSAIDs. Because each study had a distinct method of measuring and categorizing NSAID exposure, it is difficult to make comparisons across studies. However, the dominant theme to emerge from a comparison of the different studies is that consistent use of drug is important to see statistically meaningful effects in cancer risk reduction [46–50]. A second theme that emerges is that continual use of drug is also important since attenuation of the chemopreventive effect is seen with as little as 1 year of non-exposure [49, 51]. Most studies that have examined duration of NSAID use have demonstrated increased protection with more years of exposure. Three studies indicated that at least 5–10 years of use was required for a protective benefit [49, 50, 52] while a fourth study indicated that some protection could be acquired with as little as 4 years of use [47]. Higher doses seemed to provide more protection in two studies [48, 52] but not in a third [47]. Finally, it should be noted that no large-scale observational studies have examined how the exposure to the new ‘coxib’ drugs affects CRC incidence. Polyp Prevention in FAP Syndrome by Non-Selective and COX-2-Selective NSAIDs The first clinical observations of the effect of NSAIDs (sulindac) on adenoma formation occurred in groups of patients who had the FAP syndrome. This disease is an autosomal dominant inherited condition marked by the development of hundreds to thousands of adenomatous polyps in the colon during early adulthood. Without colectomy the vast majority of these patients will develop cancer in their 30s or 40s. In many centers the standard of care is
Cyclooxygenase-2 and Colorectal Cancer
131
a ‘subtotal colectomy’, an operation that leaves behind the most distal colonic segment (rectum) in order to preserve normal bowel function. Because cancer may occur in this segment, these patients are usually followed with surveillance sigmoidoscopies during which polyps are identified and removed. Waddell et al. [53] noted a paucity of polyps on a routine sigmoidoscopic surveillance exam in a single FAP patient who had received sulindac for other reasons. Because of this observation, subsequent randomized controlled trials of polyp prevention in patients with FAP were carried out. The trial reported by Giardiello et al. [54] is the most notable and served as a paradigm for subsequent studies. In this trial, sulindac or placebo was given to FAP patients who had previously undergone subtotal colectomy and who were engaged in a rectal screening program. Patients receiving sulindac had a decrease in both the number and average size of their rectal polyps on exams at 3 and 6 months. When the drug was stopped, the polyps increased in size and started to recur at the pre-treatment rate. There were no such effects among those treated with placebo. Recently, the COX-2-selective drug, celecoxib, was tested as a potential chemopreventive agent in patients with FAP [55]. This drug offers a potential advantage because it causes fewer serious gastrointestinal side effects than traditional NSAIDs such as sulindac. The trial utilized celecoxib at two doses compared to placebo in cohorts of patients with FAP who were engaged in screening regimens. There was a 28% reduction in polyp formation among those who took celecoxib at the higher dose (400 mg bid). Those receiving celecoxib at the lower dose (100 mg bid) had a smaller (11%) non-significant decrease in the number of polyps at the follow-up endoscopy. Thus both traditional NSAIDs and at least one of the newer ‘coxib drugs’ decrease polyp formation in patients with FAP. It should be noted that authorities have recommended that NSAID therapy be considered only as a possible adjunct to established screening practices. The absolute decrease in polyps in the treatment groups is modest and is not sufficient to replace standard screening practices and surgical intervention in FAP. Alarmingly, there are case reports of cancer of the rectum occurring in patients with FAP who were on sulindac therapy [56]. Future Clinical Studies of COX-2 Inhibitors and CRC The current CRC chemoprevention trials underway are designed to test the efficacy of selective COX-2 inhibitors in preventing adenoma recurrence after polypectomy. These studies are not yet completed and their design is similar to previously published studies of calcium carbonate in preventing polyp recurrence [57]. In these studies, patients with sporadic polyps removed at an index
DuBois
132
colonoscopy are recruited from practices that provide screening colonoscopic examinations. The standard practice for most patients with adenomatous polyps would be to have a repeat colonoscopy in 3 years. Many of the chemoprevention studies incorporate a colonoscopy at 1 year after the index exam to ensure that the colon has been ‘cleared’ of polyps and that any polyps found on subsequent exams are ‘incident’ polyps. Thus, the endpoint of interest is the recurrence rate of polyps at 3 or 4 years after the index date. The agents currently being tested in polyp prevention trials include aspirin and two of the coxib drugs, celecoxib and rofecoxib. No clinical data exists regarding the efficacy of COX-2 blockade therapy in the treatment of established CRCs. However, a number of clinical experiments are underway that tests the ability of COX-2 inhibitors to potentiate the effects of traditional anti-cancer regimens. There is emerging evidence from pre-clinical studies that combination therapy with COX-2 inhibitors and drugs that target other oncogenic pathways can lead to improved clinical outcomes. For example, two studies have recently reported enhanced anti-tumor efficacy in combination regimens consisting of an NSAID and an inhibitor that targets the ErbB/HER family of growth factor receptors [58, 59].
Conclusion
Although there is now compelling evidence from both animal and cell culture systems to suggest that targeted inhibition of COX-2 is a viable approach for CRC prevention and/or treatment, many questions remain unanswered. Future studies examining intestinal polyp susceptibility in mice with targeted deletions in specific PG synthases and receptors should help clarify the mechanisms by which COX-2 promotes tumorigenesis. The precise signaling pathways and direct target genes that COX-2-derived PGs modulate in CRC cells are also largely unknown and must be identified. Collectively, these experiments may lead to the discovery of novel and more potent inhibitors of CRC cell growth.
Acknowledgements This work is supported in part from the United States Public Health Services Grants RO1DK 47279 (RND), P030 ES-00267–29 (RND) and P01CA-77839 (RND). RND is a recipient of a VA Research Merit Grant and is the Mina C. Wallace Professor of Cancer Prevention. We also thank the T.J. Martell Foundation and the NCCRA for generous support.
Cyclooxygenase-2 and Colorectal Cancer
133
This chapter is an adaptation of a review entitled ‘Colorectal Cancer Prevention and Treatment by Inhibition of Cyclooxygenase-2’, Nat Rev Cancer 2001;1:11–21.
References 1 2 3 4 5
6
7
8 9 10 11
12
13
14
15
16
17
Greenlee RT, Murray T, Bolden S, Wingo PA: Cancer statistics, 2000. CA Cancer J Clin 2000; 50:7–33. Smith WL, DeWitt DL, Garavito RM: Cyclooxygenases: Structural, cellular, and molecular biology. Annu Rev Biochem 2000;69:145–182. DuBois RN, Abramson SB, Crofford L, Gupta RA, Simon LS, Van De Putte LB, Lipsky PE: Cyclooxygenase in biology and disease. FASEB J 1998;12:1063–1073. Isakson P, Seibert K, Masferrer J, Salvemini D, Lee L, Needleman P: Discovery of a better aspirin. Adv Prostaglandin Thromboxane Leukot Res 1995;23:49–54. Masferrer JL, Zweifel BS, Manning PT, Hauser SD, Leahy KM, Smith WG, Isakson PC, Seibert K: Selective inhibition of inducible cyclooxygenase 2 in vivo is antiinflammatory and nonulcerogenic. Proc Natl Acad Sci USA 1994;91:3228–3232. Kurumbail RG, Stevens AM, Gierse JK, McDonald JJ, Stegeman RA, Pak JY, Gildehaus D, Miyashiro JM, Penning TD, Seibert K, Isakson PC, Stallings WC: Structural basis for selective inhibition of cyclooxygenase-2 by anti-inflammatory agents. Nature 1996;384: 644–648. Silverstein FE, Faich G, Goldstein JL, Simon LS, Pincus T, Whelton A, Makuch R, Eisen G, Agrawal NM, Stenson WF, Burr AM, Zhao WW, Kent JD, Lefkowith JB, Verburg KM, Geis GS: Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: The CLASS study: A randomized controlled trial. Celecoxib long-term arthritis safety study. JAMA 2000;284:1247–1255. Kalgutkar AS, Crews BC, Rowlinson SW, Garner C, Seibert K, Marnett LJ: Aspirin-like molecules that covalently inactivate cyclooxygenase-2. Science 1998;280:1268–1270. Williams C, Shattuck-Brandt RL, DuBois RN: The role of COX-2 in intestinal cancer. Ann NY Acad Sci 1999;889:72–83. Rigas B, Goldman IS, Levine L: Altered eicosanoid levels in human colon cancer. J Lab Clin Med 1993;122:518–523. Qiao L, Kozoni V, Tsioulias GJ, Koutsos MI, Hanif R, Shiff SJ, Rigas B: Selected eicosanoids increase the proliferation rate of human colon carcinoma cell lines and mouse colonocytes in vivo. Biochim Biophys Acta 1995;1258:215–223. Eberhart CE, Coffey RJ, Radhika A, Giardiello FM, Ferrenbach S, DuBois RN: Up-regulation of cyclooxygenase 2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 1994;107:1183–1188. Sano H, Kawahito Y, Wilder RL, Hashiramoto A, Mukai S, Asai K, Kimura S, Kato H, Kondo M, Hla T: Expression of cyclooxygenase-1 and -2 in human colorectal cancer. Cancer Res 1995;55: 3785–3789. Oshima M, Dinchuk JE, Kargman SL, Oshima H, Hancock B, Kwong E, Trzaskos JM, Evans JF, Taketo MM: Suppression of intestinal polyposis in Apc delta716 knockout mice by inhibition of cyclooxygenase 2 (COX-2). Cell 1996;87:803–809. Liu CH, Chang SH, Narko K, Trifan OC, Wu MT, Smith E, Haudenschild C, Lane TF, Hla T: Overexpression of cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem 2001;276:18563–18569. Neufang G, Furstenberger G, Heidt M, Marks F, Muller-Decker K: Abnormal differentiation of epidermis in transgenic mice constitutively expressing cyclooxygenase-2 in skin. Proc Natl Acad Sci USA 2001;98:7629–7634. Zhang X, Morham SG, Langenbach R, Young DA: Malignant transformation and antineoplastic actions of nonsteroidal antiinflammatory drugs (NSAIDs) on cyclooxygenase-null embryo fibroblasts. J Exp Med 1999;190:445–450.
DuBois
134
18 19
20 21 22
23
24 25 26 27
28
29
30
31
32 33
34 35
36
37
38
Yin MJ, Yamamoto Y, Gaynor RB: The anti-inflammatory agents aspirin and salicylate inhibit the activity of I(kappa)B kinase-beta. Nature 1998;396:77–80. Lehmann JM, Lenhard JM, Oliver BB, Ringold GM, Kliewer SA: Peroxisome proliferatoractivated receptors alpha and gamma are activated by indomethacin and other non-steroidal anti-inflammatory drugs. J Biol Chem 1997;272:3406–3410. Yamamoto Y, Yin MJ, Lin KM, Gaynor RB: Sulindac inhibits activation of the NF-kappaB pathway. J Biol Chem 1999;274:27307–27314. Zhang L, Yu J, Park BH, Kinzler KW, Vogelstein B: Role of BAX in the apoptotic response to anticancer agents. Science 2000;290:989–992. Williams CS, Watson AJ, Sheng H, Helou R, Shao J, DuBois RN: Celecoxib prevents tumor growth in vivo without toxicity to normal gut: Lack of correlation between in vitro and in vivo models. Cancer Res 2000;60: 6045–6051. Sheng H, Shao J, Kirkland SC, Isakson P, Coffey R, Morrow J, Beauchamp RD, DuBois RN: Inhibition of human colon cancer cell growth by selective inhibition of cyclooxygenase-2. J Clin Invest 1997;99:2254–2259. Tsujii M, DuBois RN: Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell 1995;83:493–501. Tsujii M, Kuwano S, DuBois RN: Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA 1997;94:3336–3340. Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN: Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705–716. Majima M, Isono M, Ikeda Y, Hayashi I, Hatanaka K, Harada Y, Katsumata O, Yamashina S, Katori M, Yamamoto S: Significant roles of inducible cyclooxygenase (COX)-2 in angiogenesis in rat sponge implants. Jpn J Pharmacol 1997;75:105–114. Majima M, Hayashi I, Muramatsu M, Katada J, Yamashina S, Katori M: Cyclo-oxygenase-2 enhances basic fibroblast growth factor-induced angiogenesis through induction of vascular endothelial growth factor in rat sponge implants. Br J Pharmacol 2000;130:641–649. Daniel TO, Liu H, Morrow JD, Crews BC, Marnett LJ: Thromboxane A2 is a mediator of cyclooxygenase-2-dependent endothelial migration and angiogenesis. Cancer Res 1999;59: 4574–4577. Masferrer JL, Leahy KM, Koki AT, Zweifel BS, Settle SL, Woerner MB, Edwards DA, Flickinger AG, Moore RJ, Seibert K: Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 2000;60:1306–1311. Jones MK, Wang H, Peskar BM, Levin E, Itani RM, Sarfeh IJ, Tarnawski AS: Inhibition of angiogenesis by nonsteroidal anti-inflammatory drugs: Insight into mechanisms and implications for cancer growth and ulcer healing. Nat Med 1999;5:1418–1423. Williams CS, Tsujii M, Reese J, Dey SK, DuBois RN: Host cyclooxygenase-2 modulates carcinoma growth. J Clin Invest 2000;105: 1589–1594. Dormond O, Foletti A, Paroz C, Ruegg C: NSAIDs inhibit alpha V beta 3 integrin-mediated and Cdc42/Rac-dependent endothelial-cell spreading, migration and angiogenesis. Nat Med 2001; 9:1041–1047. Sugimoto Y, Narumiya S, Ichikawa A: Distribution and function of prostanoid receptors: Studies from knockout mice. Prog Lipid Res 2000;39:289–314. Morita I, Schindler M, Regier MK, Otto JC, Hori T, DeWitt DL, Smith WL: Different intracellular locations for prostaglandin endoperoxide H synthase-1 and -2. J Biol Chem 1995;270: 10902–10908. Bhattacharya M, Peri KG, Almazan G, Ribeiro-da-Silva A, Shichi H, Durocher Y, Abramovitz M, Hou X, Varma DR, Chemtob S: Nuclear localization of prostaglandin E2 receptors. Proc Natl Acad Sci USA 1998;95:15792–15797. Giardiello FM, Spannhake EW, DuBois RN, Hylind LM, Hubbard WC, Hamilton SR, Yang VW: Prostaglandin levels in human colorectal mucosa: Effects of sulindac in patients with familial adenomatous polyposis. Gastroenterology 1997;112:A568. Sheng H, Shao J, Morrow J, Beauchamp RD, DuBois RN: Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells. Cancer Res 1998;58: 362–366.
Cyclooxygenase-2 and Colorectal Cancer
135
39 40
41
42
43
44 45 46 47
48
49
50 51 52
53 54
55
56 57
Sheng H, Shao J, Washington MK, DuBois RN: Prostaglandin E2 increases growth and motility of colorectal carcinoma cells. J Biol Chem 2001;12:12. Adam L, Mazumdar A, Sharma T, Jones TR, Kumar R: A three-dimensional and temporo-spatial model to study invasiveness of cancer cells by heregulin and prostaglandin E2. Cancer Res 2001;61:81–87. Watanabe K, Kawamori T, Nakatsugi S, Ohta T, Ohuchida S, Yamamoto H, Maruyama T, Kondo K, Ushikubi F, Narumiya S, Sugimura T, Wakabayashi K: Role of the prostaglandin E receptor subtype EP1 in colon carcinogenesis. Cancer Res 1999;59: 5093–5096. Watanabe K, Kawamori T, Nakatsugi S, Ohta T, Ohuchida S, Yamamoto H, Maruyama T, Kondo K, Narumiya S, Sugimura T, Wakabayashi K: Inhibitory effect of a prostaglandin E receptor subtype EP(1) selective antagonist, ONO-8713, on development of azoxymethane-induced aberrant crypt foci in mice. Cancer Lett 2000;156:57–61. Sonoshita M, Takaku K, Sasaki N, Sugimoto Y, Ushikubi F, Narumiya S, Oshima Z, Taketo MM: Acceleration of intestinal polyposis through prostaglandin receptor EP2 in Apc(Delta 716) knockout mice. Nat Med 2001;9:1048–1051. DuBois RN, Giardiello FM, Smalley WE: Nonsteroidal anti-inflammatory drugs, eicosanoids, and colorectal cancer prevention. Gastroenterol Clin North Am 1996;25: 773–791. Paganini-Hill A, Hsu G, Ross RK, Henderson BE: Aspirin use and incidence of large-bowel cancer in a California retirement community. J Natl Cancer Inst 1991;83: 1182–1183. Thun MJ, Namboodiri MM, Heath CW Jr: Aspirin use and reduced risk of fatal colon cancer. N Engl J Med 1991;325:1593–1596. Smalley W, Ray WA, Daugherty J, Griffin MR: Use of nonsteroidal anti-inflammatory drugs and incidence of colorectal cancer: A population-based study. Arch Intern Med 1999;159: 161–169. Peleg II, Maibach HT, Brown SH, Wilcox CM: Aspirin and nonsteroidal anti-inflammatory drug use and the risk of subsequent colorectal cancer. Arch Intern Med 1994;154: 394–399. Rosenberg L, Palmer J, Zauber A, Warshauer M, Stolley P, Shapiro S: A hypothesis: Nonsteroidal anti-inflammatory drugs reduce the incidence of large-bowel cancer. J Natl Cancer Inst 1991; 83:355–358. Giovannucci E, Egan KM, Hunter DJ, Stampfer MJ, Colditz GA, Willett WC, Speizer FE: Aspirin and the risk of colorectal cancer in women. N Engl J Med 1995;333:609–614. Williams CS, Smalley W, DuBois RN: Aspirin use and potential mechanisms for colorectal cancer prevention. J Clin Invest 1997;100:1–5. Collet JP, Sharpe C, Belzile E, Boivin JF, Hanley J, Abenhaim L: Colorectal cancer prevention by non-steroidal anti-inflammatory drugs: Effects of dosage and timing. Br J Cancer 1999;81: 62–68. Waddell WR, Ganser GF, Cerise EJ, Loughry RW: Sulindac for polyposis of the colon. Am J Surg 1989;157:175–179. Giardiello FM, Hamilton SR, Krush AJ, Piantadosi S, Hylind LM, Celano P, Booker SV, Robinson CR, Offerhaus GJ: Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med 1993;328:1313–1316. Steinbach G, Lynch PM, Phillips RKS, Wallace MH, Hawk E, Gordon GB, Wakabayashi N, Saunders B, Shen Y, Fujimura T, Su LK, Levin B, Godio L, Patterson S, Rodriguez-Bigas MA, Jester SL, King KL, Schumacher M, Abbruzzese J, DuBois RN, Hittelman WN, Zimmerman S, Sherman JW, Kelloff G: The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–1952. Matsuhashi N, Nakajima A, Shinohara K, Oka T, Yazaki Y: Rectal cancer after sulindac therapy for a sporadic adenomatous colonic polyp. Am J Gastroenterol 1998; 93:2261–2266. Baron JA, Beach M, Mandel JS, van Stolk RU, Haile RW, Sandler RS, Rothstein R, Summers RW, Snover DC, Beck GJ, Bond JH, Greenberg ER: Calcium supplements for the prevention of colorectal adenomas. Calcium polyp prevention study group. N Engl J Med 1999;340: 101–107.
DuBois
136
58
59
Torrance CJ, Jackson PE, Montgomery E, Kinzler KW, Vogelstein B, Wissner A, Nunes M, Frost P, Discafani CM: Combinatorial chemoprevention of intestinal neoplasia. Nat Med 2000;6: 1024–1028. Mann M, Sheng H, Shao J, Williams CS, Pisacane PI, Sliwkowski MX, DuBois RN: Targeting cyclooxygenase 2 and HER-2/neu pathways inhibits colorectal carcinoma growth. Gastroenterology 2001;120:1713–1719.
Raymond N. DuBois, MD, PhD Department of Medicine/GI; MCN C-2104 Vanderbilt University Medical Center 1161 21st Ave. South, Nashville, TN 37232-2279 (USA) Tel. ⫹1 615 3225200, Fax ⫹1 615 343622, E-Mail
[email protected]
Cyclooxygenase-2 and Colorectal Cancer
137
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 138–162
Cyclooxygenase-2 in Lung Cancer Steven M. Dubinett, Sherven Sharma, Min Huang, Mariam Dohadwala, Mehis Pold, Jenny T. Mao Lung Cancer Research Program, UCLA Jonsson Comprehensive Cancer Center and Department of Medicine, Geffen School of Medicine at UCLA, Los Angeles, Calif., USA
Introduction
Lung cancer accounts for more than 28% of all cancer deaths each year and is the leading cause of cancer-related mortality in the USA [1]. Despite focused research in conventional therapies, the 5-year survival rate remains 14% and has improved only minimally in the past 25 years. Newly discovered molecular mechanisms in the pathogenesis of lung cancer provide novel opportunities for targeted therapies of non-small cell lung cancer (NSCLC) [2, 3]. These investigations in the molecular pathogenesis of lung cancer have presented translational researchers with new targets that specifically impact carcinogenesis [4]. Cyclooxygenase (COX)-2 is one of the novel targets under evaluation for lung cancer therapy and chemoprevention. COX (also referred to as prostaglandin endoperoxidase or prostaglandin G/H synthase) is the rate-limiting enzyme for the production of prostaglandins (PGs) and thromboxanes from free arachidonic acid [5]. The enzyme is bifunctional, with fatty acid COX (producing PGG2 from arachidonic acid) and PG hydroperoxidase activities (converting PGG2 to PGH2). Two forms of COX have now been described: a constitutively expressed enzyme, COX-1, present in most cells and tissues, and an inducible isoenzyme, COX-2 (also referred to as PGS-2), expressed in response to cytokines, growth factors and other stimuli [6, 7]. COX-2 has been reported to be constitutively overexpressed in a variety of malignancies [8–13] and is frequently constitutively elevated in human NSCLC [14–17]. In 1998 three studies provided the seminal documentation of the constitutive overexpression of COX-2 in human NSCLC [14–16]. In the first report, Huang et al. [14] utilized antibodies specific for human
COX-2 to evaluate NSCLC and normal adjacent lung resection specimens by immunohistochemistry. All of the 15 tumor specimens (8 adenocarcinomas and 7 squamous cell carcinomas) showed cytoplasmic staining for COX-2 in tumor cells. In contrast, adjacent normal lung showed no COX-2 staining in the alveolar lining epithelium, but demonstrated positive cytoplasmic staining often in alveolar macrophages and occasionally in bronchiolar epithelium (table 1). In a subsequent study by Wolff et al. [16], immunohistochemistry showed COX-2 staining in 19 of 21 adenocarcinomas and in all 11 squamous cell carcinomas studied, although the level of staining seemed to be less than that in the adenocarcinomas. Four small cell lung cancer specimens were reported to stain with a relatively weak intensity. Interestingly, abnormal alveolar epithelium in lung sections from patients with asbestosis or idiopathic fibrosing alveolitis expressed COX-2 protein. Patients with these pulmonary fibrotic disorders have an increased incidence of lung cancer [18, 19]. Hida et al. [15] reported that COX-2 overexpression was seen in approximately 70% of lung adenocarcinomas. In addition, COX-2 expression was documented in one third of atypical adenomatous hyperplasias and carcinoma in situ. This study also reported a greater proportion of lung cancer cells staining positively in lymph node metastases compared to the corresponding primary tumor [15]. COX-2 activity can be detected throughout the progression of a pre-malignant lesion to the metastatic phenotype [15]. Markedly higher COX-2 expression was observed in lung cancer lymph node metastasis compared to primary adenocarcinoma [15]. Recently, other studies have corroborated and expanded on the initial findings documenting the importance of COX-2 expression in lung cancer [16, 17, 20–25] (table 1). Khuri et al. [26] reported that COX-2 overexpression detected by in situ hybridization with riboprobes appears to portend a shorter survival among patients with early stage NSCLC. COX-2 expression in specimens from 160 patients with stage I NSCLC was evaluated. The strength of COX-2 expression was associated with a decreased overall survival rate (p ⫽ 0.001) and a diminished disease-free survival rate (p ⫽ 0.022). A preliminary report by West et al. [27] using immunohistochemistry assessment of COX-2 also suggests more aggressive behavior in NSCLC and decreased survival in those expressing COX-2. These reports, together with studies documenting an increase in COX-2 expression in precursor lesions [16, 17], suggest the involvement of COX-2 in the pathogenesis of lung cancer. Epidemiological studies that indicate a decreased incidence of lung cancer in subjects who regularly use aspirin have been interpreted as supporting this hypothesis [28–30]. Intriguing preliminary clinical studies suggest that COX-2 inhibitors may be active in resolution of carcinoma in situ [Mao et al., unpubl. data, 2002]. Mounting evidence indicates that tumor COX-2 activity has a multifaceted role in conferring the malignant and metastatic phenotypes. Although multiple
COX-2 in Lung Cancer
139
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
Table 1. COX-2 in human lung cancer Reference
COX-2 expression
Huang [14]
Adeno: Squamous:
8/8a 7/7a
First report of COX-2 expression in human lung cancer
Wolff [16]
Adeno: Squamous: Small cell weak (⫹):
19/21a 11/11a 4/4a
Abnormal epithelium from patients with asbestosis or pulmonary fibrosis who are at elevated risk for lung cancer also stained positively for COX-2
Hida [15]
Adeno: Squamous: Large cell: Adenosquamous:
7 low, 16 highb 19 low, 3 highb 3 low, 0 highb 1 low, 1 highb
Markedly higher and more homogeneous expression in lymph node metastases in comparison to primary tumors
Achiwa [21]
Adeno:
93/130a
Strong relationship between elevated COX-2 expression and shortened patient survival observed in stage I patients
Watkins [22]
Adeno: Large cell:
10, highestc 11, intermediate and variablec 13, low or absentc
Tumor cell COX-2 rather than COX-1 expression may account for the variable prostanoid production seen in NSCLC, and primary lung adenocarcinoma expresses the highest levels of COX-2
Squamous:
Comments
140
Ochiai [23]
COX-2 expression by RT-PCR elevated in 29 NSCLC specimens relative to normal lung
In 9/9 cases metastatic lymph nodes were not significantly different from primary adenocarcinoma
Hosomi [17]
COX-2 overexpression was detected in over 80% of CCH, AAH, BAC and I-Ad
Increased COX-2 expression found in possible precursors to adenocarcinoma
Soslow [25]
Adeno ⫹ squamous:
Further possibility of COX-2 paracrine effect is observed in adjacent non-neoplastic epithelium
18/20a
COX-2 in Lung Cancer
Khuri [26]
Strong: Intermediate: Weak: Not detectable:
03/160d 69/160d 24/160d 64/160d
COX-2 expression was associated with a worse overall survival rate. The median survival times for the strong, intermediate or weak, and null COX-2 expressors were 1.04, 5.50 and 8.54 years, respectively
Hasturk [24]
Adeno: Squamous: Smoker tumor: Non-smoker tumor:
21/51a 9/46a 29/91a 1/10a
Utility of COX-2 as a chemoprevention marker questioned Concluded that COX-2 status may be useful in designing treatment strategies
Brabender [20]
COX-2 mRNA: Expression in curatively resected NSCLC
89/89
Significant relationship between COX-2 expression and inferior survival rate observed
Adeno ⫽ Adenocarcinoma; AAH ⫽ atypical adenomatous hyperplasia; BAC ⫽ bronchioloalveolar carcinoma; CCH ⫽ cuboidal cell hyperplasia; COX-2 ⫽ cyclooxygenase-2; I-Ad ⫽ invasive adenocarcinoma; Large cell ⫽ large cell carcinoma; mRNA ⫽ messenger ribonucleic acid; NSCLC ⫽ non-small cell lung cancer; Small cell weak ⫽ small cell lung cancer specimens of relatively weak intensity on immunohistochemistry staining; Squamous ⫽ squamous cell carcinoma. a Number staining positively by immunohistochemistry/total. b Intensity of staining. c In comparison to other tumor types. d COX-2 expression by in situ hybridization in stage I NSCLC specimens.
141
genetic alterations are necessary for lung cancer invasion and metastasis, COX-2 may be a central element in orchestrating this process [14–17, 21]. Studies indicate that overexpression of COX-2 is associated with apoptosis resistance [31–34], angiogenesis [11, 35–38], decreased host immunity [14, 39], and enhanced invasion and metastasis [40]. Thus, COX-2 can impact multiple mechanistic pathways in lung cancer carcinogenesis. COX-2 activities are particularly relevant in the pathogenesis of lung cancer for the following reasons: COX-2 can activate tobacco smoke carcinogens such as benzo[a]pyrene (B[a]P) [41, 42]. One example of this activity is the capacity of COX-2 to catalyze the conversion of B[a]P-7,8-dihydrodiol to B[a]P-diol epoxide which binds to DNA [43]. Thus, the capacity of COX-2 to activate environmental carcinogens including polycyclic hydrocarbons suggests that it plays an important role in tobacco-induced carcinogenesis [41]. In addition, B[a]P itself has also been demonstrated to potentially upregulate epithelial cell COX-2 expression and PGE2 production [44]. COX-2 is readily inducible by a variety of stimuli that are often highly represented in the pulmonary microenvironment of those at risk for lung cancer. For example, these factors include TGF-, IL-1, hypoxia, B[a]P, and epidermal growth factor (EGF) [7]. Whereas IL-10 can downregulate COX-2 in host inflammatory cells, this capacity for IL-10 signaling is lost in human NSCLC [45]. In addition to growth factor and cytokine induction of COX-2, the enzyme may be constitutively upregulated by virtue of oncogene expression and mutational events in lung cancer development. For example, wild-type, but not mutant p53, suppresses COX-2 transcription, thus suggesting that p53 status in lung cancer may be one of the determinants of COX-2 expression [46]. Other mutational events associated with elevated COX-2 expression include K-ras and -catenin [47–50]. Wardlaw et al. [51] documented a critical role for the CCAAT/ enhancer-binding protein (C/EBP) and activating transcription factor/cAMP response element-binding protein (ATF/CREB) in the regulation of basal COX-2 expression in murine lung carcinomas. It was suggested that C/EBP and ATF/CREB may serve as new targets for downregulating COX-2 expression in lung cancer [51].
COX-2-Dependent Regulation of Immunity in Lung Cancer
Lung cancer cells elaborate immunosuppressive mediators including type 2 cytokines, PGE2 and transforming growth factor- (TGF-) that may interfere directly with cell-mediated anti-tumor immune responses [14, 52–54]. In addition to producing their own suppressive factors, tumor cells may also direct surrounding inflammatory cells to release suppressive cytokines in the tumor
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
142
milieu [52, 55]. Tumor-derived PGE2 is one mediator that orchestrates an imbalance in the production of suppressive and immune potentiating cytokines by lymphocytes and macrophages in the tumor environment [14, 39]. COX-2-Dependent Regulation of Cytokine Balance Studies by Huang et al. [14] showing COX-2 overexpression in human lung cancer, found that tumor-derived, high-level PGE2 production mediated dysregulation of host immunity by altering the balance of interleukin (IL)-10 and IL-12. IL-10 and IL-12 are critical regulatory elements of cell-mediated immunity. While IL-10 inhibits cellular immunity, IL-12 induces type 1 cytokine production and effective cell-mediated immunity [56, 57]. IL-10 overproduction at the tumor site has been implicated in tumor-mediated immunosuppression [58–62], enhanced angiogenesis [63], and appears to be an indicator of poor prognosis in NSCLC [64–67]. In contrast, IL-12 is critical for effective antitumor immunity [68, 69]. Tumor models indicate that the tumor-bearing state induces lymphocyte and macrophage IL-10 production but inhibits macrophage IL-12 [39, 70]. Thus, whereas IL-12 is the key inducer of type 1 cytokines, IL-10 production at the tumor site may inhibit type 1 cytokine production and cell-mediated anti-tumor immunity. Importantly, lung cancerderived PGE2 has been found to induce a 10- to 100-fold increase in lymphocyte IL-10 production [52]. It was hypothesized that high level PGE2 production by lung tumor cells is dependent on tumor COX-2 expression. PGE2 production by A549 NSCLC cells was found to be elevated up to 50-fold in response to IL-1. Reversal of IL-1-induced PGE2 production in A549 cells was achieved by specific pharmacologic or antisense oligonucleotide inhibition of COX-2 activity or expression. In contrast, specific COX-1 inhibition was not effective. Consistent with these findings, IL-1 induced COX-2 mRNA expression and protein production in A549 cells. Specific inhibition of COX-2 abrogated the capacity of IL-1-stimulated A549 cells to induce IL-10 in lymphocytes and macrophages. Furthermore, specific inhibition of A549 COX-2 reversed the tumor-derived PGE2-dependent inhibition of macrophage IL-12 production when whole blood was cultured in tumor supernatants. These results indicate that lung tumorderived PGE2 plays a pivotal role in promoting lymphocyte and macrophage IL-10 induction, while simultaneously inhibiting macrophage IL-12 production [14]. Thus, these studies demonstrated functional COX-2 expression by NSCLC cells and the definition of a pathway whereby tumor COX-2 expression and high-level PGE2 production mediate profound alteration in cytokine balance in the lung cancer microenvironment [14]. To evaluate lung tumor COX-2 modulation of anti-tumor immunity in vivo, Stolina et al. [39] studied the effect of specific genetic or pharmacological
COX-2 in Lung Cancer
143
inhibition of COX-2 in a murine Lewis lung carcinoma (3LL) model. Specific inhibition of COX-2 led to significant tumor reduction in vivo in murine lung cancer models. Both specific genetic or pharmacologic inhibition of COX-2 led to marked lymphocytic infiltration of the tumor and reduced tumor growth. Treatment of mice with anti-PGE2 mAb replicated the growth reduction seen in tumor-bearing mice treated with COX-2 inhibitors. COX-2 inhibition was accompanied by a significant decrement in IL-10 and a concomitant restoration of IL-12 production by antigen-presenting cells (APCs). Because PGE2 is a potent inducer of IL-10, it was hypothesized that COX-2 inhibition led to antitumor responses by downregulating production of this potent immunosuppressive cytokine. In support of this concept, transfer of IL-10 transgenic T lymphocytes that overexpress IL-10 under control of the IL-2 promoter [61, 62] reversed the COX-2 inhibitor-induced anti-tumor response [39]. COX-2-Dependent Regulation of APC Function Anti-tumor immune responses require the coordinate activities of lymphocyte effectors and professional APC [71]. Dendritic cells (DCs) are professional APC that are pivotal participants in the initiation of T-cell responses [72]. DCs acquire Ag in the periphery and subsequently transport it to lymphoid organs where they prime specific immune responses [72]. The tumor microenvironment can adversely affect DC maturation and function [73]. Studies indicate that COX-2 metabolites can play a major role in tumor-induced inhibition of DC differentiation [74]. Prostanoids have been found to mediate these effects by both IL-10-dependent [75, 76] and independent pathways [74]. To define the pathways limiting DC function in the tumor environment, bone marrow-derived DCs were cultured in murine lung cancer tumor supernatants (TSN) [77]. Although pulsed with tumor-specific peptides these DCs were incapable of generating anti-tumor immune responses in vivo. When injected into established murine lung cancer, DCs generated in TSN caused immunosuppressive effects that correlated with enhanced tumor growth. Genetic or pharmacological inhibition of murine lung cancer COX-2 expression restored DC function and effective anti-tumor immune responses. Functional analyses indicated that TSN caused a decrement in DC capacity to (1) process and present antigens; (2) induce alloreactivity, and (3) secrete IL-12. These limitations in DC activity were prevented when DCs were cultured in SN from COX-2-inhibited tumors. Whereas TSN DCs showed a significant reduction in cell surface expression of CD11c, DEC205, MHC class I, MHC class II, CD80, CD86, as well as a reduction in the transporter-associated proteins, TAP1 and TAP2, these changes were not evident when DCs were cultured in SN from COX-2-inhibited tumors. Thus, inhibition of COX-2 expression or activity can prevent tumor-induced suppression of DC activities.
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
144
COX-2-Dependent Regulation of Angiogenesis in Lung Cancer
In vivo, expansion and maintenance of a functional vascular network serving the tumor is required for propagation, invasion and subsequent metastasis [78]. Thus, angiogenesis is requisite for tumor growth [79] and has been specifically implicated in the pathogenesis and prognosis of lung cancer [80–86]. Several growth factors and cytokines have been implicated in tumor-related angiogenesis in lung cancer including vascular endothelial growth factor (VEGF), transforming growth factors ␣ and , basic fibroblast growth factor (FGF) [87–90] and chemokines such as IL-8 [91]. Tumor suppressor genes, oncogenes and immune responses have been implicated in complex regulation of these proteins involved in the angiogenic process [92–94]. COX-2 has been shown to promote angiogenesis in vitro [95, 96] and in vivo [36, 97]. Masferrer et al. [36] found that FGF-2-induced angiogenesis is dependent on COX-2 expression. Consistent with findings from other groups [39, 97], Masferrer et al. found that the COX-2 inhibitor celecoxib significantly decreased tumor growth in the Lewis lung carcinoma model (LLC). In this model, predominant COX-2 expression in the tumor-associated vasculature was noted. These findings suggested that COX-2-derived PGs contribute to tumor growth by inducing neovascularization. Thus, COX-2 inhibition may contribute to the anti-tumor response by downregulating angiogenic activities in lung cancer. To assess the role of host-derived COX-2 in the LLC model, Williams et al. [97] studied the lung cancer growth in COX-2⫺/⫺ mice. In contrast to C57BL/6 wild-type or COX-1⫺/⫺ mice in which LLC tumors developed and grew rapidly, tumor growth in COX-2⫺/⫺ mice was significantly attenuated. Compared to wild-type or COX-1⫺/⫺ mice, fibroblasts from COX-2⫺/⫺ mice were found to have a greater than 90% reduction in VEGF. These findings strongly support the concept that host stromal elements can enhance tumor growth, promoting a ‘landscape’ in which vasculature is maintained and expanded [98, 99]. Thus, in addition to the tumor cells’ contribution, host-derived COX-2 appears to regulate important angiogenic mediators in the tumor milieu by contributing products such as PGs whose paracrine effects impact potently promote tumor growth. Recent studies by Leahy et al. [100] add further support and insight into the contribution of stromal elements in COX-2-dependent tumor growth. Regardless of tumor cell expression of COX-2, the neovascular cells associated with tumors consistently demonstrated heightened COX-2 expression [100, 101]. In both cornea and tumor angiogenic models, Leahy et al. [100] found that COX-2 is expressed in vascular endothelial cells. In keeping with these findings, specific COX-2 inhibition significantly limited angiogenesis. These pre-clinical studies are consistent with the study of Marrogi et al. [102]; immunostaining for COX-2 correlated positively with VEGF status in human NSCLC sections.
COX-2 in Lung Cancer
145
COX-2-Dependent Regulation of Invasion in Lung Cancer
The complex events associated with tumor cell invasion include the active movement of cells across the extracellular matrix and spread to distant organ sites [103]. To assess the impact of COX-2 expression in lung cancer invasiveness, NSCLC cell lines were transduced with a retroviral vector expressing the human COX-2 cDNA in the sense (COX-2-S) and antisense (COX-2-AS) orientations [104]. COX-2-S clones expressed significantly more COX-2 protein, produced 10-fold more PGE2 and demonstrated an enhanced invasive capacity compared to control vector-transduced or parental cells. CD44 is a cell surface receptor for hyaluronate, a major glycosaminoglycan component of extracellular matrix. Adhesion to extracellular matrix, a critical initial step in the metastatic process, has been found to be CD44dependent in several malignancies [105–108]. CD44 was overexpressed in COX-2-S cells, and specific blockade of CD44 significantly decreased tumor cell invasion. In contrast, COX-2-AS clones had a very limited capacity for invasion and showed diminished expression of CD44. These findings indicate that a COX-2-mediated, CD44-dependent pathway is operative in NSCLC invasion [104]. Subsequent studies focus on the role of tumor-derived PGE2 in modulating COX-2-dependent NSCLC invasion [109]. PGE2, produced at heightened levels in COX-2 overexpressing tumor cells, affects target cells through interaction with G-protein-coupled EP receptors of four distinct subtypes. The pathways whereby autocrine/paracrine production of PGE2 could impact the invasive phenotype via EP receptor signaling in NSCLC have been studied. In addition to CD44, matrix metalloproteinase (MMP) production may be critical in lung cancer invasion. CD44 is known to induce co-clustering with MMPs and can therefore promote MMP activity, tumor invasion and angiogenesis [105, 110]. Antibody-mediated blockade of tumor-derived PGE2 decreased CD44 and MMP-2 expression as well as invasion. In addition, exposure of NSCLC cells to exogenous PGE2 upregulates CD44, EP4 receptor and MMP-2 expression and potently enhances invasion [109]. These studies indicate an important autocrine/paracrine role for PGE2 in the regulation of CD44 and MMP-2dependent invasion in human NSCLC. These findings in NSCLC cells are consistent with investigations in monocytes in which PGE2 was found to activate membrane type 1-MMP (MT1-MMP) and thereby promote activation of MMP-2 [111]. The role of MT1-MMP in the COX-2-dependent regulation of NSCLC cells has not yet been investigated. In contrast to the mechanisms described in the studies above, Pan et al. [112] reported that NSAIDs, including the COX-2 inhibitor NS-398, could inhibit MMP-2 transcription in lung cancer cells by suppressing promoter activity.
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
146
In the evaluation of a highly metastatic human lung cancer cell line, Kozaki et al. [113] found that the level of COX-2 expression correlated with motility and invasion in vitro as well as metastatic potential in vivo. In accord with these findings, COX-2 inhibitors were found to decrease lung cancer metastases in vivo [113]. In related studies the expression of COX-2 and laminin 5 were found to be frequently co-expressed in early stage lung adenocarcinomas [114]. Laminin 5, an extracellular matrix protein involved in cell migration and invasion, has been found to be frequently expressed in several different malignancies [115–118]. Immunohistochemical analysis of 102 lung adenocarcinomas which were ⱕ2 cm in size revealed that COX-2 and laminin 5 were expressed in 95 and 80%, respectively. An overall significant correlation was found between expression levels of COX-2 and laminin 5 and frequent co-expression of these proteins was noted at the lung adenocarcinoma invasive front [114]. Common pathways may induce these proteins; EGF receptor signaling has been shown to induce COX-2 [119] as well as laminin 5 [114, 120]. In fact, in squamous carcinoma cell lines the expression level of laminin 5 correlates with gene amplification of EGFR [120]. The notion that EGFR signaling is a common upstream regulator of COX-2 and laminin 5 is supported by the fact that lung adenocarcinomas that overexpress EGFR and erbB2 have been shown to have higher levels of COX-2 and laminin 5 compared to those without concomitant overexpression of the receptors [114]. Another possibility that has not yet been assessed is that COX-2-derived PGs may modulate laminin 5 levels. Alternatively, as suggested by the recent studies of Pai et al. [121], PGE2 has the capacity to transactivate EGFR and thus may impact laminin 5 expression as well as other proteins via this pathway.
COX-2-Dependent Regulation of Apoptosis in Lung Cancer
Dysregulation of apoptosis is intimately involved in carcinogenesis and a broad variety of anti-cancer agents mediate their effects by induction of apoptosis [122]. Thus, heightened apoptosis resistance may be responsible for drug or radiation resistance in lung cancer therapy. Apoptosis induction has been widely investigated and consistently supported in studies that seek to define the potential anti-neoplastic mechanisms of COX-2 inhibition. In the landmark studies of Tsujii and DuBois [31], forced expression of COX-2 was found to increase apoptosis resistance and Bcl-2 expression. Subsequent studies in a variety of tumors suggest that both Bcl-2-dependent [123] and independent [33] pathways may be operative. COX-2-selective inhibitors have been shown to induce apoptosis in several different types of tumors [34, 123–126] including
COX-2 in Lung Cancer
147
lung cancer [127–129]. Lin et al. [130] found that either overexpression of COX-2 or exposure to PGE2 can increase the apoptosis threshold in lung adenocarcinoma cells by upregulation of the mcl-1 gene in a PI3K/Akt-dependent manner. The practical implications of apoptosis regulation by NSAIDs include their potential use in combination with chemotherapy and radiation therapy. Thus, cells overexpressing COX-2 have been found to resist apoptosis [31] and COX-2 inhibition can promote apoptosis in these cells [32–34, 124, 128, 129, 131–134]. Hida et al. [127] found that the COX-2 inhibitor nimesulide can induce apoptosis in NSCLC cell lines. In vitro evaluation of nimesulide as an adjunct to chemotherapy revealed that the IC50 values of various anticancer agents including etoposide and cisplatin were significantly reduced [127]. These in vitro studies have recently also been verified in vivo [135]. A distinct benefit of combining COX-2 inhibition with chemotherapy is the possibility of limiting chemotherapy-induced COX-2 expression by tumor cells. Subbaramaiah et al. [136] found that microtubule-interfering agents such as taxol have the capacity to stimulate COX-2 transcription via ERK and p38 MAP kinase pathways. Thus, tumor cells that escape apoptosis induction by microtubuleinterfering agents such as taxol may become promoters of angiogenesis, invasion, apoptosis resistance and immune dysregulation as a function heightened COX-2 expression. These findings are consistent with those of Moos et al. [137] and Cassidy et al. [138] who found that taxanes and their analogues increased macrophage COX-2 expression. These latter studies present a potential pathway for paracrine PGE2 production to negatively impact lung cancer chemotherapy regimens. These findings promote the rationale for clinical evaluation of COX-2 inhibitors in combination with microtubule-interfering agents in therapy for NSCLC. Radiation-induced apoptosis can be significantly enhanced by COX-2 inhibitors [139] and this increase in apoptosis has been demonstrated in lung cancer models [127, 140]. Milas et al. [141, 142] were the first to suggest that NSAID-induced radiation responsiveness may be related to neoangiogenesis and host immune competence. Several studies have now documented that COX-2 inhibitors potentiate radiation therapy in model systems by enhancing radiation-induced apoptosis [143–146]. While certain COX-2 inhibitors primarily induce apoptosis, others may predominantly induce growth arrest [147]. In addition, an individual NSAID may induce anti-tumor effects via different mechanistic pathways in different types of tumors. For example, sulindac, a nonselective COX-2 inhibitor, acts to induce apoptosis in the human colon cancer cell line HT-29 but induces predominantly growth arrest in human lung cancer cells [148, 149]. Culture conditions, including the NSAID concentration, may also impact which mode of
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
148
action predominates [128, 147]. Thus, in parallel to the induction of apoptosis, COX-2 inhibitor-induced alterations in cell cycle progression have been documented [150, 151]. Mammalian cell cycle progression is governed by cyclins, cyclin-dependent kinases (CDKs) and their inhibitors (CDKIs) [152]. By binding CDKs, cyclins activate kinase activity and promote cell cycle progression. In contrast, CDKIs, including members of the kinase inhibitor protein family which bind CDK-cyclin complexes, inhibit cell cycle progression. In a study specifically addressing cell cycle regulation by COX-2 inhibitors in human lung cancer cells, Hung et al. [151] found that p27KIP1 was upregulated in response to the COX-2 inhibitor NS398. Normal epithelial cells, including lung epithelium, usually express high levels of p27KIP1, however, a decrease of this tumor suppressor gene product is found in NSCLC; more that 70% of NSCLC tumors show reduced p27KIP1 and the intracellular level of this protein is predominantly regulated by translational or post-translational mechanisms [153–156]. Because NS398 inhibited the degradation of p27KIP1, it was suggested that the regulation of this protein constitutes another mechanistic pathway of COX-2 inhibitor-mediated tumor growth-suppressive effects in lung cancer [151]. In addition to COX-2-dependent effects, NSAIDs can also act to induce apoptosis by COX-2-independent pathways [157–159]. In support of the concept of COX-2-independent effects, NSAIDs induce apoptosis in cancer cells that do not express COX-2 [160, 161]. In addition, the dose of NSAIDs used in some studies to induce apoptosis exceeds the amount required to inhibit COX-2 enzymatic activity [157]. NSAIDs could act in a COX-2-independent manner through PPAR-␦ [162], NFB [163], AP-1 [164], or arachidonic acid [165]. Delineation of these additional COX-2-independent pathways will facilitate the rational use of COX-2 inhibitors in lung cancer therapy and prevention [158]. COX-2 inhibitors have been found to enhance photodynamic therapy (PDT) [166], another apoptosis inducing modality that may be helpful in NSCLC [167–169]. In PDT, the systemic administration of tumor-localizing photosensitizer is followed by focal light activation resulting in generation of cytotoxic reactive oxygen species within the tumor [170]. The rationale for examining PDT and COX-2 inhibition in combination was based on previous studies that documented the release of PGE2 from cells following PDT [171] and the suggestion that COX products may play a role in PDT-induced vascular effects [172]. Ferrario et al. [166] found that PDT upregulated COX-2 expression due, in part, to transcriptional activation. The combination of PDT and COX-2 inhibitors resulted in greater anti-tumor efficacy than did either therapy alone. These findings raise the possibility that COX-2 inhibitors may enhance the efficacy of PDT in NSCLC.
COX-2 in Lung Cancer
149
COX-2 Inhibition in Treatment and Chemoprevention of Lung Cancer
Because pre-clinical data strongly suggest that overexpression of COX-2 plays a critical role in tumor-mediated angiogenesis, invasion and metastasis, apoptosis resistance and immune dysregulation, clinical trials are currently evaluating COX-2 inhibition in the context of treatment and chemoprevention of lung cancer. This section reviews several of these studies as examples of the lung cancer clinical areas under investigation. In order to prevent chemotherapy-induced overexpression of COX-2 [136–138] and take advantage of a relatively non-toxic intervention that could have additive or synergistic effects with conventional therapies, investigations have begun assessing COX-2 inhibitors combined with chemotherapy. Csiki et al. [173] recently reported preliminary results in the evaluation of COX-2 inhibition plus docetaxel in recurrent NSCLC. This phase II trial utilizes celecoxib (400 mg bid) and docetaxel (75 mg/m2 i.v. every 3 weeks). There was a significant decline of intratumoral PGE2 levels following treatment. Thus, these preliminary data indicate that celecoxib in combination with chemotherapy can significantly decrease PGE2 within tumor tissues, suggesting that COX-2-dependent expression of genes that are deleterious to the anti-tumor response may also be decreased. Based on results of the BLOT trial of neoadjuvant paclitaxel/carboplatin alone [174], investigators are evaluating the role of celecoxib plus paclitaxel/ carboplatin in the preoperative setting. Altorki et al. [175] assessed the role of a selective COX-2 inhibitor, celecoxib, as an adjunct to preoperative chemotherapy in patients with resectable NSCLC. Twenty-six patients completed a phase II trial of preoperative chemotherapy and celecoxib for stage IB-IIIA NSCLC. Two cycles of paclitaxel (225 mg/m2) and carboplatin were given 3 weeks apart. Celecoxib was given orally at 400 mg bid from day 1 until the day of surgery. Surgery was performed on days 42–56. Preoperative clinical stages included patients with IB (13), IIB (3) or IIIA (10) disease. All 26 patients completed induction therapy and 25 patients were evaluable for response. There was no significant unexpected chemotherapy-related toxicity. Overall response rate was 68% (48% PR, 20% CR). All 25 patients were explored and resected. There were no complete pathological responses, but 7 (28%) had minimal residual microscopic disease. It was concluded that in comparison to historically reported response rates, the addition of a selective COX-2 inhibitor may enhance the response to preoperative paclitaxel/ carboplatin in NSCLC [175]. Confirmatory trials are planned. To assess the role of COX-2 inhibitors as potential radiation sensitizers in NSCLC, Carbone et al. [176] initiated a phase II trial using celecoxib 400 mg
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
150
p.o. bid until progression plus concurrent weekly paclitaxel (50 mg/m2)/carboplatin (AUC 2) and chest radiation therapy (63 Gy) for stage III NSCLC patients. Nine patients with stage III NSCLC had been enrolled at the time of the preliminary report. Three patients out of these 9 developed grade 3 esophagitis, and 1 patient had grade 3 pneumonitis. Three out of 5 evaluable patients had objective CR or PR. Blood and urine specimens were obtained pre-, post-celecoxib and every 2 months for VEGF and PGE-M (the major urinary metabolite of PGE2) assays. VEGF levels from the first 8 patients showed that post-celecoxib, serum/plasma levels of VEGF do not show a consistent pattern whereas VEGF levels fall in the months following treatment. These preliminary data indicate that COX-2 inhibition may lead to changes in serum/plasma VEGF.
COX-2 Inhibition in Chemoprevention of Lung Cancer
Several lung cancer chemoprevention trials have not demonstrated a decrease in the incidence of lung cancer, however, unexpected adverse effects have been noted [177–179]. Generally, these phase III trials were designed based on epidemiological or animal data, without the use of systematic pilot phase I/II trials [177–179]. Thus, a systematic approach is needed, with welldesigned pilot trials to determine feasibility, to evaluate promising chemopreventive agents. According to the field carcinogenesis theory, malignant transformation may occur throughout the respiratory epithelium, at multiple independent sites simultaneously [180, 181]. Thus, systemic therapy targeting the process of tumorigenesis may reverse the progression of pre-malignancy and prevent the development of lung cancer. Pre-clinical models show that COX-2 expression is abundant in alveolar type II cells in lung cancer-sensitive mouse strains and in pre-malignant lesions [182]. Inhibition of COX in these models, with NSAIDs or COX-2-specific inhibitors, slows tumorigenesis [183]. COX-2 expression and PG levels appear to be key factors contributing to lung carcinogenesis. Animal models demonstrate that COX-2-specific inhibitors protect against tumorigenesis resulting from exposure to the tobacco-specific nitrosamine, NNK, through bioactivation of NNK [183]. In addition, overexpression of COX-2 and increased PG levels are associated with several well-established risk factors for lung cancer, including upregulation of EGFR [184], epithelial cell proliferation [185], microvascular angiogenesis [36], and resistance to apoptosis [186]. Upregulation of PG synthesis in the lung microenvironment also produces immunosuppressive effects which may interfere with anti-tumor immunity and promote tumor growth [14, 187–189].
COX-2 in Lung Cancer
151
Airflow obstruction OR resected stage I NSCLC
Randomize
n ⫽ 180 Smoking history ⱖ30 pack/years
Celecoxib bid ⫻ 6 months
Placebo bid ⫻ 6 months
Placebo bid ⫻ 6 months
Celecoxib bid ⫻ 6 months
Fig. 1. Randomized phase II trial of celecoxib for chemoprevention of NSCLC: study design.
Ongoing Phase II Trials of COX-2 Inhibition for Chemoprevention of NSCLC
Two ongoing phase II trials at UCLA are evaluating the role of celecoxib in chemoprevention of NSCLC. The objective of both trials is to determine the feasibility of celecoxib treatment for chemoprevention of lung cancer in populations at high risk of developing primary or secondary lung cancers. The trials are evaluating the effect of celecoxib on cellular and molecular events associated with lung carcinogenesis, including (1) modulation of biomarkers; (2) regulation of arachidonic acid metabolism; (3) anti-tumor immunity, and (4) angiogenesis in the lung microenvironment. The safety and long-term effects of treatment will also be monitored. A pilot, single-arm trial is enrolling smokers at high risk for developing lung cancer, defined as age ⱖ45 years and smoking history of ⱖ20 pack-years with or without evidence of airflow obstruction (FEV1 (80%)). Trial objectives are to determine the efficacy and feasibility as well as to evaluate the safety and long-term side effects of celecoxib treatment in active smokers. Treatment comprises celecoxib orally twice daily for 6 months and patients are evaluated at 2 weeks and then 6 months. A randomized phase II trial will further evaluate celecoxib in a larger population of 180 former smokers (ⱖ30 pack-years) with either evidence of airflow obstruction or a history of successful surgical resection for stage I NSCLC. Patients are randomized (1-to-1, double-blind, placebo-controlled, crossover design) to treatment as summarized in figure 1. A chemoprevention trial evaluating COX-2 inhibition in current and former smokers is also underway at M.D. Anderson Cancer Center [160]. Molecular epidemiologic studies are needed to identify reliable biomarkers that are highly predictive of lung cancer risk. This ‘molecular profiling’ will efficiently characterize the highest-risk population for enrollment in trials of
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
152
chemoprevention, permitting smaller sample size, therapeutic stratification, and shorter trial duration. In addition, reliable surrogate endpoint markers need to be identified and validated to monitor the therapeutic efficacy of lung cancer chemoprevention strategies including those evaluating COX-2 inhibition. The selection of biomarkers to be evaluated and additional targets in lung cancer prevention and therapy may be facilitated by COX-2-dependent gene discovery programs [113, 190]. The transformation to malignancy can occur throughout the respiratory epithelium. Systemic therapy targeting molecular processes involved in lung cancer carcinogenesis, such as COX-2-specific inhibition, has the potential to slow tumorigenesis [183]. Ongoing trials of COX-2 inhibitors in chemoprevention of NSCLC will determine the feasibility of this approach, as well as the effects of COX-2 inhibitors in lung carcinogenesis.
Acknowledgement Supported by the UCLA SPORE in Lung Cancer (National Institutes of Health Grant P50 CA90388).
References 1 2 3 4 5 6 7 8 9
10
11 12
Greenlee RT, Murray T, Bolden S, Wingo PA: Cancer statistics, 2000. CA Cancer J Clin 2000; 50:7–33. Dy GK, Adjei AA: Novel targets for lung cancer therapy. I. J Clin Oncol 2002;20:2881–2894. Dy GK, Adjei AA: Novel targets for lung cancer therapy. II. J Clin Oncol 2002;20:3016–3028. Hirsch FR, Franklin WA, Bunn PA Jr: Expression of target molecules in lung cancer: Challenge for a new treatment paradigm. Semin Oncol 2002;29:2–8. Herschman H: Review: Prostaglandin synthase 2. Biochim Biophys Acta 1996;1299:125–140. DuBois RN, Abramson SB, Crofford L, Gupta RA, Simon LS, Van De Putte LB, Lipsky PE: Cyclooxygenase in biology and disease. Faseb J 1998;12:1063–1073. Smith WL, DeWitt DL, Garavito RM: Cyclooxygenases: Structural, cellular and molecular biology. Annu Rev Biochem 2000;69:145–182. Soslow R, Dannenberg A, Rush D, Woerner B, Khan K, Masferrer J, Koki A: COX-2 is expressed in human pulmonary, colonic and mammary tumors. Cancer 2000;89:2637–2645 Yip-Schneider MT, Barnard DS, Billings SD, Cheng L, Heilman DK, Lin A, Marshall SJ, Crowell PL, Marshall MS, Sweeney CJ: Cyclooxygenase-2 expression in human pancreatic adenocarcinomas. Carcinogenesis 2000;21:139–146. Sheng H, Shao J, Dixon DA, Williams CS, Prescott SM, DuBois RN, Beauchamp RD: Transforming growth factor-1 enhances Ha-ras-induced expression of cyclooxygenase-2 in intestinal epithelial cells via stabilization of mRNA. J Biol Chem 2000;275:6628–6635. Liu X, Kirschenbaum A, Yao S, Lee R, Holland J, Levine A: Inhibition of cyclooxygenase-2 suppresses angiogenesis and the growth of prostate cancer in vivo. J Urol 2000;164:820–825. Shamma A, Yamamoto H, Doki Y, Okami J, Kondo M, Fujiwara Y, Yano M, Inoue M, Matsuura N, Shiozaki H, Monden M: Up-regulation of cyclooxygenase-2 in squamous carcinogenesis of the esophagus. Clin Cancer Res 2000;6:1229–1238.
COX-2 in Lung Cancer
153
13
14
15
16 17
18 19 20
21
22 23 24
25 26
27
28 29 30 31 32 33
34
Chan G, Boyle JO, Yang EK, Zhang F, Sacks PG, Shah JP, Edelstein D, Soslow RA, Koki AT, Woerner BM, Masferrer JL, Dannenberg AJ: Cyclooxygenase-2 expression is up-regulated in squamous cell carcinoma of the head and neck. Cancer Res 1999;59:991–994. Huang M, Stolina M, Sharma S, Mao J, Zhu L, Miller P, Wollman J, Herschman H, Dubinett S: Non-small cell lung cancer cyclooxygenase-2-dependent regulation of cytokine balance in lymphocytes and macrophages: Up-regulation of interleukin-10 and down-regulation of interleukin-12 production. Cancer Res 1998;58:1208–1216. Hida T, Yatabe Y, Achiwa H, Muramatsu H, Kozaki KI, Nakamura S, Ogawa M, Mitsudomi T, Sugiura T, Takahashi T: Increased expression of cyclooxygenase-2 occurs frequently in human lung cancers, specifically in adenocarcinomas. Cancer Res 1998;58:3761–3764. Wolff H, Saukkonen K, Anttila S, Karjalainen A, Vainio H, Ristimaki A: Expression of cyclooxygenase-2 in human lung carcinoma. Cancer Res 1998;58:4997–5001. Hosomi Y, Yokose T, Hirose Y, Nakajima R, Nagai K, Nishiwaki Y, Ochiai A: Increased cyclooxygenase-2 expression occurs frequently in precursor lesions of human adenocarcinoma of the lung. Lung Cancer 2000;30:73–81. Bouros D, Hatzakis K, Labrakis H, Zeibecoglou K: Association of malignancy with diseases causing interstitial pulmonary changes. Chest 2002;121:1278–1289. Hillerdal G, Henderson DW: Asbestos, asbestosis, pleural plaques and lung cancer. Scand J Work Environ Health 1997;23:93–103. Brabender J, Park J, Metzger R, Schneider PM, Lord RV, Holscher AH, Danenberg KD, Danenberg PV: Prognostic significance of cyclooxygenase-2 mRNA expression in non-small cell lung cancer. Ann Surg 2002;235:440–443. Achiwa H, Yatabe Y, Hida T, Kuroishi T, Kozaki K, Nakamura S, Ogawa M, Sugiura T, Mitsudomi T, Takahashi T: Prognostic significance of elevated cyclooxygenase-2 expression in primary, resected lung adenocarcinomas. Clin Cancer Res 1999;5:1001–1005. Watkins DN, Lenzo JC, Segal A, Garlepp MJ, Thompson PJ: Expression and localization of cyclo-oxygenase isoforms in non-small cell lung cancer. Eur Respir J 1999;14:412–418. Ochiai M, Oguri T, Isobe T, Ishioka S, Yamakido M: Cyclooxygenase-2 mRNA expression levels in normal lung tissues and non-small cell lung cancers. Jpn J Cancer Res 1999;90:1338–1343. Hasturk S, Kemp B, Kalapurakal SK, Kurie JM, Hong WK, Lee JS: Expression of cyclooxygenase-1 and cyclooxygenase-2 in bronchial epithelium and non-small cell lung carcinoma. Cancer 2002;94:1023–1031. Soslow RA, Dannenberg AJ, Rush D, Woerner BM, Khan KN, Masferrer J, Koki AT: COX-2 is expressed in human pulmonary, colonic and mammary tumors. Cancer 2000;89:2637–2645. Khuri F, Wu H, Lee J, Kemp B, Lotan R, Lippman S, Feng L, Hong W, Xu XC: Cyclooxygenase-2 overexpression is a marker of poor prognosis in stage I non-small cell lung cancer. Clin Cancer Res 2001;7:861–867. West H, Yaziji H, Giarrita S, Livingston R, Vallieres E, Wood D, Gown A: Cyclooxygenase-2 overexpression by immunohistochemistry is associated with more aggressive biologic behavior of non-small cell lung cancer tumors. ASCO 2002;21:1235. Schreinemachers D, Everson R: Aspirin use and lung, colon, breast cancer incidence in a prospective study. Epidemiology 1994;5:138–146. Harris RE, Beebe-Donk J, Schuller HM: Chemoprevention of lung cancer by non-steroidal anti-inflammatory drugs among cigarette smokers. Oncol Rep 2002;9:693–695. Akhmedkhanov A, Toniolo P, Zeleniuch-Jacquotte A, Koenig KL, Shore RE: Aspirin and lung cancer in women. Br J Cancer 2002;87:49–53. Tsujii M, DuBois R: Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase-2. Cell 1995;83:493–501. Liu XH, Yao S, Kirschenbaum A, Levine AC: NS398, a selective cyclooxygenase-2 inhibitor, induces apoptosis and down-regulates bcl-2 expression in LNCaP cells. Cancer Res 1998;58:4245–4249. Hsu AL, Ching TT, Wang DS, Song X, Rangnekar VM, Chen CS: The cyclooxygenase-2 inhibitor celecoxib induces apoptosis by blocking Akt activation in human prostate cancer cells independently of Bcl-2. J Biol Chem 2000;275:11397–11403. Sheng H, Shao J, Morrow J, Beauchamp R, DuBois R: Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells. Cancer Res 1998;58:362–366.
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
154
35 36
37
38 39
40 41 42 43
44
45 46 47
48
49
50 51
52
53
54
Leahy K, Koki A, Masferrer J: Role of cyclooxygenases in angiogenesis. Curr Med Chem 2000; 7:1163–1170. Masferrer JL, Leahy KM, Koki AT, Zweifel BS, Settle SL, Woerner BM, Edwards DA, Flickinger AG, Moore RJ, Seibert K: Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 2000;60:1306–1311. Uefuji K, Ichikura T, Mochizuki H: Cyclooxygenase-2 expression is related to prostaglandin biosynthesis and angiogenesis in human gastric cancer. Clin Cancer Res 2000;6: 135–138. Gately S: The contributions of cyclooxygenase-2 to tumor angiogenesis. Cancer Metastasis Rev 2000;19:19–27. Stolina M, Sharma S, Lin Y, Dohadwala M, Gardner B, Luo J, Zhu L, Kronenberg M, Miller PW, Portanova J, Lee JC, Dubinett SM: Specific inhibition of cyclooxygenase-2 restores antitumor reactivity by altering the balance of IL-10 and IL-12 synthesis. J Immunol 2000; 164:361–370. Tsujii M, Kawano S, DuBois R: Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA 1997;94:3336–3340. Wiese F, Thompson P, Kadlubar F: Carcinogen substrate specificity of human COX-1 and COX-2. Carcinogenesis 2001;22:5–10. Eling TE, Thompson DC, Foureman GL, et al: Prostaglandin H synthase and xenobiotic oxidation. Annu Rev Pharmacol Toxicol 1990;30:1–45. Ho I, Lee T: Arsenite pretreatment attenuates benzo[a]pyrene cytotoxicity in a human lung adenocarcinoma cell line by decreasing cyclooxygenase-2 levels. J Toxicol Environ Health A 2002;65:245–263. Kelley D, Mestre JR, Subbaramaiah K, Sacks P, Schantz S, Tanabe T, Inoue H, Ramonetti J, Dannenberg A: Benzo[a]pyrene up-regulates cyclooxygenase-2 gene expression in oral epithelial cells. Carcinogenesis 1997;18:795–799. Berg D, Zhang J, Lauricella D, Moore S: IL-10 is a central regulator of cyclooxygenase-2 expression and prostaglandin production. J Immunol 2001;166:2674–2680. Subbaramaiah K, Altorki N, Chung WJ, Mestre JR, Sampat A, Dannenberg AJ: Inhibition of cyclooxygenase-2 gene expression by p53. J Biol Chem 1999;274:10911–10915. Fujita M, Fukui H, Kusaka T, Morita K, Fujii S, Ueda Y, Chiba T, Sakamoto C, Kawamata H, Fujimori T: Relationship between cyclooxygenase-2 expression and K-ras gene mutation in colorectal adenomas. J Gastroenterol Hepatol 2000;15:1277–1281. Subbaramaiah K, Telang N, Ramonetti JT, Araki R, DeVito B, Weksler BB, Dannenberg AJ: Transcription of cyclooxygenase-2 is enhanced in transformed mammary epithelial cells. Cancer Res 1996;56:4424–4429. Bissonnette M, Khare S, von Lintig FC, Wali RK, Nguyen L, Zhang Y, Hart J, Skarosi S, Varki N, Boss GR, Brasitus TA: Mutational and nonmutational activation of p21ras in rat colonic azoxymethane-induced tumors: Effects on mitogen-activated protein kinase, cyclooxygenase-2 and cyclin D1. Cancer Res 2000;60:4602–4609. Gilhooly E, Rose D: The association between a mutated ras gene and cyclooxygenase-2 expression in human breast cancer cell lines. Int J Oncol 1999;15:267–270. Wardlaw SA, Zhang N, Belinsky SA: Transcriptional regulation of basal cyclooxygenase-2 expression in murine lung tumor-derived cell lines by CCAAT/enhancer-binding protein and activating transcription factor/cAMP response element-binding protein. Mol Pharmacol 2002; 62:326–333. Huang M, Sharma S, Mao JT, Dubinett SM: Non-small cell lung cancer-derived soluble mediators and prostaglandin E2 enhance peripheral blood lymphocyte IL-10 transcription and protein production. J Immunol 1996;157:5512–5520. Huang M, Wang J, Lee P, Sharma S, Mao JT, Meissner H, Uyemura K, Modlin R, Wollman J, Dubinett SM: Human non-small cell lung cancer cells express a type 2 cytokine pattern. Cancer Res 1995;55:3847–3853. Neuner A, Schindel M, Wildenberg U, Muley T, Lahm H, Fischer JR: Cytokine secretion: Clinical relevance of immunosuppression in non-small cell lung cancer. Lung Cancer 2001; 34:S79–S82.
COX-2 in Lung Cancer
155
55
56
57 58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
Alleva DG, Burger CJ, Elgert KD: Tumor-induced regulation of suppressor macrophage nitric oxide and TNF-␣ production: Role of tumor-derived IL-10, TGF- and prostaglandin E2. J Immunol 1994;153:1674. D’Andrea A, Aste-Amezaga M, Valiante N, Ma X, Kubin M, Trinchieri G: Interleukin-10 inhibits human lymphocyte interferon-␥ production by suppressing natural killer cell stimulatory factor/IL-12 synthesis in accessory cells. J Exp Med 1993;178:1041–1048. Wu CY, Demeure C, Kiniwa M, Gately M, Delespesse G: IL-12 induces the production of IFN-␥ by neonatal human CD4 T cells. J Immunol 1993;151:1938–1949. Kim J, Modlin RL, Moy RL, Dubinett SM, McHugh T, Nickoloff BJ, Uyemura K: IL-10 production in cutaneous basal and squamous cell carcinomas: A mechanism for evading the local T-cell immune response. J Immunol 1995;155:2240–2247. Qin L, Yaohong D, Pahud D, Robson N, Shaked A, Bromberg J: Adenovirus-mediated gene transfer of viral interleukin-10 inhibits the immune response to both alloantigen and adenoviral antigen. Hum Gene Ther 1997;8:1365–1374. Halak BK, Maguire HC Jr, Lattime EC: Tumor-induced interleukin-10 inhibits type 1 immune responses directed at a tumor antigen as well as a non-tumor antigen present at the tumor site. Cancer Res 1999;59:911–917. Sharma S, Stolina M, Lin Y, Gardner B, Miller PW, Kronenberg M, Dubinett SM: T-cell-derived IL-10 promotes lung cancer growth by suppressing both T-cell and APC function. J Immunol 1999;163:5020–5028. Hagenbaugh A, Sharma S, Dubinett S, Wei SHY, Aranda R, Cheroutre H, Fowell D, Binder S, Tsao B, Locksley R, Moore K, Kronenberg M: Altered immune responses in IL-10 transgenic mice. J Exp Med 1997;185:2101–2110. Hatanaka H, Abe Y, Naruke M, Tokunaga T, Oshika Y, Kawakami T, Osada H, Nagata J, Kamochi J, Tsuchida T, Kijima H, Yamazaki H, Inoue H, Inoue H, Ueyama Y, Nakamura MN: Significant correlation between interleukin-10 expression and vascularization through angiopoietin/TIE2 networks in non-small cell lung cancer. Clin Cancer Res 2001;7:1287–1292. Urosevic M, Kurrer MO, Kamarashev J, Mueller B, Weder W, Guenter B, Stahel RA, Dummer R, Trojan A: Human leukocyte antigen G up-regulation in lung cancer associates with high-grade histology, human leukocyte antigen class I loss and interleukin-10 production. Am J Pathol 2001; 159:817–824. Naruke M, Abe Y, Hatanaka H, Ohnishi Y, Yamazaki H, Kijima H, Nakamura M, Ueyama Y, Kobayashi K: Interleukin-10 expression is correlated with growth fraction in human non-small cell lung cancer xenografts. Int J Oncol 2001;18:1213–1217. De Vita F, Orditura M, Galizia G, Romano C, Roscigno A, Lieto E, Catalano G: Serum interleukin-10 levels as a prognostic factor in advanced non-small cell lung cancer patients. Chest 2000;117:365–373. Hatanaka H, Abe Y, Kamiya T, Morino F, Nagata J, Tokunaga T, Oshika Y, Suemizu H, Kijima H, Tsuchida T, Yamazaki H, Inoue H, Nakamura M, Ueyama Y: Clinical implications of IL-10 induced by non-small-cell lung cancer. Ann Oncol 2000;11:815–819. Bianchi R, Grohmann U, Belladonna M, Silla S, Fallarino F, Ayroldi E, Fioretti M, Puccetti P: IL-12 is both required and sufficient for initiating T-cell reactivity to a class I-restricted tumor peptide (P815AB) following transfer of P815AB-pulsed dendritic cells. J Immunol 1996;157: 1589–1597. Colombo M, Vagliani M, Spreafico F, Parenza M, Chiodoni C, Melani C, Stoppacciaro A: Amount of interleukin-12 available at the tumor site is critical for tumor regression. Cancer Res 1996; 56:2531–2534. Handel-Fernandez ME, Ching X, Herbert LM, Lopez DM: Down-regulation of IL-12, not a shift from a T-helper-1 to a T-helper-2 phenotype, is responsible for impaired IFN-␥ production in mammary tumor-bearing mice. J Immunol 1997;158:280–286. Huang AYC, Golumbek P, Ahmadzadeh M, Jaffee E, Pardoll D, Levitsky H: Role of bone marrow-derived cells in presenting MHC class I-restricted tumor antigens. Science 1994;264: 961–965. Banchereau J, Steinman RM: Dendritic cells and the control of immunity. Nature 1998;392: 245–252.
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
156
73
74
75
76 77 78 79 80
81 82
83
84
85
86
87
88
89
90
91
92
Almand B, Resser JR, Lindman B, Nadaf S, Clark JI, Kwon ED, Carbone DP, Gabrilovich DI: Clinical significance of defective dendritic cell differentiation in cancer. Clin Cancer Res 2000; 6:1755–1766. Sombroek CC, Stam AG, Masterson AJ, Lougheed SM, Schakel MJ, Meijer CJ, Pinedo HM, van den Eertwegh AJ, Scheper RJ, de Gruijl TD: Prostanoids play a major role in the primary tumor-induced inhibition of dendritic cell differentiation. J Immunol 2002;168:4333–4343. Harizi H, Juzan M, Pitard V, Moreau JF, Gualde N: Cyclooxygenase-2-issued prostaglandin E2 enhances the production of endogenous IL-10, which down-regulates dendritic cell functions. J Immunol 2002;168:2255–2263. Kambayashi T, Wallin RP, Ljunggren HG: cAMP-elevating agents suppress dendritic cell function. J Leukoc Biol 2001;70:903–910. Stolina M, Sharma S, Zhu L, Dubinett SM: Lung cancer cyclooxygenase-2-dependent inhibition of dendritic cell maturation and function. Proc Am Assoc Cancer Res, 2000, pp 1–619. Folkman J: Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1995; 1:27–31. Folkman J: Angiogenesis-dependent diseases. Semin Oncol 2001;28:536–542. Harpole DH Jr, Richards WG, Herndon JE 2nd, Sugarbaker DJ: Angiogenesis and molecular biologic substaging in patients with stage I non-small cell lung cancer. Ann Thorac Surg 1996;61: 1470–1476. Macchiarini P, Fontanini G, Hardin MJ, Squartini F, Angeletti CA: Relation of neovascularization to metastasis of non-small cell lung cancer. Lancet 1992;340:145–146. Duarte IG, Bufkin BL, Pennington MF, Gal AA, Cohen C, Kosinski AS, Mansour KA, Miller JI: Angiogenesis as a predictor of survival after surgical resection for stage I non-small cell lung cancer. J Thorac Cardiovasc Surg 1998;115:652–659. Giatromanolaki A, Koukourakis M, O’Byrne K, Fox S, Whitehouse R, Talbot DC, Harris AL, Gatter KC: Prognostic value of angiogenesis in operable non-small cell lung cancer. J Pathol 1996; 179:80–88. Fontanini G, Bigini D, Vignati S, Basolo F, Mussi A, Lucchi M, Chine S, Angeletti CA, Harris AL, Bevilacqua G: Microvessel count predicts metastatic disease and survival in non-small cell lung cancer. J Pathol 1995;177:57–63. D’Amico TA, Massey M, Herndon JE 2nd, Moore MB, Harpole DH Jr: A biologic risk model for stage I lung cancer: Immunohistochemical analysis of 408 patients with the use of ten molecular markers. J Thorac Cardiovasc Surg 1999;117:736–743. Dazzi C, Cariello A, Maioli P, Solaini L, Scarpi E, Rosti G, Lanzanova G, Marangolo M: Prognostic and predictive value of intratumoral microvessel density in operable non-small cell lung cancer. Lung Cancer 1999;24:81–88. Guddo F, Fontanini G, Reina C, Vignola AM, Angeletti A, Bonsignore G: The expression of basic fibroblast growth factor in tumor-associated stromal cells and vessels is inversely correlated with non-small cell lung cancer progression. Hum Pathol 1999;30:788–794. Ohta Y, Shridhar V, Bright RK, Kalemkerian GP, Du W, Carbone M, Watanabe Y, Pass HI: VEGF and VEGF type C play an important role in angiogenesis and lymphangiogenesis in human malignant mesothelioma tumours. Br J Cancer 1999;81:54–61. Wikstrom P, Stattin P, Franck-Lissbrant I, Damber JE, Bergh A: Transforming growth factor-1 is associated with angiogenesis, metastasis and poor clinical outcome in prostate cancer. Prostate 1998;37:19–29. Koukourakis MI, Giatromanolaki A, O’Byrne KJ, Comley M, Whitehouse RM, Talbot DC, Gatter KC, Harris AL: Platelet-derived endothelial cell growth factor expression correlates with tumour angiogenesis and prognosis in non-small cell lung cancer. Br J Cancer 1997;75: 477–481. Arenberg DA, Kunkel SL, Polverini PJ, Glass M, Burdick MD, Strieter RM: Inhibition of interleukin-8 reduces tumorigenesis of human non-small cell lung cancer in SCID mice. J Clin Invest 1996;97:2792–2802. Konishi T, Huang CL, Adachi M, Taki T, Inufusa H, Kodama K, Kohno N, Miyake M: The K-ras gene regulates vascular endothelial growth factor gene expression in non-small cell lung cancers. Int J Oncol 2000;16:501–511.
COX-2 in Lung Cancer
157
93 Ravi R, Mookerjee B, Bhujwalla ZM, Sutter CH, Artemov D, Zeng Q, Dillehay LE, Madan A, Semenza GL, Bedi A: Regulation of tumor angiogenesis by p53-induced degradation of hypoxiainducible factor 1␣. Genes Dev 2000;14:34–44. 94 Ambs S, Bennett WP, Merriam WG, Ogunfusika MO, Oser SM, Khan MA, Jones RT, Harris CC: Vascular endothelial growth factor and nitric oxide synthase expression in human lung cancer and the relation to p53. Br J Cancer 1998;78:233–239. 95 Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois R: Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705–716. 96 Daniel TO, Liu H, Morrow JD, Crews BC, Marnett LJ: Thromboxane A2 is a mediator of cyclooxygenase-2-dependent endothelial migration and angiogenesis. Cancer Res 1999;59:4574–4577. 97 Williams CS, Tsujii M, Reese J, Dey SK, DuBois RN: Host cyclooxygenase-2 modulates carcinoma growth. J Clin Invest 2000;105:1589–1594. 98 Hanahan D, Weinberg RA: The hallmarks of cancer. Cell 2000;100:57–70. 99 Kinzler KW, Vogelstein B: Landscaping the cancer terrain. Science 1998;280:1036–1037. 100 Leahy KM, Ornberg RL, Wang Y, Zweifel BS, Koki AT, Masferrer JL: Cyclooxygenase-2 inhibition by celecoxib reduces proliferation and induces apoptosis in angiogenic endothelial cells in vivo. Cancer Res 2002;62:625–631. 101 Koki AT, Leahy KM, Masferrer JL: Potential utility of COX-2 inhibitors in chemoprevention and chemotherapy. Expert Opin Investig Drugs 1999;8:1623–1638. 102 Marrogi AJ, Travis WD, Welsh JA, Khan MA, Rahim H, Tazelaar H, Pairolero P, Trastek V, Jett J, Caporaso NE, Liotta LA, Harris CC: Nitric oxide synthase, cyclooxygenase-2, and vascular endothelial growth factor in the angiogenesis of non-small cell lung carcinoma. Clin Cancer Res 2000;6:4739–4744. 103 Chambers AF, Groom AC, MacDonald IC: Metastasis: Dissemination and growth of cancer cells in metastatic sites. Nat Rev Cancer 2002;2:563–572. 104 Dohadwala M, Luo J, Zhu L, Lin Y, Dougherty G, Sharma S, Huang M, Pold M, Batra R, Dubinett S: Non-small cell lung cancer cylooxygenase-2-dependent invasion is mediated by CD44. J Biol Chem 2001;276:20809–20812. 105 Yu Q, Stamenkovic I: Localization of matrix metalloproteinase 9 to the cell surface provides a mechanism for CD44-mediated tumor invasion. Genes Dev 1999;13:35–48. 106 Lamb RF, Hennigan RF, Turnbull K, Katsanakis KD, MacKenzie ED, Birnie GD, Ozanne BW: AP-1-mediated invasion requires increased expression of the hyaluronan receptor CD44. Mol Cell Biol 1997;17:963–976. 107 Bartolazzi A, Peach R, Aruffo A, Stamenkovic I: Interaction between CD44 and hyaluronate is directly implicated in the regulation of tumor development. J Exp Med 1994;180:53–66. 108 Seiter S, Arch R, Reber S, Komitowski D, Hofmann M, Ponta H, Herrlich P, Matzku S, Zoller M: Prevention of tumor metastasis formation by anti-variant CD44. J Exp Med 1993;177:443–455. 109 Dohadwala M, Luo J, Lin Y, Sharma S, Zhu L, Lee P, Pold M, Dubinett S. M: The role of PGE2 in cyclooxygenase-2 regulated, CD44-depedent invasion in non-small cell lung cancer. Proc Am Assoc Cancer Res 2002;43:371. 110 Yu Q, Stamenkovic I: Cell surface-localized matrix metalloproteinase-9 proteolytically activates TGF- and promotes tumor invasion and angiogenesis. Genes Dev 2000;14:163–176. 111 Shankavaram UT, Lai WC, Netzel-Arnett S, Mangan PR, Ardans JA, Caterina N, StetlerStevenson WG, Birkedal-Hansen H, Wahl LM: Monocyte membrane type 1-matrix metalloproteinase. Prostaglandin-dependent regulation and role in metalloproteinase-2 activation. J Biol Chem 2001;276:19027–19032. 112 Pan MR, Chuang LY, Hung WC: Non-steroidal anti-inflammatory drugs inhibit matrix metalloproteinase-2 expression via repression of transcription in lung cancer cells. FEBS Lett 2001;508: 365–368. 113 Kozaki K, Koshikawa K, Tatematsu Y, Miyaishi O, Saito H, Hida T, Osada H, Takahashi T: Multifaceted analyses of a highly metastatic human lung cancer cell line NCI-H460-LNM35 suggest mimicry of inflammatory cells in metastasis. Oncogene 2001;20:4228–4234. 114 Niki T, Kohno T, Iba S, Moriya Y, Takahashi Y, Saito M, Maeshima A, Yamada T, Matsuno Y, Fukayama M, Yokota J, Hirohashi S: Frequent co-localization of Cox-2 and laminin-5 ␥2 chain at the invasive front of early-stage lung adenocarcinomas. Am J Pathol 2002;160:1129–1141.
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
158
115 Pyke C, Romer J, Kallunki P, Lund LR, Ralfkiaer E, Dano K, Tryggvason K: The ␥2 chain of kalinin/laminin-5 is preferentially expressed in invading malignant cells in human cancers. Am J Pathol 1994;145:782–791. 116 Pyke C, Salo S, Ralfkiaer E, Romer J, Dano K, Tryggvason K: Laminin-5 is a marker of invading cancer cells in some human carcinomas and is coexpressed with the receptor for urokinase plasminogen activator in budding cancer cells in colon adenocarcinomas. Cancer Res 1995;55: 4132–4139. 117 Sordat I, Bosman FT, Dorta G, Rousselle P, Aberdam D, Blum AL, Sordat B: Differential expression of laminin-5 subunits and integrin receptors in human colorectal neoplasia. J Pathol 1998; 185:44–52. 118 Skyldberg B, Salo S, Eriksson E, Aspenblad U, Moberger B, Tryggvason K, Auer G: Laminin-5 as a marker of invasiveness in cervical lesions. J Natl Cancer Inst 1999;91:1882–1887. 119 Coffey RJ, Hawkey CJ, Damstrup L, Graves-Deal R, Daniel VC, Dempsey PJ, Chinery R, Kirkland SC, DuBois RN, Jetton TL, Morrow JD: Epidermal growth factor receptor activation induces nuclear targeting of cyclooxygenase-2, basolateral release of prostaglandins and mitogenesis in polarizing colon cancer cells. Proc Natl Acad Sci USA 1997;94:657–662. 120 Ono Y, Nakanishi Y, Gotoh M, Sakamoto M, Hirohashi S: Epidermal growth factor receptor gene amplification is correlated with laminin-5 ␥2 chain expression in oral squamous cell carcinoma cell lines. Cancer Lett 2002;175:197–204. 121 Pai R, Soreghan B, Szabo IL, Pavelka M, Baatar D, Tarnawski AS: Prostaglandin E2 transactivates EGF receptor: A novel mechanism for promoting colon cancer growth and gastrointestinal hypertrophy. Nat Med 2002;8:289–293. 122 Ferreira CG, Epping M, Kruyt FA, Giaccone G: Apoptosis: Target of cancer therapy. Clin Cancer Res 2002;8:2024–2034. 123 Liu XH, Yao S, Kirschenbaum A, Levine AC: NS398, a selective cyclooxygenase-2 inhibitor, induces apoptosis and down-regulates bcl-2 expression in LNCaP cells. Cancer Res 1998; 58:4245–4249. 124 Hara A, Yoshimi N, Niwa M, Ino N, Mori H: Apoptosis induced by NS-398, a selective cyclooxygenase-2 inhibitor, in human colorectal cancer cell lines. Jpn J Cancer Res 1997;88: 600–604. 125 Erickson BA, Longo WE, Panesar N, Mazuski JE, Kaminski DL: The effect of selective cyclooxygenase inhibitors on intestinal epithelial cell mitogenesis. J Surg Res 1999;81:101–107. 126 Sawaoka H, Kawano S, Tsuji S, Tsujii M, Gunawan ES, Takei Y, Nagano K, Hori M: Cyclooxygenase-2 inhibitors suppress the growth of gastric cancer xenografts via induction of apoptosis in nude mice. Am J Physiol 1998;274:G1061–G1067. 127 Hida T, Kozaki K, Muramatsu H, Masuda A, Shimizu S, Mitsudomi T, Sugiura T, Ogawa M, Takahashi T: Cyclooxygenase-2 inhibitor induces apoptosis and enhances cytotoxicity of various anticancer agents in non-small cell lung cancer cell lines. Clin Cancer Res 2000;6:2006–2011. 128 Chang HC, Weng CF: Cyclooxygenase-2 level and culture conditions influence NS398-induced apoptosis and caspase activation in lung cancer cells. Oncol Rep 2001;8:1321–1325. 129 Yao R, Rioux N, Castonguay A, You M: Inhibition of COX-2 and induction of apoptosis: Two determinants of nonsteroidal anti-inflammatory drugs’ chemopreventive efficacies in mouse lung tumorigenesis. Exp Lung Res 2000;26:731–742. 130 Lin MT, Lee RC, Yang PC, Ho FM, Kuo ML: Cyclooxygenase-2 inducing Mcl-1-dependent survival mechanism in human lung adenocarcinoma CL1.0 cells. Involvement of phosphatidylinositol 3-kinase/Akt pathway. J Biol Chem 2001;276:48997–49002. 131 Hsueh CT, Chiu CF, Kelsen DP, Schwartz GK: Selective inhibition of cyclooxygenase-2 enhances mitomycin-C-induced apoptosis. Cancer Chemother Pharmacol 2000;45:389–396. 132 Elder DJ, Halton DE, Hague A, Paraskeva C: Induction of apoptotic cell death in human colorectal carcinoma cell lines by a cyclooxygenase-2 (COX-2)-selective nonsteroidal anti-inflammatory drug: Independence from COX-2 protein expression. Clin Cancer Res 1997;3:1679–1683. 133 Crew TE, Elder DJ, Paraskeva C: A cyclooxygenase-2 (COX-2) selective non-steroidal anti-inflammatory drug enhances the growth inhibitory effect of butyrate in colorectal carcinoma cells expressing COX-2 protein: Regulation of COX-2 by butyrate. Carcinogenesis 2000;21: 69–77.
COX-2 in Lung Cancer
159
134 135
136
137
138
139 140
141
142
143
144
145 146
147 148
149
150 151
152 153
Dannenberg AJ, Altorki NK, Boyle JO, Lin DT, Subbaramaiah K: Inhibition of cyclooxygenase-2: An approach to preventing cancer of the upper aerodigestive tract. Ann NY Acad Sci 2001;952:109–115. Hida T, Kozaki K, Ito H, Miyaishi O, Tatematsu Y, Suzuki T, Matsuo K, Sugiura T, Ogawa M, Takahashi T: Significant growth inhibition of human lung cancer cells both in vitro and in vivo by the combined use of a selective cyclooxygenase-2 inhibitor, JTE-522, and conventional anticancer agents. Clin Cancer Res 2002;8:2443–2447. Subbaramaiah K, Hart JC, Norton L, Dannenberg AJ: Microtubule-interfering agents stimulate the transcription of cyclooxygenase-2. Evidence for involvement of ERK1/2 and p38 mitogenactivated protein kinase pathways. J Biol Chem 2000;275:14838–14845. Moos PJ, Muskardin DT, Fitzpatrick FA: Effect of taxol and taxotere on gene expression in macrophages: Induction of the prostaglandin H synthase-2 isoenzyme. J Immunol 1999;162: 467–473. Cassidy PB, Moos PJ, Kelly RC, Fitzpatrick FA: Cyclooxygenase-2 induction by paclitaxel, docetaxel, and taxane analogues in human monocytes and murine macrophages: structure-activity relationships and their implications. Clin Cancer Res 2002;8:846–855. Milas L, Kishi K, Hunter N, Mason K, Masferrer JL, Tofilon PJ: Enhancement of tumor response to ␥-radiation by an inhibitor of cyclooxygenase-2 enzyme. J Natl Cancer Inst 1999;91:1501–1504. Pyo H, Choy H, Amorino GP, Kim JS, Cao Q, Hercules SK, DuBois RN: A selective cyclooxygenase-2 inhibitor, NS-398, enhances the effect of radiation in vitro and in vivo preferentially on the cells that express cyclooxygenase-2. Clin Cancer Res 2001;7:2998–3005. Milas L, Furuta Y, Hunter N, Nishiguchi I, Runkel S: Dependence of indomethacin-induced potentiation of murine tumor radioresponse on tumor host immunocompetence. Cancer Res 1990;50: 4473–4477. Milas L, Hunter N, Furuta Y, Nishiguchi I, Runkel S: Antitumour effects of indomethacin alone and in combination with radiotherapy: Role of inhibition of tumour angiogenesis. Int J Radiat Biol 1991;60:65–70. Petersen C, Petersen S, Milas L, Lang FF, Tofilon PJ: Enhancement of intrinsic tumor cell radiosensitivity induced by a selective cyclooxygenase-2 inhibitor. Clin Cancer Res 2000;6: 2513–2520. Kishi K, Petersen S, Petersen C, Hunter N, Mason K, Masferrer JL, Tofilon PJ, Milas L: Preferential enhancement of tumor radioresponse by a cyclooxygenase-2 inhibitor. Cancer Res 2000;60:1326–1331. Milas L: Cyclooxygenase-2 enzyme inhibitors as potential enhancers of tumor radioresponse. Semin Radiat Oncol 2001;11:290–299. Shah T, Ryu S, Lee HJ, Brown S, Kim JH: Pronounced radiosensitization of cultured human cancer cells by COX inhibitor under acidic microenvironment. Int J Radiat Oncol Biol Phys 2002; 53:1314–1318. Thun M, Henley J, Patrono C: Nonsteroidal anti-inflammatory drugs as anticancer agents: Mechanistic, pharmacologic and clinical issues. J Natl Cancer Inst 2002;94:252–266. Shiff SJ, Qiao L, Tsai LL, Rigas B: Sulindac sulfide, an aspirin-like compound, inhibits proliferation, causes cell cycle quiescence, and induces apoptosis in HT-29 colon adenocarcinoma cells. J Clin Invest 1995;96:491–503. Heasley L, Thaler S, Nicks M, Price B, Skorecki K, Nemenoff RA: Induction of cytosolic phospholipase A2 by oncogenic Ras in human non-small cell lung cancer. J Biol Chem 1997;272: 14501–14504. DuBois R, Shao J, Tsujii M, Sheng H, Beauchamp R: G1 delay in cells overexpressing prostaglandin endoperoxide synthase-2. Cancer Res 1996;56:733–737. Hung WC, Chang HC, Pan MR, Lee TH, Chuang LY: Induction of p27(KIP1) as a mechanism underlying NS398-induced growth inhibition in human lung cancer cells. Mol Pharmacol 2000; 58:1398–1403. Malumbres M, Barbacid M: To cycle or not to cycle: A critical decision in cancer. Nat Rev Cancer 2001;1:222–231. Shirane M, Harumiya Y, Ishida N, Hirai A, Miyamoto C, Hatakeyama S, Nakayama K, Kitagawa M: Down-regulation of p27(Kip1) by two mechanisms, ubiquitin-mediated degradation and proteolytic processing. J Biol Chem 1999;274:13886–13893.
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
160
154 Catzavelos C, Tsao MS, DeBoer G, Bhattacharya N, Shepherd FA, Slingerland JM: Reduced expression of the cell cycle inhibitor p27Kip1 in non-small cell lung carcinoma: A prognostic factor independent of Ras. Cancer Res 1999;59:684–688. 155 Esposito V, Baldi A, De Luca A, Groger AM, Loda M, Giordano GG, Caputi M, Baldi F, Pagano M, Giordano A: Prognostic role of the cyclin-dependent kinase inhibitor p27 in non-small cell lung cancer. Cancer Res 1997;57:3381–3385. 156 Yatabe Y, Masuda A, Koshikawa T, Nakamura S, Kuroishi T, Osada H, Takahashi T, Mitsudomi T: p27KIP1 in human lung cancers: Differential changes in small cell and non-small cell carcinomas. Cancer Res 1998;58:1042–1047. 157 Hawk ET, Viner JL, Dannenberg A, DuBois RN: COX-2 in cancer – A player that’s defining the rules. J Natl Cancer Inst 2002;94:545–546. 158 Hwang DH, Fung V, Dannenberg AJ: National Cancer Institute workshop on chemopreventive properties of nonsteroidal anti-inflammatory drugs: Role of COX-dependent and -independent mechanisms. Neoplasia 2002;4:91–97. 159 Song X, Lin HP, Johnson AJ, Tseng PH, Yang YT, Kulp SK, Chen CS: Cyclooxygenase-2, player or spectator in cyclooxygenase-2 inhibitor-induced apoptosis in prostate cancer cells. J Natl Cancer Inst 2002;94:585–591. 160 Xu XC: COX-2 inhibitors in cancer treatment and prevention, a recent development. Anticancer Drugs 2002;13:127–137. 161 Grosch S, Tegeder I, Niederberger E, Brautigam L, Geisslinger G: COX-2-independent induction of cell cycle arrest and apoptosis in colon cancer cells by the selective COX-2 inhibitor celecoxib. Faseb J 2001;15:2742–2744. 162 He TC, Chan TA, Vogelstein B, Kinzler KW: PPAR-␦ is an APC-regulated target of nonsteroidal anti-inflammatory drugs. Cell 1999;99:335–345. 163 Frantz B, O’Neill EA: The effect of sodium salicylate and aspirin on NF-B. Science 1995;270: 2017–2019. 164 Huang C, Ma WY, Hanenberger D, Cleary MP, Bowden GT, Dong Z: Inhibition of ultraviolet B-induced activator protein-1 (AP-1) activity by aspirin in AP-1-luciferase transgenic mice. J Biol Chem 1997;272:26325–26331. 165 Cao Y, Pearman AT, Zimmerman GA, McIntyre TM, Prescott SM: Intracellular unesterified arachidonic acid signals apoptosis. Proc Natl Acad Sci USA 2000;97:11280–11285. 166 Ferrario A, Von Tiehl K, Wong S, Luna M, Gomer CJ: Cyclooxygenase-2 inhibitor treatment enhances photodynamic therapy-mediated tumor response. Cancer Res 2002;62:3956–3961. 167 Lee P, Kupeli E, Mehta AC: Therapeutic bronchoscopy in lung cancer. Laser therapy, electrocautery, brachytherapy, stents and photodynamic therapy. Clin Chest Med 2002;23:241–256. 168 Metz JM, Friedberg JS: Endobronchial photodynamic therapy for the treatment of lung cancer. Chest Surg Clin North Am 2001;11:829–839. 169 Sutedja G, Postmus PE: The role of photodynamic therapy in the management of stage I/II NSCLC. Lung Cancer 2001;34(suppl 3):35–38. 170 Dougherty TJ, Gomer CJ, Henderson BW, Jori G, Kessel D, Korbelik M, Moan J, Peng Q: Photodynamic therapy. J Natl Cancer Inst 1998;90:889–905. 171 Henderson BW, Donovan JM: Release of prostaglandin E2 from cells by photodynamic treatment in vitro. Cancer Res 1989;49:6896–6900. 172 Reed MW, Wieman TJ, Doak KW, Pietsch CG, Schuschke DA: The microvascular effects of photodynamic therapy: Evidence for a possible role of cyclooxygenase products. Photochem Photobiol 1989;50:419–423. 173 Csiki I, Gonzalez D, Sandler A, Carbone D, Choy H, Campbell N, Garcia B, Morrow J, Johnson D: Cyclooxygenase-2 inhibition ⫹ docetaxel in recurrent non-small cell lung cancer: Preliminary results of a phase II trial (THO-0054). ASCO 2002;21:1187. 174 Pisters KM, Ginsberg RJ, Giroux DJ, Putnam JB Jr, Kris MG, Johnson DH, Roberts JR, Mault J, Crowley JJ, Bunn PA Jr: Induction chemotherapy before surgery for early-stage lung cancer: A novel approach. Bimodality Lung Oncology Team. J Thorac Cardiovasc Surg 2000;119:429–439. 175 Altorki N, Keresztes R, Port J, Flieder D, Ferrara C, Pasmantier M, Libby D, Yankelevitz D, Dannenberg A: Celecoxib (Celebrex), a selective COX-2 inhibitor, enhances the response to preoperative paclitaxel/carboplatin in early stage non-small cell lung cancer. ASCO 2002;21:101.
COX-2 in Lung Cancer
161
176 Carbone D, Choy H, Csiki I, Dang T, Campbell N, Garcia B, Morrow J, Saha D, Johnson D, Sandler A: Serum/plasma VEGF level changes with cyclooxygenase-2 inhibition in combined modality therapy in state III non-small cell lung cancer: Preliminary results of a phase II trial (THO-0059). ASCO 2002;21:1270. 177 Omenn G, Goodman GE, Thornquist MD, Balmes J, Cullen MR, Glass A, Keogh JP, Meyskens FL, Valanis B, Williams JH, Barnhart S, Hammar S: Effects of a combination of -carotene and vitamin A on lung cancer and cardiovascular disease. N Engl J Med 1996;334:1150–1155. 178 Lippman SM, Lee JJ, Karp DD, Vokes EE, Benner SE, Goodman GE, Khuri FR, Marks R, Winn RJ, Fry W, Graziano SL, Gandara DR, Okawara G, Woodhouse CL, Williams B, Perez C, Kim HW, Lotan R, Roth JA, Hong WK: Randomized phase III intergroup trial of isotretinoin to prevent second primary tumors in stage I non-small cell lung cancer. J Natl Cancer Inst 2001; 93:605–618. 179 Hennekens CH, Buring JE, Manson JE, Stampfer M, Rosner B, Cook NR, Belanger C, LaMotte F, Gaziano JM, Ridker PM, Willett W, Peto R: Lack of effect of long-term supplementation with -carotene on the incidence of malignant neoplasms and cardiovascular disease. N Engl J Med 1996;334:1145–1149. 180 Fong KM, Sekido Y, Minna JD: Molecular pathogenesis of lung cancer. J Thorac Cardiovasc Surg 1999;118:1136–1152. 181 Hirsch F, Franklin W, Gazdar A, Bunn P Jr: Early detection of lung cancer: Clinical perspectives of recent advances in biology and radiology. Clin Cancer Res 2001;7:5–22. 182 Wardlaw SA, March TH, Belinsky SA: Cyclooxygenase-2 expression in abundant in alveolar type II cells in lung cancer-sensitive mouse strains and in premalignant lesions. Carcinogenesis 2000;21:1371–1377. 183 Rioux N, Castonguay A: Prevention of NNK-induced lung tumorigenesis in A/J mice by acetylsalicylic acid and NS-398. Cancer Res 1998;58:5354–5360. 184 Kinoshita T, Takahashi Y, Sakashita T, Inoue H, Tanabe T, Yoshimoto T: Growth stimulation and induction of epidermal growth factor receptor by overexpression of cyclooxygenases-1 and -2 in human colon carcinoma cells. Biochim Biophys Acta 1999;1438:120–130. 185 Murata H, Kawano S, Tsuji S, Tsuji M, Sawaoka H, Kimura Y, Shiozaki H, Hori M: Cyclooxygenase-2 overexpression enhances lymphatic invasion and metastasis in human gastric carcinoma. Am J Gastroenterol 1999;94:451–455. 186 Hida T, Leyton J, Makheja A, Ben-Av P, Hla T, Martinez A, Mulshine J, Malkani S, Chung P, Moody T: Non-small cell lung cancer cyclooxygenase activity and proliferation are inhibited by non-steroidal anti-inflammatory drugs. Anticancer Res 1998;18:775–782. 187 Kalinski P, Hilkens CM, Snijders A, Snijdewint FG, Kapsenberg ML: Dendritic cells, obtained from peripheral blood precursors in the presence of PGE2, promote Th2 responses. Adv Exp Med Biol 1997;417:363–367. 188 Swisher S, Kiertscher S, Golub S, Holmes E, Roth M: Pulmonary macrophages suppress the proliferation and cytotoxicity of tumor-infiltrating lymphocytes. Am J Respir Cell Mol Biol 1993; 8:486–492. 189 Roth MD, Golub SH: Human pulmonary macrophages utilize prostaglandins and transforming growth factor-1 to suppress lymphocyte activation. J Leukoc Biol 1993;53:366–371. 190 Pold M, Dohadwala M, Luo J, Lin Y, Dubinett S: Microarray identifies cyclo-oxygenase-2dependent modulation of insulin-like growth factor binding protein-3 in non-small cell lung cancer cells. Chest 2002;121:29S–30S.
Steven M. Dubinett, MD Lung Cancer Research Program 37-131 CHS, Division of Pulmonary and Critical Care Medicine Department of Medicine, Geffen School of Medicine at UCLA 10833 LeConte Avenue, Los Angeles, CA 90025 (USA) Tel. ⫹1 310 7946566, Fax ⫹1 310 2672829, E-Mail
[email protected]
Dubinett/Sharma/Huang/Dohadwala/Pold/Mao
162
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 163–178
COX-2 Inhibitors and Other NSAIDs in Bladder and Prostate Cancer Anita L. Sabichi, Scott M. Lippman Department of Clinical Cancer Prevention, The University of Texas M.D. Anderson Cancer Center, Houston, Tex., USA
Introduction
A basic hypothesis underpins the rationale for use of non-steroidal antiinflammatory drugs (NSAIDs) in preventing and treating bladder and prostate cancer – NSAIDs (and other agents, e.g. lipoxygenase (LOX) modulators) may suppress carcinogenesis through their effects on fatty acid metabolism and inflammation. The roles and relationship of fatty acids and inflammation in genitourinary carcinogenesis are active areas of research. Although the clinical data on NSAIDs in treating and preventing bladder and prostate cancers are limited. Encouraging NSAID studies in animal models, recent provocative studies of cyclooxygenase-2 (COX-2) expression patterns and novel NSAID mechanisms in bladder and prostate cancer have generated tremendous interest in the use of these agents for therapy and prevention. This chapter reviews the epidemiology, biology, molecular pathology, preclinical and clinical studies of NSAIDs in genitourinary carcinogenesis. Bladder Cancer
Overview Bladder cancer is common throughout the developed world. In the USA it is the fourth most commonly diagnosed cancer in men, and the tenth in women. Lifetime bladder cancer risk is approximately 1 in 29 (3.45%) for men and 1 in 89 for women (1.12%); an estimated 56,500 new cases and 12,600 deaths from bladder cancer estimated to occur in 2002 [1]. Cigarette smoke is a rich source of carcinogenic aromatic amines, and smoking is thought to be the cause from half to a third of all bladder cancers [2, 3]. Bladder carcinogens
cause multicentric areas of genetic damage in the urothelium that can lead to cancer after typically a 15- to 20-year latency period. The most common type of bladder cancer in the USA is transitional cell carcinoma (TCC), whereas squamous cell carcinomas (SCCs) are more common in other parts of the world [4]. Treatment of bladder neoplasia depends on the grade and stage of disease at diagnosis. Cystectomy typically is reserved for invasive disease. Superficial TCC is managed effectively by transurethral excisional biopsy (TURBT). Tumors with a high risk of recurrence or progression, i.e. carcinoma in situ (CIS) or lamina propria invasion are treated with TURBT followed by intravesical adjuvant therapy which usually consists of immuno- or chemotherapy [5]. Bacillus Calmette-Guérin (BCG) is FDA-approved adjuvant therapy for CIS, valrubicin for BCG-refractory disease [6]. Despite standard therapy with TURBT and intravesical BCG, patients remain at high risk for relapse; approximately 45% of tumors recur, over 30% of them recur within a year of surgery. Therefore, novel treatment and preventive approaches are being investigated. The Epidemiology of NSAIDs and Bladder Cancer There is little epidemiological evidence for an association between NSAIDs and decreased bladder cancer incidence. No large study evaluating such an association has been conducted. A case-control study in the United Kingdom found that NSAID prescriptions written for persons newly diagnosed with bladder cancer (cases) and persons without bladder cancer (controls) did not differ significantly in number in the 13- to 36-month period prior to diagnosis (of the cases) (p ⫽ 0.14) [7]. A USA population-based study involving 1,514 incident bladder cancer cases (and an equal number of appropriately matched non-cancer controls), however, found that the regular use of NSAIDs was associated with a nearly 20% reduction in risk of bladder cancer (OR ⫽ 0.81; CI ⫽ 0.68–0.96) [8]. In this study, all classes of NSAIDs, except pyrazolon derivatives, were associated with a decreased bladder cancer risk, and the data suggested that the protective effect varied according to subcategories of formulation – the strongest protective effect associated with acetic acids, the weakest with aspirin and oxicam [8]. SCC of the urinary bladder is less common than TCC in the USA. Data have indicated that SCC of the bladder usually arises in the setting of chronic inflammation of the urinary tract [9]. In the USA, SCC is most often diagnosed in spinal cord-injured patients who require a chronic indwelling bladder catheter. Over time, irritation from the catheter leads to chronic bladder inflammation and ultimately to cancer. In parts of the world where SCC has a high incidence, such as Africa, parasitic infection with Schistosoma haematobium is common [9]. Molecular and cellular mechanisms associated with chronic inflammation leading to cancer of the bladder are being explored. One potential mechanism that may contribute to
Sabichi/Lippman
164
the development of bladder SCC in this setting is inducible nitric oxide synthetase (iNOS), which is expressed in inflammatory macrophages in chronically inflamed areas of the bladder mucosa. iNOS can lead to the sustained production of nitric oxide, which oxidizes urinary amines, forming carcinogenic nitrosamines. When these locally produced carcinogens are in direct prolonged contact with bladder epithelium, they may cause irreversible damage to the bladder mucosa [10]. COX Expression in Human Bladder Cancer Bladder tissue from cystectomy specimens has been analyzed for COX expression by immunohistochemical staining. In normal urothelium, COX-1 expression is low and COX-2 is undetectable, whereas in invasive lesions, COX-2 is strongly expressed [11–18]. COX-2 expression is localized to neoplastic (TCC) and inflammatory cells and is strongest in the invading cells [12, 13]. There is a strong correlation between high COX-2 expression and advanced grade (degree of differentiation) [12, 17] and stage (degree of invasion) [14, 17] of bladder TCC in humans. The frequency of COX-2 immunostaining in invasive TCC is from 31 to 80% [12, 14–16, 18]. Superficially invasive tumors have a much lower rate of COX-2 expression (20–30%) [13, 14], as do low-grade papillary TCCs [13, 18]. COX-2 expression is frequent, however, in premalignant lesions such as dysplasia and CIS, the latter which frequently progresses to invasive TCC. In one study, 48% of dysplastic urothelial lesions stained positively for COX-2 [14]. 65–93% of high-grade CIS lesions express COX-2, exceeding the COX-2 staining frequency in invasive bladder disease [13, 14]. SCCs and squamous metaplasia (a SCC precursor) also appear to express COX-2 [11]. Whether associated with schistosomiasis or not, SCC overexpresses COX-2 (but not COX-1) and the degree of COX-2 overexpression has been directly correlated with tumor grade [11, 15]. Based on the data of COX-2 expression in human cancers, it can be postulated that COX-2 plays a role in TCC development and progression to invasive disease, and so COX-2 inhibitors may be effective alone or combined with other agents for chemoprevention or treatment of TCC [19]. Although COX-2 is expressed frequently in human bladder TCC and in premalignant lesions at high risk of progressing to invasive disease (such as CIS), it has not been found to be an independent prognostic factor for survival. A multivariate logistic regression analysis of 108 cystectomy specimens by Shirahama et al. [16] showed that local invasion and lymph node metastases were independent prognostic factors, but COX-2 expression was not. NSAIDs in Animal Bladder Cancer Models As in human bladder TCC, COX-2 is overexpressed in spontaneously developing canine TCC and in other bladder cancer animal models [19, 20].
COX-2 Inhibitors and Other NSAIDs in Bladder and Prostate Cancer
165
COX-2 is the rate-limiting enzyme in the conversion of arachadonic acid to prostaglandins (PGs). It is rapidly inducible by a host of growth factors and tumor promoters and is thought to contribute to carcinogenesis and malignant phenotype in that it increases PG production, modulates immune function, inhibits apoptosis, and promotes angiogenesis [21]. COX-2 also is implicated in the conversion of pro-carcinogens to carcinogens and in increasing the invasive potential of malignant cells. Many of these mechanisms may be directly relevant to the development of bladder cancer. The mechanism(s) involved in anti-tumor effects of COX inhibitors are not defined, but COX inhibitors and other NSAIDs have been well studied in animal bladder cancer models (table 1) [22–44]. In animal models, NSAIDs suppress bladder cancer development, reduce tumor size and inhibit carcinogenesis by carcinogen inactivation [37]. NSAID mechanisms may also include antiangiogenic effects and apoptosis induction. A study of dogs with spontaneous invasive TCC of the urinary bladder showed that piroxicam, which inhibits COX-1 and COX-2, reduced tumor volume in association with a significant increase in apoptosis (p ⫽ 0.015) and reduction in urine bFGF, a potent angiogenic factor [44]. Studies using COX inhibitors show that they may also potentiate standard chemotherapeutic agent activity in bladder cancer [42]. Other preclinical studies have tested the preventive efficacy of NSAIDs in rodent models of carcinogen-induced urinary bladder cancer. Selective COX-2 inhibitors (e.g., celecoxib, NS398, nimulside) or non-selective COX inhibitors (e.g., aspirin, indomethacin, piroxicam) have also had consistent and significant preventive and therapeutic effects in these rodent models (table 1). Celecoxib inhibited tumor size and multiplicity and prolonged survival in the N-butyl-N(4-hydroxybutyl)nitrosamine (OH-BBN) mouse carcinogenesis model and significantly reduced incidence of urinary bladder cancers in OH-BBN-exposed Fischer-344 rats [41]. The non-selective inhibitor indomethacin is active in prevention of rodent bladder carcinogenesis [25, 30, 33] and in treatment of carcinogen-induced rodent bladder tumors [23, 24, 37, 38]. Taken together, these data implicate involvement of the COX pathway in NSAID effects. Recent data, however, suggest that a COX-independent pathway may be involved. Piazza et al. [43], tested efficacy of sulindac sulfone, a novel pro-apoptotic NSAID that does not inhibit COX-1 or COX-2, in a well-designed study using the OH-BBN rat urinary bladder tumorigenesis model. Their results showed that rats administered 800, 1,000 and 1,200 mg/kg of sulindac sulfone in the diet resulted in a dose-dependent reduction in tumor multiplicity (by 36, 47 and 64%) and tumor incidence (by 31, 38 and 61%) for the respectively stated doses of drug [43]. These findings are extremely provocative and further studies to decipher anti-tumorigenic mechanisms of sulindac sulfone and other NSAID derivatives in bladder cancer will undoubtedly be undertaken.
Sabichi/Lippman
166
Table 1. In vivo NSAID studies in bladder cancer Group (first author)
Year
NSAIDs
Inhibits COX-1/ COX-2
Animal model
Study design (prevention (P)/ treatment (T))
Study result
Cohen [22] Droller [23] Droller [24] Droller [25] Hasegawa [26] Murasaki [27] Sakata [28] Cohen [29] Grubbs [30] Klan [31] Moon [32] Shibata [33] Knapp [34] Holmang [35] Rao [36]
1981 1982 1982 1984 1984 1984 1986 1989 1993 1993 1993 1993 1994 1995 1996
⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹ ⫹/⫹
Rats/FANFT Rats/FANFT Rats/FANFT Rats/FANFT Rats/FANFT Rats/FANFT Rats/FANFT Rats/FANFT Mice/BDF Rats/OH-BBN Mice/OH-BBN Rats/OH-BBN Dogs Rats/FANFT BDF mice/ OH-BBN
P T T P P P P P P P P P T T P
⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹Ⲑ⫺ ⫺ ⫺* ⫹⫹⫺
ObrechtPflumios [37] Ozaki [38]
1996
⫹/⫹
⫹
⫹/⫹
P
⫺
Okajima [39] Okajima [40] Grubbs [41]
1997 1998 2000
Piroxicam Nimesulide Celecoxib
⫺Ⲑ⫹ ⫺Ⲑ⫹ ⫺Ⲑ⫹
P P P
⫹ ⫹ ⫹
Knapp [42]
2000
Platinum and proxicam
⫹/⫹
T
⫹
Piazza [43]
2001
⫺/⫺
P
⫹
Mohammed [44]
2002
Sulindac sulfone Piroxicam
Mice/ ochratoxin A Rats/OH-BBN and NMU Rats/OH-BBN Rats/OH-BBN Rats/OH-BBN, BDF mice Dogs (spontaneous TCC) Rats/OH-BBN
P
1997
ASA Indomethacin Indomethacin Indomethacin ASA ASA ASA ASA Indomethacin ASA Piroxicam Indomethacin Piroxicam Indomethacin Sulindac, ketoprofen or ASA ASA or indomethacin Indomethacin
T
⫹
⫹/⫹
Dogs (spontaneous TCC)
ASA ⫽ Acetyl salicyclic acid; NMU ⫽ N-nitroso-N-methylurea; OH-BBN ⫽ N-butyl-N(4-hydroxybutyl)nitrosamine; FANFT ⫽ N-[4-(5-nitro-2-furyl)-2-thiazole] formamide; BDF ⫽ B6D2F1 hybrid strain mice; TCC ⫽ transitional cell carcinoma. *Enhanced tumor formation.
COX-2 Inhibitors and Other NSAIDs in Bladder and Prostate Cancer
167
NSAID Activity in Human Bladder Cancer Cell Lines Mechanistic studies exploring NSAIDs specific mechanism(s) of action have been undertaken in human bladder cancer cell lines. The human bladder cancer cell lines RT4, 5637 and T24 were shown to express high levels of COX-2 mRNA and to have high basal levels of prostaglandin E2 (PGE2) [45]. Indomethacin, which inhibits both COX-1 and COX-2, inhibited synthesis of PGE2 production in these three bladder cell lines [45]. Sulindac sulfone (which does not inhibit either COX enzyme), a metabolite of sulindac, also inhibited growth (in the human bladder tumor cell line HT1376), suggesting a COX-independent mechanism [43]. Although sulindac sulfone’s mechanism of growth inhibition is unknown, this drug induced apoptosis (assessed with DNA fragmentation, caspase activation, and morphology) and inhibited the phosphodiesterase (PDE) isozymes PDE5 and PDE4 [43]. There is evidence from bladder cancer cell lines that NSAIDs may inhibit bladder carcinogenesis by reducing conversion of pro-carcinogens to carcinogens. One study showed that sulindac (a non-selective COX inhibitor) inhibits expression and activity of arylamine N-acetyltransferase (NAT) in a dosedependent manner in a human bladder cancer cell line T24 [46]. NAT is an important enzyme involved in the biotransformation of various xenobiotics to carcinogens that are etiologic in the formation of bladder cancer. In T24, NAT converts a pro-carcinogen to a carcinogenic metabolite that covalently binds DNA to form cell-damaging DNA adducts. Yang et al. [46] demonstrated that sulindac also inhibited DNA-carcinogen adduct formation in these human bladder tumor cells. BCG immunotherapy, often used as an adjuvant treatment for bladder cancer, stimulates production of cytokines (interferon (IFN), tumor necrosis factor (TNF), etc.) and macrophpage-mediated cytotoxic activity within the bladder [47]. BCG-activated macrophages, however, also produce PGE2, a product of COX-2 that negatively regulates the cytotoxic activity of macrophages. PGE2, can be down-regulated by COX-2 inhibitors. The effects of COX inhibitors on BCG-induced macrophage cytotoxicity have been studied in the human bladder cancer cell line MBR-2. COX-2 inhibition by NS398 (a selective COX-2 inhibitor) and indomethacin (a non-selective COX inhibitor) enhanced BCGinduced cytotoxic activity and IFN-␥ and TNF-␣ production by reducing production of PGE2 [47]. These data suggest that COX inhibitors such as NS398 and indomethacin may potentially enhance BCG-induced anti-tumor activity. To date, there are no completed clinical trials determining efficacy of COX-2 inhibitors in bladder cancer. However, COX-2 inhibitors in bladder carcinogenesis are currently under clinical study. An NCI-sponsored prospective phase III trial is testing the ability of the selective COX-2 inhibitor celecoxib (200 mg twice a day for 1–2 years) (versus placebo) to prevent tumor recurrence
Sabichi/Lippman
168
in superficial bladder cancer patients at high risk of recurrence following TURBT and adjuvant BCG. Important secondary endpoints of this clinical trial include measures of quality of life and the correlations between the modulation of one or more urothelial biomarkers and TCC recurrence, which may validate these biomarkers in this setting.
Prostate Cancer
Overview Prostate cancer is the most common visceral cancer in US men, who have a 1 in 6 lifetime risk of prostate cancer development and a 1 in 30 lifetime risk of prostate cancer death. Over 330,000 US men died of this disease between 1989 and 1998 [48]. Increased screening with prostate-specific antigen (PSA) has increased the fraction of men diagnosed with localized prostate cancer, and prostate cancer mortality rates have begun to fall. Despite this progress, it was estimated that 189,000 cases and 30,200 deaths from prostate cancer would occur in 2002 [1]. Prostate cancer incidence and mortality rates are markedly different between various ethnicities, African-American men having the highest rates in the world. These data underscore our relatively poor understanding of the genetic, environmental, nutritional and biologic variables of this disease. Besides incidence and mortality, prostate cancer burden also involves important psychosocial and quality of life consequences of the disease and its treatment, such as the >50% impotence rate following radical prostatectomy or definitive radiation [49, 50]. Inflammation and Fatty Acids Epidemiological studies suggest that prostate infections may increase prostate cancer risk [51, 52]. Inflammatory cells respond to infections by producing activated oxygen (e.g., superoxide) and nitrogen (e.g., nitric oxide), which may cause prostate cell and genome damage. iNOS is expressed at high levels by inflammatory cells in the prostate and is a potential target for prostate cancer prevention. A recently described prostate lesion, proliferative inflammatory atrophy (PIA), may provide a pathologic link between prostate inflammation and prostate cancer. PIA is a focal chronic inflammatory lesion containing proliferating epithelial cells that fail to fully differentiate into columnar secretory cells. These lesions are usually found in the periphery of the gland, where prostate cancers commonly arise and are often adjacent to prostatic intraepithelial neoplasia (PIN) and prostate cancers. PIA is characterized by an increased proliferative index, increased bcl-2, low apoptosis, decreased p27 and increased GST and ␣, and gene alterations similar to PIN
COX-2 Inhibitors and Other NSAIDs in Bladder and Prostate Cancer
169
and cancer. It is not clear whether PIA lesions arise from or are the cause of prostatic inflammation [53]. A number of epidemiological studies suggest that a diet high in fat, especially red meat, is a risk factor for prostate cancer. The Health Professionals Follow-Up Study, Physicians Health Study, and a large cohort study in Hawaii indicated that the consumption of red meats was consistently associated with increased prostate cancer risk [54]. These epidemiological data suggest that polyunsaturated fats in the diet may influence prostate carcinogenesis. If this hypothesis is correct, it may be rational to focus prostate cancer prevention strategies on agents such as NSAIDs and other COX or LOX modulators that target molecular alterations within polyunsaturated fatty acid metabolic pathways. There are 9 published epidemiological studies involving NSAIDs and prostate cancer. Five of these studies (2 case-control, 3 cohort) suggested that NSAID intake and prostate cancer risk are inversely associated. The most recent study was conducted in a multiracial US cohort of over 90,000 men, of whom 2,574 developed prostate cancer. The relative risk of prostate cancer associated with use of over 6 aspirin tablets almost every day was 0.75 (95% CI ⫽ 0.60–0.98). Prostate cancer relative risk did not differ significantly by age (at diagnosis), stage or race [55]. COX-2 Expression in Human Prostate Cancer In contrast to COX-2 in bladder cancer, there are conflicting reports of COX-2 expression in prostate cancer. In several studies of clinical specimens, COX-2 expression was increased in tumor versus normal tissue [56]. Three recent reports, however, present conflicting data. Subbarayan et al. [57] found that in vitro prostate cancer cells had very low COX-2 expression and had an abnormal response (disorganized subcellular COX-2 distribution) to cytokine stimulation. Shappell et al. [58] reported that COX-2 expression was not generally increased in prostate cancer (versus benign prostate tissue), although they reported more COX-2 expression in higher grade tumors. In the most recent and comprehensive report of this issue, Zha et al. [59] confirmed and extended the in vitro findings of Subbarayan et al. They also found that the levels of COX-2 expression in prostate cancer and high-grade PIN specimens from 144 human prostate cancer cases were low and comparable with expression levels in adjacent normal prostate tissue. The extent of positive COX-2 immunohistochemical staining did not correlate with prostate cancer grade or stage. In contrast to expression in normal, PIN or prostate cancer tissue, COX-2 expression levels in PIA were high, due possibly to cellular stress induced by inflammatory oxidants and cytokines. These provocative data suggest that COX-2 inhibitors may exert preventive effects at the PIA stage of prostate carcinogenesis.
Sabichi/Lippman
170
Table 2. In vivo NSAID studies in prostate cancer Group (first author)
Year
NSAID
Inhibits COX-1/ COX-2
Animal model
Study design (prevention (P)/ treatment (T))
Study result
Pollard [60]
1977
⫹/⫹
Wistar rats
T
⫹
Drago [61]
1983
Indomethacin or ASA Indomethacin
⫹/⫹
T
⫹
Drago [62]
1984
Indomethacin
⫹/⫹
T
⫹
Drago [63]
1984
Indomethacin
⫹/⫹
T
⫹
Pollard [64] Goluboff [65] Wechter [66] Liu [67] Asea [68]
1986 1999 2000 2000 2001
⫹/⫹ ⫺/⫺ ⫺/⫺ ⫺/⫹ ⫹/⫹
T T P, T T T
⫹ ⫹ ⫹ ⫹ ⫹
Goluboff [69]
2001
Piroxicam Sulindac sulfone R-fluroprofen NS398 Ibuprofen, sulindac, hyperthermia and XRT Sulindac sulfone vs. placebo
Noble rats xenografts/AS Noble rats xenografts/AI Noble rats xenografts/AI Rat/xenograft Xenografts/AS TRAMP Xenograft/AI Xenograft/AI
P
⫹*
⫺/⫺
D0 prostate cancer patients
TRAMP ⫽ Transgenic adenocarcinoma of mouse prostate; AI ⫽ androgen-independent prostate cancer cell line; AS ⫽ androgen-sensitive prostate cancer cell line; XRT ⫽ radiotherapy. *Stabilization/reduction of PSA.
Preclinical NSAID Studies in Prostate Cancer A number of NSAID studies in various animal prostate cancer models [60–68] and one in humans [69] have been published (see table 2). Selective COX-2-inhibitors, non-selective COX-inhibitors and non-COX-inhibitors have major activity in prostate cancer treatment and prevention. The role of COX-2 as a target of NSAID activity in the prostate is further complicated by the unique COX-2 expression patterns (i.e., low in human PIN and prostate cancer) discussed above. The COX-2 inhibitors are active against prostate cancer in vivo and in vitro. NS398 suppressed tumor growth in the androgen-independent PC3 mouse xenograft model by directly inducing apoptosis and down-regulating tumor VEGF with decreased angiogenesis [67]. Non-selective COX inhibiting NSAIDs can sensitize mouse prostate tumors to hyperthermia and radiation [68]. Celecoxib induces apoptosis in prostate cancer cells by interfering with
COX-2 Inhibitors and Other NSAIDs in Bladder and Prostate Cancer
171
multiple signaling targets, including Akt, ERK2, and endoplasmic reticulum Ca-ATPases [70]. Celecoxib affects apoptosis independently of androgen responsiveness, the level of bcl-2, or the functional status of p53. Other selective COX-2 inhibitors (e.g., rofecoxib and NS398) are much less effective than celecoxib in inducing apoptosis. Celecoxib derivatives that lacked COX-2 inhibitory activity, however, were as potent as celecoxib itself in inducing apoptosis in PC3 cells [70]. NSAIDs with no COX-1 or -2 inhibitory activity were tested in three recent animal studies. Wechter et al. [66] tested R-flurbiprofen, an NSAID that does not inhibit COX activity but may down-regulate COX-2 mRNA transcripts, in the transgenic adenocarcinoma mouse prostate (TRAMP) model. R-flurbiprofen treatment began at week 12, when PIN develops in the TRAMP model. R-flurbiprofen produced a dose-response preventive effect (reduced rate of conversion from PIN to prostate cancer) and therapeutic effect (reduced rate of metastases) [66]. Sulindac sulfone, an NSAID that does not inhibit COXs and may target cGMP PDE, has major in vitro activity against androgensensitive and -insensitive human prostate cancer cells [65]. Based on the in vitro work, Goluboff et al. [65] tested sulindac sulfone in a LNCaP nude mouse xenograft model. They observed an 80–90% growth inhibition of prostate cancer in association with a significant increase in apoptosis. This preclinical work led to a recently reported multicenter placebo-controlled, double blind, phaseIII trial in patients with rising prostate specific antigen (PSA) levels following radical prostatectomy (stage D0 disease). The trial included 96 patients randomized to sulindac sulfone (500 mg/day) versus placebo for 12 months [69]. The NSAID suppressed PSA levels overall (p ⫽ 0.017) and in the high-risk subset of men (p ⫽ 0.0003) and prolonged PSA doubling time (p ⫽ 0.048). Polyunsaturated fatty acid metabolic pathways involving 5-, 12- and 15-LOX also have been implicated in prostate carcinogenesis [71]. 5-LOX converts arachadonic acid to 5-S-hydroxyeicosatetraenoic acid (HETE), which in turn is converted to leukotriene A4 (LTA4) and then to leukotriene B4 (LTB4) by LTA-4 hydrolase. This process requires the activity of the additional enzyme 5-LOX-activating protein (FLAP), which activates 5-LOX and facilitates its translocation into the nuclear membrane. 5-LOX overexpression has been documented in human prostate cancer tissue [71, 72], and 5-S-HETE formation and inhibition respectively promote and inhibit growth of prostate cancer cells. 5-S-HETE but not other HETE products can inhibit apoptosis induction by MK-886 (a specific FLAP inhibitor) in prostate cancer cell lines [73]. 12-LOX has been shown to regulate the growth, angiogenesis and metastasis of prostate cancer. The degree of 12-S-LOX overexpression in human prostate cancer correlates with tumor grade and stage [56]. A role for 12-S-LOX in tumor metastasis also is supported by higher 12-S-LOX expression
Sabichi/Lippman
172
levels in metastatic prostate cancer cells (DU145) than in non-metastatic prostate cancer cells (PC3). 12-S-LOX inhibition markedly reduced the metastatic potential of DU145 cells [56]. Overexpression of exogenous 12-S-LOX enhances tumorigenesis by promoting angiogenesis in human prostate cancer cells [74]. 12-LOX inhibition produces growth inhibition associated with a specific G1 arrest, followed by induction of apoptosis through caspase and bcl-mediated mechanisms [75]. 15-LOX-2 converts arachidonic acid mainly into 15-S-HETE. 15-LOX-2 expression is reduced in human prostate cancer [76] and high-grade PIN [77]. There are conflicting in vitro data, however, regarding the role of 15-HETE in carcinogenesis. A recent study indicated that 15-S-HETE inhibits proliferation in PC3 cells possibly via activation of peroxisome proliferator-activated receptor-␥ [78]. The effects of NSAIDs on 15-LOX-2 expression are under evaluation.
Conclusions
COX-2 inhibitors and other NSAIDs are very promising for the prevention and treatment of genitourinary cancers. In contrast to the strong NSAID epidemiology in colon cancer, there are few epidemiological data on the relationship between NSAIDs and the incidences of bladder or prostate cancers, and these data are mixed. A substantial body of preclinical in vitro and animal model data, however, shows consistent activity of NSAIDs in treating and in some cases preventing these cancers (tables 1, 2). Selective-COX-inhibitor, non-selective-COX-inhibitor and non-COX-inhibitor NSAIDs have all shown activity against bladder and prostate cancers, which implicates involvement of non-COX-2-mediated pathway in bladder carcinogenesis, where COX-2 overexpression is clear, and in prostate carcinogenesis, where it is not. Although the mechanisms involved in NSAID anti-tumor effects are not fully elucidated, the anti-tumor effects are linked consistently with induction of apoptosis in studies of bladder and prostate cancer cell lines. Clinical results have shown that combinations of chemotherapy [42] or radiotherapy with NSAID/COX-2 inhibitors [68] are more effective in treating bladder and prostate cancers than is either modality alone. Therefore, these agents may be most effective when combined with other treatment modalities or agents. This concept will require further evaluation in clinical trials. The only reported completed NSAID study in prostate cancer indicated that a nonCOX-inhibiting NSAID, sulindac sulfone, was active in preventing PSA progression in patients with D0 prostate cancer (elevated PSA in absence of clinically detectable prostate cancer). The selective COX-2 inhibitor, celecoxib, is a very promising NSAID currently being tested in an NCI-sponsored
COX-2 Inhibitors and Other NSAIDs in Bladder and Prostate Cancer
173
phase III prevention trial in the bladder and in a pharmacodynamic trial in preprostatectomy patients. A large number of preclinical and clinical NSAID studies are underway in bladder, prostate, skin, esophageal, pancreatic, breast, lung and cervical cancers. These studies are addressing many unanswered questions, such as what are the best type (selective or non-selective COX inhibitors or non-COX inhibitors), dose and duration of NSAIDs. Future studies likely will focus on NSAID mechanisms and combinations for both treating and preventing genitourinary cancers.
References 1 2 3 4 5 6 7
8 9
10
11 12
13
14 15 16
17
Jemal A, Thomas A, Murray T, Thun M: Cancer statistics, 2002. CA Cancer J Clin 2002;52:23–47. Ross RK, Jones PA, Yu MC: Bladder cancer epidemiology and pathogenesis. Semin Oncol 1996;23:536–545. Vineis P, Esteve J, Hartge P, Hoover R, Silverman DT, Terracini B: Effects of timing and type of tobacco in cigarette-induced bladder cancer. Cancer Res 1988;48:3849–3852. Serretta V, Pomara G, Piazza F, Gange E: Pure squamous cell carcinoma of the bladder in western countries. Report on 19 consecutive cases. Eur Urol 2000;37:85–89. Amling CL: Diagnosis and management of superficial bladder cancer. Curr Probl Cancer 2001; 25:219–278. Baselli EC, Greenberg RE: Intravesical therapy for superficial bladder cancer. Oncology (Huntingt) 2000;14:719–729. Langman MJ, Cheng KK, Gilman EA, Lancashire RJ: Effect of anti-inflammatory drugs on overall risk of common cancer: Case-control study in general practice research database. BMJ 2000; 320:1642–1646. Castelao JE, Yuan JM, Gago-Dominguez M, Yu MC, Ross RK: Non-steroidal anti-inflammatory drugs and bladder cancer prevention. Br J Cancer 2000;82:1364–1369. Hodder SL, Mahmoud AA, Sorenson K, Weinert DM, Stein RL, Ouma JH, Koech D, King CH: Predisposition to urinary tract epithelial metaplasia in Schistosoma haematobium infection. Am J Trop Med Hyg 2000;63:133–138. Wall BM, Dmochowski RR, Malecha M, Mangold T, Bobal MA, Cooke CR: Inducible nitric oxide synthase in the bladder of spinal cord injured patients with a chronic indwelling urinary catheter. J Urol 2001;165:1457–1461. El-Sheikh SS, Madaan S, Alhasso A, Abel P, Stamp G, Lalani EN: Cyclooxygenase-2: A possible target in schistosoma-associated bladder cancer. BJU Int 2001;88:921–927. Komhoff M, Guan Y, Shappell HW, Davis L, Jack G, Shyr Y, Koch MO, Shappell SB, Breyer MD: Enhanced expression of cyclooxygenase-2 in high-grade human transitional cell bladder carcinomas. Am J Pathol 2000;157:29–35. Ristimaki A, Nieminen O, Saukkonen K, Hotakainen K, Nordling S, Haglund C: Expression of cyclooxygenase-2 in human transitional cell carcinoma of the urinary bladder. Am J Pathol 2001; 158:849–853. Shirahama T: Cyclooxygenase-2 expression is up-regulated in transitional cell carcinoma and its preneoplastic lesions in the human urinary bladder. Clin Cancer Res 2000;6:2424–2430. Shirahama T, Sakakura C: Overexpression of cyclooxygenase-2 in squamous cell carcinoma of the urinary bladder. Clin Cancer Res 2001;7:558–561. Shirahama T, Arima J, Akiba S, Sakakura C: Relation between cyclooxygenase-2 expression and tumor invasiveness and patient survival in transitional cell carcinoma of the urinary bladder. Cancer 2001;92:188–193. Yoshimura R, Sano H, Mitsuhashi M, Kohno M, Chargui J, Wada S: Expression of cyclooxygenase-2 in patients with bladder carcinoma. J Urol 2001;165:1468–1472.
Sabichi/Lippman
174
18
19
20 21 22
23
24
25 26
27
28
29
30
31 32 33 34
35
36
37
Mohammed SI, Knapp DW, Bostwick DG, Foster RS, Khan KN, Masferrer JL, Woerner BM, Snyder PW, Koki AT: Expression of cyclooxygenase-2 in human invasive transitional cell carcinoma of the urinary bladder. Cancer Res 1999;59:5647–5650. Dannenberg AJ, Altorki NK, Boyle JO, Dang C, Howe LR, Weksler BB, Subbaramaiah K: Cyclooxygenase-2: A pharmacological target for the prevention of cancer. Lancet Oncol 2001;2: 544–551. Khan KN, Knapp DW, Denicola DB, Harris RK: Expression of cyclooxygenase-2 in transitional cell carcinoma of the urinary bladder in dogs. Am J Vet Res 2000;61:478–481. Dubois RN, Abramson SB, Crofford L, Gupta RA, Simon LS, Van De Putte LB, Lipsky PE: Cyclooxygenase in biology and disease. FASEB J 1998;12:1063–1073. Cohen SM, Zenser TV, Murasaki G, Fukushima S, Mattammal MB, Rapp NS, Davis BB: Aspirin inhibition of N-[4-(5-nitro-2-furyl)-2-thiazolyl]formamide-induced lesions of the urinary bladder correlated with inhibition of metabolism by bladder prostaglandin endoperoxide synthetase. Cancer Res 1981;41:3355–3359. Droller MJ, Gomolka D: Inhibition of tumor growth in association with modification of in vivo immune response by indomethacin and polyinosinic:polycytidylic acid. Cancer Res 1982;42: 5038–5045. Droller MJ, Gomolka D: Interferon induction and prostaglandin synthetase inhibition in the in vitro and in vivo manipulation of immune response expression in an animal model of bladder cancer. Cell Immunol 1982;72:1–13. Droller MJ, Gomolka D: Indomethacin and poly I:C in the inhibition of carcinogen-induced bladder cancer in an experimental animal model. J Urol 1984;131:1212–1217. Hasegawa R, St John M, Murasaki G, Fukushima S, Cohen SM: Effect of aspirin on N-[4-(5-nitro2-furyl)-2-thiazolyl]formamide-induced epithelial proliferation in the urinary bladder and forestomach of the rat. Cancer Lett 1984;21:269–275. Murasaki G, Zenser TV, Davis BB, Cohen SM: Inhibition by aspirin of N-[4-(5-nitro-2-furyl)-2thiazolyl]formamide-induced bladder carcinogenesis and enhancement of forestomach carcinogenesis. Carcinogenesis 1984;5:53–55. Sakata T, Hasegawa R, Johansson SL, Zenser TV, Cohen SM: Inhibition by aspirin of N-[4-(5nitro-2-furyl)-2-thiazolyl]formamide initiation and sodium saccharin promotion of urinary bladder carcinogenesis in male F344 rats. Cancer Res 1986;46:3903–3906. Cohen SM, Hasegawa R, Sakata T, Johansson SL: Effect of aspirin on urinary bladder carcinogenesis initiated with N-[4-(5-nitro-2-furyl)-2-thiazolyl]formamide in rats. Cancer Res 1989;49: 372–377. Grubbs CJ, Juliana MM, Eto I, Casebolt T, Whitaker LM, Canfield GJ, Manczak M, Steele VE, Kelloff GJ: Chemoprevention by indomethacin of N-butyl-N-(4-hydroxybutyl)nitrosamineinduced urinary bladder tumors. Anticancer Res 1993;13:33–36. Klan R, Knispel HH, Meier T: Acetylsalicylic acid inhibition of n-butyl-(4-hydroxybutyl) nitrosamine-induced bladder carcinogenesis in rats. J Cancer Res Clin Oncol 1993;119:482–485. Moon RC, Kelloff GJ, Detrisac CJ, Steele VE, Thomas CF, Sigman CC: Chemoprevention of OHBBN-induced bladder cancer in mice by piroxicam. Carcinogenesis 1993;14:1487–1489. Shibata MA, Hasegawa R, Shirai T, Takesada Y, Fukushima S: Chemoprevention by indomethacin of tumor promotion in a rat urinary bladder carcinogenesis model. Int J Cancer 1993;55:1011–1017. Knapp DW, Richardson RC, Chan TC, Bottoms GD, Widmer WR, DeNicola DB, Teclaw R, Bonney PL, Kuczek T: Piroxicam therapy in 34 dogs with transitional cell carcinoma of the urinary bladder. J Vet Intern Med 1994;8:273–278. Holmang S, Cano M, Grenabo L, Hedelin H, Johansson SL: Effect of indomethacin on N-[4-(5nitro-2-furyl)-2-thiazolyl]formamide-induced urinary tract carcinogenesis. Carcinogenesis 1995; 16:1493–1498. Rao KV, Detrisac CJ, Steele VE, Hawk ET, Kelloff GJ, McCormick DL: Differential activity of aspirin, ketoprofen and sulindac as cancer chemopreventive agents in the mouse urinary bladder. Carcinogenesis 1996;17:1435–1438. Obrecht-Pflumio S, Grosse Y, Pfohl-Leszkowicz A, Dirheimer G: Protection by indomethacin and aspirin against genotoxicity of ochratoxin A, particularly in the urinary bladder and kidney. Arch Toxicol 1996;70:244–248.
COX-2 Inhibitors and Other NSAIDs in Bladder and Prostate Cancer
175
38
39
40
41
42
43
44
45
46
47
48 49
50
51
52 53 54 55
Ozaki M, Shibata MA, Takahashi S, Orita SI, Kawabe M, Shirai T: Lack of involvement of p53 gene mutations in N-methyl-N-nitrosourea-induced bladder tumor progression in N-butyl-N-(4hydroxybutyl)nitrosamine-treated rats and no suppression by indomethacin. Cancer Lett 1997; 115:249–255. Okajima E, Ozono S, Endo T, Majima T, Tsutsumi M, Fukuda T, Akai H, Denda A, Hirao Y, Nishino H, Nir Z, Konishi Y: Chemopreventive efficacy of piroxicam administered alone or in combination with lycopene and -carotene on the development of rat urinary bladder carcinoma after N-butyl-N-(4-hydroxybutyl)nitrosamine treatment. Jpn J Cancer Res 1997;88:543–552. Okajima E, Denda A, Ozono S, Takahama M, Akai H, Sasaki Y, Kitayama W, Wakabayashi K, Konishi Y: Chemopreventive effects of nimesulide, a selective cyclooxygenase-2 inhibitor, on the development of rat urinary bladder carcinomas initiated by N-butyl-N-(4-hydroxybutyl) nitrosamine. Cancer Res 1998;58:3028–3031. Grubbs CJ, Lubet RA, Koki AT, Leahy KM, Masferrer JL, Steele VE, Kelloff GJ, Hill DL, Seibert K: Celecoxib inhibits N-butyl-N-(4-hydroxybutyl)nitrosamine-induced urinary bladder cancers in male B6D2F1 mice and female Fischer-344 rats. Cancer Res 2000;60: 5599–5602. Knapp DW, Glickman NW, Widmer WR, DeNicola DB, Adams LG, Kuczek T, Bonney PL, DeGortari AE, Han C, Glickman LT: Cisplatin versus cisplatin combined with piroxicam in a canine model of human invasive urinary bladder cancer. Cancer Chemother Pharmacol 2000;46: 221–226. Piazza GA, Thompson WJ, Pamukcu R, Alila HW, Whitehead CM, Liu L, Fetter JR, Gresh WE Jr, Klein-Szanto AJ, Farnell DR, Eto I, Grubbs CJ: Exisulind, a novel proapoptotic drug, inhibits rat urinary bladder tumorigenesis. Cancer Res 2001;61:3961–3968. Mohammed SI, Bennett PF, Craig BA, Glickman NW, Mutsaers AJ, Snyder PW, Widmer WR, DeGortari AE, Bonney PL, Knapp DW: Effects of the cyclooxygenase inhibitor, piroxicam, on tumor response, apoptosis, and angiogenesis in a canine model of human invasive urinary bladder cancer. Cancer Res 2002;62:356–358. Bostrom PJ, Aaltonen V, Soderstrom KO, Uotila P, Laato M: Expression of cyclooxygenase-1 and -2 in urinary bladder carcinomas in vivo and in vitro and prostaglandin E2 synthesis in cultured bladder cancer cells. Pathology 2001;33:469–474. Yang CC, Chung JG: Sulindac-inhibited gene expression and activity of arylamine N-acetyltransferase and DNA-2-aminofluorene adduct formation in T24 human bladder tumor cells. Urol Res 2001;29:406–411. Yamada H, Kuroda E, Matsumoto S, Matsumoto T, Yamada T, Yamashita U: Prostaglandin E2 down-regulates viable Bacille Calmette-Guérin-induced macrophage cytotoxicity against murine bladder cancer cell MBT-2 in vitro. Clin Exp Immunol 2002;128:52–58. US Mortality Public Use Data Tapes 1989–1998, in National Center for Health Statistics, Centers for Disease Control and Prevention, 2000. Stanford JL, Feng Z, Hamilton AS, Gilliland FD, Stephenson RA, Eley JW, Albertsen PC, Harlan LC, Potosky AL: Urinary and sexual function after radical prostatectomy for clinically localized prostate cancer: The Prostate Cancer Outcomes Study. JAMA 2000;283:354–360. Potosky AL, Legler J, Albertsen PC, Stanford JL, Gilliland FD, Hamilton AS, Eley JW, Stephenson RA, Harlan LC: Health outcomes after prostatectomy or radiotherapy for prostate cancer: Results from the Prostate Cancer Outcomes Study. J Natl Cancer Inst 2000;92:1582–1592. Hayes RB, Pottern LM, Strickler H, Rabkin C, Pope V, Swanson GM, Greenberg RS, Schoenberg JB, Liff J, Schwartz AG, Hoover RN, Fraumeni JF Jr: Sexual behaviour, STDs and risks for prostate cancer. Br J Cancer 2000;82:718–725. Dennis LK, Dawson DV: Meta-analysis of measures of sexual activity and prostate cancer. Epidemiology 2002;13:72–79. Nelson WG, De Marzo AM, DeWeese TL: The molecular pathogenesis of prostate cancer: Implications for prostate cancer prevention. Urology 2001;57:39–45. Kolonel LN, Nomura AM, Cooney RV: Dietary fat and prostate cancer: Current status. J Natl Cancer Inst 1999;91:414–428. Habel LA, Zhao W, Stanford JL: Daily aspirin use and prostate cancer risk in a large, multiracial cohort in the US. Cancer Causes Control 2002;13:427–434.
Sabichi/Lippman
176
56 57
58
59
60 61 62 63 64 65
66
67 68
69
70
71 72 73 74
75
76 77
Nie D, Che M, Grignon D, Tang K, Honn KV: Role of eicosanoids in prostate cancer progression. Cancer Metastasis Rev 2001;20:195–206. Subbarayan V, Sabichi AL, Llansa N, Lippman SM, Menter DG: Differential expression of cyclooxygenase-2 and its regulation by tumor necrosis factor-␣ in normal and malignant prostate cells. Cancer Res 2001;61:2720–2726. Shappell SB, Manning S, Boeglin WE, Guan YF, Roberts RL, Davis L, Olson SJ, Jack GS, Coffey CS, Wheeler TM, Breyer MD, Brash AR: Alterations in lipoxygenase and cyclooxygenase-2 catalytic activity and mRNA expression in prostate carcinoma. Neoplasia 2001;3:287–303. Zha S, Gage WR, Sauvageot J, Saria EA, Putzi MJ, Ewing CM, Faith DA, Nelson WG, De Marzo AM, Isaacs WB: Cyclooxygenase-2 is up-regulated in proliferative inflammatory atrophy of the prostate, but not in prostate carcinoma. Cancer Res 2001;61:8617–8623. Pollard M, Chang CF, Luckert PH: Investigations on prostatic adenocarcinomas in rats. Oncology 1977;34:129–132. Drago JR, Weed P, Fralisch A: Nb rat prostate adenocarcinoma model androgen-sensitive tumor: Metastasis control. J Surg Oncol 1983;24:33–35. Drago JR, Al-Mondhiry HA: The effect of prostaglandin modulators on prostate tumor growth and metastasis. Anticancer Res 1984;4:391–394. Drago JR, Murray C: Control of metastases in the Nb rat prostatic adenocarcinoma model. J Androl 1984;5:265–268. Pollard M, Luckert PH: The beneficial effects of diphosphonate and piroxicam on the osteolytic and metastatic spread of rat prostate carcinoma cells. Prostate 1986;8:81–86. Goluboff ET, Shabsigh A, Saidi JA, Weinstein IB, Mitra N, Heitjan D, Piazza GA, Pamukcu R, Buttyan R, Olsson CA: Exisulind (sulindac sulfone) suppresses growth of human prostate cancer in a nude mouse xenograft model by increasing apoptosis. Urology 1999;53:440–445. Wechter WJ, Leipold DD, Murray ED Jr, Quiggle D, McCracken JD, Barrios RS, Greenberg NM: E-7869 (R-flurbiprofen) inhibits progression of prostate cancer in the TRAMP mouse. Cancer Res 2000;60:2203–2208. Liu XH, Kirschenbaum A, Yao S, Lee R, Holland JF, Levine AC: Inhibition of cyclooxygenase-2 suppresses angiogenesis and the growth of prostate cancer in vivo. J Urol 2000;164:820–825. Asea A, Mallick R, Lechpammer S, Ara G, Teicher BA, Fiorentino S, Stevenson MA, Calderwood SK: Cyclooxygenase inhibitors are potent sensitizers of prostate tumours to hyperthermia and radiation. Int J Hyperthermia 2001;17:401–414. Goluboff ET, Prager D, Rukstalis D, Giantonio B, Madorsky M, Barken I, Weinstein IB, Partin AW, Olsson CA: Safety and efficacy of exisulind for treatment of recurrent prostate cancer after radical prostatectomy. J Urol 2001;166:882–886. Song X, Lin HP, Johnson AJ, Tseng PH, Yang YT, Kulp SK, Chen CS: Cyclooxygenase-2, player or spectator in cyclooxygenase-2 inhibitor-induced apoptosis in prostate cancer cells. J Natl Cancer Inst 2002;94:585–591. Shureiqi I, Lippman SM: Lipoxygenase modulation to reverse carcinogenesis. Cancer Res 2001;61:6307–6312. Gupta S, Srivastava M, Ahmad N, Bostwick DG, Mukhtar H: Over-expression of cyclooxygenase-2 in human prostate adenocarcinoma. Prostate 2000;42:73–78. Ghosh J, Myers CE: Inhibition of arachidonate 5-lipoxygenase triggers massive apoptosis in human prostate cancer cells. Proc Natl Acad Sci USA 1998;95:13182–13187. Nie D, Hillman GG, Geddes T, Tang K, Pierson C, Grignon DJ, Honn KV: Platelet-type 12-lipoxygenase in a human prostate carcinoma stimulates angiogenesis and tumor growth. Cancer Res 1998;58:4047–4051. Pidgeon GP, Kandouz M, Meram A, Honn KV: Mechanisms controlling cell cycle arrest and induction of apoptosis after 12-lipoxygenase inhibition in prostate cancer cells. Cancer Res 2002;62:2721–2727. Shappell SB, Boeglin WE, Olson SJ, Kasper S, Brash AR: 15-Lipoxygenase-2 is expressed in benign prostatic epithelium and reduced in prostate adenocarcinoma. Am J Pathol 1999;155:235–245. Jack GS, Brash AR, Olson SJ, Manning S, Coffey CS, Smith JA Jr, Shappell SB: Reduced 15-lipoxygenase-2 immunostaining in prostate adenocarcinoma: Correlation with grade and expression in high-grade prostatic intraepithelial neoplasia. Hum Pathol 2000;31:1146–1154.
COX-2 Inhibitors and Other NSAIDs in Bladder and Prostate Cancer
177
78
Shappell SB, Gupta RA, Manning S, Whitehead R, Boeglin WE, Schneider C, Case T, Price J, Jack GS, Wheeler TM, Matusik RJ, Brash AR, Dubois RN: 15S-hydroxyeicosatetraenoic acid activates peroxisome proliferator-activated receptor ␥ and inhibits proliferation in PC3 prostate carcinoma cells. Cancer Res 2001;61:497–503.
Anita L. Sabichi Department of Clinical Cancer Prevention, The University of Texas M.D. Anderson Cancer Center, Houston, TX 77030 (USA) Tel. ⫹1 713 7454928, Fax ⫹1 713 7944679, E-Mail
[email protected]
Sabichi/Lippman
178
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 179–192
Therapeutic Potential of Selective Cyclooxygenase-2 Inhibitors in the Management of Tumor Angiogenesis Stephen Gately a, Robert Kerbel b a b
NeoPharm Inc., Department of Translational Medicine, Lake Forest, Ill., USA and Sunnybrook and Women’s College Health Sciences Center, Molecular and Cellular Biology Research and Department of Medical Biophysics, University of Toronto, Toronto, Canada
Therapeutic strategies for the treatment and management of neoplastic diseases have largely focused on the cancer cell. These approaches have achieved only limited success with nearly half of all diagnosed cancer patients dying as a result of their disease despite aggressive treatment with highly toxic regimens. The hypothesis that tumor growth is dependent upon angiogenesis [1], and the subsequent genetic confirmation of this hypothesis [2, 3], provide the impetus for developing therapeutic strategies that target the microvasculature.
Angiogenesis
Tumor angiogenesis results from a cascade of molecular and cellular events [4], often initiated by the release of angiogenic growth factors [5–7]. Tumor cells express a variety of proangiogenic growth factors that diffuse in the direction of pre-existing blood vessels. These growth factors activate the normally quiescent vascular cells inducing: (a) proteolytic degradation of the basement membrane [8, 9]; (b) migration of endothelial cells towards the angiogenic stimulus; (c) endothelial cell proliferation; (d) lumen formation [10]; (e) pericyte capping and (f) production of a new basement membrane [11, 12]. Tumors also can acquire a vasculature through vasculogenesis, the formation of blood vessels from progenitor endothelial cells, or angioblasts. Vasculogenesis was thought to be restricted to the embryo, however, circulating
stem cells in peripheral blood that can differentiate into endothelial cells and contribute to angiogenesis in the adult have been detected [13–18].
Angiogenesis and Cancer
It generally is accepted that solid tumor growth and metastases are dependent upon the acquisition of an adequate blood supply [19, 20]. Pharmacological targeting of the microvasculature in patients with cancer represents an attractive therapeutic approach because inhibition of angiogenesis has been shown to prevent growth [21], and induce regression of experimental solid tumors [22]. Furthermore, antiangiogenic therapy could be utilized as adjuvant treatment for microscopic metastases, preventing the emergence of a vascular supply, and maintaining tumor dormancy [23, 24]. Because of the normally slow renewal of vascular stroma, even in tissues with high epithelial turnover, such as skin and jejunum, antiangiogenic therapies promise to be free of the hematopoietic, gastrointestinal and other toxicities of standard antiproliferative therapies [25]. Recent data suggest that selective inhibitors of COX-2 are potent inhibitors of angiogenesis [26, 27], suggesting the potential utility of these agents in oncology.
Cyclooxygenase-1 and -2
Cyclooxygenases (COX-1 and COX-2) are the two enzymes that convert arachidonic acid to prostaglandins [28]. COX-1 is constitutively expressed and is responsible for normal kidney and platelet function as well as the maintenance of the gastrointestinal mucosa [29]. In contrast, the COX-2 enzyme can be induced by a variety of proinflammatory cytokines such as interleukin-1 [30, 31], growth factors such as epidermal growth factor (EGF) [32, 33], transforming growth factor- (TGF-) [33–35], tumor necrosis factor-␣ (TNF-␣) [36], and ultraviolet B light (UVB) [37]. The enthusiasm for using nonselective COX inhibitors in oncology has been limited because these agents are associated with renal and gastrointestinal side effects in addition to suppressing platelet function. By contrast, selective COX-2 inhibitors appear to be safe enough to permit large-scale testing as possible anticancer agents.
Cyclooxygenase-2
The significant contribution of COX-2 in cancer promotion was elegantly demonstrated experimentally in a model of human familial adenomatous
Gately/Kerbel
180
polyposis, where mice, genetically predisposed for polyp formation by a targeted truncation deletion in the APC tumor suppressor gene (Apc knockout mice), were crossed with COX-2⫺/⫺ mice [38]. The tumor burden of the double mutant offspring was reduced significantly by the genetic knockout of COX-2. Polyp formation was also reduced by treating the COX-2⫹/⫹ mice with a selective COX-2 inhibitor [38]. More recently, it was shown in the Apc/COX-2 knockout mice that stromal expression of COX-2 was required for induction of VEGF and subsequent tumor angiogenesis [39]. The pharmacologic application of selective COX-2 inhibitors has also been demonstrated to significantly reduce the growth of a variety of experimental tumors [40].
Cellular Expression of COX-2
The elevated expression of COX-2 in the cancer setting has been localized to the neoplastic epithelium [41] within the microvasculature [27], within infiltrating immune cells [42], and within stromal fibroblasts [43]. Tumor Associated COX-2 Expression The potential importance of tumor-associated COX-2 expression for angiogenesis and tumor growth was demonstrated when COX-2-expressing tumor cells were found to form larger tumors and produce more angiogenic factors than tumor cells that lacked COX-2 expression [41]. Additionally, tumor growth and angiogenesis could be suppressed by selective COX-2 inhibitors only if the tumor cells expressed COX-2 [44]. Together these data demonstrate that tumor-associated COX-2 is important for angiogenesis and tumor growth, and suggest that the potential effectiveness of selective COX-2 inhibitors is determined by expression of COX-2 in the tumor cell. One potential mechanism for the elevated tumor-associated COX-2 expression could be related to the marked repression of transcription of the COX-2 gene by wild-type p53 [45]. Loss of wild-type p53 would then be associated with the increased expression of COX-2 mRNA and protein [45], increased secretion of basic fibroblast growth factor (bFGF) [46, 47], VEGF [48] and decreased production of an endogenous inhibitor of angiogenesis, thrombospondin-1 [49], resulting in the acquisition of a proangiogenic phenotype. Microvascular COX-2 Expression Previous studies suggested a role for endothelial cell-derived COX-1 in angiogenesis [41, 44]. However, a recent study demonstrated that selective COX-2 inhibitors could suppress growth factor-induced angiogenesis, and
COX-2 and Tumor Angiogenesis
181
suggested that endothelial-derived COX-2 is essential in directly regulating angiogenesis [50]. Normal quiescent vascular endothelial cells express COX-1 but not COX-2 [26]. A recent histochemical evaluation of human colon, breast, prostate and lung tumor biopsies demonstrated that COX-1 immunoreactivity is broadly distributed within normal and neoplastic tissues [27]. By contrast, COX-2 but not COX-1 appears to be induced and expressed in the vasculature of these tumors [27]. In this same study, Masferrer et al. [27] demonstrated the inhibition of experimental colon and lung tumor growth and bFGF-induced corneal angiogenesis using a selective inhibitor of COX-2, but not using a selective inhibitor of COX-1. In other studies, vascular endothelial cell-derived COX-2 has been shown to be important in the angiogenesis cascade [51]. These data confirm the important contribution of endothelial cell COX-2 in angiogenesis and suggest that the target of selective COX-2 inhibitors could be the microvasculature. It also suggests that COX-2 inhibitors may be effective inhibitors of tumor angiogenesis and consequently of tumor growth regardless of the COX-2 status of the neoplastic epithelial cell. Stromal Cell COX-2 Expression Williams et al. [43] addressed the relative contribution of tumor cell or host-derived COX-2 in the regulation of angiogenesis. Tumor growth was attenuated when implanted into COX-2⫺/⫺ knockout mice, but not COX-1⫺/⫺ or wild-type mice, suggesting host-derived COX-2 was required for successful tumor growth and angiogenesis. In this study, COX-1 did not play a significant role in either tumor growth or angiogenesis [43]. These data confirmed an important role for host-derived COX-2 in tumor growth and angiogenesis particularly by stromal fibroblasts. The tumor microenvironment is complex and includes not only cancer cells, endothelial cells and ‘normal’ stromal fibroblasts, but immune cells as well [52]. Inflammatory cells attracted to cancer sites contribute to sustained tumor growth through a variety of pathways that includes the production and/or release of angiogenic cytokines [52–54]. The selective COX-2 inhibitors celecoxib and rofecoxib were originally developed and approved for their anti-inflammatory properties [55], and may be indirectly antiangiogenic by inhibition of inflammatory cell activity.
Contributions of COX-2 in Angiogenesis
Prostaglandins and VEGF Expression In human vascular endothelial cells, the expression of the COX-2 gene is regulated by hypoxia [56–59]. Hypoxia also induces the expression of
Gately/Kerbel
182
proangiogenic molecules including VEGF [60, 61], inducible nitric oxide synthase (iNOS), [62] interleukin-6 [63], interleukin-8 [64] and TIE-2 [65]. A relationship between COX-2 expression and VEGF was demonstrated recently [43]. The expression pattern of COX-2 and VEGF mRNA in experimental tumor isografts were similar, with high levels of expression detected within the stromal cell compartment [43], consistent with that reported for VEGF previously [66]. Quantitation of VEGF protein levels in COX-2⫺/⫺ fibroblasts demonstrated a 94% reduction in VEGF levels when compared to wild-type and COX-1⫺/⫺ fibroblasts [43]. Tumors grown in COX-2⫺/⫺ mice demonstrated decreased expression of VEGF mRNA, as well as a 30% decrease in vascular density [43]. Pharmacologic exposure of wild-type fibroblasts to a selective COX-2 inhibitor resulted in a 92% reduction in VEGF production [43]. Other data suggest that COX-2-generated prostaglandins can enhance bFGF-induced angiogenesis through the induction of VEGF [67]. These data are consistent with reports demonstrating that prostaglandins can regulate VEGF expression [39, 68–74], and demonstrate an important link between COX-2 activity and VEGF in the stimulation of tumor angiogenesis [43]. Vascular endothelial cell COX-2 and generated prostaglandins can now be added to the growing list of bioactive products that can affect adjacent tumor cells in a paracrine manner [75]. Eicosanoid Products The cyclooxygenases catalyze the first two steps in the biosynthesis of the prostaglandins from arachidonic acid [28]. Prostaglandin H2 is transformed into the primary prostanoids PGE2, PGF2␣, PGD2, PGI2 and thromboxane A2 [28] that could be capable of functioning as intermediaries of angiogenesis. Thromboxane A 2 (TXA2 ) Because COX-2 has been associated with endothelial cell motility and the ability to form capillary-like structures [41, 76], studies were undertaken to determine the contribution of COX-2 to these phenotypes [51]. Endothelial cell migration was inhibited in the presence of a selective COX-2 inhibitor, however, addition of a TXA2 mimetic reconstituted the endothelial cell migratory phenotype [51]. These data suggest that inhibition of endothelial cell migration by COX-2 inhibition is mediated by downregulation of endothelial cell TXA2 production. The contribution of TXA2 to in vivo angiogenesis was confirmed in the mouse corneal angiogenesis assay, where local delivery of TXA2 receptor antagonists inhibited bFGF-induced corneal angiogenesis [51]. Furthermore, a TXA2 agonist could reverse the COX-2 inhibitor-mediated suppression of bFGF-induced corneal angiogenesis [51]. Additional studies demonstrated that inhibition of TXA2 synthesis could suppress bFGF and VEGF-induced human
COX-2 and Tumor Angiogenesis
183
endothelial cell migration [77]. Furthermore, inhibition of thromboxane synthase was capable of suppressing experimental tumor metastases [77], an angiogenesis dependent process. PGI2 /PGE2 A variety of eicosanoids are produced from PGH2 depending on the enzymatic machinery present in particular cell types [29], with vascular endothelial cells primarily producing prostacyclin, PGI2 [78]. Prostacyclin release can be increased when endothelial cells are exposed to the angiogenic growth factors bFGF, VEGF [79, 80], angiogenin [81] or by engagement of human plateletendothelial cell adhesion molecule-1 (PECAM) [82]. Prostacyclin and nitric oxide release appear to be key regulators of VEGF-induced vascular permeability [83]. Prostaglandins of the E series have been linked to experimental angiogenesis [84–89]. PGE2 has been directly associated with the expression and regulation of VEGF [39, 90–92], and was shown to induce capillary sprouting from veins [93]. COX-2 and Apoptosis The increased tumorigenic potential of COX-2 overexpressing cells is thought to be mediated in part by resistance to apoptosis [76]; treatment with selective COX-2 inhibitors induces apoptosis in a variety of cancer cells [94, 95]. The overexpression of COX-2 can lead to the increased production of the antiapoptotic protein Bcl-2 [96]. One potential mechanism for the proapoptotic activity of COX-2 inhibitors has been the downregulation of Bcl-2 [97]. The subcellular localization of COX-2 is nuclear as well as microsomal [98] providing a possible link between prostaglandins produced by COX-2 and Bcl-2 transcription, and the regulation of gene expression in general [28]. Although the precise link between prostaglandin production and Bcl-2 synthesis has not been elucidated, it is interesting to speculate on the potential role of COX-2 in the increased expression of Bcl-2 that results in vascular endothelial cell survival. Human microvascular endothelial cells that overexpress Bcl-2 are refractory to the apoptotic and angiosuppressive properties of thrombospondin-1 and appear to participate in a more vigorous and sustained angiogenic response [99, 100]. COX-2 appears to play a role in the activation of the survival serine threonine kinase, Akt [94, 101, 102]. Activation of Akt has important implications for angiogenesis as the Akt system is a critical signaling pathway for vascular endothelial cell survival [103, 104]. Selective COX-2 inhibitors have been found to induce apoptosis by blocking Akt activation [94, 101, 105], independent of Bcl-2 [106].
Gately/Kerbel
184
Summary and Conclusions
It is clear that COX-2 plays an important role in tumor and endothelial cell biology. Increased expression of COX-2 occurs in multiple cells within the tumor microenvironment that can impact on angiogenesis. COX-2 appears to: (a) play a key role in the release and activity of proangiogenic proteins; (b) result in the production of eicosanoid products TXA2, PGI2, PGE2 that directly stimulate endothelial cell migration and angiogenesis in vivo, and (c) result in enhanced tumor cell, and possibly, vascular endothelial cell survival by upregulation of the antiapoptotic proteins Bcl-2 and/or activation of PI3K-Akt. Selective pharmacologic inhibition of COX-2 represents a viable therapeutic option for the treatment of malignancies. Agents that selectively inhibit COX-2 appear to be safe, and well tolerated suggesting that chronic treatment for angiogenesis inhibition is feasible [107–110]. Because these agents inhibit angiogenesis, they should have at least additive benefit in combination with standard chemotherapy [111] and radiation therapy [24, 112]. In preclinical models, a selective inhibitor of COX-2 was shown to potentiate the beneficial antitumor effects of ionizing radiation with no increase in normal tissue cytotoxicity [113–115]. More recently, metronomic dosing regimens of standard chemotherapeutic agents without extended rest periods were shown to target the microvasculature in experimental animal models and result in significant antitumor activity [116–118]. This antiangiogenic chemotherapy regimen could be enhanced by the concurrent administration of an angiogenesis inhibitor [116–119]. Trials that will evaluate continuous low dose cyclophosphamide in combination with celecoxib are underway in patients with metastatic renal cancer, and non-Hodgkin’s lymphoma [120]. Given the safety and tolerability of the selective COX-2 inhibitors, and the potent antiangiogenic properties of these agents, the combination of antiangiogenic chemotherapy with a COX-2 inhibitor warrants clinical evaluation [118, 121, 122]. The effects of selective COX-2 inhibitors on angiogenesis may also be due, in part, to COX-independent mechanisms [123–125]. Several reports have confirmed COX-independent effects of celecoxib, at relatively high concentrations (50 M), where apoptosis is stimulated in cells that lack both COX-1 and COX-2 [126]. More recently, Song et al. [127] described structural modifications to celecoxib that revealed no association between the COX-2 inhibitory and proapoptotic activities of celecoxib [125]. Some of the COX-independent mechanisms for NSAIDs and selective COX-2 inhibitors include activation of protein kinase G, inhibition of NF-B activation, downregulation of the antiapoptotic protein Bcl-XL, inhibition of PPAR␦, and activation of PPAR␥. One or more of these COX-independent effects could contribute to the antiangiogenic properties of NSAIDs and selective COX-2 inhibitors. In order to take
COX-2 and Tumor Angiogenesis
185
advantage of both the COX-dependent and COX-independent benefits of NSAIDs and selective COX-2 inhibitors, will require evaluation of these agents in neoplastic disease settings, using cancer-specific biomarkers. In conclusion, the contribution of COX-2 at multiple points in the angiogenic cascade makes it an ideal target for pharmacologic inhibition. The reported success of selective COX-2 inhibitors in cancer prevention could be related to angiogenesis inhibition [109]. As premalignant lesions progress towards malignancy, there is a switch to the angiogenic phenotype that is subsequently followed by rapid tumor growth [128, 129]. Intervention with angiogenesis inhibitors at this early stage of carcinogenesis has been shown to attenuate tumor growth in transgenic mouse models [130, 131]. The continued dependence on angiogenesis for later stages of tumorigenesis suggests that COX-2 inhibitors also will have clinical utility in the management of advanced cancers.
Acknowledgement The authors gratefully acknowledge the excellent editorial assistance of Jeanne Pauvlik.
References 1 2
3
4 5
6 7 8 9
10
Folkman J: Tumor angiogenesis: Therapeutic implications. N Engl J Med 1971;285:1182–1186. Lyden D, Hattori K, Dias S, Costa C, Blaikie P, Butros L, Chadburn A, Heissig B, Marks W, Witte L, Wu Y, Hicklin D, Zhu Z, Hackett NR, Crystal RG, Moore MA, Hajjar KA, Manova K, Benezra R, Rafii S: Impaired recruitment of bone-marrow-derived endothelial and hematopoietic precursor cells blocks tumor angiogenesis and growth. Nat Med 2001;7:1194–1201. Lyden D, Young AZ, Zagzag D, Yan W, Gerald W, O’Reilly R, Bader BL, Hynes RO, Zhuang Y, Manova K, Benezra R: Id1 and Id3 are required for neurogenesis, angiogenesis and vascularization of tumour xenografts. Nature 1999;401:670–677. Hanahan D, Folkman J: Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell 1996;86:353–364. Kandel J, Bossy-Wetzel E, Radvanyi F, Klagsbrun M, Folkman J, Hanahan D: Neovascularization is associated with a switch to the export of bFGF in the multistep development of fibrosarcoma. Cell 1991;66:1095–1104. Reinmuth N, Stoeltzing O, Liu W, Ahmad SA, Jung YD, Fan F, Parikh A, Ellis LM: Endothelial survival factors as targets for antineoplastic therapy. Cancer J 2001;7(suppl 3):109–119. Ellis LM, Liu W, Ahmad SA, Fan F, Jung YD, Shaheen RM, Reinmuth N: Overview of angiogenesis: Biologic implications for antiangiogenic therapy. Semin Oncol 2001;28:94–104. Pepper MS, Montesano R: Proteolytic balance and capillary morphogenesis. Cell Differ Dev 1990;32:319–327. Subbaramaiah K, Telang N, Bansal MB, Weksler BB, Dannenberg AJ: Cyclooxygenase-2 gene expression is upregulated in transformed mammary epithelial cells. Ann NY Acad Sci 1997;833: 179–185. Mestre JR, Subbaramaiah K, Sacks PG, Schantz SP, Tanabe T, Inoue H, Dannenberg AJ: Retinoids suppress epidermal growth factor-induced transcription of cyclooxygenase-2 in human oral squamous carcinoma cells. Cancer Res 1997;57:2890–2895.
Gately/Kerbel
186
11 12 13 14
15 16 17
18
19 20 21 22
23
24
25 26
27
28 29 30 31
32 33
34
Paku S, Paweletz N: First steps of tumor-related angiogenesis. Lab Invest 1991;65:334–346. Paku S: Current concepts of tumor-induced angiogenesis. Pathol Oncol Res 1998;4:62–75. Risau W, Flamme I: Vasculogenesis. Annu Rev Cell Dev Biol 1995;11:73–91. Rafii S, Meeus S, Dias S, Hattori K, Heissig B, Shmelkov S, Rafii D, Lyden D: Contribution of marrow-derived progenitors to vascular and cardiac regeneration. Semin Cell Dev Biol 2002; 13:61–67. Benezra R, Rafii S, Lyden D: The Id proteins and angiogenesis. Oncogene 2001;20:8334–8341. Rafii S: Circulating endothelial precursors: mystery, reality, and promise. J Clin Invest 2000; 105:17–19. Shi Q, Rafii S, Wu MH, Wijelath ES, Yu C, Ishida A, Fujita Y, Kothari S, Mohle R, Sauvage LR, Moore MA, Storb RF, Hammond WP: Evidence for circulating bone marrow-derived endothelial cells. Blood 1998;92:362–367. Peichev M, Naiyer AJ, Pereira D, Zhu Z, Lane WJ, Williams M, Oz MC, Hicklin DJ, Witte L, Moore MA, Rafii S: Expression of VEGFR-2 and AC133 by circulating human CD34(+) cells identifies a population of functional endothelial precursors. Blood 2000;95:952–958. Folkman J: Looking for a good endothelial address. Cancer Cell 2002;1:113–115. Folkman J: Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1995; 1:27–31. Lannutti BJ, Gately ST, Quevedo ME, Soff GA, Paller AS: Human angiostatin inhibits murine hemangioendothelioma tumor growth in vivo. Cancer Res 1997;57:5277–5280. O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS, Flynn E, Birkhead JR, Olsen BR, Folkman J: Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell 1997;88: 277–285. Gately S, Twardowski P, Stack MS, Cundiff DL, Grella D, Castellino FJ, Enghild J, Kwaan HC, Lee F, Kramer RA, Volpert O, Bouck N, Soff GA: The mechanism of cancer-mediated conversion of plasminogen to the angiogenesis inhibitor angiostatin. Proc Natl Acad Sci USA 1997; 94:10868–10872. Mauceri HJ, Hanna NN, Beckett MA, Gorski DH, Staba MJ, Stellato KA, Bigelow K, Heimann R, Gately S, Dhanabal M, Soff GA, Sukhatme VP, Kufe DW, Weichselbaum RR: Combined effects of angiostatin and ionizing radiation in antitumour therapy. Nature 1998;394:287–291. Boehm T, Folkman J, Browder T, O’Reilly MS: Antiangiogenic therapy of experimental cancer does not induce acquired drug resistance. Nature 1997;390:404–407. Leahy KM, Ornberg RL, Wang Y, Zweifel BS, Koki AT, Masferrer JL: Cyclooxygenase-2 inhibition by celecoxib reduces proliferation and induces apoptosis in angiogenic endothelial cells in vivo. Cancer Res 2002;62:625–631. Masferrer JL, Leahy KM, Koki AT, Zweifel BS, Settle SL, Woerner BM, Edwards DA, Flickinger AG, Moore RJ, Seibert K: Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 2000;60:1306–1311. Vane JR, Bakhle YS, Botting RM: Cyclooxygenases 1 and 2. Annu Rev Pharmacol Toxicol 1998; 38:97–120. DuBois RN, Abramson SB, Crofford L, Gupta RA, Simon LS, van de Putte LB, Lipsky PE: Cyclooxygenase in biology and disease. FASEB J 1998;12:1063–1073. Ristimaki A, Garfinkel S, Wessendorf J, Maciag T, Hla T: Induction of cyclooxygenase-2 by interleukin-1␣. Evidence for post-transcriptional regulation. J Biol Chem 1994;269:11769–11775. Yan X, Wu XC, Sun M, Tsang BK, Gibb W: Nuclear factor B activation and regulation of cyclooxygenase type-2 expression in human amnion mesenchymal cells by interleukin-1. Biol Reprod 2002;66:1667–1671. Kulkarni S, Rader JS, Zhang F, Liapis H, Koki AT, Masferrer JL, Subbaramaiah K, Dannenberg AJ: Cyclooxygenase-2 is overexpressed in human cervical cancer. Clin Cancer Res 2001;7:429–434. Saha D, Datta PK, Sheng H, Morrow JD, Wada M, Moses HL, Beauchamp RD: Synergistic induction of cyclooxygenase-2 by transforming growth factor-1 and epidermal growth factor inhibits apoptosis in epithelial cells. Neoplasia 1999;1:508–517. Roman CD, Morrow J, Whitehead R, Beauchamp RD: Induction of cyclooxygenase-2 and invasiveness by transforming growth factor-1 in immortalized mouse colonocytes expressing oncogenic Ras. J Gastrointest Surg 2002;6:304–309.
COX-2 and Tumor Angiogenesis
187
35 36
37 38
39
40
41 42 43 44 45 46
47
48
49 50
51
52 53 54
55 56
Fong CY, Pang L, Holland E, Knox AJ: TGF-1 stimulates IL-8 release, COX-2 expression, and PGE2 release in human airway smooth muscle cells. Am J Physiol 2000;279:L201–L207. Mark KS, Trickler WJ, Miller DW: Tumor necrosis factor-␣ induces cyclooxygenase-2 expression and prostaglandin release in brain microvessel endothelial cells. J Pharmacol Exp Ther 2001;297: 1051–1058. Fosslien E: Molecular pathology of cyclooxygenase-2 in neoplasia. Ann Clin Lab Sci 2000; 30:3–21. Oshima M, Dinchuk JE, Kargman SL, Oshima H, Hancock B, Kwong E, Trzaskos JM, Evans JF, Taketo MM: Suppression of intestinal polyposis in Apc⌬716 knockout mice by inhibition of cyclooxygenase 2. Cell 1996;87:803–809. Seno H, Oshima M, Ishikawa TO, Oshima H, Takaku K, Chiba T, Narumiya S, Taketo MM: Cyclooxygenase 2- and prostaglandin E2 receptor EP2-dependent angiogenesis in Apc⌬716 mouse intestinal polyps. Cancer Res 2002;62:506–511. Dannenberg AJ, Altorki NK, Boyle JO, Lin DT, Subbaramaiah K: Inhibition of cyclooxygenase-2: An approach to preventing cancer of the upper aerodigestive tract. Ann NY Acad Sci 2001; 952:109–115. Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN: Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705–716. Kirschenbaum A, Liu X, Yao S, Levine AC: The role of cyclooxygenase-2 in prostate cancer. Urology 2001;58:127–131. Williams CS, Tsujii M, Reese J, Dey SK, DuBois RN: Host cyclooxygenase-2 modulates carcinoma growth. J Clin Invest 2000;105:1589–1594. Sawaoka H, Tsuji S, Tsujii M, Gunawan ES, Sasaki Y, Kawano S, Hori M: Cyclooxygenase inhibitors suppress angiogenesis and reduce tumor growth in vivo. Lab Invest 1999;79:1469–1477. Subbaramaiah K, Altorki N, Chung WJ, Mestre JR, Sampat A, Dannenberg AJ: Inhibition of cyclooxygenase-2 gene expression by p53. J Biol Chem 1999;274:10911–10915. Galy B, Creancier L, Prado-Lourenco L, Prats AC, Prats H: p53 directs conformational change and translation initiation blockade of human fibroblast growth factor 2 mRNA. Oncogene 2001; 20:4613–4620. Ueba T, Nosaka T, Takahashi JA, Shibata F, Florkiewicz RZ, Vogelstein B, Oda Y, Kikuchi H, Hatanaka M: Transcriptional regulation of basic fibroblast growth factor gene by p53 in human glioblastoma and hepatocellular carcinoma cells. Proc Natl Acad Sci USA 1994;91: 9009–9013. Yuan A, Yu CJ, Luh KT, Kuo SH, Lee YC, Yang PC: Aberrant p53 expression correlates with expression of vascular endothelial growth factor mRNA and interleukin-8 mRNA and neoangiogenesis in non-small-cell lung cancer. J Clin Oncol 2002;20:900–910. Dameron KM, Volpert OV, Tainsky MA, Bouck N: Control of angiogenesis in fibroblasts by p53 regulation of thrombospondin-1. Science 1994;265:1582–1584. Jones MK, Wang H, Peskar BM, Levin E, Itani RM, Sarfeh IJ, Tarnawski AS: Inhibition of angiogenesis by nonsteroidal anti-inflammatory drugs: Insight into mechanisms and implications for cancer growth and ulcer healing. Nat Med 1999;5:1418–1423. Daniel TO, Liu H, Morrow JD, Crews BC, Marnett LJ: Thromboxane A2 is a mediator of cyclooxygenase-2-dependent endothelial migration and angiogenesis. Cancer Res 1999;59: 4574–4577. Hanahan D, Weinberg RA: The hallmarks of cancer. Cell 2000;100:57–70. Cordon-Cardo C, Prives C: At the crossroads of inflammation and tumorigenesis. J Exp Med 1999;190:1367–1370. Coussens LM, Raymond WW, Bergers G, Laig-Webster M, Behrendtsen O, Werb Z, Caughey GH, Hanahan D: Inflammatory mast cells up-regulate angiogenesis during squamous epithelial carcinogenesis. Genes Dev 1999;13:1382–1397. Michalowski J: COX-2 inhibitors: Cancer trials test new uses for pain drug. J Natl Cancer Inst 2002;94:248–249. Jones MK, Szabo IL, Kawanaka H, Husain SS, Tarnawski AS: Von Hippel-Lindau tumor suppressor and HIF-1␣: New targets of NSAIDs inhibition of hypoxia-induced angiogenesis. FASEB J 2002;16:264–266.
Gately/Kerbel
188
57 58 59
60 61 62
63
64
65
66
67
68
69
70
71
72
73
74
75 76
Chiarugi V, Magnelli L, Gallo O: Cox-2, iNOS and p53 as play-makers of tumor angiogenesis (review). Int J Mol Med 1998;2:715–719. Ji YS, Xu Q, Schmedtje JF Jr: Hypoxia induces high-mobility-group protein I(Y) and transcription of the cyclooxygenase-2 gene in human vascular endothelium. Circ Res 1998;83:295–304. Schmedtje JF Jr, Ji YS, Liu WL, DuBois RN, Runge MS: Hypoxia induces cyclooxygenase-2 via the NF-B p65 transcription factor in human vascular endothelial cells. J Biol Chem 1997;272: 601–608. Maxwell PH, Ratcliffe PJ: Oxygen sensors and angiogenesis. Semin Cell Dev Biol 2002; 13:29–37. Shweiki D, Itin A, Soffer D, Keshet E: Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature 1992;359:843–845. Melillo G, Musso T, Sica A, Taylor LS, Cox GW, Varesio L: A hypoxia-responsive element mediates a novel pathway of activation of the inducible nitric oxide synthase promoter. J Exp Med 1995;182:1683–1693. Yan SF, Tritto I, Pinsky D, Liao H, Huang J, Fuller G, Brett J, May L, Stern D: Induction of interleukin 6 (IL-6) by hypoxia in vascular cells. Central role of the binding site for nuclear factor-IL-6. J Biol Chem 1995;270:11463–11471. Hirani N, Antonicelli F, Strieter RM, Wiesener MS, Ratcliffe PJ, Haslett C, Donnelly SC: The regulation of interleukin-8 by hypoxia in human macrophages – A potential role in the pathogenesis of the acute respiratory distress syndrome. Mol Med 2001;7:685–697. Willam C, Koehne P, Jurgensen JS, Grafe M, Wagner KD, Bachmann S, Frei U, Eckardt KU: TIE-2 receptor expression is stimulated by hypoxia and proinflammatory cytokines in human endothelial cells. Circ Res 2000;87:370–377. Fukumura D, Xavier R, Sugiura T, Chen Y, Park EC, Lu N, Selig M, Nielsen G, Taksir T, Jain RK, Seed B: Tumor induction of VEGF promoter activity in stromal cells. Cell 1998;94: 715–725. Majima M, Hayashi I, Muramatsu M, Katada J, Yamashina S, Katori M: Cyclo-oxygenase-2 enhances basic fibroblast growth factor-induced angiogenesis through induction of vascular endothelial growth factor in rat sponge implants. Br J Pharmacol 2000;130:641–649. Mukutmoni M, Hubbard NE, Erickson KL: Prostaglandin E2 modulation of vascular endothelial growth factor production in murine macrophages. Prostaglandins Leukot Essent Fatty Acids 2001; 65:123–131. Pai R, Szabo IL, Soreghan BA, Atay S, Kawanaka H, Tarnawski AS: PGE2 stimulates VEGF expression in endothelial cells via ERK2/JNK1 signaling pathways. Biochem Biophys Res Commun 2001;286:923–928. Tokuda H, Kozawa O, Miwa M, Uematsu T: p38 mitogen-activated protein (MAP) kinase but not p44/p42 MAP kinase is involved in prostaglandin E1-induced vascular endothelial growth factor synthesis in osteoblasts. J Endocrinol 2001;170:629–638. Inoue H, Takamori M, Shimoyama Y, Ishibashi H, Yamamoto S, Koshihara Y: Regulation by PGE2 of the production of interleukin-6, macrophage colony-stimulating factor, and vascular endothelial growth factor in human synovial fibroblasts. Br J Pharmacol 2002;136:287–295. Cianchi F, Cortesini C, Bechi P, Fantappie O, Messerini L, Vannacci A, Sardi I, Baroni G, Boddi V, Mazzanti R, Masini E: Up-regulation of cyclooxygenase 2 gene expression correlates with tumor angiogenesis in human colorectal cancer. Gastroenterology 2001;121:1339–1347. Sonoshita M, Takaku K, Sasaki N, Sugimoto Y, Ushikubi F, Narumiya S, Oshima M, Taketo MM: Acceleration of intestinal polyposis through prostaglandin receptor EP2 in Apc⌬716 knockout mice. Nat Med 2001;7:1048–1051. Murono S, Inoue H, Tanabe T, Joab I, Yoshizaki T, Furukawa M, Pagano JS: Induction of cyclooxygenase-2 by Epstein-Barr virus latent membrane protein 1 is involved in vascular endothelial growth factor production in nasopharyngeal carcinoma cells. Proc Natl Acad Sci USA 2001;98:6905–6910. Rak J, Filmus J, Kerbel RS: Reciprocal paracrine interactions between tumour cells and endothelial cells: The ‘angiogenesis progression’ hypothesis. Eur J Cancer 1996;32A:2438–2450. Tsujii M, DuBois RN: Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell 1995;83:493–501.
COX-2 and Tumor Angiogenesis
189
77
78 79 80
81 82
83
84 85 86 87 88
89 90
91 92
93 94
95
96
97
98
Nie D, Lamberti M, Zacharek A, Li L, Szekeres K, Tang K, Chen Y, Honn KV: Thromboxane A2 regulation of endothelial cell migration, angiogenesis and tumor metastasis. Biochem Biophys Res Commun 2000;267:245–251. Herschman HR, Xie W, Reddy S: Inflammation, reproduction, cancer and all that.... The regulation and role of the inducible prostaglandin synthase. Bioessays 1995;17:1031–1037. Zachary I: Signaling mechanisms mediating vascular protective actions of vascular endothelial growth factor. Am J Physiol Cell Physiol 2001;280:C1375–C1386. Krishnamurthy P, Bird IM, Sheppard C, Magness RR: Effects of angiogenic growth factors on endothelium-derived prostacyclin production by ovine uterine and placental arteries. Prostaglandins Other Lipid Mediat 1999;57:1–12. Bicknell R, Vallee BL: Angiogenin stimulates endothelial cell prostacyclin secretion by activation of phospholipase A2. Proc Natl Acad Sci USA 1989;86:1573–1577. Gurubhagavatula I, Amrani Y, Pratico D, Ruberg FL, Albelda SM, Panettieri RA Jr: Engagement of human PECAM-1 (CD31) on human endothelial cells increases intracellular calcium ion concentration and stimulates prostacyclin release. J Clin Invest 1998;101:212–222. Murohara T, Horowitz JR, Silver M, Tsurumi Y, Chen D, Sullivan A, Isner JM: Vascular endothelial growth factor/vascular permeability factor enhances vascular permeability via nitric oxide and prostacyclin. Circulation 1998;97:99–107. Ziche M, Jones J, Gullino PM: Role of prostaglandin E1 and copper in angiogenesis. J Natl Cancer Inst 1982;69:475–482. Gullino PM, Ziche M, Alessandri G: Gangliosides, copper ions and angiogenic capacity of adult tissues. Cancer Metastasis Rev 1990;9:239–251. Ziche M, Alessandri G, Gullino PM: Gangliosides promote the angiogenic response. Lab Invest 1989;61:629–634. Ziche M, Morbidelli L, Alessandri G, Gullino PM: Angiogenesis can be stimulated or repressed in vivo by a change in GM3:GD3 ganglioside ratio. Lab Invest 1992;67:711–715. Ziche M, Morbidelli L, Parenti A, Ledda F: Nitric oxide modulates angiogenesis elicited by prostaglandin E1 in rabbit cornea. Adv Prostaglandin Thromboxane Leukot Res 1995;23: 495–497. Form DM, Auerbach R: PGE2 and angiogenesis. Proc Soc Exp Biol Med 1983;172:214–218. Ben-Av P, Crofford LJ, Wilder RL, Hla T: Induction of vascular endothelial growth factor expression in synovial fibroblasts by prostaglandin E and interleukin-1: A potential mechanism for inflammatory angiogenesis. FEBS Lett 1995;372:83–87. Harada S, Rodan SB, Rodan GA: Expression and regulation of vascular endothelial growth factor in osteoblasts. Clin Orthop 1995;76–80. Harada S, Nagy JA, Sullivan KA, Thomas KA, Endo N, Rodan GA, Rodan SB: Induction of vascular endothelial growth factor expression by prostaglandin E2 and E1 in osteoblasts. J Clin Invest 1994;93:2490–2496. Diaz-Flores L, Gutierrez R, Valladares F, Varela H, Perez M: Intense vascular sprouting from rat femoral vein induced by prostaglandins E1 and E2. Anat Rec 1994;238:68–76. Arico S, Pattingre S, Bauvy C, Gane P, Barbat A, Codogno P, Ogier-Denis E: Celecoxib induces apoptosis by inhibiting 3-phosphoinositide-dependent protein kinase-1 activity in the human colon cancer cell line HT-29. J Biol Chem 2002;277:27613–27621. Elder DJ, Halton DE, Playle LC, Paraskeva C: The MEK/ERK pathway mediates COX-2-selective NSAID-induced apoptosis and induced COX-2 protein expression in colorectal carcinoma cells. Int J Cancer 2002;99:323–327. Liu CH, Chang SH, Narko K, Trifan OC, Wu MT, Smith E, Haudenschild C, Lane TF, Hla T: Overexpression of cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem 2001;276:18563–18569. Sheng H, Shao J, Morrow JD, Beauchamp RD, DuBois RN: Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells. Cancer Res 1998;58: 362–366. Parfenova H, Parfenov VN, Shlopov BV, Levine V, Falkos S, Pourcyrous M, Leffler CW: Dynamics of nuclear localization sites for COX-2 in vascular endothelial cells. Am J Physiol 2001;281:C166–C178.
Gately/Kerbel
190
99 Nor JE, Christensen J, Liu J, Peters M, Mooney DJ, Strieter RM, Polverini PJ: Up-regulation of Bcl-2 in microvascular endothelial cells enhances intratumoral angiogenesis and accelerates tumor growth. Cancer Res 2001;61:2183–2188. 100 Nor JE, Christensen J, Mooney DJ, Polverini PJ: Vascular endothelial growth factor-mediated angiogenesis is associated with enhanced endothelial cell survival and induction of Bcl-2 expression. Am J Pathol 1999;154:375–384. 101 Lin MT, Lee RC, Yang PC, Ho FM, Kuo ML: Cyclooxygenase-2 inducing Mcl-1-dependent survival mechanism in human lung adenocarcinoma CL1.0 cells. Involvement of phosphatidylinositol 3-kinase/Akt pathway. J Biol Chem 2001;276:48997–49002. 102 Sheng H, Shao J, DuBois RN: K-Ras-mediated increase in cyclooxygenase 2 mRNA stability involves activation of the protein kinase B1. Cancer Res 2001;61:2670–2675. 103 Tran J, Master Z, Yu JL, Rak J, Dumont DJ, Kerbel RS: A role for survivin in chemoresistance of endothelial cells mediated by VEGF. Proc Natl Acad Sci USA 2002;99:4349–4354. 104 Dimmeler S, Zeiher AM: Akt takes center stage in angiogenesis signaling. Circ Res 2000;86:4–5. 105 Patti R, Gumired K, Reddanna P, Sutton LN, Phillips PC, Reddy CD: Overexpression of cyclooxygenase-2 (COX-2) in human primitive neuroectodermal tumors: Effect of celecoxib and rofecoxib. Cancer Lett 2002;180:13–21. 106 Hsu AL, Ching TT, Wang DS, Song X, Rangnekar VM, Chen CS: The cyclooxygenase-2 inhibitor celecoxib induces apoptosis by blocking Akt activation in human prostate cancer cells independently of Bcl-2. J Biol Chem 2000;275:11397–11403. 107 Leese PT, Hubbard RC, Karim A, Isakson PC, Yu SS, Geis GS: Effects of celecoxib, a novel cyclooxygenase-2 inhibitor, on platelet function in healthy adults: A randomized, controlled trial. J Clin Pharmacol 2000;40:124–132. 108 Clemett D, Goa KL: Celecoxib: a review of its use in osteoarthritis, rheumatoid arthritis and acute pain. Drugs 2000;59:957–980. 109 Steinbach G, Lynch PM, Phillips RK, Wallace MH, Hawk E, Gordon GB, Wakabayashi N, Saunders B, Shen Y, Fujimura T, Su LK, Levin B: The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–1952. 110 Schnitzer TJ, Truitt K, Fleischmann R, Dalgin P, Block J, Zeng Q, Bolognese J, Seidenberg B, Ehrich EW: The safety profile, tolerability, and effective dose range of rofecoxib in the treatment of rheumatoid arthritis. Phase II Rofecoxib Rheumatoid Arthritis Study Group. Clin Ther 1999; 21:1688–1702. 111 Teicher BA, Sotomayor EA, Huang ZD: Antiangiogenic agents potentiate cytotoxic cancer therapies against primary and metastatic disease. Cancer Res 1992;52:6702–6704. 112 Gorski DH, Mauceri HJ, Salloum RM, Gately S, Hellman S, Beckett MA, Sukhatme VP, Soff GA, Kufe DW, Weichselbaum RR: Potentiation of the antitumor effect of ionizing radiation by brief concomitant exposures to angiostatin. Cancer Res 1998;58:5686–5689. 113 Milas L: Cyclooxygenase-2 enzyme inhibitors as potential enhancers of tumor radioresponse. Semin Radiat Oncol 2001;11:290–299. 114 Petersen C, Petersen S, Milas L, Lang FF, Tofilon PJ: Enhancement of intrinsic tumor cell radiosensitivity induced by a selective cyclooxygenase-2 inhibitor. Clin Cancer Res 2000;6:2513–2520. 115 Kishi K, Petersen S, Petersen C, Hunter N, Mason K, Masferrer JL, Tofilon PJ, Milas L: Preferential enhancement of tumor radioresponse by a cyclooxygenase-2 inhibitor. Cancer Res 2000;60:1326–1331. 116 Klement G, Baruchel S, Rak J, Man S, Clark K, Hicklin DJ, Bohlen P, Kerbel RS: Continuous lowdose therapy with vinblastine and VEGF receptor-2 antibody induces sustained tumor regression without overt toxicity. J Clin Invest 2000;105:R15–R24. 117 Browder T, Butterfield CE, Kraling BM, Shi B, Marshall B, O’Reilly MS, Folkman J: Antiangiogenic scheduling of chemotherapy improves efficacy against experimental drugresistant cancer. Cancer Res 2000;60:1878–1886. 118 Gately S, Kerbel R: Antiangiogenic scheduling of lower dose cancer chemotherapy. Cancer J 2001;7:427–436. 119 Man S, Bocci G, Francia G, Green SK, Jothy S, Hanahan D, Bohlen P, Hicklin DJ, Bergers G, Kerbel RS: Antitumor effects in mice of low-dose (metronomic) cyclophosphamide administered continuously through the drinking water. Cancer Res 2002;62:2731–2735.
COX-2 and Tumor Angiogenesis
191
120 Garber K: Could less be more? Low-dose chemotherapy goes on trial. J Natl Cancer Inst 2002; 94:82–84. 121 Gasparini G: Metronomic scheduling: The future of chemotherapy? Lancet Oncol 2001;2: 733–740. 122 Kerbel RS, Klement G, Pritchard KI, Kamen B: Continuous low-dose anti-angiogenic/ metronomic chemotherapy: From the research laboratory into the oncology clinic. Ann Oncol 2002;13:12–15. 123 Marx J: Cancer research. Anti-inflammatories inhibit cancer growth – but how? Science 2001; 291:581–582. 124 Hwang DH, Fung V, Dannenberg AJ: National Cancer Institute workshop on chemopreventive properties of nonsteroidal anti-inflammatory drugs: role of COX-dependent and -independent mechanisms. Neoplasia 2002;4:91–97. 125 Hawk ET, Viner JL, Dannenberg A, DuBois RN: COX-2 in cancer – a player that’s defining the rules. J Natl Cancer Inst 2002;94:545–546. 126 Williams CS, Watson AJ, Sheng H, Helou R, Shao J, DuBois RN: Celecoxib prevents tumor growth in vivo without toxicity to normal gut: lack of correlation between in vitro and in vivo models. Cancer Res 2000;60:6045–6051. 127 Song X, Lin HP, Johnson AJ, Tseng PH, Yang YT, Kulp SK, Chen CS: Cyclooxygenase-2, player or spectator in cyclooxygenase-2 inhibitor-induced apoptosis in prostate cancer cells. J Natl Cancer Inst 2002;94:585–591. 128 Hanahan D, Folkman J: Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell 1996;86:353–364. 129 Folkman J, Watson K, Ingber D, Hanahan D: Induction of angiogenesis during the transition from hyperplasia to neoplasia. Nature 1989;339:58–61. 130 Parangi S, O’Reilly M, Christofori G, Holmgren L, Grosfeld J, Folkman J, Hanahan D: Antiangiogenic therapy of transgenic mice impairs de novo tumor growth. Proc Natl Acad Sci USA 1996;93:2002–2007. 131 Bergers G, Javaherian K, Lo KM, Folkman J, Hanahan D: Effects of angiogenesis inhibitors on multistage carcinogenesis in mice. Science 1999;284:808–812.
Stephen Gately NeoPharm Inc., Department of Translational Medicine 150 Field Drive, Suite 195, Lake Forest, IL 60045 (USA) Tel. +1 847 2958678, Fax +1 847 2958854, E-Mail
[email protected]
Gately/Kerbel
192
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 193–209
Potential for Combined Modality Therapy of Cyclooxygenase Inhibitors and Radiation Debabrata Saha, Hak Choy Department of Radiation Oncology, Vanderbilt University Medical Center, Nashville, Tenn., USA
Radiation Therapy and Cancer Treatment
The goal of radiation therapy is to eradicate cancer cells with as little risk as possible to normal cells. It has proven to be an effective agent in the ongoing battle against cancer. It is presumed that the essential target for radiation is the cellular DNA where it acts through the formation of free radicals to directly or indirectly cause double-stranded breaks. It is these double-stranded breaks in the DNA that are responsible for the death of malignant cells following therapeutic radiation. While oncologists and researchers have often tried to cure cancers with radiation alone or with various chemotherapeutic strategies, in general these have been met with limited success for many reasons. The strategy of integrating different treatment modalities into a more comprehensive approach for both local control and the treatment of micrometastatic disease often referred to as combined modality therapy has been more successful. In the majority of solid tumors the cure rates, especially those seen with locally advanced tumors, remain poor despite the concurrent integration of chemotherapy and radiation treatment. It is the hope that new-targeted therapy might improve existing treatment paradigms and allow for more patients to be cured of their malignancies. This chapter will focus on radiation enhancing effects of cyclooxygenase-2 (COX-2) inhibitors as a combined modality therapy with new-targeted therapy. It will attempt to set the background with an examination of the rationale for concurrent targeted therapy with radiation. It will discuss the molecular and biological changes with radiation and the role of COX-2 in cancer biology.
Beyond this it will illustrate some of preclinical evidence for the radiation enhancing effects of COX-2 inhibitors. Finally it will focus on the potential for the clinical use that is based on an increased understanding of the molecular response to COX-2 inhibitors as well as the response of COX-2 inhibitors to therapy with concurrent radiation and chemotherapy. The integration of new agents that are aimed at more specific cellular targets than either radiation or traditional cytotoxic chemotherapy alone may significantly influence the success of combined modality therapy.
Radiation-Induced Molecular and Biological Responses
Ionizing radiation (IR) initiates multiple cellular and biological effects by either direct interaction with DNA or through the formation of free radical species leading to DNA damage [1–5]. These effects include induction of cell cycle-specific growth arrest, repair of the damaged DNA, modulation of radical-scavenging proteins, induction of gene mutations, transcriptional and translational changes, malignant transformation and cell death. Growing evidence suggests that irradiation of cells stimulates a series of biochemical and molecular signals. Various components of the IR-inducible signal transduction cascade function as either survival factors participating in cell growth/proliferation and cellular protection, or as cell death-related factors. A balance between the anti-apoptotic and pro-apoptotic pathways seems to determine the fate of the cells. Furthermore, tumor cells may respond to cytotoxic agents by launching a compensatory defense system, and resistance to therapy may be acquired at several points along the apoptotic pathway.
COX-2 in Cancer
COX-2 is the enzyme that converts arachidonic acid to prostaglandins (PGs) and other eicosanoids. COX-2 is an inducible enzyme that is upregulated by a variety of factors, which include cytokines, growth factors and tumor promoters [6]. PGs are known to possess diverse biologic capabilities that include vasoconstriction, vasodilation, stimulation or inhibition of platelet aggregation and immunomodulation [7, 8]. This enzyme is overexpressed in a variety of different tumors, including colon, pancreatic, prostate, lung, and head and neck cancers. COX-2 is expressed within human tumor neovasculature, suggesting that COX-2-derived PGs may contribute to tumor growth by inducing formation of
Saha/Choy
194
new blood vessels. Up to 90% of non-small cell lung cancers (NSCLC) have been shown to express COX-2 at a moderate to strong level [9–11]. Although Hida et al. [12] have reported only a 14% incidence of COX-2 overexpression in squamous cell carcinoma, other investigators report much higher expression levels in squamous cell carcinoma. Soslow et al. [10] reported COX-2 expression in 67% of squamous cell carcinomas. Wolff et al. [9] reported COX-2 expression in 11 of 11 squamous cell carcinomas evaluated. The expression level in NSCLC has been shown to be significantly higher than in normal lung tissue for both adenocarcinoma and squamous cell carcinoma [9]. In stage I NSCLC, increased expression of COX-2 has been shown to correlate with shortened survival [13]. Adenocarcinoma cells metastatic to mediastinal lymph nodes express COX-2 in a higher proportion and level than their corresponding primary tumors. These data suggest that COX-2 overexpression may enhance metastatic potential [11, 14]. In fact, it was recently reported that upregulation of COX-2 is associated with increased tumor cell invasiveness and migration [15]. This phenotypic change was accompanied by increased EtNA levels of the membrane-type metalloproteinase (MT-MMP), activation of matrix metalloproteinase-2 (MMP-2) and increased prostaglandin E2(PGE2) production. This increased PGE2 production subsequently initiates a cascade of events that favor tumor growth and dissemination. These events include increased production of interleukins IL-6 and IL-10, increased bcl-2 protein, and promotion of various proangiogenic factors including VEGF, bFGF, PDGF and TGF-. Inhibition of COX-2 (with and without classical cytotoxic drugs) favorably modulates each of these permissive events and results in dose-dependent growth retardation of lung and colon tumor xenografts [14, 16–25]. Therefore, upregulation of COX-2 represents a potentially important therapeutic target in lung cancer. Celecoxib, a COX-2 inhibitor, is also a potent inhibitor of angiogenesis and has been shown to inhibit neoangiogenic vasculature proliferation by 40–60% in these tumors [23]. Hida et al. [12] report that a selective COX-2 inhibitor, nimesulide, can inhibit proliferation of NSCLC cell lines in vitro in a dose-dependent manner in clinically-achievable low concentrations. These and other data suggest that COX-2-dependent angiogenesis plays a major role in development of cancer. The ability of celecoxib to block neoangiogenesis and tumor proliferation, regardless of the expression of the enzyme in the cancer cells, suggests the potential utility of this anti-inflammatory drug in the treatment of human cancer.
The Role of COX-2 and PGs in Early Tumor Development
Cyclooxygenases may contribute to tumor initiation by catalyzing the oxidation of arachidonic acid to ultimately yield mutagenic metabolites such as
Potential for Combined Modality Therapy of COX Inhibitors and Radiation
195
malondialdehyde. The cyclooxygenases and their prostanoid products are likely to contribute importantly to carcinogenesis in several other ways. Cells overexpressing COX-2 escape apoptosis, acquire invasive phenotypes and promote tumor angiogenesis [26]. PGs may impair host immunosurveillance mechanisms to favor tumor progression [27]. Tsujii and DuBois [28] found that COX-2 overexpression may contribute to the tumorigenic potential of dysplastic intestinal epithelial cells by enhancing adhesion to extracellular matrix and by inhibition of apoptosis. COX-2 overexpression may protect cells from apoptosis through the induction of Bcl-2 by PGE2 [28, 29]. In addition, Tsujii et al. [14] recently reported that forced COX-2 expression in Caco-2 cells (that do not ordinarily express detectable COX) results in the induction of vascular endothelial tubular morphogenesis when co-cultured with human vascular endothelial cells. This angiogenic induction is abrogated by treatment of the Caco-2 cells with NSAIDS. This is consistent with a previous study which has shown that the anti-tumor effect of COX-2 inhibitor diclofenac was due to an anti-angiogenic effect [24]. Thus, induction of angiogenesis may be another important tumor-promoting action of COX-2. COX-2 is also a downstream target of known oncogenic pathways including epidermal growth factor receptor (EGFR). One of the members of this family is HER-2/neu and is overexpressed in 20–30% of human breast cancers. Danenberg et al. [30] reported that COX-2 was overexpressed in 14 of 15 HER-2/neu-positive breast cancers, but in only 4 of 14 HER-2/neu-negative tumors. PG synthesis was also increased dramatically in human mammary epithelial cells overexpressing HER-2/neu. It is reported that COX-2 overexpression in these cells was a result of increased ras signaling through AP1 transcription factor. COX-2 expression is suppressed by p53 [30]. The mutant forms of this tumor suppressor were unable to reduce COX-2 expression in a murine embryo fibroblast system. It is also observed that COX-2 overexpression has been associated with missense p53 mutations in gastric cancer [31].
COX-2 Expression and Outcome of Patients Treated with Radiation
Khuri et al. [32] reported the COX-2 expression in specimens from 160 patients with stage I NSCLC. Of these, 3 specimens had strong COX-2 expression, 69 had intermediate expression of COX-2, 24 had weak expression, and 64 had no detectable COX-2. The strength of COX-2 expression was associated with a worse overall survival rate (p ⫽ 0.001) and a worse disease-free survival rate (p ⫽ 0.022). The median survival times for the strong, intermediate or weak, and null COX-2 expressors were 1.04, 5.50 and 8.54 years, respectively.
Saha/Choy
196
They also reported that all 3 specimens with strong COX-2 expression came from patients who died within 18 months [32]. Recently, Achwia et al. [13] demonstrated that elevated expression of COX-2 is frequently seen in lung adenocarcinomas, and is possibly associated with invasion and metastasis. The prognostic significance of elevated COX-2 expression was evaluated in a cohort of 130 adenocarcinoma patients who had consecutively undergone potentially curative resections. Immunohistological examination showed the presence of tumor cells with markedly increased COX-2 immunoreactivity in 93 of 130 (72%) cases. However, no relationship was found between the increase in COX-2 expression and clinical outcomes when the entire cohort was considered (p ⫽ 0.099). A significant relationship between elevated COX-2 expression and shortened patient survival was observed only in a cohort of patients with stage I disease (p ⫽ 0.034). These findings suggest that an increase in COX-2 expression may be clinically significant for the prognosis of patients undergoing surgical resection of early-stage adenocarcinomas and, thus, warrant additional studies involving a larger cohort. The expression of COX-2 has clinical relevance in the outcome of patients receiving radiation therapy. Gaffney et al. [33] performed a clinicopathological study on 24 patients with carcinoma of the cervix that were treated with radiation to determine if COX-2 expression correlated with overall survival (fig. 1a, b). In this series, the only clinical or pathologic factor that correlated with survival was COX-2 distribution. Patients with less than 10% of tumor cells staining for COX-2 had an improved overall survival (p ⫽ 0.021) and disease-free survival (p ⫽ 0.015) compared to patients with 10% or more of the tumor cells expressing COX-2. In another study, Steinauer et al. [34] demonstrated the expression of COX-2 is upregulated following irradiation and PGE2 levels are subsequently increased. They reported that radiation doses as small as 5 Gy increased the expression of COX-2 in human PC-3 cells (prostate) 6 h post-irradiation [34]. The PGE2 level of irradiated cells was higher than in controls (1,512 ⫾ 157.5 vs. 973.7 ⫾ 54.2 pg PGE2/ml; p ⬍ 0.005, n ⫽ 4) while cells irradiated in the presence of NS-398 had reduced PGE2 levels (218.8 ⫾ 80.1 pg PGE2/ml; p ⬍ 0.005; n ⫽ 4) (fig. 2A, B).
Recent Developments in Combined Treatment of COX-2 Inhibitors and Radiotherapy
Mouse Tumor Models The first study to demonstrate that specific COX-2 inhibitors can improve the radiation response was by Milas et al. [35]. The NFSA sarcoma,
Potential for Combined Modality Therapy of COX Inhibitors and Radiation
197
1.0
1.0
(75%, n ⫽ 12)
0.8 Fraction surviving
Fraction surviving
0.8
(92%, n⫽12)
0.6 (35%, n⫽ 12)
0.4
p ⫽0.0215
0.2
0.6
p⫽0.0152
0.4 (28%, n⫽12) 0.2
0.0
0.0 50
0
a
100
150
200
250
50
0
100
150
200
250
b
Months
Months
750,000
c
1,000,000 750,000
500,000 250,000
1,500
250,000
0
0 Control 5Gy 10Gy 15Gy
15Gy 15Gy ⫹NS-398
b
d 72 kD
72 kD Control 5 Gy 10Gy 15Gy
1,250 1,000 750 500 250
15Gy 15Gy ⫹NS-398
A
Control 15Gy 15Gy⫹NS-398
1,705
500,000 PGE2 (pg/ml)
a
COX-2 protein (Au)
Fig. 1. a Overall survival as a function of COX-2 distribution scores. The solid line and dashed line refer to patients whose tumors displayed ⬍10% and 10% of the cells expressing COX-2, respectively. The number in parentheses indicates 5-year survival values. b Diseasefree survival as a function of COX-2 distribution scores. The solid line and dashed line refer to patients whose tumors displayed ⬍10% and 10% of the cells expressing COX-2, respectively. The number in parentheses indicates 5-year survival values [adapted from 33].
0 Control
15Gy 15Gy⫹NS-398
Fig. 2. A a COX-2 protein expression in PC-3 cells 6 h after a single dose of radiation with 0, 5, 10 or 15 Gy. b Western blot analysis (corresponding to a) of COX-2 protein (molecular weight 72 kD) 6 h after irradiation with 0, 5, 10 or 15 Gy. c COX-2 expression in 6 h after radiation with and without 100 mol NS-398 measured in arbitrary units (AU). The level of COX-2 protein was increased 24.2% after treatment with NS-398 and irradiation compared to irradiation alone. d Western blot corresponding to c. B PGE2 levels 7 h after irradiation (0 or 15 Gy), with or without 100 mol COX-2 inhibitor NS-398, demonstrating an increase in the level of PGE2 after irradiation compared with unirradiated control cells (1,512 ⫾ 157.5 vs. 973.7 ⫾ 54.2 g PGE2/ml; p ⬍ 0.005) [adapted from 34].
Saha/Choy
198
B
Tumor size (mm)
22 20 18
Tumor size (mm)
16
12 10 8 6 0
5
10
15
Days after SC-236
14 12 10 8 6 4 2 0
5
10
15
20
25
30
35
40
45
Days after irradiation
Fig. 3. Effect of SC-236 alone (insert) or in combination with local tumor irradiation on tumor growth. Tumors were generated in the muscles of the right hind legs of mice by injection of 3 ⫻ 105 viable NFSA tumor cells. Treatment with SC-236 (6 mg/kg b.w.) or vehicle (0.05% Tween-20 and 0.95% polyethylene glycol) given in the drinking water was continued for 10 consecutive days. Growth of tumors treated with vehicle (䊊), SC-236 (䉭), radiation 30 Gy (䊉) and SC-236 plus 30 Gy (䉱). Vertical bars represent 95% confidence intervals [adapted from 35].
a non-immunogenic and PG-producing tumor that spontaneously formed in C3Hf/Kam mice was used in the study (fig. 3). The NFSA tumor model was previously used in radiation experiments with NSAIDs. It was demonstrated that the radiation response of this tumor was increased if mice were pretreated with indomethacin before radiation therapy [36]. However, long-term in vivo experiments with indomethacin resulted in significant gastrointestinal toxicity. A selective inhibitor of the COX-2 enzyme, SC-236 (a laboratory equivalent of celecoxib) [37] was given in the drinking water (6 mg/kg b.w.) of mice for 10 consecutive days when tumors reached 6 mm in diameter. A single dose of radiation was given when tumors reached 8 mm in diameter. Two endpoints were investigated, tumor growth delay (TGD) and tumor control dose 50 (TCD50, dose at which 50% of tumors are cured). Treatment with SC-236 significantly inhibited tumor growth. Radiation in combination with SC-236 was even more effective in increasing TGD and TCD50. TGD was increased 2.5 days by drug alone, 9 days by radiation alone, and 38.9 days by drug plus radiation.
Potential for Combined Modality Therapy of COX Inhibitors and Radiation
199
The TCD50 was reduced from 69.2 to 39.2 Gy in the drug plus radiation group. The mechanisms responsible for the increase in radiation response were not fully elucidated. However, it was demonstrated that the COX-2 inhibitor had some effects on angiogenesis. An intradermal injection of tumor cells was used to visualize and count newly forming vessels. It was clearly demonstrated that neovascularization preceded tumor growth. SC-236 effectively reduced neovascularization of the tumor and inhibited tumor growth. A follow-up study by Kishi et al. [38] investigated the specificity of SC-236 in tumor versus normal tissue in relation to the radiation response. The mouse FSA tumor was used in this study. FSA is an immunogenic sarcoma induced by methylcholanthrene that produces high levels of PGE2 [36]. As in the previous study by this group [35], mice bearing FSA were given SC-236 in their drinking water (6 mg/kg b.w.) for 10 consecutive days when tumors reached 6 mm in diameter. A single dose of radiation was given when tumors reached 8 mm in diameter. The effect of drug on the TGD and TCD50 was similar to the previous study. SC-236 alone caused TGD and radiation plus SC-236 was more than additive. A jejunal crypt cell assay [39] was used to evaluate acute normal tissue response to radiation, and a leg contracture assay was used to evaluate the late responses [40]. To perform the crypt assay, mice were given drug or vehicle and exposed to whole-body irradiation (WBI). Jejunum were removed from mice and prepared for histological examination and the regenerating crypts were counted (fig. 4). WBI reduced crypt survival, while treatment with SC-236 resulted in a slight but significant decrease in crypt survival. The decrease was small and may not be significant. WBI increased leg contracture, however, there was no further increase in contracture by the addition of the drug. The mechanism of increased radiation sensitivity was also investigated in this study. It was determined that apoptosis played no role in the increased radiation response. Treatment of mice with SC-236 did not affect COX-2 expression, however, PGE2 was significantly decreased. PGE2 can stimulate tumor growth, induce angiogenesis and protect against radiation damage. As in the previous study using NFSA tumors, inhibition of angiogenesis preceded reduced tumor growth and is the probable mechanism for enhanced TGD. However, it is important to note that PGE2 is also a vasoactive compound that can affect tumor perfusion, an important factor in radiation response. This report demonstrated that an inhibitor of COX-2 enhanced the tumoricidal effects of radiotherapy without an increase in normal tissue toxicity. The concept of combined modality treatment controlling tumor growth, especially in the animal models, brought great excitement to the field of radiation oncology and led to the design of numerous clinical trials.
Saha/Choy
200
100 50 14 20 12
10 5
10
2
8
1 0
a
5
10
15
20
25
30
8
9
Days after irradiation
10
11
12
13
Radiation (Gy)
Fig. 4. a Effect of combination of SC-236 treatment and local tumor irradiation on tumor growth of FSA tumors. Mice were treated with vehicle (䊉), SC-236 (䊊), 25-Gy single-dose irradiation (䉲), or SC-236 plus 25-Gy single-dose irradiation (䉮). The treatment with SC-236, 6 mg/kg in drinking water, was started when tumors were 6 mm in diameter and was continued for 10 days. b Radiation single-dose survival curves for jejunum crypt cells in mice treated with SC-236 (䊊) or with vehicle (䊉). Intercepts of survival curves were significantly different at p ⬍ 0.0007 [adapted from 38].
Human Cancer Cells and Tumor Models The findings by the MD Anderson group using rodent models prompted Petersen et al. [41] to investigate the effect of COX-2 inhibitors in human cancer cell lines and human tumor models. U251 human glioblastoma cells were treated with SC-236. Treatment of cells with SC-236 had no effect on the expression of COX-2, however, cell survival decreased with exposure time. Cell death resulted from apoptosis and detachment. The U251 human glioblastoma was also grown as a tumor xenograft and treated with radiation. SC-236 was given in the drinking water (6 mg/kg b.w.) of mice for 10 consecutive days when tumors reached 5 mm in diameter. Radiation was initiated when the tumors reached 7 mm in diameter. Treatment alone reduced tumor growth, as did radiation. When radiation and drug were combined, a more than additive TGD was observed. The effect of COX-2 inhibitors on the radiation response was also investigated in human H460 and HCT-116 cells and tumors [42]. Human H460 lung cancer cells which constitutively express COX-2, and HCT-116 human colon cancer cells which do not express COX-2, were treated with the selective COX-2
Potential for Combined Modality Therapy of COX Inhibitors and Radiation
201
14
15
b
inhibitor NS-398. Treatment with NS-398 enhanced the effect of radiation in H460 cells by a factor of 1.8, and protected HCT-116 cells by a factor of 0.83. Apoptosis was evaluated in these cell lines and was a significant factor in the response of H460 cells. Irradiation and drug increased apoptosis in H460 cells in a more than additive manner, while the apoptosis was additive in HCT-116 cells [42]. We [42] also demonstrated that NS-398 in combination with radiation significantly delayed the tumor growth in the human lung cancer xenograft model. In this study, H460 lung carcinoma cells were implanted into the nude mice and treated with either NS-398 (36 mg/kg/day) or a fractionated dose of radiation (2 Gy/day) alone or in combination with NS-398 and radiation for 5 consecutive days. Treatment with NS-398 delayed tumor growth by 0.086 days. Fractioned radiation treatment produced a growth delay (GD) of 3.8 days. In contrast, the combined treatment of NS-398 and radiation produced a GD of 9.34 days after normalization for drug alone. This resulted in an enhancing factor (EF) of 2.5, indicating a more than additive effect for the combination. In another xenograft experiment with HCT-116 human colon cancer cells lacking COX-2 expression, similar treatment with the combination of NS-398 and radiation resulted in an EF of 1.04. These results indicate that the COX-2-selective inhibitor, NS-398, efficiently enhanced the effect of radiation on COX-2 expressing cells but not in cells lacking COX-2 (fig. 5a, b). COX-2 Inhibitors and Anti-Angiogenesis Agents: Their Impact on Radiation Sensitization As discussed earlier, COX-2 upregulation plays a key role in the growth and metastatic potential of solid tumors. The increased production of proangiogenic factors such as VEGF, along with the increased production of key matrix metalloproteinases that accompany COX-2 upregulation, indicate that COX-2 may represent a useful therapeutic target in many solid tumors. A selective COX-2 inhibitor, celecoxib, was found to inhibit [43] tumor growth and another specific inhibitor of COX-2, SC-58125 yielded more than a 90% reduction in VEGF production in a rodent carcinoma model [44]. Most recent studies suggest that radiation-induced VEGF expression may play a role in delayed increase in lung cancer blood flow [45, 46]. Therefore, blocking the effects of VEGF with an anti-angiogenic agent may enhance the therapeutic effects of IR. Studies done by Gorski et al. [45] reveal that anti-VEGF antibodies prolong the GD achieved by radiation, however, endothelial cell proliferation and survival after in vitro irradiation are enhanced by the supplementation of VEGF.
Saha/Choy
202
2.5
2.5
2.0 Tumor volume (cm3)
Tumor volume (cm3)
2.0
1.5
1.0
1.0
0.5
0.5
0.0
0.0 0
a
1.5
4
8
12
16
20
24
0
4
8
12
Time (days)
16
20
24
28
Fig. 5. Combined effect of fractionated NS-398 and radiotherapy on the tumor growth delay of H460 (a) and HCT-116 (b) human tumor xenografts in nude mice. 䊊 ⫽ Vehicle treatment alone; 䊐 ⫽ NS-398 treatment alone; 䊉 ⫽ radiation plus vehicle treatment; 䊏 ⫽ radiation plus NS-398 treatment. EFs were 2.5 on H460 and 1.04 on HCT-116 tumors [adapted from 42].
Evidence That Inhibitors of COX-2 Have Specific Anti-Angiogenic Activity Dicker et al. [47] recently demonstrated the effects of rofecoxib (Vioxx), another selective inhibitor of COX-2, on angiogenesis. In this report, they established the clinically relevant doses for use in the treatment of arthritis in combination with IR. The endpoints for the assays of angiogenesis include: the effect of rofecoxib on the proliferation, attachment and differentiation of cultured human umbilical vein endothelial cells (HUVEC) in vitro, and capillary sprouting of rat aortic ring explants embedded in collagen (ex vivo). They reported a dose-dependent inhibition of endothelial cell proliferation. Inhibition of proliferation was statistically significant with as little as 0.25 nM of rofecoxib (p ⬍ 0.001), and maximal at 1 M (73% inhibition). HUVEC differentiation was inhibited by the treatment of rofecoxib whereas it was potentiated by radiation. Cells treated with 0.25 M of rofecoxib, in absence of radiation, resulted in a 60% inhibition of tube formation compared to the control. Irradiated cells, pretreated with 0.25 M of rofecoxib, resulted in a further inhibition of tube formation as compared to control and cells treated with rofecoxib alone (no radiation). This difference in tube formation was approximately 40–73% and was determined to be statistically significant (p ⬍ 0.001) and was also associated with the significant decrease in the area of tube formation (fig. 6).
Potential for Combined Modality Therapy of COX Inhibitors and Radiation
32
36
b
Time (days)
203
0.25 M 0 Gy
0.25M 2 Gy
Tube area/mm2
0.12 0 Gy 2 Gy
0.10
*p⫽0.001 0.08 0.06 0.04
*
0.02 0.0
a
0.2
0.3
Concentration (M)
b
0.1
Fig. 6. a, b Combination of rofecoxib and radiation on HUVEC differentiation (tube formation). Endothelial cells treated with rofecoxib resulted in a dose-dependent inhibition of differentiation. Treatment of endothelial cells with rofecoxib resulted in a dose-dependent inhibition of tube formation, with maximal inhibition at 1 mol/l. Untreated endothelial cells irradiated at 2 Gy resulted in a 30% inhibition of tube formation, whereas pretreated endothelial cells resulted in further inhibition of tube formation, suggesting that radiation potentiates the inhibitory effect of rofecoxib on HUVEC differentiation (b). Representative photomicrographs are depicted in a [adapted from 47].
They also evaluated the angiogenesis in an ex vivo model using rat aortic sprouts. The addition of 0.25 M of rofecoxib resulted in a 60% inhibition of sprout differentiation and further inhibition was observed at higher drug doses. Capillary sprouts pretreated with rofecoxib and irradiated at 2 Gy showed an even further increase in sprout inhibition (80–90% inhibition) as compared to the untreated and unirradiated sprouts. Irradiated (2 Gy) sprouts, pretreated with 0.25–0.75 M of rofecoxib, resulted in a 50–80% inhibition of sprout differentiation as compared to those with no radiation. These results suggest that the combination of radiation and rofecoxib had a profound impact on processes of angiogenesis and on aortic sprout differentiation.
COX-2 Inhibitors and Combined Chemoradiotherapy
Despite unceasing efforts by clinicians to develop better therapies for lung cancer, improvements in treatment efficacy have been unsatisfactory. It is clear that new treatment strategies are needed. Many types of lung cancer cells express COX-2, and we and other researchers have shown that COX-2 inhibitors can enhance the effect of chemotherapy and radiation therapy,
Saha/Choy
204
respectively, on these tumor cells but not on normal tissues. Thus, we hypothesize that selective COX-2 inhibitors may enhance the therapeutic effect of concurrent treatment of paclitaxel, carboplatin and radiation on lung cancers in the patients. Celecoxib, a selective COX-2 inhibitor, is an ideal agent to combine with chemoradiotherapy for several reasons. Celecoxib is a 1,5-diaryl-pyrazole-based compound that inhibits recombinant COX-2 with an IC50 of 4 ⫻ 10⫺8 mol/l compared with 1.5 ⫻ 10⫺5 mol/l for COX-1 [48]. It is 375-fold selective for COX-2, based on human recombinant enzyme assays [48]. Phase II studies in osteoarthritis and rheumatoid arthritis have established dose ranges of 100–800 mg/day as safe and effective. In large, randomized, multicenter, placebo-controlled, double-blind trials conducted in patients with rheumatoid arthritis, celecoxib proved to be less toxic than non-selective inhibitors of COX-1 and COX-2, and no more toxic than a placebo [48]. Of note, a high-dose of celecoxib (600 mg bid) has no effect on serum thromboxane or platelet function [49]. This is obviously important in patients receiving myelosuppressive drugs since we normally avoid agents that impair platelet function. It is currently felt that traditional phase I trials to establish the parameters of maximum-tolerated dose and dose-limiting toxicities are not necessarily warranted in an era where new agents being tested are not necessarily hypothesized to exert their action in the traditional cytotoxic manner. As such, the traditional phase I trial design may not always be relevant. It is unlikely that the addition of a selective COX-2 inhibitor, celecoxib, at doses well tolerated in other medical conditions, will add to the toxicity of established chemoradiation schedules in NSCLC. However, several clinical trials are evaluating the toxicities in an initial cohort of patients to ensure that this combination is safe and well tolerated before they proceed to the targeted accrual. To assess the role of COX-2 inhibitors as a potential radiation sensitizer in NSCLC, we initiated another phase II trial (THO-0059) using celecoxib 400 mg p.o. bid plus concurrent weekly paclitaxel (50 mg/m2)/carboplatin (AUC 2) and chest radiation therapy (63 Gy) for stage III NSCLC patients. The main objectives of the trials are to determine the response rate, overall and relapse-free survival and toxicities with this combination. Serum and plasma VEGF levels and PGE2-M levels in urine specimens are being measured as well. We will investigate the correlation of these measurements with tumor response rate and survival parameters of patients. Nine patients with stage III NSCLC have been enrolled thus far. VEGF levels from the first 8 patients decreased after celecoxib treatment compared to the pretreatment level. Post-celecoxib, serum/plasma levels of VEGF do not show a consistent pattern whereas VEGF levels fell in the months following
Potential for Combined Modality Therapy of COX Inhibitors and Radiation
205
treatment. The preliminary data suggest that concurrent administration of celecoxib with chemoradiation therapy for locally advanced NSCLC appears to be feasible, serum and plasma VEGF levels decreased following 5-day administration of celecoxib and declined further over several months. Serum and plasma VEGF levels may serve as a tumor marker in response to the chemoradiotherapy and/or celecoxib.
Effect of Selective COX-2 Inhibitor in Combination with Docetaxel in Lung Cancer Trial
As mentioned before, COX-2 is upregulated in many malignancies including NSCLC and may play a role in the growth and development of cancer. PGE2, a downstream product of COX-2, can promote tumor growth and invasion via stimulation of vascular endothelial growth factor (VEGF), inhibition of immune surveillance and upregulation of Bcl-2 and various MMPs. To assess the role of COX-2 inhibition in recurrent NSCLC, we have initiated a phase II trial combining celecoxib 400 mg p.o. bid and Txt 75 mg/m2 i.v. q3wk. Our preliminary results indicated that COX-2 is overexpressed in 75% of analyzed tumors, that COX-2 activity increased, and that there was also significant COX-1 activity compared to normal lung tissue. Celecoxib inhibits intratumoral COX-2 activity and intra-tumoral PGE2 levels decreased considerably (to normal tissue values in most samples) following celecoxib administration; pre- and post-celecoxib intra-tumoral tissue PGE2 levels demonstrated a marked decline (100.7 vs. 18.1 ng/g respectively). Urine PGE-M levels are currently being analyzed. These preliminary data indicate celecoxib as given in this trial inhibits intra-tumoral COX-2 leading to changes in intra-tumoral PGE2 activity. Further data are needed to more fully understand the role of COX-2 inhibition in NSCLC. Based on the initial responses observed, enrollment will continue to a planned accrual of 54 patients [ASCO 2002, Vol. 21 (Part 1); p297a, #1187].
Summary and Conclusions
In conclusion, COX-2 inhibitors have potent anti-tumorigenic activity. Results from animal studies strongly indicate that the likely mechanism for enhanced TGD and TCD50 in tumors treated with radiation and COX-2 inhibitors was the inhibition of angiogenesis. In our recent findings we observed that the antagonists of angiogenesis also inhibited the endogenous as well as phorbol-ester-mediated induction of COX-2 expression in human lung
Saha/Choy
206
cancer cell lines and that in the xenograft model a combination of angiogenic antagonists and radiation significantly delayed tumor growth [ASCO 2002, Vol. 21 (Part 1); p445a, #1779]. In human tumor models, apoptosis was another mechanism of cell death. Furthermore, it was demonstrated that COX-2 inhibitors could change the intrinsic radiosensitivity of human cancer cells [41]. Therefore, inhibition of angiogenesis by COX-2 inhibitors may be the major mechanism for increased radiation effects in tumors. However, other mechanisms such as changes in tumor perfusion, apoptosis, and an increase in intrinsic radiation sensitivity must also be considered. Inhibitors of COX-2 in combination with radiation therapy may be an alternative strategy that can be tested in clinical trials. The combination of COX-2 inhibitors and radiation suggest a complementary strategy to target angiogenesis while potentially minimizing the impact on quality of life. Currently, the Radiation Therapy Oncology Group [www.rtog.org] is just one of the National Cancer Institute sponsored cooperative groups conducting clinical trials in cervix cancer, lung cancer and brain tumors, using inhibitors of COX-2 in combination with chemotherapy and radiation therapy. These clinical trials will help elucidate the role of this interesting class of agents in combination with cytotoxic therapy for the treatment of cancer.
References 1 2 3 4
5 6 7 8
9 10 11
Fornace AJ Jr: Mammalian genes induced by radiation: Activation of genes associated with growth control. Annu Rev Genet 1992;26:507–526. Hartwell LH, Kastan MB: Cell cycle control and cancer. Science 1994;266:1821–1828. Riley PA: Free radicals in biology: Oxidative stress and the effects of ionizing radiation. Int J Radiat Biol 1994;65:27–33. McKenna WG, Bernhard EJ, Markiewicz DA, Rudoltz MS, Maity A, Muschel RJ: Regulation of radiation-induced apoptosis in oncogene-transfected fibroblasts: Influence of H-ras on the G2 delay. Oncogene 1996;12:237–245. Morgan SE, Kastan MB: p53 and ATM: Cell cycle, cell death, and cancer. Adv Cancer Res 1997;71:1–25. Williams CS, Mann M, DuBois RN: The role of cyclooxygenases in inflammation, cancer and development. Oncogene 1999;18:7908–7916. Oates JA, FitzGerald GA, Branch RA, Jackson EK, Knapp HR, Roberts LJ 2nd: Clinical implications of prostaglandin and thromboxane A2 formation (2). N Engl J Med 1988;319:761–767. Oates JA, Brash A, Fitzgerald G: Comparison of the effects of selective inhibition of thromboxane synthase with those of inhibition of the cyclooxygenase enzyme in man. Trans Am Clin Climatol Assoc 1983;95:157–162. Wolff H, Saukkonen K, Anttila S, Karjalainen A, Vainio H, Ristimaki A: Expression of cyclooxygenase-2 in human lung carcinoma. Cancer Res 1998;58:4997–5001. Soslow RA, Dannenberg AJ, Rush D, Woerner BM, Khan KN, Masferrer J, Koki AT: COX-2 is expressed in human pulmonary, colonic and mammary tumors. Cancer 2000;89:2637–2645. Hida T, Yatabe Y, Achiwa H, Muramatsu H, Kozaki K, Nakamura S, Ogawa M, Mitsudomi T, Sugiura T, Takahashi T: Increased expression of cyclooxygenase-2 occurs frequently in human lung cancers, specifically in adenocarcinomas. Cancer Res 1998;58:3761–3764.
Potential for Combined Modality Therapy of COX Inhibitors and Radiation
207
12
13
14 15 16 17
18
19
20
21
22 23
24
25
26 27 28 29 30
31
32
Hida T, Kozaki K, Muramatsu H, Masuda A, Shimizu S, Mitsudomi T, Sugiura T, Ogawa M, Takahashi T: Cyclooxygenase-2 inhibitor induces apoptosis and enhances cytotoxicity of various anticancer agents in non-small cell lung cancer cell lines. Clin Cancer Res 2000;6:2006–2011. Achiwa H, Yatabe Y, Hida T, Kuroishi T, Kozaki K, Nakamura S, Ogawa M, Sugiura T, Mitsudomi T, Takahashi T: Prognostic significance of elevated cyclooxygenase-2 expression in primary, resected lung adenocarcinomas. Clin Cancer Res 1999;5:1001–1005. Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN: Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998;93:705–716. Tsujii M, Kawano S, DuBois RN: Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA 1997;94:3336–3340. Kalgutkar AS, Crews BC, Rowlinson SW, Garner C, Seibert K, Marnett LJ: Aspirin-like molecules that covalently inactivate cyclooxygenase-2. Science 1998;280:1268–1270. Crew TE, Elder DJ, Paraskeva C: A cyclooxygenase-2 (COX-2)-selective non-steroidal antiinflammatory drug enhances the growth inhibitory effect of butyrate in colorectal carcinoma cells expressing COX-2 protein: Regulation of COX-2 by butyrate. Carcinogenesis 2000;21:69–77. Elder DJ, Halton DE, Crew TE, Paraskeva C: Apoptosis induction and cyclooxygenase-2 regulation in human colorectal adenoma and carcinoma cell lines by the cyclooxygenase-2-selective non-steroidal anti-inflammatory drug NS-398. Int J Cancer 2000;86:553–560. Elder DJ, Halton DE, Hague A, Paraskeva C: Induction of apoptotic cell death in human colorectal carcinoma cell lines by a cyclooxygenase-2 (COX-2)-selective nonsteroidal anti-inflammatory drug: Independence from COX-2 protein expression. Clin Cancer Res 1997;3:1679–1683. Hsu AL, Ching TT, Wang DS, Song X, Rangnekar VM, Chen CS: The cyclooxygenase-2 inhibitor celecoxib induces apoptosis by blocking Akt activation in human prostate cancer cells independently of Bcl-2. J Biol Chem 2000;275:11397–11403. Liu XH, Yao S, Kirschenbaum A, Levine AC: NS398, a selective cyclooxygenase-2 inhibitor, induces apoptosis and down-regulates bcl-2 expression in LNCaP cells. Cancer Res 1998;58: 4245–4249. Masferrer JL, Koki A, Seibert K: COX-2 inhibitors. A new class of antiangiogenic agents. Ann NY Acad Sci 1999;889:84–86. Masferrer JL, Leahy KM, Koki AT, Zweifel BS, Settle SL, Woerner BM, Edwards DA, Flickinger AG, Moore RJ, Seibert K: Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 2000;60:1306–1311. Seed MP, Brown JR, Freemantle CN, Papworth JL, Colville-Nash PR, Willis D, Somerville KW, Asculai S, Willoughby DA: The inhibition of colon-26 adenocarcinoma development and angiogenesis by topical diclofenac in 2.5% hyaluronan. Cancer Res 1997;57:1625–1629. Sheng H, Shao J, Kirkland SC, Isakson P, Coffey RJ, Morrow J, Beauchamp RD, DuBois RN: Inhibition of human colon cancer cell growth by selective inhibition of cyclooxygenase-2. J Clin Invest 1997;99:2254–2259. Tsuji S, Tsujii M, Kawano S, Hori M: Cyclooxygenase-2 upregulation as a perigenetic change in carcinogenesis. J Exp Clin Cancer Res 2001;20:117–129. Goodwin JS, Ceuppens J: Regulation of the immune response by prostaglandins. J Clin Immunol 1983;3:295–315. Tsujii M, DuBois RN: Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell 1995;83:493–501. Sheng H, Shao J, Morrow JD, Beauchamp RD, DuBois RN: Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells. Cancer Res 1998;58:362–366. Subbaramaiah K, Norton L, Gerald W, Dannenberg AJ: Cyclooxygenase-2 is overexpressed in HER-2/neu-positive breast cancer. Evidence for involvement of AP-1 and PEA3. J Biol Chem 2002;277:18649–18657. Leung WK, To KF, Ng YP, Lee TL, Lau JY, Chan FK, Ng EK, Chung SC, Sung JJ: Association between cyclo-oxygenase-2 overexpression and missense p53 mutations in gastric cancer. Br J Cancer 2001;84:335–339. Khuri FR, Wu H, Lee JJ, Kemp BL, Lotan R, Lippman SM, Feng L, Hong WK, Xu XC: Cyclooxygenase-2 overexpression is a marker of poor prognosis in stage I non-small cell lung cancer. Clin Cancer Res 2001;7:861–867.
Saha/Choy
208
33
34 35 36 37
38
39 40 41
42
43 44 45
46
47 48 49
Gaffney DK, Holden J, Davis M, Zempolich K, Murphy KJ, Dodson M: Elevated cyclooxygenase-2 expression correlates with diminished survival in carcinoma of the cervix treated with radiotherapy. Int J Radiat Oncol Biol Phys 2001;49:1213–1217. Steinauer KK, Gibbs I, Ning S, French JN, Armstrong J, Knox SJ: Radiation induces upregulation of cyclooxygenase-2 protein in PC-3 cells. Int J Radiat Oncol Biol Phys 2000;48:325–328. Milas L, Kishi K, Hunter N, Mason K, Masferrer JL, Tofilon PJ: Enhancement of tumor response to ␥-radiation by an inhibitor of cyclooxygenase-2 enzyme. J Natl Cancer Inst 1999;91:1501–1504. Furuta Y, Hunter N, Barkley T Jr, Hall E, Milas L: Increase in radioresponse of murine tumors by treatment with indomethacin. Cancer Res 1988;48:3008–3013. Penning TD, Talley JJ, Bertenshaw SR, Carter JS, Collins PW, Docter S, Graneto MJ, Lee LF, Malecha JW, Miyashiro JM, Rogers RS, Rogier DJ, Yu SS, Anderson GD, Burton EG, Cogburn JN, Gregory SA, Koboldt CM, Perkins WE, Seibert K, Veenhuizen AW, Zhang YY, Isakson PC: Synthesis and biological evaluation of the 1,5-diarylpyrazole class of cyclooxygenase-2 inhibitors: Identification of 4-[5-(4-methylphenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]benzenesulfonamide (SC-58635, celecoxib). J Med Chem 1997;40:1347–1365. Kishi K, Petersen S, Petersen C, Hunter N, Mason K, Masferrer JL, Tofilon PJ, Milas L: Preferential enhancement of tumor radioresponse by a cyclooxygenase-2 inhibitor. Cancer Res 2000;60:1326–1331. Withers HR, Elkind MM: Radiosensitivity and fractionation response of crypt cells of mouse jejunum. Radiat Res 1969;38:598–613. Stone HB: Leg contracture in mice: An assay of normal tissue response. Int J Radiat Oncol Biol Phys 1984;10:1053–1061. Petersen C, Petersen S, Milas L, Lang FF, Tofilon PJ: Enhancement of intrinsic tumor cell radiosensitivity induced by a selective cyclooxygenase-2 inhibitor. Clin Cancer Res 2000;6: 2513–2520. Pyo H, Choy H, Amorino GP, Kim JS, Cao Q, Hercules SK, DuBois RN: A selective cyclooxygenase-2 inhibitor, NS-398, enhances the effect of radiation in vitro and in vivo preferentially on the cells that express cyclooxygenase-2. Clin Cancer Res 2001;7:2998–3005. Gallo O: Re: Enhancement of tumor response to ␥-radiation by an inhibitor of cyclooxygenase-2 enzyme. J Natl Cancer Inst 2000;92:346–347. Williams CS, Tsujii M, Reese J, Dey SK, DuBois RN: Host cyclooxygenase-2 modulates carcinoma growth. J Clin Invest 2000;105:1589–1594. Gorski DH, Beckett MA, Jaskowiak NT, Calvin DP, Mauceri HJ, Salloum RM, Seetharam S, Koons A, Hari DM, Kufe DW, Weichselbaum RR: Blockage of the vascular endothelial growth factor stress response increases the antitumor effects of ionizing radiation. Cancer Res 1999;59: 3374–3378. Kakeji Y, Maehara Y, Ikebe M, Teicher BA: Dynamics of tumor oxygenation, CD31 staining and transforming growth factor- levels after treatment with radiation or cyclophosphamide in the rat 13762 mammary carcinoma. Int J Radiat Oncol Biol Phys 1997;37:1115–1123. Dicker AP, Williams TL, Grant DS: Targeting angiogenic processes by combination rofecoxib and ionizing radiation. Am J Clin Oncol 2001;24:438–442. Hawkey CJ: COX-2 inhibitors. Lancet 1999;353:307–314. Leese PT, Hubbard RC, Karim A, Isakson PC, Yu SS, Geis GS: Effects of celecoxib, a novel cyclooxygenase-2 inhibitor, on platelet function in healthy adults: A randomized, controlled trial. J Clin Pharmacol 2000;40:124–132.
Dr. Hak Choy Department of Radiation Oncology B902 TVC Vanderbilt University Medical Center 22nd Avenue at Pierce, Nashville, TN 37232-5671 (USA) Tel. ⫹1 615 3227253, Fax ⫹1 615 3437218, E-Mail
[email protected]
Potential for Combined Modality Therapy of COX Inhibitors and Radiation
209
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 210–242
Non-Steroidal Anti-Inflammatory and Cyclooxygenase-2-Selective Inhibitors in Clinical Cancer Prevention Trials Ernest T. Hawk, Jaye L. Viner, Asad Umar Gastrointestinal & Other Cancers Research Group, National Cancer Institute, Division of Cancer Prevention, Bethesda, Md., USA
Introduction
Cyclooxygenase (COX), and specifically COX-2, may be an important molecular target for cancer prevention. Initial support for this hypothesis comes from several lines of evidence, including mechanistic insights into the role of COX in carcinogenesis, in vitro and in vivo studies of COX inhibitors, and epidemiologic and experimental studies of COX inhibitors in humans. More than 30 non-steroidal anti-inflammatory drugs (NSAIDs) or COX-2-selective inhibitors (COXIBs) are currently marketed in the USA (table 1). This article reviews published data on NSAIDs and COXIBs that have been tested in cancer prevention trials, highlights ongoing research, and considers the public health impact this class of preventive agents may have on cancer and other common chronic diseases of aging.
Molecular Cancer Prevention
Cancer is a late stage of a process referred to as carcinogenesis. This process is characterized by successive aberrations occurring at the molecular, cellular, tissue, organ and organism levels of organization. These changes typically require years, providing time and targets for preventive interventions prior to cancer development, which by pathologic convention occurs upon invasion into the basement membrane. Carcinogenesis can be detected prior to the development of clinical symptoms, typically through screening for hallmarks of
Table 1. NSAID and COXIB inhibitors marketed in the USA (generic and marketed names) – 2002 Generic name Propionic acids Carprofen
Trade name
IsoptoCarpine Ocu-Carpine
Generic name Acetic acids Diclofenac
Trade name
Arthrotec Cataflam Voltaren
Fenoprofen
Nalfon
Flurbiprofen
Ansaid
Indomethacin
Indocin
Advil Dolgesic Genpril Haltran Ibuprin Medipren Menadol Motrin Nuprin
Ketorolac
Toradol
Nabumetone
Relafen
Sulindac
Clinoril
Tolmetin
Tolectin
Ibuprofen
Indoprofen
Endyne
Ketoprofen
Orudis Oruvail
Ketorolac
Toradol
Loxoprofen
Loxatane Loxapine
Naproxen
Anaprox Naprelan Naprosyn
Oxaprozin
Oxicams Meloxicam
Mobic
Piroxicam
Feldene
Fenamic acids Meclofenamate
Mefenamic acid Pyrazolones Oxyphenbutazone
Daypro Phenylbutazone
Salicylates Acetylsalicylic acid
Aspirin Ecotrin
Meclodium Meclofenaf Meclomen Ponstel
Daricon Enarax Vistarax Butatab
COX-2 selective inhibitors Celecoxib Celebrex
Diflunisal
Dolobid
Etodolac
Lodine
Salsalate
Disalcid Salflex
Etoricoxib
Arcoxiba
Rofecoxib
Vioxx
Trilisate
Valdecoxib
Bextra
Choline magnesium trisalicylate a
In development.
NSAIDs/COXIBs in Prevention Trials
211
evolving pathology. Generally, subclinical signatures of carcinogenesis occurring immediately proximal to clinically-evident cancer are regarded as more predictive of cancer development; whereas the predictive value of more remote biologic events is less certain. Most cancer prevention studies, because of the considerable time and resources required to evaluate changes in cancer incidence or mortality, focus on changes in intraepithelial neoplasia (IEN) (e.g., adenomas, actinic keratoses) or reasonably quantitative and reliable markers of risk (e.g., cellular proliferation, apoptosis, or characteristic biochemical/molecular alterations). The association between IEN and a particular cancer, although probabilistic rather than absolute, potentially provides early and important information regarding impending development of more advanced pathology. As proven in other medical disciplines, reasonably validated intermediate markers (e.g., reduced blood pressure in hypertensives as a surrogate for cardiovascular risk reductions) can advance preventive practice in terms of risk assessment, trial design, and regulatory approval of effective interventions [1]. Indeed, the regulatory utility of IEN as a reasonable surrogate for druginduced neoplasia regression (i.e., therapy) or prevention was recently affirmed by a committee of experts [2]. Prevention is primarily dependent on the identification and application of effective interventions. Perhaps equally important is the relative safety of such interventions in an intended risk cohort or patient population. Indeed, while some very high risk cohorts, such as those with germline aberrations in key tumor suppressor genes, may tolerate significant toxicity, individuals at lower risk may only rarely find an agent with mild or infrequent side effects acceptable. As a result, chemopreventive agent identification and development matches candidate agents with appropriate risk cohorts such that the therapeutic index (TI) or benefit:risk ratio is likely to be acceptable.
The Premise for Clinical Prevention Trials with NSAIDs/COXIBs
COX as a Molecular Target The TI for a preventive agent is determined largely by the prevalence, distribution and function of its molecular target [3]. The best molecular targets for cancer prevention are: (1) consistently and preferentially expressed in preinvasive neoplastic, versus normal tissues; (2) biologically functional; (3) critical for the maintenance and/or progression of an aberrant cell (or clonal population of cells); (4) pharmacologically accessible, and (5) modulable by a specific pharmacologic or nutriceutical agent. COX, or the COX-2 isoform in particular, satisfies most of these criteria and is an excellent molecular target for cancer prevention in several organs [4].
Hawk/Viner/Umar
212
COX-2 is overexpressed in neoplastic lesions of the bladder, breast, cervix, central nervous system, colorectum, esophagus, head and neck, liver, lung, pancreas, prostate, skin and stomach [5]. Mechanistically, COX contributes to carcinogenesis in a variety of ways. For example, COX may directly activate carcinogens via its peroxidase activity, indirectly promote carcinogenesis through prostanoid-stimulated cell proliferation, inhibit cell differentiation, inhibit apoptosis, promote cancer metastasis and suppress immunosurveillance [6].
NSAIDs and COXIBs in Preclinical and Epidemiologic Studies
More than 150 published in vivo studies (data not shown) have tested NSAIDs/COXIBs in carcinogen-induced or genetic models of human cancer. The vast majority of these studies show that these agents reduce IEN/cancer incidence, multiplicity, burden/volume; or increase the latency to neoplasia, suggesting their potential for preventive efficacy in a wide range of organs [5]. For example, in vivo efficacy has been demonstrated in intestine, skin, lung, mammary, oral, esophagus and bladder carcinogenesis. In addition, COX inhibitors block the growth of a variety of human tumor xenografts. More recently, COXIBs including JTE-522, NS-398, MF tricyclic, nimesulide, celecoxib, and rofecoxib were effective in rodent carcinogenesis models of oral cavity [7], esophagus [8], colorectal [9–20], mammary [21, 22], skin [23–25], bladder [26, 27] and lung [28] cancers. More than 30 epidemiologic studies have examined possible associations between aspirin/NSAID use and neoplasia. Data from most of these studies, which started in the late 1980s, suggest a 40–50% reduction in colorectal adenomas, cancer and cancer-associated mortality. Activity against other tumors, however, appears to be less consistent and less compelling. Nevertheless, epidemiologic data generally suggest strong preventive effects against cancer of the upper gastrointestinal tract and several other sites (discussed in detail elsewhere in this text) [5, 29]. In summary, compelling mechanistic, preclinical and observational data support the testing and development of NSAIDs and COXIBs as candidate chemopreventives. Nevertheless, their safety must be assessed and deemed acceptable in the context of chronic administration for cancer prevention. Because many physiologic roles are mediated by COX products including prostaglandins and thromboxanes, most NSAIDs exert multiple effects on a wide range of organs. For example, NSAIDs are known to contribute to sodium and water retention, inhibit uterine contractility, inhibit platelet aggregation, reduce cellular proliferation and/or promote apoptosis, and reduce inflammation. The clinical context dictates whether such effects are desirable or
NSAIDs/COXIBs in Prevention Trials
213
Risks
Effects
Benefits
Hemorrhage
Block platelet aggregation
Reduce risk of thrombosis
Exacerbate hypertension congestive failure peripheral edema
Sodium and water retention
Treat salt-wasting nephropathy
Inhibit prematue labor
Reduce uterine tone
Prolong labor, reduce menstrual cramps
Induce gastroduodenal ulceration
Induce apoptosis
Induce apoptosis in neoplasia
Mask infection
Reduce inflammation
Treat inflammatory arthritis and injury
Mask infection
Reduce fever
Treat fever
Mask a sign of tissue injury
Reduce pain
Treat pain
Block neoangiogenesis in response to physiologic need
Reduce neoangiogenesis
Block neoangiogenesis beyond physiologic need
Impair wound healing
Reduce proliferation
Control cell growth beyond physiologic needs
Fig. 1. The Yin-Yang of anti-COX pharmacodynamics.
problematic; in the context of cancer prevention, such toxicities are generally regarded as problematic (fig. 1).
Clinical Trials of Aspirin and NSAIDs in Cancer Prevention
While the premise for NSAIDs in cancer chemoprevention is compelling based on mechanistic, in vivo and observational data, clinical trials are necessary to definitively establish the TI of an agent in an intended risk cohort. NSAIDs in Colorectal Neoplasia of Patients with FAP The first wave of chemoprevention efficacy data on this class of agents included clinical anecdotes and small case series that showed drastic reductions in the adenoma burden of familial adenomatous polyposis (FAP) patients who were treated with sulindac or another NSAID (table 2). The first study described almost complete regression of colorectal adenomas in 4 patients administered sulindac for 4–12 months [30]. Subsequently, 11 uncontrolled case reports [31] or case series [32–40] confirmed significant reductions in colorectal adenoma size and/or number in more than 110 patients treated with sulindac in doses ranging from 150 to 450 mg/day for 2–92 months. One report recently highlighted
Hawk/Viner/Umar
214
Table 2. NSAIDs and COXIBs in treatment or prevention of colorectal neoplasia in familial adenomatous polyposis NSAIDs/COXIBs in Prevention Trials 215
NSAID
Dose and duration, months
Trial design
Sample size
Primary result(s)
Reference
Celecoxib
100, 400 mg po bid ⫻ 6
DBRCT
77 (⫹ 6 w/ duodenal disease only)
Focal and global adenoma number and burden – reduced* (400 mg bid dose vs. placebo)
Steinbach, 2000 [76]
Indomethacin
50 mg ir qd-bid ⫻ 1–3 50 mg qd-bid ⫻ 1–2
CS CS
2 7
Adenoma number – reduced Adenoma number – reduced*; proliferative index – increased
Hirata, 1994 [46] Hirota, 1996 [47]
Nimesulide
2 mg/kg qd po ⫻ 2.5
CS
7
Endoscopic effect – none; proliferative index – unchanged
Dolara, 1999 [75]
Sulindac
200–300 mg po qd ⫻ 1–8
CS
8
Microadenoma number – reduced
158 mg po qd on average ⫻ 63.4 ⫾ 31.3 150 mg po qd-bid ⫻ 1.5–3 150 mg po bid ⫻ 9
CS
12
CS DBRCT
3 22
150 mg po bid ⫻ 3
CS
22
Adenoma number and size – reduced; high-grade adenomas – reduced Adenoma number – reduced Adenoma number and size – reduced*; mucosal prostanoids – reduced* a Adenoma number and size – reduced*
75–150 mg po bid ⫻ 48
DBRCT
41
Dose not specified ⫻ 12
CR
1
150 mg po qd ⫻ 3
CS
10
Charneau, 1990 [32] Cruz-Correa, 2002 [40] Friend, 1990 [33] Giardiello, 1993 [54] Giardiello, 1998 [34] Giardiello, 2002 [55] Gonzaga, 1985 [31] Iwama, 1991 [45]
100 mg po tid ⫻ 4
DBRCXT
9
Adenoma number, size, incidence – not reduced Adenoma number – reduced Adenoma number and size – no overall effect; mild reduction in selected subjects Adenoma number and size – reduced*
Labayle, 1991 [52]
Table 2 (continued) Hawk/Viner/Umar
NSAID
Sulindac sulfone
Dose and duration, months
Trial design
Sample size
Primary result(s)
Reference
100 mg po tid ⫻ 4 150 mg po tid ⫻ 16 200 mg po bid ⫻ 6
CS CR DBRCT
10 1 14
200 mg po bid ⫻ 6
CS
4
Muller, 1994 [35] Niv, 1994 [41] Nugent, 1993 [53] Rigau, 1991 [36]
100 mg po bid ⫻ 2
CS
20
100 mg po bid ⫻ 6
CS
13
Adenoma number and size – reduced Adenoma burden – no overall effecta Adenoma burden – reduced*; proliferative index – reduced* Adenoma number and size – reduced; prostanoids reduced* Adenoma number or size – reduced*; proliferative index – no effect Adenoma number and size – reduceda
75–150 mg po bid ⫻ 4–12
CS
4
Adenoma number – reduced
75–200 mg po bid ⫻ 21–92
CS
10
Adenoma number – reduced
25–150 mg ir bid ⫻ 3–48
CCTRL
38
Adenoma number and size – reduced*; proliferative index – reduced*; prostanoids – reduced
200–400 mg po bid ⫻ 6
CS
18
Adenoma number and size – stable; apoptosis within lesions – increased
Spagnesi, 1994 [37] Tonelli, 1994 [38, 39] Waddell, 1983 [30] Waddell, 1989 [129] Winde, 1997 [49–51] Van Stolk, 2000 [48]
216
*Statistically significant result (p ⬍ 0.05). Sample size is the number of subjects evaluated at study completion. Intervention is the duration of agent administration until described effect. po ⫽ Administered by mouth; ir ⫽ administered per rectum; qd ⫽ once per day; bid ⫽ twice per day; tid ⫽ three times per day. a One subject with adenocarcinoma on extended follow-up. CR ⫽ Case report; CS ⫽ case series; DBRCT ⫽ double-blind, randomized, controlled trial; DBRCXT ⫽ double-blind, randomized, controlled, cross-over trial; CCTRL ⫽ case-control.
the tumor-suppressive effects of sulindac, noting significant reductions in adenoma burden after 12 months administration, and sustained efficacy over a mean of 63 months of treatment [40]. Interestingly, significant prevention of higher grade adenomas was reported. Of note, 1 patient in the study developed a stage III rectal cancer after 35 months on sulindac, which sadly is not an isolated event – 4 other cases of colorectal cancer have been reported in patients taking sulindac [34, 38, 41–44]. The frequency and clinical significance of these cancers are difficult to interpret without more information on the total number of comparable patients treated with this agent, and the duration of treatment. One case report [41] and one case series of 10 patients [45] failed to demonstrate significant adenoma regressive effects with sulindac. Interestingly, three other case series have noted preventive efficacy with other agents such as indomethacin [46, 47] and sulindac sulfone [48], suggesting the generalizability of promising data with sulindac to other NSAIDs as well. A second wave of data from several controlled studies – one non-randomized series of trials [49–51] and four randomized trials [52–55] – confirmed NSAIDassociated reductions in adenoma number and size. These corroborative studies demonstrated that within a few months NSAIDs regressed a large proportion of prevalent adenomas in a cohort of 85 patients; these lesions, however, recurred shortly after agent cessation [49–54]. In addition, few patients experienced complete adenoma regression, and those that did typically experienced adenoma recurrences upon discontinuation of the drug. The most recent trial was a randomized, placebo-controlled trial of sulindac at 75–150 mg twice a day administered for 48 months in 41 genotype-positive, phenotype-negative patients with FAP [55]. While compliance with the regimen exceeded 76% and mucosal prostanoid levels were reduced in individuals administered sulindac, incident adenomas were not significantly suppressed in terms of size, number, or rates. It is hypothesized that low doses of drug or the small sample size may account for this study’s negative result. Nevertheless, in aggregate the data suggest that sulindac’s efficacy against incident adenomas in FAP cohorts is likely to be mild. Sulindac’s effects against duodenal neoplasia in patients with FAP have been less encouraging (table 3) [53, 56–59]. While an early uncontrolled study [58] suggested reductions in duodenal adenomas, a subsequent case series failed to confirm this effect and actually reported an increase in adenoma burden over 6–24 months of treatment [57]. In one controlled trial, sulindac achieved non-significant reductions in duodenal adenoma burden [53], primarily in adenomas 2 mm in diameter or smaller [59]. NSAIDs in Sporadic Colorectal Neoplasia The efficacy of NSAIDs against sporadic neoplasia was suggested by several small studies, which demonstrated a reduction in rectal mucosal
NSAIDs/COXIBs in Prevention Trials
217
Table 3. NSAIDs and COXIBs in the treatment of duodenal adenomas in familial adenomatous polyposis Hawk/Viner/Umar
Agent
Sizea
Durationb
Design
Results
Dose
Study
Celecoxib
83
6
DBRCT
Polyp number/size reductionc
100–400 mg bid
Phillips et al., 2002 [77]
Sulindac
24
6
DBRCT
Polyp number/size reductionc (only in ⱕ 2mm)
200 mg bid
Debinski et al.,1995 [59]
Sulindac
23
6
DBRCT
Polyp number/size reductionc; proliferation markers reduction
200 mg bid
Nugent et al., 1993 [54]
Sulindac
1
Not reported
CRPT
Polyp number/size reduction
150 mg bid
Parker et al., 1993 [58]
Sulindac
6
6–24
CSER
Polyp number/size increase
150 mg bid
Richard et al., 1997 [57]
Sulindac calcium ⫹ calciferol
15
6
DBRCT
Proliferation markers showed no change in either treatment
300 mg qd (380 mg ⫹ 500 mg) qd
Seow-Choen et al., 1996 [56]
CRPT ⫽ case report; CSER ⫽ case series; DBRCT ⫽ double-blind, randomized, controlled trial. Number of subjects evaluated. b Duration of agent administration in months. c Statistically significant (p ⬍ 0.05). a
218
prostanoids following short- to moderate-term oral treatment with piroxicam [60], ibuprofen [61], or aspirin (table 4) [62, 63]. Indeed, aspirin doses as low as 80 mg/day reliably reduced mucosal prostanoid concentrations. These effects, however, were transient and prostanoid concentrations reverted to baseline within 3 months. NSAIDs’ efficacy in regressing sporadic colorectal polyps is less established because there have been fewer studies, and the results from these studies are generally less consistent than those in FAP cohorts. Even so, two small case series have reported modest effects in the setting of sporadic adenomas [64, 65]. In the first study, polyp regression occurred in 1 of 7 patients treated with either sulindac or piroxicam for 6 months [64]. In a second open-label study of 15 patients, sulindac administered for 4 months regressed 13 of 20 polyps [65]. By contrast, in a double-blind, placebo-controlled randomized trial of 44 subjects, 4 months of sulindac treatment did not achieve a clinically significant regression of sporadic colonic polyps [66]. Three larger clinical trials evaluating the effects of NSAIDs in colorectal adenoma or cancer prevention have been reported thus far [67–69]. Baron et al. [67] studied the effects of aspirin at 80 or 325 mg/day versus placebo over 48 months in 1,121 patients at moderate risk for colorectal cancer, based on a history of prior adenomas. This study showed an interesting inverse dose response, with 19 and 4% reductions in recurrent adenomas – and more significantly, 40 and 19% reductions in recurrent advanced adenomas – in patients administered 80 and 325 mg/day, respectively. Benamouzig et al. [68] recently reported interim 1-year results from a randomized, placebo-controlled trial of aspirin (given at 160–300 mg/day over 48 months) in 291 patients with prior adenomas. Preliminary data show a 44% reduction in recurrent adenomas, however, it will be several years before the trial will generate mature data. Finally, in a large group of physicians at average risk (though physicians may be at less than average risk, based on their sociodemographic status), Gann et al. [69] reported non-significant effects of aspirin 325 mg every other day over 60 months on both incident adenomas/in situ cancers (RR = 0.86; 95% CI = 0.49–1.32) and invasive cancers (RR = 1.15; 95% CI = 0.80–1.65). In sum, trials of aspirin in patients at moderate cancer risk have demonstrated reductions in the risk of adenoma recurrence; but trials in persons at average risk for colorectal cancer have not shown significant reductions for incident adenomas or cancer. NSAIDs in Skin Neoplasia Both non-melanoma skin cancer and melanomas overexpress COX-2 and produce elevated levels of prostanoids (particularly PGE2) [5]. Emerging data show that COX-2 overexpression typically occurs during preinvasive stages of carcinogenesis, suggesting that it may be a good target for skin cancer
NSAIDs/COXIBs in Prevention Trials
219
Hawk/Viner/Umar
Table 4. Published clinical trials involving NSAIDs in the prevention/treatment of sporadic colorectal adenomas or colorectal cancer prevention
220
NSAID
Dose and duration, months
Trial design
Sample size
Primary result(s)
Reference
Aspirin
81 mg po qd vs. placebo ⫻ 3
DBRCXT
10
Reduced mucosal prostanoid concentrations*; reduced TGF␣ staining*; both effects returned to baseline levels
Barnes, 1999 [62]
Aspirin
80 mg po qd vs. 325 mg po qd vs. placebo ⫻ 48
DBRCT
1,121
Adenoma recurrence – all and advanced 80 mg: RR(all) ⫽ 0.81* (0.68–0.96); RR(advanced) ⫽ 0.60 (0.35–1.03) 325 mg: RR(all) ⫽ 0.96 (0.82–1.13); RR(advanced) ⫽ 0.81 (0.49–1.32)
Baron, 2002 [67]
Aspirin
160–300 mg po qd vs. placebo ⫻ 48
DBRCT
291
Interim adenoma recurrence at 1 year (preliminary result): OR ⫽ 0.56* (95% CI ⫽ 0.31–1.01)
Benamouzig, 2002 [68]
Aspirin
325 mg po qod vs. placebo ⫻ 60
DBRCT
22,071
CRC RR ⫽ 1.15 (95% CI ⫽ 0.80–1.65); Gann, 1994 [69] In situ cancer or adenoma RR ⫽ 0.86 (95% CI ⫽ 0.68–1.10)
Aspirin
40.5, 81, 162, 324, or 648 mg po qd vs. placebo ⫻ 2 wks
DBRCT
65
Reduced mucosal PGE2* with 81 mg po qd; reduced mucosal PGF2␣ with 40.5 mg po qd
Ruffin, 1997 [63]
Ibuprofen
300 or 600 mg po qd vs. placebo ⫻ 1
DBRCT
27
Reduced mucosal prostanoid concentrations*; 300 mg po qd recommended for future studies
Chow, 2000 [61]
NSAIDs/COXIBs in Prevention Trials
Piroxicam
7.5 mg po qd vs. placebo ⫻ 24
Piroxicam
Step 1–10 mg po qd or qod Steps 1 & 2 – CS Step 2–10 mg po qod ⫹ DFMO 0.5 gm/m2 ⫻ 6
Sulindac
150 mg po bid vs. placebo ⫻ 4
Sulindac Sulindac or piroxicam
DBRCT
96
Reduced mucosal prostanoid concentrations*; significant gastrointestinal side effects*
Calaluce, 2000 [60]
Step 1 ⫽ 12 Step 2 ⫽ 31
Step 1 – 10 mg qod tolerable; no changes in ODC or urinary polyamine concentrations Step 2 – Combination is tolerable; possible synergistic effects
Carbone, 1998 [130]
DBRCT
44
8% probability of ⬎50% sporadic adenoma regression
Ladenheim, 1995 [66]
300 mg po qd ⫻ 4
CS
15
13 of 20 polyps in 15 patients shrank or disappeared*
Matsuhashi, 1997 [65]
200 mg po bid (sulindac) or 20 mg po qd (piroxicam) ⫻ 6
CS
9
Discontinued due to toxicity in 2; piroxicam more toxic than sulindac; no dramatic adenoma regression
Hixson, 1993 [64]
*Statistically significant result (p ⬍ 0.05). Sample size is the number of subjects evaluated at study completion. Intervention is the duration of agent administration until described effect. po ⫽ Administered by mouth; qd ⫽ once per day; qod ⫽ every other day; bid ⫽ twice per day; CS ⫽ case series; DBRCT ⫽ double-blind, randomized, controlled trial; DBRCXT ⫽ double-blind, randomized, controlled, cross-over trial; ODC ⫽ ornithine decarboxylase; RR ⫽ relative risk; OR ⫽ odds ratio; CI ⫽ confidence interval; CRC ⫽ colorectal cancer; DFMO ⫽ difluoromethylornithine.
221
prevention as well. For example, Pentland et al. [24] have recently shown that COX-2 is upregulated and prostaglandin E2 (PGE2) concentrations are increased in precancerous lesions (i.e., actinic keratoses) in mice treated with UV radiation. In addition, preclinical carcinogenesis studies have shown that COX-2 overexpression in basal keratinocytes of mice transgenic for the COX-2 gene results in hyperplastic and dysplastic cutaneous changes consistent with neoplastic development [70]. Finally, reductions in tumor incidence, multiplicity and size, and increases in tumor latency have been reported in association with NSAID/COXIB treatment in several different rodent models of skin carcinogenesis [23–25, 71, 72]. In terms of human investigations, a recent uncontrolled study of 5 patients with chronic lymphocytic leukemia and Bowen’s disease (squamous cell carcinoma in situ) reported clinical resolution and complete histologic clearing after a brief period of treatment with topical imiquimod and oral sulindac [73]. In addition, a recent open-label trial of topical diclofenac (Solaraze™, SkyePharma) 3% in a 2.5% hyaluronic acid gel, applied twice daily for up to 6 months in 29 patients with actinic keratoses revealed major responses in more than 90% of patients [74]. Indeed, of 27 evaluable patients following discontinuation of the drug, 81% had a complete response and an additional 15% showed marked clinical improvement in remaining actinic keratoses. Additional randomized trials with topical diclofenac have been performed, although not yet published anywhere except in the label. On the strength of these data, the FDA approved this agent in 2000 for the management of actinic keratoses [http://www.fda.gov/cder/approval/index.htm].
Clinical Trials of COXIBs in Cancer Prevention
COXIBs have considerable promise in FAP. Although one small case series of nimesulide administered over 10 weeks found no effect on rectal adenoma burden in 7 FAP patients [75], celecoxib administered over 6 months to 83 FAP subjects significantly reduced colorectal adenoma number and size (table 2) [76]. In this double-blinded, randomized, placebo-controlled trial, no significant side effects attributable to celecoxib treatment were observed, and a dose of 400 mg twice a day resulted in a significant 28% reduction in the mean adenoma number from selected mucosal areas. In addition, celecoxib improved the global endoscopic appearance of both the colorectum and the duodenum (tables 2, 3) [77], suggesting that this agent may reduce neoplastic risk in both organs. Based on these findings and a commitment to further investigations, in 2000 the FDA granted accelerated approval for the use of celecoxib in FAP as a complement to standard management (i.e., surveillance and prophylactic surgery).
Hawk/Viner/Umar
222
In addition, investigators are now evaluating the efficacy of another COXIB, rofecoxib, in patients with FAP, the results of which are forthcoming [N. Arber, pers. commun. 2002]. More recently, a 2 ⫻ 2 factorial trial of celecoxib in patients with prevalent esophageal squamous dysplasia was reported [78]. In this trial, 360 residents of Linxian, China, with mild to moderate esophageal squamous dysplasia, were randomized to treatment with celecoxib 200 mg bid, selenomethionine 200 mg qd, both agents, or matched placebos. After 10 months, repeat esophagogastroduodenoscopy showed no effects of celecoxib, but improvement in patients with mild dysplasia treated with selenomethionine.
Ongoing Clinical Trials of NSAIDs and COXIBs in Cancer Prevention
As discussed above, several NSAIDs are efficacious in reducing the burden of precancerous lesions in persons with FAP, prior colorectal adenomas, or actinic keratoses. COXIBs may benefit persons with these disorders as well, but most studies evaluating its effects are not yet completed. Several short- and long-term studies have been initiated in an attempt to corroborate preliminary evidence of NSAIDs/COXIBs efficacy in the colorectum, explore their usefulness in other organs, and more fully evaluate their safety in the setting of cancer prevention (table 5). Many of these trials have been initiated through collaborative agreements between the National Cancer Institute (NCI), and pharmaceutical companies such as Pharmacia/Pfizer and Merck. NSAIDs in Cancer Prevention Trials Sulindac was tested in some of the earliest human cancer prevention trials, as reviewed earlier in this chapter. COXIBs are a new research thrust in chemoprevention research, however, NSAIDs are still under intense evaluation in hypothesis-driven clinical trials – typically as a positive control. For example, a new phase II trial is evaluating aspirin or sulindac against prevalent colorectal ACF in order to confirm and extend prior observations regarding the potential chemopreventive efficacy of sulindac against ACF [79]. If this study confirms short-term effects of NSAIDs against ACF, and if longer-term trials of NSAIDs also demonstrate reductions in adenoma recurrence, then ACF might be useful as early monitors of chemopreventive response. If this is the case, ACF may serve as meaningful endpoints in phase II trials and thereby improve the efficiency of agent identification and prioritization for longer and costlier phase III trials focused on colorectal adenoma and cancer prevention [67, 68].
NSAIDs/COXIBs in Prevention Trials
223
Hawk/Viner/Umar
Table 5. Ongoing clinical chemoprevention trials with COX inhibitors and/or NSAID derivatives as of July 2002
224
Intervention(s)
Cohort
Control
Primary objective
Principal investigator(s)
Phase II trials Aspirin ⫹ calcium Budesonide Celecoxib Celecoxib Celecoxib Celecoxib Celecoxib Celecoxib Celecoxib Celecoxib Celecoxib ⫾ selenium Ketorolac Sulindac sulfone Sulindac, aspirin, urosodiol Sulindac, plant phenolics
Colorectum – Sporadic Lung Breast – genetic risk Colorectum – HNPCC Esophagus – Barrett’s Lung Lung Oral dysplasia Prostate Skin – genetic risk Esophagus – squamous Oral leukoplakia Duodenum – FAP Colorectum – sporadic Colorectum – sporadic
Placebo Placebo None Placebo Placebo Placebo Placebo Placebo Placebo Placebo Placebo Placebo Placebo Placebo None
Adenoma prevention Dysplasia regression/progression Biomarker modulation Biomarker modulation Dysplasia regression/prevention Biomarker modulation Dysplasia regression/progression Dysplasia regression/progression Biomarker modulation Cancer suppression Dysplasia regression/prevention Leukoplakia regression Adenoma regression ACF regression Biomarker modulation
Wargovich/Sinicrope Lam Fabian Lynch Forastiere/Heath Kurie Mao Boyle Carducci Epstein/Bickers Dawsey/Wang Mulshine DiSario Sinicrope Shiff
Phase II/III trials Celecoxib Celecoxib Celecoxib Celecoxib ⫹ eflornithine Sulindac ⫹ DFMO Sulindac ⫹ eflornithine
Bladder Colorectum/duodenum – FAP Skin – sporadic Colorectum/duodenum – FAP Colorectum – sporadic Colorectum – sporadic
Placebo Placebo Placebo Celecoxib Placebo Placebo
Cancer prevention Adenoma prevention AK regression/prevention Adenoma regression Adenoma prevention Adenoma prevention
Sabichi Lynch Elmets Sinicrope Meyskens Meyskens
NSAIDs/COXIBs in Prevention Trials
Phase III trials Aspirin Aspirin ⫾ folate Aspirin ⫾ folic acid Aspirin ⫾ resistant starch Aspirin ⫾ resistant starch Celecoxib Celecoxib ⫾ selenium Piroxicam ⫾ calcium Rofecoxib
Colorectum – sporadic Colorectum – sporadic Colorectum – sporadic (previous adenomas) Colorectum – FAP Colorectum – HNPCC Colorectum – sporadic Colorectum – sporadic Colorectum – sporadic Colorectum – sporadic
Placebo Placebo Placebo
Adenoma prevention Adenoma prevention Adenoma prevention
Sandler Baron Logan
Placebo Placebo Placebo Placebo Placebo Placebo
Adenoma prevention Adenoma prevention Adenoma prevention Adenoma prevention Adenoma prevention Adenoma prevention
Burn Burn Bertagnolli Alberts Berkel Mortensen
ACF ⫽ Aberrant crypt foci; AK ⫽ actinic keratosis; FAP ⫽ familial adenomatous polyposis; HNPCC ⫽ hereditary non-polyposis colorectal cancer; sporadic ⫽ sporadic colorectal cancer; squamous ⫽ squamous cell cancer of the esophagus.
225
Preliminary phase III data of Baron and Benamouzig are building the premise for the utility of NSAIDs in cancer prevention. Another important phase III trial of aspirin (325 mg qd) versus placebo in adenoma prevention recently completed accrual. This CALGB Consortium trial is studying a cohort of nearly 900 patients with definitively-treated, early stage colorectal cancer. Three additional aspirin studies are ongoing in Europe. The first of these is a factorial trial of aspirin with or without folic acid versus placebo in 1,000 patients with prior adenomas. Two other trials are testing aspirin combined with resistant starch in patients harboring APC or HNPCC-related mutations [29]. Looking beyond single agents, two phase III trials are evaluating the efficacy of NSAID combinatorial chemoprevention. In one placebo-controlled adenoma prevention study, investigators are evaluating the activity and safety of sulindac as a single agent or combined with eflornithine, an ornithine decarboxylase inhibitor. Both ornithine decarboxylase [80, 81] and COX-2 [82, 83] are overexpressed in colorectal neoplasia, and in vitro studies suggest significant preventive effects of inhibitors directed against each target [84]. In addition, five in vivo studies have demonstrated additive or supra-additive effects when various NSAIDs are co-administered with an ornithine decarboxylase inhibitor, even with significant dose reductions of one or both agents [85–89]. Based on the independent effects of NSAIDs and calcium against colorectal neoplasia, a phase III, 2 ⫻ 2 factorial trial of piroxicam and/or calcium in persons with prior sporadic adenomas has recently been initiated. Finally, NSAIDs are being evaluated in several other phase II studies. For example, DiSario is leading an effort to study the efficacy of sulindac sulfone versus placebo in reducing the duodenal adenoma burden of approximately 100 patients with FAP. Two other phase IIb trials are exploring the effects and toxicity of topically-delivered anti-inflammatory drugs. At the NCI, investigators are evaluating a topical (i.e., swish and spit) solution of ketorolac in persons with oral leukoplakia. Yet another NCI-contracted trial is testing the efficacy and tolerability of inhaled budesonide (Pulmicort™ a glucocorticoid marketed by AstraZeneca that effectively inhibits COX expression and activity) versus placebo in regressing and/or preventing progression of prevalent bronchial dysplasia in current or former smokers. COXIBs in Colorectal Cancer Prevention Trials To corroborate and extend the clinical database on celecoxib in persons with FAP, three additional studies are underway. The first is a small phase I celecoxib dose-finding study that will also assess short-term agent safety in genotype-positive children with FAP. A subsequent, larger phase II/III trial will evaluate whether celecoxib can suppress or delay the phenotypic expression of colorectal neoplasia in more than 200 FAP genotype-positive, phenotype-negative
Hawk/Viner/Umar
226
children. If celecoxib is shown to suppress or delay FAP phenotypic expression, this may allow deferment of colectomy until adulthood or reduce the extent of this surgery. The third trial is being conducted in adults with FAP, and investigates the combined effect of celecoxib administered with eflornithine (see above) on colorectal and duodenal adenomas. This combination may improve the efficacy and/or safety of one or both agents in the duodenum and/or colorectum, and promote the identification and development of newer agents and combinations that might reduce or delay the need for primary or secondary surgeries, lengthen surveillance intervals, and possibly reduce cancer incidence and mortality. An NCI, DCP-sponsored phase II trial in a hereditary non-polyposis colorectal cancer (HNPCC) cohort is evaluating the safety and biomarker-based efficacy of celecoxib administered daily for 12 months. While the trial is primarily focused on molecular markers, it may also generate data on the effects of celecoxib against prevalent ACF, and in a few instances, on prevalent adenomas, thereby permitting comparisons between cohorts with primary defects in the APC gene versus defects in mismatch repair, which are among the best characterized and distinctive pathophysiologies underlying colorectal carcinogenesis. The chemopreventive efficacy of daily dose celecoxib against sporadic colorectal carcinogenesis is under evaluation in two ongoing phase III NCIsponsored studies. The first study is evaluating approximately 2,000 patients with a personal history of colorectal adenoma(s). The second, a 2 ⫻ 2 factorial, placebo-controlled study, compares the effects of celecoxib, selenomethionine, or both agents in combination with regard to adenoma recurrence. In addition, both Pharmacia/Pfizer and Merck are conducting independent, phase III adenoma prevention trials with celecoxib or rofecoxib, respectively. In aggregate, these trials should provide a robust assessment of the preventive value of COXIBs against colorectal neoplasia. COXIBs in Extracolonic Cancer Prevention Trials As reviewed in other chapters, the chemopreventive efficacy of selective COXIBs has been demonstrated in a wide range of preclinical models of cancer prevention, including mammary [21, 22], skin [23–25], bladder [26, 27], esophagus [8], lung [28] and oral cavity [7] cancer. In addition, COX-2 is overexpressed in many human cancers and epidemiologic data suggest that COX inhibitors may have protective effects against carcinogenesis in several of these extracolonic sites. Ongoing NCI, DCP-sponsored phase II studies are building on these preclinical and observational data by examining the efficacy of celecoxib in a variety of target organs (table 5). Celecoxib (200–400 mg bid) will be administered
NSAIDs/COXIBs in Prevention Trials
227
for periods ranging from 6 to 24 months in nine different studies, most of which are placebo-controlled. The at-risk cohorts include patients with prior bladder cancer, actinic keratoses, basal cell nevi, bronchial meta-/dysplasia, oral precancers, and Barrett’s dysplasia. Other phase II trials include a biomarker study conducted in presurgical prostate cancer patients, and another biomarker study in UV radiation-induced skin lesions. Another investigation will examine the effect of celecoxib on proliferation-related biomarkers in breast carcinogenesis.
The Potential and Implications of NSAIDs and COXIBs in Clinical Prevention
The utility of medical interventions is defined by clinical demonstration of intended efficacy. However, broader considerations factor in as well, such as the agent’s ancillary benefits, toxicities, availability; and its medical, surgical and behavioral costs relative to – or in addition to – other available options. Given the ubiquity of COX and prostanoids in human tissues (and their promiscuity of physiologic and pathophysiologic actions), the demonstration of clinical utility of NSAIDs and COXIBs for cancer prevention will compel complex and challenging assessments of their aggregate risks and benefits in individual patients. Efficacy and Safety: Inseparable Considerations for Agent Development The primary goal of drug development is to identify plausible agents that modulate disease or disease symptoms. Once this is established, the utility of an agent (or agent combinations) is defined by the therapeutic ratio (benefit-risk equation) in a particular clinical setting. While it is ideal to identify agents with potent efficacy and little or no toxicity, even agents with middling activity may prove clinically useful if they modulate cancer risk with almost no toxicity or cost. By contrast, unacceptable toxicity (in terms of type, grade, frequency, or duration) or inordinate cost may render even highly effective agents useless for standard clinical applications. In other words, the goal of drug development is to identify effective, affordable, and well-tolerated agents. Each of these factors is a function of the proposed clinical application (e.g., the intended cohort and their level of risk, threshold for tolerating drug side effects or inconvenience, ability to afford the intervention, etc.). Toxicities of NSAIDs The gastrointestinal toxicities of NSAIDs are associated with substantial morbidity and mortality [90]. For example, an estimated 100,000 hospitalizations and 16,500 deaths per year result from NSAID use in the USA [91].
Hawk/Viner/Umar
228
Approximately 10–20% of NSAID users develop dyspepsia [92], and another 1% have serious gastrointestinal complications [93]. NSAIDs as a class have also been associated with an increased risk of bleeding, edema, hypertension and renal insufficiency [94, 95]. Most of these toxicities tend to occur in the elderly, particularly in the contexts of higher NSAID doses, concomitant use of corticosteroids or anticoagulants, and co-morbid conditions [96]. Toxicities of COXIBs – Established and Emerging COXIBs were designed to be safer NSAID alternatives – in a sense recapitulating aspirin’s development as a safer form of salicylic acid one century earlier. Even so, the side effect profile of COXIBs is akin to that of traditional NSAIDs, including renal, hepatic and gastrointestinal toxicities [97]. Nevertheless, the safety profile of COXIBs – in particular, the risk of gastrointestinal side effects – appears much improved as compared to that of nonselective NSAIDs [97]. The most definitive data for this claim derive from two large clinical trials – the VIGOR (Vioxx Gastrointestinal Outcomes Research) [98] and the CLASS (Celecoxib Long-Term Arthritis Safety Study) trials [99]. The VIGOR trial compared the incidence of gastrointestinal perforation, hemorrhage, or symptomatic peptic ulcer associated with rofecoxib (50 mg qd) or naproxen (500 mg bid) in more than 8,000 rheumatoid arthritis patients treated for a median of 9 months. While the drugs exhibited comparable efficacy, rofecoxib was associated with a 54% reduction in the safety-related events of interest (i.e., 2.1 vs. 4.5 per 100 patient years). An overview analysis of gastrointestinal side effects in osteoarthritis patients treated with rofecoxib vs. NSAIDs reported similar results [100]. The CLASS trial compared ulcer-related complications in patients with osteoarthritis or rheumatoid arthritis who were treated with celecoxib (400 mg bid), diclofenac (75 mg bid), or ibuprofen (800 mg tid) and followed for 12–15 months. While the primary analysis showed no significant difference between treatment arms with regard to the occurrence of ulcer-related complications, analysis of ulcer-related complications over a shorter interval (the first 6 months), and yet another analysis focused on ulcer complications plus symptomatic ulcers revealed significant 40–50% reductions in patients treated with celecoxib [99, 101]. A notable difference between the CLASS and VIGOR trials was that 21% of CLASS study participants also took aspirin (up to 325 mg/day). Post-marketing data suggest that celecoxib’s renal side effect profile is comparable to that of non-selective NSAIDs, with a 2.1 and 0.8% incidence of edema and hypertension, respectively [94]. Exacerbation of baseline hypertension has also been reported in 0.6% of patients treated with celecoxib. Emerging post-marketing data on rofecoxib also suggest renal side effects,
NSAIDs/COXIBs in Prevention Trials
229
as evident from a 3.8% incidence of peripheral edema [reviewed in 97]. Several cases of acute renal failure associated with COXIBs have been reported, though in some instances this effect resolved upon discontinuation of the agent [102, 103]. As with any chronically administered agent, the potential for drugdrug interactions presents additional safety concerns. For example, celecoxib inhibits cytochrome P450 CYP2C9, which is known to metabolize a number of drugs, including warfarin. Recent reports have described an interaction between warfarin and celecoxib [104, 105], and warfarin is contraindicated with rofecoxib. Lithium and fluconazole can potentially interact with celecoxib. As additional post-marketing data accrue, a more complete profile of drugs that should not be used in combination with COXIBs will likely emerge. More recently, concerns have been raised that COXIBs might have the potential to promote cardiovascular events. These concerns are based on mechanistic hypotheses that COXIBs may inhibit the production of prostacyclin, but not thromboxane, creating an imbalance between vasodilatory and vasoconstrictive influences [97]. Indeed, in the VIGOR study, myocardial infarction (MI) was four times more common with rofecoxib use as compared to naproxen (0.4 vs. 0.1%; RR ⫽ 0.2, 95% CI, 0.1–0.7), prompting the FDA and Merck to strengthen sections of the Vioxx™ (Merck, Whitehouse Station, N.J., USA) label with regard to patients with a medical history of ischemic heart disease [see http://www.fda.gov/medwatch/SAFETY/2002/safety02.htm#vioxx]. In addition, a recent comparison of cardiovascular data from the CLASS and VIGOR studies [106] to data from placebo patients in a meta-analysis of primary cardiovascular prevention trials with aspirin [107] found that both COXIBs were associated with a 40–50% increase in the annualized rates of MIs (i.e., celecoxib ⫽ 0.8 (p ⫽ 0.02); rofecoxib ⫽ 0.74 (p ⫽ 0.04); versus placebo ⫽ 0.52). These results are best viewed as hypothesis-generating, rather than a definitive test of a hypothesis, however, due to a variety of study limitations (i.e., the small absolute number of cardiovascular events (⬍70), substantial heterogeneity among the three study populations (with regard to aspirin use, rheumatoid versus osteoarthritis patients versus primary prevention trial participants), and issues with the statistical methods that were employed). Clearly, the spectrum and magnitude of cardiovascular risk associated with COXIBs, and the potential mitigating effects of low-dose aspirin, require further investigation [108]. Data on COXIBs from placebo-controlled trials – including ongoing adenoma prevention trials – will improve our ability to characterize and quantify their potential for increasing cardiovascular risk. In addition, Merck has initiated a large trial specifically intended to evaluate the cardiovascular safety of rofecoxib [109].
Hawk/Viner/Umar
230
Approaches to Improve the TI of NSAIDs and COXIBs One of the major challenges to NSAID and COXIB development for cancer prevention involves matching up efficacious and safe preventive regimens with patients that are most likely to benefit from them. Key determinants of success in this endeavor include the relative pharmacokinetic, phamacodynamic and mechanistic properties of the agents – including a propensity for collateral effects on unintended targets – and the needs, susceptibilities and tolerances of the cohorts in which they are administered. Practical means for improving the TI of NSAIDs and COXIBs include: (1) identifying and applying agents with greater anti-neoplastic specificity; (2) combining them with other classes of agents that mutually boost efficacy (via serial or parallel synergies), reduce toxicities, and/or allow for marked dose reduction with comparable efficacy of one or both agents; (3) administering them locally or regionally, rather than systemically, and (4) restricting their use to cohorts that are more likely derive benefit from them and/or more tolerant of their toxicities. Each of these four strategies has the potential to optimize and expedite the development, testing and application of candidate agents in appropriate risk cohorts. Two clinical anecdotes of this principle are described below. NSAIDs and COXIBs in Chemopreventive Combinations Combinatorial therapy is particularly appealing when serial or parallel synergies in efficacy, or reductions in toxicity, can be anticipated. In a model of serial synergy, one might apply NSAIDs or COXIBs in combination with an agent that targets up- or downstream components of the COX pathway in order to further suppress COX expression or activity. For example, an agent given to promote serial synergism might reduce the concentration of arachidonic acid (AA) available to COX by inhibiting phospholipase A2 (PLA2). Indeed, dietary supplementation with eicosapentaenoic acid competes with AA for incorporation into certain phospholipids, and has been shown to reduce adenoma number by 68% in ApcMIN mice [110]. These antitumor effects, however, were reversed by additional dietary AA. Similarly, knocking out cytoplasmic PLA2 in Apc D716 mice reduced the size, although not the number, of adenomas [111]. Alternatively, serial synergism could be accomplished by blocking the interactions of COX-derived prostanoids with their receptors, although it is unclear exactly which EP receptors might be critical to carcinogenesis. Recently, homozygous deletion of the gene encoding the EP2 cell-surface receptor for PGE2 was shown to decrease the number and size of intestinal adenomas in Apc D716 mice [112]. EP2 receptor gene mutation also attenuated microvessel density in Apc D716 mice [113]. Alternatively, recent in vitro studies have suggested a role for EP4 in the increased cell proliferation and motility of
NSAIDs/COXIBs in Prevention Trials
231
colon cancer cells following PGE2 treatment [114]. Finally, the EP1 receptor may also be important, as demonstrated by a recent experiment in which an EP1 receptor antagonist, ONO-8711, decreased ACF incidence in AOM-treated C57BL/6J mice in a dose-dependent manner, and reduced intestinal adenoma formation by 43% in ApcMIN mice [115]. Comparable results were found with ONO-8713, a more selective EP1 receptor antagonist [116]. Parallel synergies may result from the co-administration of NSAIDs/ COXIBs with effective chemopreventives directed against entirely different molecular targets. For example, the combination of COXIBS with agents that exhibit complementary mechanisms of action (e.g., antiproliferatives, inducible nitric oxide synthase inhibitors [117]) has shown considerable preclinical promise. For the last 15 years, a series of preclinical studies have demonstrated significant cooperative chemopreventive effects of NSAIDs placed in combination with eflornithine, an irreversible inhibitor of ornithine decarboxylase [85–89]. As noted above, these results are now being translated into clinic trials (table 5). Finally, NSAIDs/COXIBs may be given in combination with drugs that block some of their toxicities in order to improve the TI. Proton pump inhibitors, such as omeprazole, are reasonably safe, inexpensive and apparently mitigate NSAID-associated upper gastrointestinal toxicity [118]. High doses of H2 blockers may also reduce the upper gastrointestinal side effects of NSAIDs. NSAIDs and COXIBs in Specific Risk Cohorts Selection of cohorts suitable for NSAID- or COXIB-based chemoprevention poses unique challenges. An acceptable risk-benefit ratio requires scrutiny of three recipient susceptibilities: (1) cancer risk; (2) toxicity risk relating to the agent, and (3) the likelihood of benefiting from the agent’s intended effects. An individual’s risk of cancer will dictate whether he or she is inclined to undertake preventive measures and tolerate their side effects. For example, a patient with a germline defect in a tumor suppressor gene, such as APC, that portends an extremely high lifetime cancer risk, substantial morbidities from prophylactic surgical procedures, and the possibility of transmitting their disease to subsequent generations, may be highly motivated to explore chemopreventive options despite significant (in terms of frequency, severity, or duration) side effects. By contrast, an individual at modestly increased risk for cancer due to prior preinvasive neoplastic lesions or to an effectively-treated primary cancer may be somewhat less motivated and/or tolerant of side effects. A patient at average risk may not be interested at all, unless the chemopreventive had little or no chance of causing side effects. Interest in NSAID- or COXIB-based chemoprevention is also a function of individual susceptibility to these agents’ toxicities. We know that NSAIDs and
Hawk/Viner/Umar
232
COXIBs may pose health risks to certain individuals, including those susceptible to renal or cardiovascular disease. The incidence of serious gastrointestinal side effects also depends upon pre-existing risk factors. In fact, a history of peptic ulcer disease and gastrointestinal bleeding increases the risk of complicated ulcer in NSAID users approximately 12-fold, from 0.4% in patients without such history to about 5% per year [119]. Finally, chemopreventive efficacy of NSAIDs and COXIBs is not uniform among individuals, or even across target organs within an individual. In the best case, available evidence from large observational studies reveals only 40–50% reductions in colorectal neoplasia incidence with NSAIDs [5]. Therefore, while some individuals may experience complete protection from neoplastic lesions, others probably derive only partial or no protection. Currently, it is impossible to predict who may and who may not benefit from these agents, therefore many patients are treated while only a few are actually protected. In a therapeutic setting, this problem can be addressed by evaluating target status. For example, only a minority of breast cancers appear to overexpress COX-2. Emerging evidence from Dannenberg et al. [reviewed in 120] suggests that breast cancers overexpressing HER-2/neu may be more likely to overexpress COX-2, potentially identifying a subset of patients that are more likely to respond to treatment with NSAIDs/COXIBs. Since the tissue of concern – preinvasive neoplastic lesions or cancer itself – is not yet present in most preventive settings, comparable estimates related to molecular targets are usually unavailable, and thus it is impossible to more effectively identify likely responders to treatment in a similar manner. New approaches characterizing polymorphisms in CYP2C9, which catalyzes the inactivation of aspirin and other NSAIDs/COXIBs, may yield valuable insights into individuals who may or may not benefit from preventive interventions [121]. Nevertheless, until such approaches are validated, the decision to use or not use an NSAID or COXIB for cancer prevention will remain an inexact and difficult proposition. The Potential Impact of NSAIDs and COXIBs in Cancer Prevention and Beyond Two recent cost-effectiveness studies highlight challenges that must be surmounted if the promise of NSAIDs/COXIBs for the chemoprevention of colorectal cancer is to be realized [122, 123]. Both of these analyses evaluated the cost-effectiveness of aspirin, versus colorectal screening alone, versus aspirin and screening in the general US population over age 50 using computer simulations (i.e., Markov models). In both analyses, screening was found to save more lives at a lower cost than aspirin alone. In addition, aspirin did not improve the benefits of screening, in large part owing to the impact of
NSAIDs/COXIBs in Prevention Trials
233
aspirin-related complications. These reports concluded that aspirin could not be substituted for screening in this setting. Of course, these sorts of analyses are inherently inexact due to the large number of variables that must be estimated. At best, they serve as models of what might happen under the prescribed conditions. Nevertheless, the value of these models is substantial. One of the best uses of cost-effectiveness analyses is exploration of the relative impact of certain variables on the outcomes of interest, which might guide future research and agent prioritization. Toward that end, the results of analyses cited above were sensitive to the magnitude of risk reduction with aspirin, aspirin-related complication rates, and screening adherence rates [122, 123]. In the near term, aspirin and other NSAID/COXIB chemopreventive regimens are unlikely to be administered in the general population, for all of the reasons identified by the Markov models. Indeed, in our estimation, aspirin is most likely to be integrated into cancer prevention much as it was in cardiovascular prevention. Aspirin was first demonstrated to reduce the risk of cardiovascular (CV) events in a secondary prevention setting involving patients with prior MIs [124]. After many additional studies, aspirin is now used for primary prophylaxis, though only in persons with a greater than 1% annual risk of having a CV event [107]. In persons with a CV risk between 0.5 and 1%, the TI is much more tenuous, and decisions regarding aspirin use for CV disease prophylaxis requires careful consideration by the patient and his/her health professional. Finally, in individuals with less than a 0.5% annual risk, the risks of using aspirin for CV disease prevention clearly outweigh its benefits. We have every reason to believe that a similar approach will eventually inform us as to which patients should or should not use aspirin with chemopreventive intent, based on definitive data on aspirin’s usefulness and safety in various clinical settings. We expect that chemopreventive applications of aspirin will initially be as adjuncts to standard management (i.e., endoscopic surveillance) of patients at moderate to high risk of cancer. If proven effective and reasonably safe in this context, aspirin may then be explored for risk reduction in the setting of primary cancer prevention. More interesting, and perhaps likely, is the possibility that aspirin (or other NSAIDs/COXIBs) might simultaneously modulate the risk of many cancers, and possibly several chronic diseases of aging. This theory derives from the fact that the risk for most cancers and serious CV events – common afflictions of Western societies – increases with age. No doubt, this situation points to an underlying molecular pathophysiology shared by these (and possibly by other) seemingly disparate diseases [125]. Because a large fraction of the population at risk for CV events are probably also at risk for cancers of various sites, and vice versa, rarely would patients be taking a prophylactic agent that exclusively
Hawk/Viner/Umar
234
targeted one disease. Indeed, as our understanding of age-related diseases improves, we are finding many common pathophysiologic mechanisms or risk factors (e.g., age, diabetes, lack of physical activity, dietary factors, and folate deficiency) that may be targeted in order to achieve preventive effects far beyond cancer alone. COX plays an important role in chronic diseases of aging common to Western societies. Because we have more than a century of experience with aspirin – both from clinical trials in patients at low to moderate risk for several of these diseases, as well as anecdotally within large segments of the population – aspirin is likely to remain the lead preventive agent for COX targeting. Of course, aggressive marketing of novel compounds with greater proprietary interest may skew agent selection away from aspirin. Even so, given the many financial pressures on medicine in the 21st century, as well as public and prescribing sector familiarity, aspirin is likely to remain an important agent for the near future. Advantages of aspirin over other NSAID/COXIB include: (1) well-established use in the prevention of CV events, and the treatment of arthritis, pain, fever and other inflammatory conditions; (2) detailed safety profiles of short- and long-term risks in various cohorts; (3) low cost; (4) potential usefulness in preventing or treating other common diseases of aging, such as Alzheimer’s disease, cataracts, osteoporosis and cancer of several different organs, and (5) knowledge of how to prophylax against its side effects using inexpensive agents (e.g., proton pump inhibitors or high doses of H2 blockers) [118]. In the future, aspirin, NSAIDs and/or COXIBs will likely be applied with preventive intent against a constellation of risks quantified by complex genomic [126] and proteomic analyses [127, 128]. The question will not be, ‘Can I prevent this cancer’ but ‘Can I reduce my risk of diseases “A through K” by taking a chemopreventive?’ For this reason, aspirin and COXIBs are poised to play key roles in the near- and long-term prevention of cancer and a broadening range of other diseases.
References 1 2
3
Staessen JA: Definition of new targets in cardiovascular prevention from young into old age. Eur Heart J 2002;23:507–509. O’Shaughnessy JA, Kelloff GJ, Gordon GB, Dannenberg AJ, Hong WK, Fabian CJ, Sigman CC, Bertagnolli MM, Stratton SP, Lam S, Nelson WG, Meyskens FL, Alberts DS, Follen M, Rustgi AK, Papadimitrakopoulou V, Scardino PT, Gazdar AF, Wattenberg LW, Sporn MB, Sakr WA, Lippman SM, Von Hoff DD: Treatment and prevention of intraepithelial neoplasia: An important target for accelerated new agent development. Clin Cancer Res 2002;8:314–346. Teicher BA: Molecular targets and cancer therapeutics: Discovery, development and clinical validation. Drug Resist Updat 2000;3:67–73.
NSAIDs/COXIBs in Prevention Trials
235
4 5 6 7
8
9 10 11
12
13 14
15
16
17
18
19
20
21 22
Marnett LJ, DuBois RN: COX-2: A target for colon cancer prevention. Annu Rev Pharmacol Toxicol 2002;42:55–80. Anderson WF, Umar A, Viner JL, Hawk ET: The role of cyclooxygenase inhibitors in cancer prevention. Curr Pharm Des 2002;8:99–110. Marnett LJ: Aspirin and the potential role of prostaglandins in colon cancer. Cancer Res 1992;52: 5575–5589. Shiotani H, Denda A, Yamamoto K, Kitayama W, Endoh T, Sasaki Y, Tsutsumi N, Sugimura M, Konishi Y: Increased expression of cyclooxygenase-2 protein in 4-nitroquinoline-1-oxide-induced rat tongue carcinomas and chemopreventive efficacy of a specific inhibitor, nimesulide. Cancer Res 2001;61:1451–1456. Li Z, Shimada Y, Kawabe A, Sato F, Maeda M, Komoto I, Hong T, Ding Y, Kaganoi J, Imamura M: Suppression of N-nitrosomethylbenzylamine-induced esophageal tumorigenesis in F344 rats by JTE-522, a selective COX-2 inhibitor. Carcinogenesis 2001;22:547–551. Reddy BS, Rao CV, Seibert K: Evaluation of cyclooxygenase-2 inhibitor for potential chemopreventive properties in colon carcinogenesis. Cancer Res 1996;56:4566–4569. Kawamori T, Rao CV, Seibert K, Reddy BS: Chemopreventive activity of celecoxib, a specific cyclooxygenase-2 inhibitor, against colon carcinogenesis. Cancer Res 1998;58:409–412. Reddy BS, Hirose Y, Lubet R, Steele V, Kelloff G, Paulson S, Seibert K, Rao CV: Chemoprevention of colon cancer by specific cyclooxygenase-2 inhibitor, celecoxib, administered during different stages of carcinogenesis. Cancer Res 2000;60:293–297. Jacoby RF, Seibert K, Cole CE, Kelloff G, Lubet RA: The cyclooxygenase-2 inhibitor celecoxib is a potent preventive and therapeutic agent in the Min mouse model of adenomatous polyposis. Cancer Res 2000;60:5040–5044. Sasai H, Masaki M, Wakitani K: Suppression of polypogenesis in a new mouse strain with a truncated Apc⌬474 by a novel COX-2 inhibitor, JTE-522. Carcinogenesis 2000;21:953–958. Oshima M, Dinchuk JE, Kargman SL, Oshima H, Hancock B, Kwong E, Trzaskos JM, Evans JF, Taketo MM: Suppression of intestinal polyposis in Apc⌬716 knockout mice by inhibition of cyclooxygenase 2. Cell 1996;87:803–809. Yoshimi N, Kawabata K, Hara A, Matsunaga K, Yamada Y, Mori H: Inhibitory effect of NS-398, a selective cyclooxygenase-2 inhibitor, on azoxymethane-induced aberrant crypt foci in colon carcinogenesis of F344 rats. Jpn J Cancer Res 1997;88:1044–1051. Fukutake M, Nakatsugi S, Isoi T, Takahashi M, Ohta T, Mamiya S, Taniguchi Y, Sato H, Fukuda K, Sugimura T, Wakabayashi K: Suppressive effects of nimesulide, a selective inhibitor of cyclooxygenase-2, on azoxymethane-induced colon carcinogenesis in mice. Carcinogenesis 1998; 19:1939–1942. Nakatsugi S, Fukutake M, Takahashi M, Fukuda K, Isoi T, Taniguchi Y, Sugimura T, Wakabayashi K: Suppression of intestinal polyp development by nimesulide, a selective cyclooxygenase-2 inhibitor, in Min mice. Jpn J Cancer Res 1997;88:1117–1120. Yoshimi N, Shimizu M, Matsunaga K, Yamada Y, Fujii K, Hara A, Mori H: Chemopreventive effect of N-(2-cyclohexyloxy-4-nitrophenyl)methane sulfonamide (NS-398), a selective cyclooxygenase-2 inhibitor, in rat colon carcinogenesis induced by azoxymethane. Jpn J Cancer Res 1999; 90:406–412. Oshima M, Murai N, Kargman S, Arguello M, Luk P, Kwong E, Taketo MM, Evans JF: Chemoprevention of intestinal polyposis in the Apc⌬716 mouse by rofecoxib, a specific cyclooxygenase-2 inhibitor. Cancer Res 2001;61:1733–1740. Lal G, Ash C, Hay K, Redston M, Kwong E, Hancock B, Mak T, Kargman S, Evans JF, Gallinger S: Suppression of intestinal polyps in Msh2-deficient and non-Msh2-deficient multiple intestinal neoplasia mice by a specific cyclooxygenase-2 inhibitor and by a dual cyclooxygenase-1/2 inhibitor. Cancer Res 2001;61:6131–6136. Harris RE, Alshafie GA, Abou-Issa H, Seibert K: Chemoprevention of breast cancer in rats by celecoxib, a cyclooxygenase-2 inhibitor. Cancer Res 2000;60:2101–2103. Matsunaga K, Yoshimi N, Yamada Y, Shimizu M, Kawabata K, Ozawa Y, Hara A, Mori H: Inhibitory effects of nabumetone, a cyclooxygenase-2 inhibitor, and esculetin, a lipoxygenase inhibitor, on N-methyl-N-nitrosourea-induced mammary carcinogenesis in rats. Jpn J Cancer Res 1998;89:496–501.
Hawk/Viner/Umar
236
23
24 25
26
27
28 29 30 31 32 33 34
35
36 37
38 39 40
41 42 43 44 45 46
Fischer SM, Lo HH, Gordon GB, Seibert K, Kelloff G, Lubet RA, Conti CJ: Chemopreventive activity of celecoxib, a specific cyclooxygenase-2 inhibitor, and indomethacin against ultraviolet light-induced skin carcinogenesis. Mol Carcinog 1999;25:231–240. Pentland AP, Schoggins JW, Scott GA, Khan KN, Han R: Reduction of UV-induced skin tumors in hairless mice by selective COX-2 inhibition. Carcinogenesis 1999;20:1939–1944. Muller-Decker K, Kopp-Schneider A, Marks F, Seibert K, Furstenberger G: Localization of prostaglandin H synthase isoenzymes in murine epidermal tumors: Suppression of skin tumor promotion by inhibition of prostaglandin H synthase-2. Mol Carcinog 1998;23:36–44. Grubbs CJ, Lubet RA, Koki AT, Leahy KM, Masferrer JL, Steele VE, Kelloff GJ, Hill DL, Seibert K: Celecoxib inhibits N-butyl-N-(4-hydroxybutyl)nitrosamine-induced urinary bladder cancers in male B6D2F1 mice and female Fischer-344 rats. Cancer Res 2000;60:5599–5602. Okajima E, Denda A, Ozono S, Takahama M, Akai H, Sasaki Y, Kitayama W, Wakabayashi K, Konishi Y: Chemopreventive effects of nimesulide, a selective cyclooxygenase-2 inhibitor, on the development of rat urinary bladder carcinomas initiated by N-butyl-N-(4-hydroxybutyl) nitrosamine. Cancer Res 1998;58:3028–3031. Rioux N, Castonguay A: Prevention of NNK-induced lung tumorigenesis in A/J mice by acetylsalicylic acid and NS-398. Cancer Res 1998;58:5354–5360. Thun MJ, Henley SJ, Patrono C: Nonsteroidal anti-inflammatory drugs as anticancer agents: Mechanistic, pharmacologic and clinical issues. J Natl Cancer Inst 2002;94:252–266. Waddell WR, Loughry RW: Sulindac for polyposis of the colon. J Surg Oncol 1983;24:83–87. Gonzaga RA, Lima FR, Carneiro S, Maciel J, Amarante Junior M: Sulindac treatment for familial polyposis coli. Lancet 1985;i:751. Charneau J, D’Aubigny N, Burtin P, Person B, Boyer J: Rectal micropolyps after total colectomy in familial polyposis. Efficacy of sulindac (in French). Gastroentérol Clin Biol 1990;14:153–157. Friend WG: Sulindac suppression of colorectal polyps in Gardner’s syndrome. Am Fam Physician 1990;41:891–894. Giardiello FM, Spannhake EW, DuBois RN, Hylind LM, Robinson CR, Hubbard WC, Hamilton SR, Yang VW: Prostaglandin levels in human colorectal mucosa: Effects of sulindac in patients with familial adenomatous polyposis. Dig Dis Sci 1998;43:311–316. Muller A, Hurlimann R, Meyenberger C, Staub P, Kobler E, Ammann R: Sulindac in familial adenomatous polyposis coli – Preliminary findings of a prospective study (in German). Schweiz Med Wochenschr 1994;124:651–654. Rigau J, Pique JM, Rubio E, Planas R, Tarrech JM, Bordas JM: Effects of long-term sulindac therapy on colonic polyposis. Ann Intern Med 1991;115:952–954. Spagnesi MT, Tonelli F, Dolara P, Caderni G, Valanzano R, Anastasi A, Bianchini F: Rectal proliferation and polyp occurrence in patients with familial adenomatous polyposis after sulindac treatment. Gastroenterology 1994;106:362–366. Tonelli F, Valanzano R, Dolara P: Sulindac therapy of colorectal polyps in familial adenomatous polyposis. Dig Dis 1994;12:259–264. Tonelli F, Valanzano R: Sulindac in familial adenomatous polyposis. Lancet 1993;342:1120. Cruz-Correa M, Hylind LM, Romans KE, Booker SV, Giardiello FM: Long-term treatment with sulindac in familial adenomatous polyposis: A prospective cohort study. Gastroenterology 2002; 122:641–645. Niv Y, Fraser GM: Adenocarcinoma in the rectal segment in familial polyposis coli is not prevented by sulindac therapy. Gastroenterology 1994;107:854–857. Thorson AG, Lynch HT, Smyrk TC: Rectal cancer in FAP patient after sulindac. Lancet 1994; 343:180. Lynch HT, Thorson AG, Smyrk T: Rectal cancer after prolonged sulindac chemoprevention. A case report. Cancer 1995;75:936–938. Matsuhashi N, Nakajima A, Shinohara K, Oka T, Yazaki Y: Rectal cancer after sulindac therapy for a sporadic adenomatous colonic polyp. Am J Gastroenterol 1998;93:2261–2266. Iwama T, Imajo M, Mishima Y: Is sulindac effective for familial adenomatous polyposis? Int J Colorectal Dis 1991;6:235. Hirata K, Itoh H, Ohsato K: Regression of rectal polyps by indomethacin suppository in familial adenomatous polyposis. Report of two cases. Dis Colon Rectum 1994;37:943–946.
NSAIDs/COXIBs in Prevention Trials
237
47
48
49
50
51
52
53
54
55
56
57 58 59 60
61
62
63
64
Hirota C, Iida M, Aoyagi K, Matsumoto T, Tada S, Yao T, Fujishima M: Effect of indomethacin suppositories on rectal polyposis in patients with familial adenomatous polyposis. Cancer 1996; 78:1660–1665. Van Stolk R, Stoner G, Hayton WL, Chan K, DeYoung B, Kresty L, Kemmenoe BH, Elson P, Rybicki L, Church J, Provencher K, McLain D, Hawk E, Fryer B, Kelloff G, Ganapathi R, Budd GT: Phase I trial of exisulind (sulindac sulfone, FGN-1) as a chemopreventive agent in patients with familial adenomatous polyposis. Clin Cancer Res 2000;6:78–89. Winde G, Gumbinger HG, Osswald H, Kemper F, Bunte H: The NSAID sulindac reverses rectal adenomas in colectomized patients with familial adenomatous polyposis: Clinical results of a dose-finding study on rectal sulindac administration. Int J Colorectal Dis 1993;8:13–17. Winde G, Schmid KW, Schlegel W, Fischer R, Osswald H, Bunte H: Complete reversion and prevention of rectal adenomas in colectomized patients with familial adenomatous polyposis by rectal low-dose sulindac maintenance treatment. Advantages of a low-dose nonsteroidal antiinflammatory drug regimen in reversing adenomas exceeding 33 months. Dis Colon Rectum 1995;38:813–830. Winde G, Schmid KW, Brandt B, Muller O, Osswald H: Clinical and genomic influence of sulindac on rectal mucosa in familial adenomatous polyposis. Dis Colon Rectum 1997;40: 1156–1169. Labayle D, Fischer D, Vielh P, Drouhin F, Pariente A, Bories C, Duhamel O, Trousset M, Attali P: Sulindac causes regression of rectal polyps in familial adenomatous polyposis. Gastroenterology 1991;101:635–639. Nugent KP, Farmer KC, Spigelman AD, Williams CB, Phillips RK: Randomized controlled trial of the effect of sulindac on duodenal and rectal polyposis and cell proliferation in patients with familial adenomatous polyposis. Br J Surg 1993;80:1618–1619. Giardiello FM, Hamilton SR, Krush AJ, Piantadosi S, Hylind LM, Celano P, Booker SV, Robinson CR, Offerhaus GJ: Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med 1993;328:1313–1316. Giardiello FM, Yang VW, Hylind LM, Krush AJ, Petersen GM, Trimbath JD, Piantadosi S, Garrett E, Geiman DE, Hubbard W, Offerhaus GJ, Hamilton SR: Primary chemoprevention of familial adenomatous polyposis with sulindac. N Engl J Med 2002;346:1054–1059. Seow-Choen F, Vijayan V, Keng V: Prospective randomized study of sulindac versus calcium and calciferol for upper gastrointestinal polyps in familial adenomatous polyposis. Br J Surg 1996;83: 1763–1766. Richard CS, Berk T, Bapat BV, Haber G, Cohen Z, Gallinger S: Sulindac for periampullary polyps in FAP patients. Int J Colorectal Dis 1997;12:14–18. Parker AL, Kadakia SC, Maccini DM, Cassaday MA, Angueira CE: Disappearance of duodenal polyps in Gardner’s syndrome with sulindac therapy. Am J Gastroenterol 1993;88:93–94. Debinski HS, Trojan J, Nugent KP, Spigelman AD, Phillips RK: Effect of sulindac on small polyps in familial adenomatous polyposis. Lancet 1995;345:855–856. Calaluce R, Earnest DL, Heddens D, Einspahr JG, Roe D, Bogert CL, Marshall JR, Alberts DS: Effects of piroxicam on prostaglandin E2 levels in rectal mucosa of adenomatous polyp patients: A randomized phase IIb trial. Cancer Epidemiol Biomarkers Prev 2000;9:1287–1292. Chow HH, Earnest DL, Clark D, Mason-Liddil N, Kramer CB, Einspahr JG, Guillen-Rodriguez JM, Roe DJ, Malone W, Crowell JA, Alberts DS: Effect of subacute ibuprofen dosing on rectal mucosal prostaglandin E2 levels in healthy subjects with a history of resected polyps. Cancer Epidemiol Biomarkers Prev 2000;9:351–356. Barnes CJ, Hamby-Mason RL, Hardman WE, Cameron IL, Speeg KV, Lee M: Effect of aspirin on prostaglandin E2 formation and transforming growth factor-␣ expression in human rectal mucosa from individuals with a history of adenomatous polyps of the colon. Cancer Epidemiol Biomarkers Prev 1999;8:311–315. Ruffin MT 4th, Krishnan K, Rock CL, Normolle D, Vaerten MA, Peters-Golden M, Crowell J, Kelloff G, Boland CR, Brenner DE: Suppression of human colorectal mucosal prostaglandins: Determining the lowest effective aspirin dose. J Natl Cancer Inst 1997;89:1152–1160. Hixson LJ, Earnest DL, Fennerty MB, Sampliner RE: NSAID effect on sporadic colon polyps. Am J Gastroenterol 1993;88:1652–1656.
Hawk/Viner/Umar
238
65 66 67 68
69 70
71 72 73
74
75
76
77
78
79
80 81
82
83
84
Matsuhashi N, Nakajima A, Fukushima Y, Yazaki Y, Oka T: Effects of sulindac on sporadic colorectal adenomatous polyps. Gut 1997;40:344–349. Ladenheim J, Garcia G, Titzer D, Herzenberg H, Lavori P, Edson R, Omary MB: Effect of sulindac on sporadic colonic polyps. Gastroenterology 1995;108:1083–1087. Baron JA, Cole BF, Mott LA, Group ftPPS: Aspirin chemoprevention of colorectal adenomas. Proc Am Acad Cancer Res 2002;43:669. Benamouzig R, Deyra J, Martin A, Girard B, Piednoir B, Coste T, Couturier D, Chaussade S, Group FtAS: Efficacy of lysine acetylsalicylate to prevent colorectal adenomas: One year results of the APACC Randomized Trial. Gastroenterology 2002;122:A70. 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;85:1220–1224. Neufang G, Furstenberger G, Heidt M, Marks F, Muller-Decker K: Abnormal differentiation of epidermis in transgenic mice constitutively expressing cyclooxygenase-2 in skin. Proc Natl Acad Sci USA 2001;98:7629–7634. Reeve VE, Matheson MJ, Bosnic M, Boehm-Wilcox C: The protective effect of indomethacin on photocarcinogenesis in hairless mice. Cancer Lett 1995;95:213–219. Haedersdal M, Poulsen T, Wulf HC: Effects of systemic indomethacin on photocarcinogenesis in hairless mice. J Cancer Res Clin Oncol 1995;121:257–261. Smith KJ, Germain M, Skelton H: Bowen’s disease (squamous cell carcinoma in situ) in immunosuppressed patients treated with imiquimod 5% cream and a COX inhibitor, sulindac: Potential applications for this combination of immunotherapy. Dermatol Surg 2001;27:143–146. Rivers JK, McLean DI: An open study to assess the efficacy and safety of topical 3% diclofenac in a 2.5% hyaluronic acid gel for the treatment of actinic keratoses. Arch Dermatol 1997;133: 1239–1242. Dolara P, Caderni G, Tonelli F: Nimesulide, a selective anti-inflammatory cyclooxygenase-2 inhibitor, does not affect polyp number and mucosal proliferation in familial adenomatous polyposis. Scand J Gastroenterol 1999;34:1168. Steinbach G, Lynch PM, Phillips RK, Wallace MH, Hawk E, Gordon GB, Wakabayashi N, Saunders B, Shen Y, Fujimura T, Su LK, Levin B: The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–1952. Phillips RK, Wallace MH, Lynch PM, Hawk E, Gordon GB, Saunders BP, Wakabayashi N, Shen Y, Zimmerman S, Godio L, Rodrigues-Bigas M, Su LK, Sherman J, Kelloff G, Levin B, Steinbach G: A randomised, double-blind, placebo-controlled study of celecoxib, a selective cyclooxygenase-2 inhibitor, on duodenal polyposis in familial adenomatous polyposis. Gut 2002; 50:857–860. Limburg P, Wei W, Ahnen D, Qiao Y, Hawk E, Wang G, Giffen C, Roth M, Korn E, Dong Z, Taylor P, Dawsey S: Chemoprevention of esophageal squamous cancer: Randomized, placebocontrolled trial in a high-risk population. Gastroenterology 2002;122:A71. Takayama T, Katsuki S, Takahashi Y, Ohi M, Nojiri S, Sakamaki S, Kato J, Kogawa K, Miyake H, Niitsu Y: Aberrant crypt foci of the colon as precursors of adenoma and cancer. N Engl J Med 1998;339:1277–1284. Giardiello FM, Hamilton SR, Hylind LM, Yang VW, Tamez P, Casero RA Jr: Ornithine decarboxylase and polyamines in familial adenomatous polyposis. Cancer Res 1997;57:199–201. Patchett SE, Alstead EM, Trzeciak L, Wocial T, Phillips RK, Farthing MJ: Rectal mucosal ornithine decarboxylase activity in familial adenomatous polyposis after ileorectal anastomosis. Gut 1994;35:1624–1626. Khan KN, Masferrer JL, Woerner BM, Soslow R, Koki AT: Enhanced cyclooxygenase-2 expression in sporadic and familial adenomatous polyposis of the human colon. Scand J Gastroenterol 2001;36:865–869. Humar B, D’Orazio D, Albrecht C, Bauerfeind P, Muller H, Dobbie Z, Bendik I: Expression of putative anticancer targets in familial adenomatous polyposis and its association with the APC mutation status. Int J Oncol 2001;19:1179–1186. Katdare M, Kopelovich L, Telang N: Chemopreventive agents inhibit aberrant proliferation of the aneuploid phenotype in a colon epithelial cell line established from Apc 1638N [+/⫺] mouse. Ann N Y Acad Sci 2001;952:169–174.
NSAIDs/COXIBs in Prevention Trials
239
85 Nigro ND, Bull AW, Boyd ME: Inhibition of intestinal carcinogenesis in rats: Effect of difluoromethylornithine with piroxicam or fish oil. J Natl Cancer Inst 1986;77:1309–1313. 86 Jacoby RF, Cole CE, Tutsch K, Newton MA, Kelloff G, Hawk ET, Lubet RA: Chemopreventive efficacy of combined piroxicam and difluoromethylornithine treatment of Apc mutant Min mouse adenomas, and selective toxicity against Apc mutant embryos. Cancer Res 2000;60:1864–1870. 87 Li H, Schut HA, Conran P, Kramer PM, Lubet RA, Steele VE, Hawk EE, Kelloff GJ, Pereira MA: Prevention by aspirin and its combination with ␣-difluoromethylornithine of azoxymethaneinduced tumors, aberrant crypt foci and prostaglandin E2 levels in rat colon. Carcinogenesis 1999; 20:425–430. 88 Rao CV, Tokumo K, Rigotty J, Zang E, Kelloff G, Reddy BS: Chemoprevention of colon carcinogenesis by dietary administration of piroxicam, ␣-difluoromethylornithine, 16␣-fluoro-5-androsten17-one, and ellagic acid individually and in combination. Cancer Res 1991;51:4528–4534. 89 Reddy BS, Nayini J, Tokumo K, Rigotty J, Zang E, Kelloff G: Chemoprevention of colon carcinogenesis by concurrent administration of piroxicam, a nonsteroidal anti-inflammatory drug with D, L-␣-difluoromethylornithine, an ornithine decarboxylase inhibitor, in diet. Cancer Res 1990;50:2562–2568. 90 Wolfe MM, Lichtenstein DR, Singh G: Gastrointestinal toxicity of nonsteroidal anti-inflammatory drugs. N Engl J Med 1999;340:1888–1899. 91 Singh G: Recent considerations in nonsteroidal anti-inflammatory drug gastropathy. Am J Med 1998;105:31S–38S. 92 Larkai EN, Smith JL, Lidsky MD, Graham DY: Gastroduodenal mucosa and dyspeptic symptoms in arthritic patients during chronic nonsteroidal anti-inflammatory drug use. Am J Gastroenterol 1987;82:1153–1158. 93 Singh G, Triadafilopoulos G: Epidemiology of NSAID-induced gastrointestinal complications. J Rheumatol 1999;26(suppl 56):18–24. 94 Whelton A, Maurath CJ, Verburg KM, Geis GS: Renal safety and tolerability of celecoxib, a novel cyclooxygenase-2 inhibitor. Am J Ther 2000;7:159–175. 95 Schafer AI: Effects of nonsteroidal anti-inflammatory therapy on platelets. Am J Med 1999;106: 25S–36S. 96 Bjorkman DJ: Current status of nonsteroidal anti-inflammatory drug use in the United States: Risk factors and frequency of complications. Am J Med 1999;107:3S–10S. 97 FitzGerald GA, Patrono C: The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med 2001;345:433–442. 98 Bombardier C, Laine L, Reicin A, Shapiro D, Burgos-Vargas R, Davis B, Day R, Ferraz MB, Hawkey CJ, Hochberg MC, Kvien TK, Schnitzer TJ: Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J Med 2000;343:1520–1528. 99 Silverstein FE, Faich G, Goldstein JL, Simon LS, Pincus T, Whelton A, Makuch R, Eisen G, Agrawal NM, Stenson WF, Burr AM, Zhao WW, Kent JD, Lefkowith JB, Verburg KM, Geis GS: Gastrointestinal toxicity with celecoxib vs. nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: A randomized controlled trial. Celecoxib LongTerm Arthritis Safety Study. JAMA 2000;284:1247–1255. 100 Langman MJ, Jensen DM, Watson DJ, Harper SE, Zhao PL, Quan H, Bolognese JA, Simon TJ: Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs. JAMA 1999;282: 1929–1933. 101 Juni P, Rutjes AW, Dieppe PA: Are selective COX-2 inhibitors superior to traditional non-steroidal anti-inflammatory drugs? BMJ 2002;324:1287–1288. 102 Perazella MA, Eras J: Are selective COX-2 inhibitors nephrotoxic? Am J Kidney Dis 2000;35: 937–940. 103 Rocha JL, Fernandez-Alonso J: Acute tubulointerstitial nephritis associated with the selective COX-2 enzyme inhibitor, rofecoxib. Lancet 2001;357:1946–1947. 104 Mersfelder TL, Stewart LR: Warfarin and celecoxib interaction. Ann Pharmacother 2000;34: 325–327. 105 Haase KK, Rojas-Fernandez CH, Lane L, Frank DA: Potential interaction between celecoxib and warfarin. Ann Pharmacother 2000;34:666–667.
Hawk/Viner/Umar
240
106 Mukherjee D, Nissen SE, Topol EJ: Risk of cardiovascular events associated with selective COX-2 inhibitors. JAMA 2001;286:954–959. 107 Sanmuganathan PS, Ghahramani P, Jackson PR, Wallis EJ, Ramsay LE: Aspirin for primary prevention of coronary heart disease: Safety and absolute benefit related to coronary risk derived from meta-analysis of randomised trials. Heart 2001;85:265–271. 108 FitzGerald GA: Cardiovascular pharmacology of nonselective nonsteroidal anti-inflammatory drugs and coxibs: Clinical considerations. Am J Cardiol 2002;89:26D–32D. 109 Reports F-D-C: Merck COX-2 cardiovascular safety studies will enroll 30,000 subjects; ‘The Pink Sheet’, 2001, vol 63, pp 1–12. 110 Petrik MB, McEntee MF, Chiu CH, Whelan J: Antagonism of arachidonic acid is linked to the antitumorigenic effect of dietary eicosapentaenoic acid in ApcMIN/⫹ mice. J Nutr 2000;130: 1153–1158. 111 Takaku K, Sonoshita M, Sasaki N, Uozumi N, Doi Y, Shimizu T, Taketo MM: Suppression of intestinal polyposis in Apc⌬716 knockout mice by an additional mutation in the cytosolic phospholipase A2 gene. J Biol Chem 2000;275:34013–34016. 112 Sonoshita M, Takaku K, Sasaki N, Sugimoto Y, Ushikubi F, Narumiya S, Oshima M, Taketo MM: Acceleration of intestinal polyposis through prostaglandin receptor EP2 in Apc⌬716 knockout mice. Nat Med 2001;7:1048–1051. 113 Seno H, Oshima M, Ishikawa TO, Oshima H, Takaku K, Chiba T, Narumiya S, Taketo MM: Cyclooxygenase 2- and prostaglandin E(2) receptor EP(2)-dependent angiogenesis in ApcD716 mouse intestinal polyps. Cancer Res 2002;62:506–511. 114 Sheng H, Shao J, Washington MK, DuBois RN: Prostaglandin E2 increases growth and motility of colorectal carcinoma cells. J Biol Chem 2001;276:18075–18081. 115 Watanabe K, Kawamori T, Nakatsugi S, Ohta T, Ohuchida S, Yamamoto H, Maruyama T, Kondo K, Ushikubi F, Narumiya S, Sugimura T, Wakabayashi K: Role of the prostaglandin E receptor subtype EP1 in colon carcinogenesis. Cancer Res 1999;59:5093–5096. 116 Watanabe K, Kawamori T, Nakatsugi S, Ohta T, Ohuchida S, Yamamoto H, Maruyama T, Kondo K, Narumiya S, Sugimura T, Wakabayashi K: Inhibitory effect of a prostaglandin E receptor subtype EP1-selective antagonist, ONO-8713, on development of azoxymethane-induced aberrant crypt foci in mice. Cancer Lett 2000;156:57–61. 117 Rao CV, Indranie C, Simi B, Manning PT, Connor JR, Reddy BS: Chemopreventive properties of a selective inducible nitric oxide synthase inhibitor in colon carcinogenesis, administered alone or in combination with celecoxib, a selective cyclooygenase-2 inhibitors. Cancer Res 2001;62: 165–170. 118 Rostom A, Wells G, Tugwell P, Welch V, Dube C, McGowan J: The prevention of chronic NSAIDinduced upper gastrointestinal toxicity: A Cochrane collaboration meta-analysis of randomized controlled trials. J Rheumatol 2000;27:2203–2214. 119 Silverstein FE, Graham DY, Senior JR, Davies HW, Struthers BJ, Bittman RM, Geis GS: Misoprostol reduces serious gastrointestinal complications in patients with rheumatoid arthritis receiving nonsteroidal anti-inflammatory drugs. A randomized, double-blind, placebo-controlled trial. Ann Intern Med 1995;123:241–249. 120 Howe LR, Subbaramaiah K, Brown AM, Dannenberg AJ: Cyclooxygenase-2: A target for the prevention and treatment of breast cancer. Endocr Relat Cancer 2001;8:97–114. 121 Bigler J, Whitton J, Lampe JW, Fosdick L, Bostick RM, Potter JD: CYP2C9 and UGT1A6 genotypes modulate the protective effect of aspirin on colon adenoma risk. Cancer Res 2001;61: 3566–3569. 122 Suleiman S, Rex DK, Sonnenberg A: Chemoprevention of colorectal cancer by aspirin: A costeffectiveness analysis. Gastroenterology 2002;122:78–84. 123 Ladabaum U, Chopra CL, Huang G, Scheiman JM, Chernew ME, Fendrick AM: Aspirin as an adjunct to screening for prevention of sporadic colorectal cancer. A cost-effectiveness analysis. Ann Intern Med 2001;135:769–781. 124 Reilly IA, FitzGerald GA: Aspirin in cardiovascular disease. Drugs 1988;35:154–176. 125 Ross JS, Stagliano NE, Donovan MJ, Breitbart RE, Ginsburg GS: Atherosclerosis and cancer: Common molecular pathways of disease development and progression. Ann NY Acad Sci 2001; 947:271–293.
NSAIDs/COXIBs in Prevention Trials
241
126 Emilien G, Ponchon M, Caldas C, Isacson O, Maloteaux JM: Impact of genomics on drug discovery and clinical medicine. QJM 2000;93:391–423. 127 Chambers G, Lawrie L, Cash P, Murray GI: Proteomics: A new approach to the study of disease. J Pathol 2000;192:280–288. 128 Liotta LA, Kohn EC, Petricoin EF: Clinical proteomics: Personalized molecular medicine. JAMA 2001;286:2211–2214. 129 Waddell WR, Ganser GF, Cerise EJ, Loughry RW: Sulindac for polyposis of the colon. Am J Surg 1989;157:175–179. 130 Carbone PP, Douglas JA, Larson PO, Verma AK, Blair IA, Pomplun M, Tutsch KD: Phase I chemoprevention study of piroxicam and ␣-difluoromethylornithine. Cancer Epidemiol Biomarkers Prev 1998;7:907–912.
Ernest T. Hawk, MD, MPH Gastrointestinal & Other Cancers Research Group National Cancer Institute, Division of Cancer Prevention EPN, Suite 2141, 6130 Executive Boulevard, Bethesda, MD 20892-7317 (USA) Tel. +1 301 5942684, Fax +1 301 4356344, E-Mail
[email protected]
Hawk/Viner/Umar
242
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 243–260
Chemotherapy with Cyclooxygenase-2 Inhibitors in the Treatment of Malignant Disease: Pre-Clinical Rationale and Preliminary Results of Clinical Trials Charles D. Blanke a, Jaime L. Masferrer b a b
Oregon Health & Science University, Portland, Oreg. and Pharmacia Corporation, Chesterfield, Mo., USA
Introduction
Epidemiological studies have long suggested that NSAIDs, prototypic inhibitors of cyclooxygenase (COX), can prevent the development of a variety of solid malignancies. An abundance of pre-clinical and early clinical data now also suggest that selective COX-2 inhibitors have a potential role as targeted therapy against existing neoplasms. High-level expression of COX-2 is induced in many tissues by pro-inflammatory cytokines (IL-1␣, TNF-␣), growth factors (EGF, PDGF, FGF, TGF-), carcinogens and oncogenes (Ras, Src, HER2, Wnt) [1, 2]. Production of prostaglandins, particularly PGE2, is an important event in carcinogenesis, and indeed the COX-2 pathway has been implicated in the initiation and promotion of a host of malignancies. In fact, COX-2 expression (protein and/or mRNA) has been demonstrated in the majority of human solid neoplasms. As discussed below, the level of COX-2 in a tumor can have prognostic implications. Pre-clinically, selective COX-2 inhibitors suppress angiogenesis, induce apoptosis, and decrease the incidence of lung metastases from implanted solid tumors. The beneficial effects of selective COX-2 inhibitors appear to be additive or synergistic with therapeutic irradiation, as well as a variety of chemotherapeutic agents. Clinically, treatment with both nonselective and selective COX inhibitors leads to regression of neoplastic lesions, such as duodenal and colonic polyps. A variety of adjuvant and advanced disease clinical
cancer trials are now underway or in the late planning stage, assessing the use of selective COX-2 inhibitors with irradiation, chemotherapy and novel antineoplastic agents. This chapter addresses emerging applications of selective COX-2 inhibitors in cancer therapy.
Rationale for Use
Expression of COX-2 and Prognosis COX-2 is a fairly ubiquitous target in human neoplasms, as its expression has been demonstrated in the majority of pre-malignant and frankly cancerous lesions studied. Eberhart et al. [3] showed that products of COX-2 gene expression were present in the majority of colorectal adenocarcinomas, as well as in a substantial fraction of colorectal polyps. Additionally, overexpression of COX-2 has been demonstrated in pre-malignant lesions such as Barrett’s esophagus, oral leukoplakia and gastric and cervical dysplasia, as well as in invasive cancers of the head and neck, breast, lung (including mesothelioma), skin and gastrointestinal (liver, pancreaticobiliary system, stomach, esophagus, colon), gynecologic (cervix, uterus) and genitourinary (bladder, prostate) systems [4]. Expression often correlates with more advanced disease, or a worse outcome in aggressively treated early-stage cancer. For example, COX-2 expression is more commonly seen in frank cancers versus early dysplastic lesions of the lung and esophagus, and higher-stage (T2–4 and N1) versus lower stage (T1, N0) gastric and bronchogenic neoplasms [5–8]. Confounding these findings are reports that greater expression of COX-2 occurs in well-differentiated, as compared to poorly differentiated, cancers of the esophagus, lung and liver, as well as one description suggesting the level of COX-2 expression in resected gastric cancer does not always correlate with disease stage, recurrence, or mortality rates [9–12]. The level of COX-2 expression appears to have prognostic implications for patients undergoing potentially curative surgery for esophageal, colon and lung cancers. High versus low expression in the resected tumor specimen significantly correlates with disease recurrence and/or overall survival in these diseases [13–15]. Overall, the preponderance of evidence suggests that COX-2 is a reasonable therapeutic target in advanced human malignancy. COX-2 and Pre-Clinical Efficacy of Selective Inhibitors, With or Without Chemotherapy Selective COX-2 inhibitors can slow cancer growth and/or proliferation by decreasing tumor-mediated angiogenesis, reversing resistance to apoptosis, and abrogating the invasive phenotype mediated through matrix metalloprotease induction [16, 17]. It is possible that COX-2 inhibition might also improve the
Blanke/Masferrer
244
Chemotherapy
Apoptosis
Upregulation of COX-2
Inhibition of apoptosis
Release of pro-inflammatory mediators
COX-2 Inhibition
Upregulation of PGE2 and VEGF Upregulation of COX-2 PGE2
VEGF
Tumor growth and survival
Neovascularization
Release of survival/ pro-angiogenic factors (VEGF, other growth factors)
Fig. 1. Upon exposure to chemotherapeutic drugs, tumor cells release pro-inflammatory mediators. These compounds trigger production of COX-2 and COX-2-derived prostaglandins, which act in a paracrine manner to promote release of survival and/or proangiogenic factors. COX-2 inhibitors can blunt this response and might improve the efficacy of cytotoxic agents.
efficacy of chemotherapy, by blocking the drug-induced responses that counter antitumor activity and contribute to enhanced growth and survival of the cancer cells (fig. 1). Upon treatment with chemotherapy, some tumor cells undergo apoptosis and release pro-inflammatory mediators, which stimulate the surrounding fibroblasts and remaining neoplastic cells to upregulate COX-2. This stimulates the release of proangiogenic and other survival factors, leading to neovascularization and promotion of tumor growth. Selective COX-2 inhibitors could block these events, leading to enhanced cell kill by the cytotoxic agents. Specific Experimental Data with Taxanes and Other Antimicrotubule Agents Agents that interfere with microtubule polymerization and depolymerization induce cell cycle arrest and apoptotic death in cycling cells. Antimicrotubule agents such as colchicine, paclitaxel, docetaxel, vincristine and vinblastine can induce COX-2 by stimulating transcription or stabilizing COX-2 mRNA [18–22]. In a study utilizing paclitaxel in a non-small cell lung cancer (NSCLC)
COX-2 Inhibitors and Cancer
245
model, responses among 31 human tumors transplanted to mice correlated with COX-2 expression [23]. Sixty-six percent (2/3) of COX-2-negative tumors, 20% (4/20) of tumors that produced moderate amounts of COX-2, and 0% of 8 tumors that produced high levels of COX-2 responded to treatment with the taxane. This study also demonstrated that in vitro treatment of high COX-2-producing NSCLC cells (A427) with non-cytotoxic concentrations (⬍50 M) of the selective COX-2 inhibitor NS-398 increased sensitivity to paclitaxel 2- to 4.6-fold. In a study with docetaxel and four NSCLC cell lines that differed in their expression of COX-2, the COX-2 inhibitor nimesulide (30 M) reduced the IC50 of docetaxel to a degree that was at least additive compared with nimesulide or docetaxel alone [24]. The effect of the COX-2 inhibitor was greatest in the cell lines with the highest expression of COX-2. Experimental Data with Cisplatin, Irinotecan and Other Chemotherapeutic Agents In the studies describing potentiation of the sensitivity of NSCLC cells to paclitaxel with nimesulide, sensitivity of the cells to etoposide, cisplatin and irinotecan also correlated with COX-2 expression and was potentiated by treatment with the COX-2 inhibitor. The strongest potentiation (77%) was seen with the combination of nimesulide and SN-38, the active metabolite of irinotecan, against the NSCLC cell line with highest COX-2 expression [24]. In a Lewis lung cancer model, delayed tumor growth following treatment with cisplatin, cyclophosphamide, melphalan, or BCNU was increased 2- to 4.5-fold when chemotherapy was combined with a lipoxygenase inhibitor (phenidone) and either sulindac or indomethacin, both nonselective COX inhibitors [25]. In other studies using the Lewis lung model, cyclophosphamide inhibited tumor growth by 35% and celecoxib inhibited tumor growth by 84% when they were used as single agents, but the combination of celecoxib and cyclophosphamide resulted in near complete (97%) inhibition of tumor growth [26]. In mice with HT-29 human colon tumor xenografts, treatment with 5-fluorouracil (5-FU) inhibited tumor growth by 35%, celecoxib inhibited growth by 69%, and the combination suppressed tumor growth by about 80%. It is of interest that tumor growth inhibition with celecoxib produced no weight loss, whereas a transient 20% decrease in weight was seen with 5-FU treatment alone [26]. In studies with celecoxib and irinotecan in mice with HT-29 human colon tumor xenografts, the combination decreased tumor growth by 91% compared with vehicle controls, whereas tumor growth was inhibited by 72% with oral celecoxib (25 mg/kg/day) and by 29% with irinotecan (30 mg/kg every 4 days) used alone. Against highly aggressive Colon-25 human colon cancer xenografts, celecoxib and irinotecan had only modest effects as single agents (⬍4 and 23%, respectively) but the 39% growth inhibition rate achieved with
Blanke/Masferrer
246
90
Diarrhea ⫽ 2 & 3
Diarrhea ⫽ 3
80 Diarrhea (%)
70 60 50 40 30 20 10 0 0
0.1
0.3
1
3
10
30
50
150
Celecoxib (mg/kg per day)
Fig. 2. Irinotecan induces both an acute early diarrhea and a late diarrhea. Rats treated with this agent (150 mg/kg per day, ⫻2 days) in the presence of increasing doses of celecoxib (0–150 mg/kg per day) show dose-related decreases in both the incidence and severity of late-occurring diarrhea.
the combination demonstrated a greater than additive effect compared with either agent alone [27]. Pre-Clinical Data Suggesting COX-2 Inhibitors Decrease Chemotherapy-Mediated Toxicity Irinotecan is highly active in colorectal cancer, but its use may be complicated by late-occurring diarrhea. Studies in rats have established an association between pathologic changes in the colon and increased production of PGE2 in the affected tissue [28]. In studies addressing the effects of celecoxib on irinotecan-induced diarrhea, dose-dependent reductions in both the incidence and severity of irinotecan-induced late diarrhea were seen in animals treated with celecoxib compared with those that received irinotecan alone (fig. 2) [27]. In addition to late-occurring diarrhea, other side effects of chemotherapy either known or suggested to be related to COX-2 activity include early diarrhea, myalgias, arthralgias, peripheral neuropathy and events related to suppression of immunologic responses. It is possible that combining the cytotoxic agents with a selective COX-2 inhibitor could mitigate some of those toxicities (see below). Not all pre-clinical investigators agree that COX inhibition would be of benefit with chemotherapy, however. Lorenz et al. [29] assessed recovery from 5-FU-mediated myeloablation in COX-2-deficient mice, which have normal baseline hematologic counts. Homozygous deficient mice had a similar chemotherapeutic nadir as heterozygotes, but day 8 white blood cell, platelet
COX-2 Inhibitors and Cancer
247
and reticulocyte counts were markedly lower. Recovery eventually occurred but was significantly delayed. The authors speculated that COX-2 activity is necessary to ‘prime’ cells in the marrow microenvironment for accelerated hematopoiesis. Given the disparate hypotheses regarding the potential of COX-2 inhibitors to affect the incidence and severity of side effects associated with standard cancer therapy, this situation can only be resolved through careful clinical investigation. COX-2 and Radiation There is strong pre-clinical rationale for combining COX-2 inhibitors with therapeutic irradiation. These drugs appear to radiosensitize some solid tumors, while radioprotecting normal gut [30]. In the clinical setting, specifically patients with invasive cervical or breast cancers treated with radiotherapy, high expression of COX-2 has been associated with reduced tumor response and decreased overall survival. In one study, 5-year survival among patients with high COX-2-expressing cervical cancers was only 35% compared with 75% among patients whose tumors produced no or low levels of COX-2 [31]. In a study with paclitaxel and radiotherapy in patients with locally advanced breast cancer, 5 complete and 2 partial pathologic responses were obtained in 7/21 patients. Correlating responses to this regimen with the molecular characteristics of the tumor, the investigators found that patients with a complete pathologic response had a 30-fold lower expression of COX-2 compared with other patients, suggesting that molecular profiling of tumors prior to therapy may identify patients likely to respond well to this treatment [32].
General Clinical Use
The nonselective COX inhibitor sulindac can decrease the intestinal polyp burden in patients with familial adenomatous polyposis [33]. Selective COX-2 inhibitors are also effective in reducing pre-invasive disease in this patient population [34, 35]. Steinbach et al. [34, 35] randomized FAP patients with known colorectal adenomas in 2:2:1 fashion to celecoxib 400 mg p.o. bid, celecoxib 100 mg p.o. bid (typical arthritis dosing), or placebo, all given for a period of 6 months. Patients underwent intense endoscopic video-documentation of polyp burden. Patients treated with celecoxib, including those given high-dose drug, did not experience significantly greater toxicity than those given placebo. Therapy with the 400 mg bid regimen led to significant regression in adenoma number and overall polyp burden, while treatment with the arthritis dose had no significant effect. This trial set the dosing standard for most studies using celecoxib in the setting of advanced cancer.
Blanke/Masferrer
248
Use of Selective COX-2 Inhibitors by Organ System and Disease Non-Small Cell Lung Cancer NSCLC is a particularly attractive tumor in which to test selective COX-2 inhibitors. These agents induce apoptosis in NSCLC cell lines, and they enhance cytotoxic activity of many chemotherapeutic agents used to treat bronchogenic carcinoma, including docetaxel, cisplatin and etoposide [24]. Investigators at the Vanderbilt-Ingram Cancer Center initiated a phase II trial of celecoxib 400 mg p.o. bid with docetaxel, 75 mg/m2 i.v. every 3 weeks [36]. Eligibility requirements included histologically confirmed NSCLC that had previously been treated with chemotherapy, as well as PS of 1 or better. Seventy-one percent of analyzed patients had high pre-treatment levels of COX-2 expression by immunohistochemistry. A report at ASCO 2002 described results in 15 treated patients (prior taxanes in 13; 54 total planned): There was no grade 3–4 thrombocytopenia, though 43% of patients experienced severe or life-threatening neutropenia [36]. Two (15.4%) patients achieved a partial remission, and 3 (20%) stable disease. Correlative scientific studies showed intra-tumoral PGE2 levels markedly declined with the COX-2 inhibitor treatment. The response rate was encouraging in this heavily pre-treated group (no responses to previous chemotherapy), and the authors concluded that the decline in PGE2 levels was evidence that celecoxib inhibited intra-tumoral COX-2. The Vanderbilt group is also conducting a study in locally advanced NSCLC patients utilizing induction celecoxib, 400 mg p.o. bid, followed by the same dose of the COX-2 inhibitor with weekly paclitaxel 50 mg/m2 and carboplatin AUC 2 ⫻ 5, plus therapeutic irradiation, to a total of 6,300 cGy [37]. This is followed by celecoxib, carboplatin and paclitaxel every 3 weeks ⫻ 2, then full-dose celecoxib alone until progression. Preliminary results detailed significant esophagitis in 33% (3/9) of patients, with 1 case of severe pneumonitis [37]. Three of the first 5 patients had evidence of an objective response. VEGF levels are being measured because of previous evidence that COX-2-derived prostaglandins can regulate the expression of VEGF. The study is ongoing. Cornell University investigators assessed celecoxib with standard neoadjuvant chemotherapy in PS 0–1 stage IB-IIIA NSCLC [38, pers. commun.]. This phase II trial utilized celecoxib 400 mg p.o. bid from day 1 until the day of surgery (performed between days 42 and 56), with two cycles of paclitaxel (225 mg/m2) and carboplatin (AUC 6) given 3 weeks apart. Twenty-nine patients (18 men and 11 women) were accrued. Stage distribution was as follows: IB-16, IIB-3 and IIIA-10. There were no significant unexpected toxicities. The preoperative response rate was high: 48% had partial responses and 17% had complete clinical remissions, for an overall response rate of 65%. Twenty-eight patients were explored and resected. There were no complete pathological responses but 7 patients (24%) had minimal residual microscopic disease.
COX-2 Inhibitors and Cancer
249
In comparison to historical data in patients treated with carboplatin-paclitaxel, these results are very encouraging (the expected clinical response rate is 56%, versus the 65% seen in this study). This modest difference is noteworthy because 34% of patients had stage IIIA disease in this celecoxib-based trial compared with 7% in the previous study. Large-scale, confirmatory studies using the Cornell regimen are planned. In addition, investigators at the Karmanos Cancer Institute (Wayne State) are conducting two phase II trials of celecoxib and docetaxel [pers. commun.]. One study tests induction celecoxib ⫻ 1 week, followed by celecoxib and the taxane given every 3 weeks, in stage IIIA, IIIB, or IV NSCLC patients who progressed on or after platinum-based chemotherapy. The other study utilizes induction celecoxib followed by celecoxib with weekly docetaxel in stage IIIB or IV patients who are elderly (ⱖ70 years of age) or who have poor PS (SWOG 2). Breast Overexpression of COX-2 is more prominent in breast cancer cell lines with a more aggressive metastatic phenotype, and COX-2 is clearly expressed in some human breast cancers [39]. In addition, there has been recent evidence linking the HER-2/neu pathway with upregulation of COX-2 in breast cancer [40]. Dang et al. [40] have commenced a trial of celecoxib 400 mg p.o. bid with trastuzumab, 2 mg/kg i.v. weekly, in HER-2-positive patients with metastatic breast cancer who failed prior trastuzumab-based therapy. Nine patients were reported upon at ASCO 2002 (44% 2⫹ expressers and 56% 3⫹); 1 patient had grade 3 abdominal pain while grade 1 diarrhea (2 patients), stomatitis (2), pruritus (2) and stomatitis (2) were also seen. One patient had stable disease at 3 months. Other ongoing or proposed studies in breast cancer combine celecoxib with standard cytotoxic chemotherapy in neoadjuvant treatment of locally advanced breast cancer, and with hormonal therapy, such as exemestane, in metastatic disease. Colorectal Cancer Colorectal cancer has perhaps the strongest pre-clinical justification for targeting with a selective COX-2 inhibitor. This rationale includes the demonstration of high levels of COX-2 in primary tumor and metastatic tissue, correlation of expression with higher stage disease and worse prognosis, known chemopreventive properties of COX-2 inhibition, and the pre-clinical interactions between COX-2 inhibitors and standard large bowel chemotherapeutic agents (5-FU and CPT-11) [26, 35, 41, 42]. A phase I trial performed in patients with advanced colorectal cancer demonstrated the safety of combining sulindac with standard chemotherapy (5-FU and levamisole) [43]. Several trials using selective COX-2
Blanke/Masferrer
250
inhibitors with a variety of chemotherapeutic compounds, plus or minus irradiation, are ongoing, and some have been published in preliminary form. Investigators at the Oregon Health & Sciences University (OHSU) Cancer Institute are conducting a phase I trial of standard North Central Cancer Treatment Group 5-FU and leucovorin, with escalating doses of celecoxib (100–800 mg bid), in patients with incurable colorectal cancer [pers. commun.]. The highest dose level of the COX-2 inhibitor was selected based on absorption issues and not concerns over toxicity, and there is a 2-week lead-in period of celecoxib alone, to try to separate potential NSAID toxicity from side effects of the chemotherapy. Platelet function is measured to assess for potential COX-1 inhibition. Two of 6 patients enrolled at level 1 had dose-limiting toxicities (one grade 3 GI bleed and one with prolonged bleeding time day 5 of chemotherapy). Six additional patients were added, with one more episode of asymptomatic platelet dysfunction. Enrollment has commenced at the 200 mg bid dose level. Two phase II trials testing celecoxib 400 mg p.o. bid with standard bolus IFL (irinotecan 125 mg/m2, 5-FU 500 mg/m2, leucovorin 20 mg/m2, all i.v. weekly ⫻ 4 every 43 days, Saltz regimen) were reported at ASCO 2002 [44, 45]. The trial by the OHSU group used a 2-week induction with the COX-2 inhibitor and includes PS ⱕ2 patients with measurable, incurable colorectal cancer that had not been treated before in the advanced disease setting [44]. In the 22 initial patients reported upon, an interesting toxicity profile emerged. When compared historically to those treated with the Saltz regimen of IFL alone, a similar percent had grade 3–4 diarrhea, but only 27% had severe or greater neutropenia compared to 54% reported by Saltz. The OHSU trial was interrupted for a prolonged period when concerns over cardiotoxicity with COX-2 inhibitors emerged in the community. A Data Safety Monitoring Board assessed all toxic events and deemed the trial safe to re-open, albeit demanding more stringent eligibility requirements (e.g., PS ⱕ1), stricter rules for dose modifications, and early use of antibiotics in the face of even moderate diarrhea. The preliminary objective response rate was 24% (95% CI 11–52), with nearly 20% inevaluable but included as non-responders under intent-to-treat analysis. The trial will accrue to its original goal of 54 patients using the new guidelines, and the true efficacy of this regimen thus remains to be determined. The Hoosier Oncology Group also combined full-dose celecoxib with bolus IFL, while adding glutamine 10 g every 8 h, in an effort to decrease the diarrhea associated with the chemotherapy [45]. At the time of interim analysis, there was no obvious decrease in severe diarrhea (rate of 39%), but only 31% of patients experienced severe neutropenia, and there were no episodes of granulocytopenic fevers. With 16 evaluable patients, the objective response rate was 31% (95% CI 11–59). This trial has now completed accrual and should be reported in final form soon.
COX-2 Inhibitors and Cancer
251
M.D. Anderson investigators retrospectively reviewed 67 first- and secondline patients with metastatic colorectal cancer treated with capecitabine alone or with the fluoropyrimidine plus celecoxib (200 mg p.o. bid in the vast majority, who were mostly given the drug for pain) [46]. The rates of serious (greater than grade 1) hand-foot syndrome and severe to life-threatening diarrhea were statistically significantly lower in the group that received the COX-2 inhibitor. No significant hematologic or nonhematologic toxicities were seen in the patients who received both capecitabine and celecoxib. Median time to progression was 6 months for the dual-treatment patients, versus 3 months for those receiving capecitabine alone, with no overlap of the 95% confidence intervals. Based on these provocative data, the same group is preparing to launch a phase II study of capecitabine 1,250 mg/m2 p.o. bid for 14 days every 3 weeks, with celecoxib 400 mg p.o. bid continuously, in PS 0–2 patients with untreated metastatic disease [pers. commun.]. Celecoxib is not the only COX-2 inhibitor to be tested in colorectal cancer. The Simmons Cancer Center has published a phase II trial of rofecoxib with standard daily ⫻ 5 bolus 5-FU and leucovorin in metastatic patients not refractory to the fluoropyrimidine [47]. At a rofecoxib dose of 50 mg p.o. daily, 2 of the first 3 patients experienced grade III upper gastrointestinal bleeding, and subsequent patients received 25 mg/day. An additional patient at the lower dose again had a serious upper GI bleed, while others experienced grade 2 stomatitis, thrombocytopenia, or diarrhea. None of the first 10 patients exhibited an objective response, so the study was terminated. The authors concluded that rofecoxib did not add to the efficacy of the chemotherapy regimen and that it increased GI toxicity. A very promising area for use of COX-2 inhibitors in colorectal cancer is in treatment of potentially curatively resected patients at high risk for relapse. Primary specimens expressing high levels of COX-2 are associated with significantly higher recurrence rates than are those expressing little COX-2 [14]. The Victor study (Vioxx in colorectal cancer therapy: Definition of optimal regimen) is a double-blind, placebo-controlled trial of rofecoxib in stage II or III colorectal cancer patients without evidence of residual disease after surgery [pers. commun.]. These patients undergo chemotherapy meeting national best practice and/or radiotherapy, and are then randomized to rofecoxib 25 mg daily or placebo, for 2 or 5 years, to determine differences between drug and no drug, and the shorter versus longer periods of therapy. Patients may also undergo double randomization after surgery alone. This study, with a primary endpoint of overall survival and the secondary objective of disease-free survival, has begun accruing toward its goal of 7,000 patients. An associated scientific protocol will try to assess the prognostic and predictive powers of COX-1, COX-2, p53 and hMLH-1 from immunohistochemistry staining of fixed colorectal cancer
Blanke/Masferrer
252
tissue, and a companion study in North America will look at the effects of rofecoxib therapy on metachronous polyp development at 24 and 60 months, as well as overall development of advanced adenomas. The PETACC (Pan-European Trials in Adjuvant Colon Cancer) Group is conducting a phase III randomized double-blind, placebo-controlled of celecoxib in patients with stage III colon cancer. Patients will be treated with 24 weeks of standard therapy (infusional 5-FU vs. bolus 5-FU vs. 5-FU with irinotecan or oxaliplatin) per national standards and randomized to celecoxib or placebo to be given concurrent with chemotherapy and continued for a total of 3 years. With 1,440 patients, the trial will be able to determine if the addition of celecoxib will improve the 5-year disease-free survival. The NSABP will compare the incidence of adenomatous polyps in patients with stage I/II (Duke’s A/B1) receiving celecoxib or placebo in a phase III randomized doubleblind, placebo-controlled trial. It is anticipated that the PETACC and NSABP trials will open early in 2003. EORTC (Study 40015) is evaluating celecoxib in metastatic colorectal cancer in a phase III randomized double-blind, placebocontrolled setting. Patients will be randomized to irinotecan/5-FU vs. irinotecan/capecitibine with or without celecoxib 400 mg bid in a 2 ⫻ 2 factorial design. The primary endpoint is to determine if the addition of celecoxib improves progression-free survival as well as to compare the chemotherapy regimens. Celecoxib will be administered during chemotherapy until disease progression or 3 years, whichever is first. Molecular markers will be correlated with outcome. A number of other colorectal cancer chemotherapy studies are in the design stage, including the testing of celecoxib with Iressa. Various combinations of celecoxib, irradiation and 5-FU or irinotecan are being considered in locally advanced rectal cancer. Pancreas Neoplastic pancreatic tissues demonstrate a 60-fold increase in COX-2 mRNA expression in parallel with upregulation of COX-2 protein, and selective COX-2 blockers inhibit proliferation and induce apoptosis in human pancreatic cancer cell lines [48–50]. Xiong et al. [50] have completed a phase I trial of celecoxib with gemcitabine, the standard of care chemotherapy for those with advanced pancreatic cancer. Patients with metastatic or locally advanced disease not treated in the stage 4 setting with chemotherapy, were treated with gemcitabine (variable dose) as a 10 mg/m2⭈min i.v. infusion weekly ⫻ 3 every 4 weeks, with celecoxib 400 mg p.o. bid starting 2 days after the chemotherapy. In the first 5 patients treated with gemcitabine 750 mg/m2, 2 patients had grade 3 neutropenia, and 1 had grade 3 thrombocytopenia, while 1 patient also had a grade 3 skin rash. Six additional patients were treated with gemcitabine at 650 mg/m2; grade 3 toxicity was seen as follows: nausea (1), vomiting (1),
COX-2 Inhibitors and Cancer
253
thrombocytopenia (2). In addition, 1 patient had grade 4 neutropenia. Two patients had ‘early deaths’ – 1 dying during sleep in the second week after therapy, and 1 from tumor-associated gastric ulcer perforation in the setting of neutropenia. The pharmacokinetics of gemcitabine triphosphate in mononuclear cells was analyzed by HPLC; on the gemcitabine 750 mg/m2 dose, levels accumulated significantly by week 2 versus week 1, but there was no change at the gemcitabine 650 mg/m2 level. There was a suggestion of potential activity even in this small cohort; 1 of 11 evaluable patients exhibited a partial response. The authors concluded that the combination of gemcitabine and celecoxib clearly had greater myelosuppression than expected from gemcitabine alone, though it was possible that this reflected the timing or infusional rate of the chemotherapy. They recommended testing the 650 mg/m2 dose with celecoxib in a phase II study. A Karmanos Cancer Institute phase II study of gemcitabine (1,000 mg/m2 i.v. as a 10 mg/m2⭈min infusion weekly ⫻ 2 every 3 weeks), with cisplatin (35 mg/m2 i.v., same schedule), and celecoxib 400 mg p.o. bid, starting 1 week before chemotherapy and continuing throughout is underway in untreated metastatic patients [pers. commun.]. The initial 7 patients all had grade 3 (6) or 4 (1) neutropenia, which does not represent an increase over that expected from the chemotherapeutic regimen alone. Studies incorporating other novel agents, in addition to the COX-2 inhibitors, are underway. The University of Virginia plans a pilot trial of celecoxib and thalidomide added to standard 30-min infusional gemcitabine. Oregon Health & Sciences University has preliminary approval to conduct a phase I/II neoadjuvant trial of rofecoxib, irradiation and continuous infusion 5-FU in patients with potentially resectable pancreatic adenocarcinoma. Cervical Cancer Up to 100% of cervical cancers express COX-2, and increased expression correlates with poorer local control and decreased survival in patients treated with therapeutic irradiation [51]. The Radiation Therapy Oncology Group has an ongoing phase I/II study of celecoxib with irradiation and chemotherapy [52]. Eligibility requirements include stage IIB–IVA PS 0–2 patients (IB or IIA accepted if the patients have biopsy-proven pelvic nodal involvement or the tumor size is ⱖ5 cm) with squamous cell, adenocarcinoma, or adenosquamous carcinomas of the uterine cervix, and no history of systemic chemotherapy or pelvic irradiation (except transvaginal radiotherapy used to control bleeding). Planned accrual is 83 patients. Patients are given cisplatin 75 mg/m2 (maximum dose 150 mg) i.v. days 1 (within 16 h of the start of irradiation), 22, and 43; 5-FU 1,000 mg/m2 daily continuous infusion ⫻ 4 days, on days 2–5, 23–26, and 44–47; pelvic RT 45 Gy as 25 1.8-Gy fractions Monday–Friday over 5 weeks
Blanke/Masferrer
254
and low- or high-dose rate brachytherapy, ⫾ a parametrial boost; and celecoxib 400 mg p.o. bid, starting on the first day of irradiation and continuing for 12 months. The primary objective is to determine toxicity rates for the regimen, while secondary endpoints include compliance evaluation, efficacy assessment (including pelvic local control, distant control, disease-free survival and overall survival) and collection of tissue for immunohistochemical analysis of angiogenesis (correlating it with clinical outcome) and to perform microarray testing evaluating gene expression. This trial opened in August 2001 and is actively accruing. CNS Tumors Though all brain tissues express COX-2 protein, high-grade gliomas express more than low-grade gliomas or normal brain tissue [53]. High COX-2 expression in patients with all grades of astrocytomas correlates with poor survival and is the strongest predictor of outcome of numerous clinicopathologic features, including age and grade of tumor [54]. Treatment with NS-398, a selective COX-2 inhibitor, reduced the proliferation, increased apoptosis and decreased the migration of glioma cells in culture [55]. Investigators at M.D. Anderson are thus conducting a phase I trial of rofecoxib with irradiation for brainstem glioma [55]. PS 0–2 patients 3–85 years of age with newly diagnosed non-focal pontine tumors are eligible. The total accrual goal is 15 patients. Prostate and Renal Cancers A number of genitourinary sites would be suitable targets for COX-2 inhibition. Some investigators have shown that COX-2 is overexpressed in human prostate cancer compared with adjacent normal tissue, and PGE2 production is nearly 10 times higher in malignant versus benign tissue [56–58]. Selective COX-2 inhibitors reduce the growth of experimental prostate cancer, possibly through induction of apoptosis or inhibition of tumor angiogenesis [59, 60]. Investigators at the OHSU Cancer Institute are studying downstream effects of COX-2 inhibition in prostate cancer patients [61, pers. commun.]. The primary objective is to determine if celecoxib administration significantly increases apoptosis, and if so, whether the process is associated with downregulation of Bcl-2. Secondary objectives are to assess select biomarkers, as well as to quantify COX-2 expression in prostate study specimens. Eligible men have potentially resectable untreated adenocarcinoma, with PS ⱕ2. Patients are randomized to celecoxib (400 mg/day p.o. bid) for 4 weeks, or placebo. Neoplastic cells as well as premalignant tissue and benign prostate will be collected and examined for markers of apoptosis, angiogenesis, VEGF expression and proliferation. Whole blood COX-2 inhibition assays and tumor prostaglandin production will also be surveyed, partly to assure compliance [pers. commun.].
COX-2 Inhibitors and Cancer
255
General Low-Dose Metronomic Chemotherapy Standard cytotoxic chemotherapy agents damage tumor endothelial cells, which could lead to an antiangiogenic effect. However, administering drugs infrequently at maximum tolerated doses allows repair during the rest periods. Administering the same agents chronically on a once-per-week schedule, at approximately one-third of the maximal dose, should preserve the antivascular effect. This method of administering chemotherapy, referred to as metronomic dosing, has been shown to increase tumor chemosensitivity, even to agents in which the neoplasms had become resistant when dosed on a ‘normal’ schedule [62]. There is strong rationale for combining metronomic chemotherapy with selective COX-2 inhibitors, as these agents also possess antiangiogenic properties in pre-clinical models. Several trials are testing this concept. Buckstein et al. [pers. commun.] combine cyclophosphamide 50 mg p.o. daily with celecoxib 400 mg p.o. bid continuously in patients with relapsed or refractory aggressive histology nonHodgkin’s lymphoma. Objectives include evaluating clinical efficacy (including stable disease rates), safety and tolerability, assessing for evidence of inhibition of angiogenesis, using plasma VEGF and lymph node microvessel counts, among other biomarkers. Tannock et al. [pers. commun.] are administering continuous oral cyclophosphamide (50 mg/day) with celecoxib, 400 mg p.o. bid, to patients with measurable metastatic renal cell cancer. Their primary endpoint is objective response and rate of ‘long-term’ (⬎6 months) stable disease. Similar studies are ongoing in pediatric solid tumors. Toxicity Reduction As discussed above, there is emerging pre-clinical evidence that selective COX-2 inhibitors can diminish toxicities associated with normal use of standard chemotherapeutic agents [27]. This has largely been born out in clinical trials. Both the OHSU and Hoosier Oncology Groups colorectal cancer studies (see above), which added full-dose celecoxib to standard bolus irinotecan, 5-FU and leucovorin (the HOG also added glutamine) reported rates of grade 3 or greater neutropenia that were approximately half that expected from the chemotherapy alone [44, 45]. The M.D. Anderson review of celecoxib added to capecitabine described a significant reduction in hand-foot syndrome (greater than grade 1) and diarrhea (greater than grade 2) [46]. In addition, emerging data suggest that celecoxib may abrogate the neurotoxicity that limits dosing of oxaliplatin. Investigators at USC have retrospectively reviewed their data in colorectal cancer patients treated in a phase II trial of oxaliplatin with 5-FU (oxaliplatin 130 mg/m2 i.v. over 2 h every 3 weeks; 5-FU 200 mg/m2 CI weekly ⫻ 10) with or without celecoxib (200–400 mg daily) [pers. commun.]. The rate of serious neuropathy (greater than grade 1) was 14% for the 175 patients
Blanke/Masferrer
256
treated without celecoxib and 4% for the 75 patients given the selective COX-2 inhibitor.
Conclusions
Since the early 1990s it has been recognized that regular NSAID use decreases the risk of colorectal cancer, reduces the number and size of colonic polyps, and can improve survival of patients with advanced cancer. The improved therapeutic profile of the selective COX-2 inhibitors provides the opportunity to safely and thoroughly evaluate these anticancer properties and may allow extension of their use to new clinical areas, particularly in adjuvant therapy following resection of high-risk neoplasms and in treatment of clinically apparent disease. Experimental data support the use of selective COX-2 inhibitors in combination with chemotherapeutic drugs and/or radiation, to enhance efficacy and possibly to mitigate the side effects of cytotoxic treatment. Large trials are underway or planned to determine whether pre-clinical data will translate to true patient benefit.
References 1 2
3
4 5 6
7
8
9
10
Turini ME, DuBois RN: Cyclooxygenase-2: A therapeutic target. Annu Rev Med 2002;53:35–57. Lipsky PE, Brooks P, Crofford LJ, DuBois R, Graham D, Simon LS, van de Putte LB, Abramson SB: Unresolved issues in the role of cyclooxygenase-2 in normal physiologic processes and disease. Arch Intern Med 2000;160:913–920. Eberhart CE, Coffey RJ, Radhika A, Giardello FM, Ferrenbach S, DuBois RN: Upregulation of cyclooxygenase-2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 1994;107:1183–1188. Blanke CD: Role of COX-2 inhibitors in cancer therapy. Cancer Invest 2002 (in press). Kazuhiko U, Ichikura T, Mochizuki H: Expression of cyclooxygenase-2 in human gastric adenomas and adenocarcinomas. J Surg Oncol 2001;76:26–30. Morris CD, Armstrong GR, Bigley G, Green H, Attwood SEA: Cyclooxygenase-2 expression in the Barrett’s metaplasia-dysplasia-adenocarcinoma sequence. Am J Gastroenterol 2001;96: 990–996. Hida T, Yatabe Y, Achiwa H, Muramatsu H, Kozaki KI, Nakamura S, Ogawa M, Mitsudomi T, Sugiura T, Takahashi T: Increased expression of cyclooxygenase-2 occurs frequently in human lung cancers, specifically in adenocarcinomas. Cancer Res 1998;58:3761–3764. Ohno R, Yoshingaga K, Fujita T, Hasegawa K, Iseki H, Tsunozaki H, Ichikawa W, Nihei Z, Sugihara K: Depth of invasion parallels increased cyclooxygenase-2 levels in patients with gastric carcinoma. Cancer 2001;91:1876–1881. Ratnasinghe D, Tangrea J, Roth MJ, Dawsey S, Hu N, Anver M, Wang QH, Taylor PR: Expression of cyclooxygenase-2 in human squamous cell carcinoma of the esophagus; An immunohistochemical survey. Anticancer Res 1999;19:171–174. Shiota G, Okubo M, Noumi T, Noguchi N, Oyama K, Takano Y, Yashima K, Kishimoto Y, Kawasaki H: Cyclooxygenase-2 expression in hepatocellular carcinoma. Hepatogastroenterology 1999;46:407–412.
COX-2 Inhibitors and Cancer
257
11 12
13
14
15
16
17 18
19 20 21
22 23
24
25 26
27 28
29
30
Wolff H, Saukkonen K, Anttila S, Karjalainen A, Vainio H, Ristimaki A: Expression of cyclooxygenase-2 in human lung carcinoma. Cancer Res 1998;58:4997–5001. Lim HY, Joo HJ, Choi JH, Yi JW, Yang MS, Cho DY, Kim HS, Nam DK, Lee KB, Kim HC: Increased expression of cyclooxygenase-2 protein in human gastric carcinoma. Clin Cancer Res 2000;6:519–525. Achiwa H, Yatabe Y, Hida T, Kuroishi T, Kozaki KI, Nakamura S, Ogawa M, Sugiura T, Mitsudomi T, Takahasi T: Prognostic significance of elevated cyclooxygenase-2 expression in primary, resected lung adenocarcinomas. Clin Cancer Res 1999;5:1001–1005. Tomozawa S, Tsuno NH, Sunami E: Cyclooxygenase-2 overexpression correlates with tumour recurrence, especially haematogenous metastasis, of colorectal cancer. Br J Cancer 2000;83: 324–328. Buskens CJ, Van Rees BP, Sivula A, Reitsma JB, Haglund C, Bosma PJ, Offerhaus GJA, Van Lanshot JJB, Ristimaki A: Prognostic significance of elevated cyclooxygenase-2 expression in patients with adenocarcinoma of the esophagus. Gastroenterology 2002;122:1800–1807. Masferrer JL, Leahy KM, Koki AT, Zweifel BS, Settle SL, Woerner BM, Edwards DA, Flickinger AG, Moore RJ, Seibert K: Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 2000;60:1306–1311. Tsujii M, Kawano S, DuBois RN: Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA 1997;94:3336–3340. Moos PJ, Fitzpatrick FA: Taxane-mediated gene induction is independent of microtubule stabilization: Induction of transcription regulators and enzymes that modulate inflammation and apoptosis. Proc Natl Acad Sci USA 1998;95:3896–3901. Gemsa D, Kramer W, Brenner M, Till G, Resch K: Induction of prostaglandin E release from macrophages by colchicine. J Immunol 1980;124:376–380. Yeh CK, Rodan GA: Microtubule disruption enhances prostaglandin E2 production in osteoblastic cells. Biochim Biophys Acta 1987;927:315–323. Subbaramaiah K, Hart JC, Norton L, Dannenberg AJ: Microtubule-interfering agents stimulate the transcription of cyclooxygenase-2. Evidence for involvement of ERK1/2 AND p38 mitogenactivated protein kinase pathways. J Biol Chem 2000;275:14838–14845. Subbaramaiah K, Marmo TP, Dannenberg A: Taxotere induces cyclooxygenase-2 by enhancing transcription and mRNA stability. Proc Am Assoc Cancer Res 2002, abstr 343. Chodkiewicz C, Gomez JE, Ling YH, Perez-Soler R: Cyclooxygenase-2 expression and in vitro and in vivo sensitivity to paclitaxel in non-small cell lung cancer. Proc Am Assoc Cancer Res 2000, abstr 508. Hida T, Kozaki K, Muramatsu H, Masuda A, Shimizu S, Mitsudomi T, Sugiura T, Ogawa M, Takahashi T: Cyclooxygenase-2 inhibitor induces apoptosis and enhances cytotoxicity of various anticancer agents in non-small cell lung cancer cell lines. Clin Cancer Res 2000;6: 2006–2011. Teicher BA, Korbut TT, Menon K, Holden SA, Ara G: Cyclooxygenase and lipoxygenase inhibitors as modulators of cancer therapies. Cancer Chemother Pharmacol 1994;33:515–522. Moore RJ, Zweifel BS, Heuvelman DM, Leahy KM, Edwards DA, Woerner BM, Ornberg RL, Seibert K, Koki AT, Masferrer JL: Enhanced antitumor activity by co-administration of celecoxib and the chemotherapeutic agents cyclophosphamide and 5-FU. Proc Am Assoc Cancer Res 2000, abstr 2600. Trifan OC, Masferrer JL: Cyclooxygenase-2 inhibition with celecoxib enhances antitumor efficacy and reduces diarrhea side effect of CPT-11. Cancer Res 2002;62:5778–5784. Kase Y, Hayakawa T, Togashi Y, Kamataki T: Relevance of irinotecan hydrochloride-induced diarrhea to the level of prostaglandin E2 and water absorption of large intestine in rats. Jpn J Pharmacol 1997;75:399–405. Lorenz M, Slaughter HS, Wescott DM, Carter SI, Schnyder B, Dinchuk JE, Car BD: Cyclooxygenase-2 is essential for normal recovery from 5-fluorouracil-induced myelotoxicity in mice. Exp Hematol 2000;28:741–742. Kishi K, Petersen S, Petersen C, Hunter N, Mason K, Masferrer JL, Tofilon PJ, Milas L: Preferential enhancement of tumor radioresponse by a cyclooxygenase-2 inhibitor. Cancer Res 2000;60:1326–1331.
Blanke/Masferrer
258
31
32
33 34
35
36
37
38
39
40
41
42
43 44
45
46
Gaffney DK, Holden J, Davis M, Zempolich K, Murphy KJ, Dodson M: Elevated cyclooxygenase-2 expression correlates with diminished survival in carcinoma of the cervix treated with radiotherapy. Int J Radiat Oncol Biol Phys 2001;49:1213–1217. Formenti SC, Spicer D, Skinner K, Cohen D, Corso F, Bettini A, Muggia F, Danenberg K, Danenberg P: Pre-operative twice weekly paclitaxel and radiation therapy in locally advanced breast cancer: Molecular determinants of pathological response. Proc Am Soc Clin Oncol 1999, abstr 282. Rigau J, Pique JM, Rubio E, Planas R, Tarrech JM, Bordas JM: Effects of long-term sulindac therapy on colonic polyposis. Ann Intern Med 1991;115:952–954. Phillips RKS, Wallace MH, Lynch PM, Hawk E, Gordon GB, Saunders BP, Wakabayashi N, Shen Y, Zimmerman S, Godio L, Rodrigues-Bigas M, Su LK, Sherman J, Kelloff G, Levin B, Steinbach G, the FAP Study Group: A randomised, double-blind, placebo-controlled study of celecoxib, a selective cyclooxygenase-2 inhibitor, on duodenal polyposis in familial adenomatous polyposis. Gut 2002;50:857–860. Steinbach G, Lynch PM, Phillips RK, Wallace MH, Hawk E, Gordon GB, Wakabayashi N, Saunders B, Shen Y, Fujimura T, Su LK, Levin B: The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–1952. Csiki I, Dang T, Gonzalez A, Sandler A, Carbone D, Choy H, Campbell N, Garcia B, Morrow J, Johnson DH: Cyclooxygenase-2 inhibition ⫹ docetaxel in recurrent non-small cell lung cancer: Preliminary results of a phase II trial (THO-0054). Proc Am Soc Clin Oncol 2002;21:297a (abstr 1187). Carbone D, Choy H, Csiki I, Dang T, Campbell N, Garcia B, Morrow J, Saha D, Johnson DH, Sandler A: Serum/plasma VEGF level changes with cyclooxygenase-2 inhibition in combined modality therapy in stage III non-small cell lung cancer: Preliminary results of a phase II trial (THO-0059). Proc Am Soc Clin Oncol 2002;21:318a (abstr 1270). Altorki NK, Keresztes RS, Port JL, Flieder DB, Ferrara CA, Pasmantier MW, Libby DM, Yankelevitz D, Dannenberg AJ: Celecoxib, a selective COX-2 inhibitor, enhances the response to preoperative paclitaxel/carboplatin in early stage non-small cell lung cancer. Proc Am Soc Clin Oncol 2002;21:26a (abstr 101). Subbaramaiah K, Telang N, Ramonetti JT, Araki R, DeVito B, Weksler BB, Dannenberg AJ: Transcription of cyclooxygenase-2 is enhanced in transformed mammary epithelial cells. Cancer Res 1996;56:4424–4429. Dang CT, Dickler MN, Moasser MM, Theodoulou M, Seidman A, Norton L, Hudis CA: Celecoxib and trastuzumab is feasible after herceptin for HER-2/neu overexpressing metastatic breast cancer patients. Proc Am Soc Clin Oncol 2002;21:48b (abstr 2003). Oshima M, Dinchuk JE, Kargman SL, Oshima H, Hancock B, Kwong E, Trzaskos JM, Evans JF, Taketo MM: Suppression of intestinal polyposis in Apc ⌬716 knockout mice by inhibition of cyclooxygenase-2. Cell 1996;87:803–809. Fujita T, Matsui M, Takaku K, Uetake H, Ichikawa W, Taketo MM: Size- and invasion-dependent increase in cyclooxygenase-2 levels in human colorectal carcinomas. Cancer Res 1998;58: 4823–4826. Sinicrope FA, Pazdur R, Levin B: Phase I trial of sulindac plus 5-fluorouracil and levamisole: Potential adjuvant therapy for colon carcinoma. Clin Cancer Res 1996;2:37–41. Blanke CD, Benson AB III, Dragovich T, Lenz HJ, Haller D, Robles C, Buchbinder A: A phase II trial of celecoxib, irinotecan, 5-fluorouracil, and leucovorin in patients with unresectable or metastatic colorectal cancer. Proc Am Soc Clin Oncol 2002;21:127a (abstr 505). Sweeney C, Seitz D, Ansari R, Chowhan N, Pletcher W, Vinson J, Stoner C, Sawi J, Loehrer PJ: A phase II trial of irinotecan, 5-fluorouracil, leucovorin, celecoxib and glutamine as first-line therapy for advanced colorectal cancer: A Hoosier Oncology Group study. Proc Am Soc Clin Oncol 2002;21:105b (abstr 2235). Lin EH, Morris J, Chau NK, Crane C, Wolff R, Janjan N, Xiong H, Delcos M, Abbruzzese JL: Celecoxib attenuated capecitabine induced hand-and-foot syndrome and diarrhea and improved time to tumor progression in metastatic colorectal cancer. Proc Am Soc Clin Oncol 2002;21:138b (abstr 2364).
COX-2 Inhibitors and Cancer
259
47
48
49 50
51
52 53
54 55 56 57 58
59
60 61 62
Becerra CR, Frenkel EP, Gaynor R: A phase II study of 5-fluorouracil and leucovorin calcium plus open-label rofecoxib in patients with metastatic colorectal cancer. Proc Am Soc Clin Oncol 2002; 21:107b (abstr 2240). Tucker ON, Dannenberg AJ, Yang EK, Zhang F, Teng L, Daly JM, Soslow RA, Masferrer JL, Woerner BM, Koki AT, Fahey TJ 3rd: Cyclooxygenase-2 expression is up-regulated in human pancreatic cancer. Cancer Res 1999;59:987–990. Ding XZ, Tong WG, Adrian TE: Blockade of cyclooxygenase-2 inhibits proliferation and induces apoptosis in human pancreatic cancer cells. Anticancer Res 2000;20:2625–2631. Xiong HQ, Du M, Wolff RA, Lenzi R, Dumas P, Lassere Y, Plunkett W, Abbruzzese JL: Pharmacology study of celecoxib in combination with gemcitabine for advanced pancreatic cancer. Proc Am Soc Clin Oncol 2002;21:113a (abstr 448). Ryu HS, Chang KH, Yang HW, Kim MS, Kwon HC, Oh KS: High cyclooxygenase-2 expression in stage IB cervical cancer with lymph node metastasis or parametrial invasion. Gynecol Oncol 2000;76:320–325. http://www.RTOG.org Joki T, Heese O, Nikas DC, Bello L, Zhang J, Kraeft SK, Seyfried NT, Abe T, Chen LB, Carroll RS, Black PM: Expression of cyclooxygenase-2 in human glioma and in vitro inhibition by a specific COX-2 inhibitor, NS-398. Cancer Res 2001;60:4926–4931. Shono T, Tofilon PJ, Bruner JM, Owolabi O, Lang FF: Cyclooxygenase-2 expression in human gliomas: Prognostic significance and molecular correlations. Cancer Res 2001;61:4375–4381. http://www.mdanderson.org Gupta S, Srivastava M, Ahmad N, Bostwick DG, Mukhtar H: Over-expression of cyclooxygenase-2 in human prostate adenocarcinoma. Prostate 2000;42:73–78. Yoshimura R, Sano H, Masuda C, Kawamura M, Tsubouchi Y, Chargui J, Yoshimura N, Hla T, Wada S: Expression of cyclooxygenase-2 in prostate carcinoma. Cancer 2000;89:589–596. Chaudry AA, Wahle KW, McClinton S, Moffat LE: Arachidonic acid metabolism in benign and malignant prostatic tissue in vitro: Effects of fatty acids and cyclooxygenase inhibitors. Int J Cancer 1994;57:176–180. Liu XH, Yao S, Kirschenbaum A, Levine AC: NS398, a selective cyclooxygenase-2 inhibitor, induces apoptosis and down-regulates bcl-2 expression in LNCaP cells. Cancer Res 1998;58: 4245–4249. Liu XH, Kirschenbaum A, Yao S, Lee R, Holland JF, Levine AC: Inhibition of cyclooxygenase-2 suppresses angiogenesis and the growth of prostate cancer in vivo. J Urol 2000;164:820–825. http://www.ohsu.edu/oci/trials.shtml Kerbel RS, Klement G, Pritchard KI, Kamen B: Continuous low-dose anti-angiogenic metronomic chemotherapy: From the research laboratory into the oncology clinic. Ann Oncol 2002; 13:12–15.
Charles D. Blanke, MD Oregon Health & Science University, OHSU Oncology, MC L-586 3181 SW Sam Jackson Park Road, Portland, OR 97232 (USA) Tel. ⫹1 503 494 1556, Fax ⫹1 503 494 3257, E-Mail
[email protected]
Blanke/Masferrer
260
Dannenberg AJ, DuBois RN (eds): COX-2. Prog Exp Tum Res. Basel, Karger, 2003, vol 37, pp 261–283
Role of COX-Independent Targets of NSAIDs and Related Compounds in Cancer Prevention and Treatment Jae-Won Soh, I. Bernard Weinstein Department of Medicine, Herbert Irving Comprehensive Cancer Center, College of Physicians & Surgeons, Columbia University, New York, N.Y., USA
Background
Nonsteroidal anti-inflammatory drugs (NSAIDs) have been shown to have antitumor effects in various systems including in vitro cell cultures of human cancer cell lines, mouse and rat models of carcinogenesis, tumor growth inhibition assays in rodents and clinical trials [for reviews, see 1, 2]. The effects of various NSAIDs on inhibition of cyclooxygenase (COX)-1 and/or COX-2 enzymatic activity and the roles of COX-1 and COX-2 in tumorigenesis are reviewed in the other chapters of this book. However, the precise mechanisms by which various NSAIDs exert their antiproliferative effects on cancer cells are not known. Emerging evidence suggests that these effects can, at least in some cases, be exerted through COX-independent mechanisms. In this chapter, we review recent progress in the identification of novel molecular targets of NSAIDs and related compounds. Hopefully, elucidation of these molecular targets and the downstream signaling pathways that inhibit cell proliferation and/or induce apoptosis will facilitate the development of more effective approaches for cancer prevention and treatment, while minimizing potential toxicities. Evidence for COX-Independent Mechanisms
Lack of Correlation between COX Expression Status and Antiproliferative Effects of Various NSAIDs in Cell Culture Systems Several independent studies found that various NSAIDs and related compounds can exert antiproliferative and/or apoptotic effects in cell lines irrespective of their levels of expression of COX-1 or COX-2. For example,
indomethacin (a nonselective COX inhibitor) induced apoptosis in both Seg-1 (COX-1/2-positive) and Flo-1 (COX-1/2-negative) esophageal adenocarcinoma cells [3]. Sulindac sulfide and sulindac sulfone also induced apoptosis in malignant melanoma cell lines independent of COX-2 expression [4]. Indomethacin and NS-398 (a selective COX-2 inhibitor) had antiproliferative activity on both COX-2-positive cell lines (HT29.Fu and HCA-7) and COX-2-negative cell lines (SW480 and HCT116) [5]. Sulindac sulfide and piroxicam induced apoptosis in both COX-2-expressing HT-29 human colon cancer cells and COX-2-deficient HCT-15 cells [6]. Treatment of HCT-15 cells with PGE2, PGF2␣, or PGI2 did not reverse the apoptotic effects of these NSAIDs. Celecoxib, a COX-2-selective inhibitor, showed antiproliferative effects for both hematopoietic and epithelial cancer cell lines regardless of the COX-2 expression status of the cell lines [7]. Indeed, most hematopoietic cell lines are COX-2-negative. COX-2-negative epithelial lines were found to have IC50s for celecoxib that were very similar to the COX-2-positive epithelial lines [7]. Furthermore, celecoxib and rofecoxib have similar potency for in vitro inhibition of the COX-2 enzyme, but celecoxib has much higher antiproliferative activity in COX-2-positive A549 epithelial cells and COX-2-negative BALL-1 hematopoietic cells than rofecoxib [7]. NS-398, a COX-2-selective inhibitor, induced apoptosis in HT-29 (COX-2-positive) and S/KS (COX-2-negative) human colorectal carcinoma cell lines with comparable IC50s [8]. Growth inhibition of various colorectal cancer cell lines (SW480, HT29/HI1, VACO235 and LT79) by various COX-1 or COX-2 inhibitors (COX-1 inhibitors SC560 and sulindac sulfide, COX-2 inhibitors SC236 and SC125) was also independent of COX-1 or COX-2 levels [9]. In addition to the above studies with a spectrum of cancer cell lines, it has also been demonstrated that cells genetically engineered to lack expression of COX-1 or COX-2 can remain sensitive to the antiproliferative effects of NSAIDs. For example, depletion of cellular COX-2 by an inducible COX-2 antisense cDNA expression did not itself induce apoptosis in PC-3 prostate cancer cells, whereas COX-2 inhibitors (celecoxib, rofecoxib, NS-398, and DuP697) did induce apoptosis in the same cell line [10]. Furthermore, sensitivity to COX-2 inhibitor-induced apoptosis was independent of the level of COX-2 expression in the antisense clones. Several celecoxib derivatives, although lacking COX-2 inhibitory activity, were as potent in eliciting apoptosis in PC-3 cells as the parent compound [10]. Mouse fibroblasts from COX-1⫺/⫺, COX-2⫺/⫺ or COX-1/2⫺/⫺ knockout mice were sensitive to apoptosis induced by various NSAIDs including indomethacin, sulindac and NS-398 [11, 12]. Evidence from a Mouse Model of Intestinal Tumorigenesis Mutant Min mice have a heterozygous mutation in the APC gene and spontaneously develop a high incidence of intestinal tumors. Administration of
Soh/Weinstein
262
sulindac to Min/⫹ mice significantly reduced the tumor number but did not alter the levels of prostaglandin E2 and leukotriene B4 in intestinal tissues. Furthermore, increasing prostaglandin E2 and leukotriene B4 levels with dietary arachidonic acid supplementation had no effect on tumor number or size [13]. Since alterations in eicosanoid formation did not correlate with tumor number or size in this mouse model, it appears that the antitumor effect of sulindac in this model system is not mediated via inhibition of COX-1 or COX-2. NSAID Derivatives That Lack COX Inhibitory Activity Can Have Antiproliferative Activity Another line of evidence for the existence of COX-independent mechanisms for the antiproliferative effects of NSAIDs comes from the finding that there are various chemical derivatives of conventional NSAIDs which lack COX inhibitory activities but still retain antiproliferative and antitumor activities. For example, the R(⫺) enantiomer of ibuprofen which is devoid of COX inhibitory activity is as potent in inhibiting PDGF-induced mitogenesis of smooth-muscle cells as the S(⫹) enantiomer of ibuprofen which has COX inhibitory activity [14]. The S-enantiomer of flurbiprofen is a potent inhibitor of both COX-1 and COX-2 enzyme activity and has anticancer activity [15]. However, the R-enantiomer of flurbiprofen, which does not inhibit COX activity, still has chemopreventive activity in the Min/⫹ mouse model of intestinal polyposis [16] and in the transgenic adenocarcinoma mouse prostate (TRAMP) cancer model system [17]. NCX-4016, a nitric oxide-releasing aspirin derivative without any COX inhibitory activity, exhibited stronger chemopreventive activity than aspirin in colonic crypt foci inhibition assays in rats [18]. Sulindac sulfone (exisulind) is the oxidative metabolite of sulindac and this compound is completely devoid of COX-1 or COX-2 inhibitory activity. However, sulindac sulfone inhibits chemical carcinogenesis in rodents, and inhibits growth and induces apoptosis in a variety of human cancer-derived cell lines [19–24]. Taken together, the above studies provide strong evidence that the antiproliferative and antitumor effects of various NSAIDs and related compounds can be mediated, at least in some systems, by mechanisms independent of the inhibition of either COX-1 or COX-2. In the following section we discuss alternative direct cellular targets for these agents, and related signaling pathways.
COX-lndependent Targets and Mechanisms
Direct Cellular Targets There is accumulating evidence that certain NSAIDs and related compounds exert various biological effects, including antitumor effects, by binding
COX-Independent Targets of NSAIDs and Related Compounds
263
directly to and inhibiting the activities of various cellular molecules other than COX-1 or COX-2. These direct targets include PDE2/5, IKK, PDK1, RSK2, Ras, PPARs and collagenase (table 1). These findings are briefly summarized below. PDE2/5 Phosphodiesterases are a large and genetically diverse family of enzymes that catalyze the hydrolysis of cAMP or cGMP to biologically less active 5⬘-nucleoside monophosphates [25, 26]. In recent studies, it was discovered that the sulindac metabolite sulindac sulfone (exisulind) inhibits guanosine 3⬘,5⬘-monophosphate (cGMP)-specific phosphodiesterases (PDE) 2 and 5, thus leading to sustained increases in cellular levels of cGMP, and thereby activation of cGMP-dependent protein kinase (PKG) [27]. New derivatives (CP248, CP461 and others) of sulindac sulfone were synthesized based on this PDE2/5 inhibitory activity and, in general, these derivatives maintain a similar rank order of potency for PDE inhibition, apoptosis induction and growth inhibition of colon cancer cell lines [27]. Furthermore, they do not inhibit COX-1 or COX-2. The IC50 values for sulindac sulfone, CP461 and CP248 for PDE2/5 inhibition are 128, 3.6 and 0.3 M, respectively. Conventional selective PDE5 inhibitors such as E4021, zaprinast and sildenafil increase intracellular levels of cGMP and lead to inhibition of vascular smooth-muscle cell proliferation, prevention of platelet aggregation and vasodilation [28, 29] but do not induce apoptosis in cancer cells, perhaps because of differences in the kinetics of PDE inhibition by these compounds. Thus, sulindac sulfone and its derivatives activate a novel mechanism of induction of apoptosis. In human colon cancer cells, treatment with sulindac sulfone is associated with downregulation of -catenin, apparently due to direct phosphorylation by PKG [27]. It also caused decreased levels of cyclin D1. These effects apparently play an important role in the apoptotic process. Additional studies indicate that sulindac sulfone and the derivatives CP248 and CP461, as well as the compound YC-1, which increases cellular levels of cGMP by activating guanylate cyclase, lead to rapid and sustained activation of JNK1 (c-Jun N-terminal kinase 1), a kinase known to play a role in the induction of apoptosis by other stress-related events. Mechanistic studies indicate the existence of a novel PKG-MEKK1-SEK1-JNK1 pathway, and that this pathway is required for the induction of apoptosis by sulindac sulfone and related derivatives [30] (fig. 1). Transient transfection of an activated mutant of PKG activates JNK1 in vivo and a purified preparation of the PKG enzyme directly phosphorylates and activates MEKK1 in vitro [31]. Once activated, JNK1 plays a role in various apoptotic signaling pathways. Thus, JNK1 can phosphorylate and inactivate the anti-apoptotic proteins Bcl-2 and Bcl-XL [32, 33], and it can
Soh/Weinstein
264
Sodium salicylate
Indomethacin
Ibuprofen
Sulindac
Sulindac sulfide
Sulindac sulfone
Celecoxib
↓ ↓
– –
↓ ↓
↓ ↓
↓ ↓
↓ ↓
– –
– ↓
References
Aspirin
Table 1. Cellular targets of NSAIDs and related compounds
COX COX-1 COX-2
COX-independent direct targets PDE2/5 IKK ↓ ↓ PKD1 RKS2 ↓ ↓ Ras PPAR␥ – PPAR␦ Collagenase Indirect target Kinase ERK1/2 JNK P38 Akt P70S6K TF NF-B AP-1 Expression 15-Lox-1 Bcl-X L NAG-1 Nur77 Bax -catenin Cyclin D1 Cyclin A p21Cip1 p27Kip1
↓
↓ ↑↓ ↑
–
–
↓
↓
↓ ↓
↓
↓
↓ –
↑ ↓
↑
↓
–
–
↓
↓
↑
↑ ↓
↓ ↓
↓
↓
↓
↓
–
↓ ↑ ↑ ↑ ↓ ↓ ↑ ↑
↓
↓
↓
↓
↓
↓ ↓
↓
↑
↑
↓ ↓ ↑ ↑
[27, 30] [47, 48] [50] [53] [55] [58] [62] [65]
[67, 68] [30, 69, 70] [70] [75] [77] [47, 48, 79, 81–85] [87, 88] [89] [110] [93] [94] [3] [27, 98] [27, 77, 104] [77] [99, 102] [99, 102]
TF ⫽ Transcription factor; ↓ ⫽ inhibition; ↑ ⫽ activation; – ⫽ no change; blank ⫽ not determined (for additional details see text).
COX-Independent Targets of NSAIDs and Related Compounds
265
YC-1
Guanylate cyclase GTP
cGMP
Other targets (VASP, HSP27)
PKG
PDE2/5
Sulindac sulfone and derivatives
Decreased -catenin
MEKK1
SEK1
JNK1
Direct inactivation of Bcl-2 and Bcl-XL and alterations in gene expression through AP-1, etc.
Downregulation of cyclin D1
Growth inhibition and apoptosis
Fig. 1. Pathways by which inhibition of PDE2/5 and activation of PKG can lead to apoptosis.
also induce the expression of proapoptotic proteins through activation of the transcription factor AP-1. Activation of PKG leads to the induction of apoptosis in several types of cancer cells [27, 34]. In addition to -catenin [27] and MEKK1 [31], other PKG substrates have been identified, which include vasodilator-stimulated phosphoprotein (VASP) [35, 36], heat-shock protein HSP27 [37], inositol 1,4,5-trisphophate receptor-associated cGMP kinase substrate (IRAG) [38], cysteine-rich protein 2 (CRP2) [39] and cGMP-specific phosphodiesterase 5 [40, 41]. The possible roles of these PKG substrates in mediating the growth inhibitory and apoptotic effects induced by sulindac sulfone and related compounds remain to be determined.
Soh/Weinstein
266
Tumor necrosis factor (TNF), inflammation, infection or cellular stress
IKK
(Inactive)
NF-B
IB
(Active)
NF-B
⫹
NSAIDs
IB
Nuclear translocation and gene expression
p
Degradation
Upregulation of proliferative genes (IL-2, GM-CSF, COX-2, c-myc, cyclin D1) and anti-apoptotic genes (c-FLIP and Bcl-XL ), thus leading to increased cell proliferation and survival
Fig. 2. Signal transduction pathways of IKK and NF-B.
IKKb/NF-kB NF-B is a transcription factor composed of p50 and p65 (Rel A) subunits which exist as inactive forms in the cytoplasm when complexed with the inhibitory regulatory protein IB-␣ (fig. 2). Phosphorylation of serines 32 and 36 of IB-␣ by the protein kinase IKK leads to ubiquitination and proteasomal degradation of IB-␣ [42, 43]. Free NF-B can then enter the nucleus and enhance cell proliferation and survival by activating several target genes, including interleukin-2 (IL-2), granulocyte-macrophage colony-stimulating factor (GM-CSF), COX-2, c-myc and cyclin D1. NF-B can also inhibit apoptosis by activating anti-apoptotic genes, including cellular inhibitors of apoptosis (cIAPs), c-FLIP and Bcl-XL. IB kinase (IKK) phosphorylates IB and the phosphorylated IB protein is then ubiquitinated and degraded by proteasome-mediated proteolysis resulting in the nuclear translocation of NK-B as illustrated in figure 2 [44, 45].
COX-Independent Targets of NSAIDs and Related Compounds
267
NF-B is comprised of a family of related transcription factors that are involved in the inducible expression of a variety of cellular genes that regulate the inflammatory response and cellular stress pathways [46]. Various NSAIDs including aspirin (5 mM), sodium salicylate (5 mM), sulindac (1 mM), sulindac sulfide (200 M) and sulindac sulfone (1 mM) can directly inhibit IKK kinase activity, in vivo and in vitro [47, 48]. The IC50 of aspirin for in vitro IKK inhibition is about 50 M. Aspirin and sodium salicylate were shown to inhibit IKK by directly binding to the ATP binding site of the enzyme. However, indomethacin (25 M) and ibuprofen (25 M) do not have IKK inhibitory activity [47]. This ability of several NSAIDs and sulindac sulfone to directly inhibit the kinase activity of IKK may explain the findings, discussed in greater detail below, indicating that these compounds can also inhibit the activation and transcriptional activity of NF-B, and thereby contribute to growth inhibition and apoptosis. PDK1 Several growth factors lead to the activation of phosphatidylinositol 3-kinase (PI3K), which then activates the protein 3-phosphoinositide-dependent kinase 1 (PDK1). This serine/threonine kinase phosphorylates and activates Akt which plays a critical role in cell proliferation and survival, as illustrated in figure 3 [49]. Evidence was recently obtained that celecoxib (100 M), a COX-2-specific inhibitor, induces apoptosis in the HT-29 colon cancer cell line by directly inhibiting the activity of PDK1 [50]. This effect correlated with inhibition of phosphorylation of the PDK1 downstream substrate Akt. Constitutively activated PDK1 prevented the inhibitory effect of celecoxib on Akt phosphorylation and protected HT-29 cells from celecoxib-induced apoptosis. The IC50 of celecoxib for PDK1 inhibition was 3.5 M when measured in a subcellular assay using inactive serum and glucocorticoid-regulated kinase (SGK) as substrates of PDK1. These data suggest that celecoxib exerts its proapoptotic effect, at least in part, by impairing the PDK1-Akt signaling pathway. RSK2 The 90-kDa ribosomal S6 kinase 2 (RSK2) is a serine/threonine protein kinase that is activated by mitogen-activated protein kinase (MAPK) [51]. Once activated, it can phosphorylate and activate c-Fos and serum response factor (SRF) [52]. RSK2 plays a critical role as an effector of the RAS-MAPK pathway and as a regulator of immediate early gene transcription. Various NSAIDs including aspirin (3 mM), sodium salicylate (20 mM), ibuprofen (2 mM), indomethacin (0.3 mM), sulindac (0.6 mM) and diclofenac (0.3 mM) were demonstrated to have inhibitory effects on the subcellular
Soh/Weinstein
268
Growth factors
PI3K
PDK1
Celecoxib
Akt
Phosphorylation of p70S6K, IB-␣, CREB; Bad, Caspase-9
Increased proliferation, inhibition of apoptosis and cell survival
Fig. 3. Signal transduction pathways related to PI3 kinase, PDK1 and Akt.
kinase activity of RSK2 [53]. In human monocytes, these NSAIDs suppressed phosphorylation of the RSK2 substrates cAMP response element binding protein (CREB) and IB-␣. Furthermore, in NIH3T3 cells sodium salicylate inhibited the phosphorylation of CREB and IB-␣ by RSK2 on residues crucial for their transcriptional activity and thereby inhibited CREB and NF-B-dependent transcriptional activity. Ras The Ras proteins are a family of small guanosine nucleotide triphosphatases (GTPases) that undergo a guanine nucleotide-dependent change in conformation [54]. Ras proteins are active in the GTP-bound state and are inactive when bound to GDP. Activation of Ras GTPase activity is mediated by the action of guanine nucleotide exchange factors (GEFs) that catalyze the exchange of GDP for GTP, whereas GTPase-activating proteins (GAPs) promote intrinsic GTP hydrolysis,
COX-Independent Targets of NSAIDs and Related Compounds
269
thus leading to the GDP-bound state. In their active GTP-bound state, Ras proteins bind to and activate several downstream effectors including c-Raf. In HEK293 cells, sulindac sulfide (250 M) was shown to bind directly to the Ras protein and inhibit its interaction with the Ras binding domain of c-Raf, thus inhibiting Ras induced activation of the kinase activity of c-Raf-1 [55]. In addition, sulindac sulfide inhibited the nucleotide exchange on Ras by p21rasGEF (CDC25) and also inhibited the acceleration of the Ras GTPase reaction by p120GAP [55]. These effects could contribute to the growth inhibitory and apoptotic effects of sulindac sulfide and related compounds. PPAR␣, ␥ and ␦ Peroxisome proliferator-activated receptors ␣, ␥ and ␦ (PPAR␣, ␥ and ␦) are members of a class of nuclear hormone receptors involved in controlling the transcription of various genes that regulate energy metabolism, cell differentiation, apoptosis and inflammation [56]. PPARs bind to sequence-specific DNA response elements as a heterodimer with the retinoic acid receptor (RXR). Ligands for PPAR␥ suppress breast carcinogenesis in experimental models and induce differentiation of human liposarcoma cells [57]. Indomethacin (100 M), flufenamic acid (100 M), fenoprofen (100 M) and ibuprofen (100 M) activate PPAR␣ and ␥ [58] in monkey kidney epithelium-derived CV-1 cells, while acetaminophen (100 M) and salicylic acid (100 M) did not show any effect. Some of these NSAIDs are likely to be direct ligands for PPAR␣ and ␥ since indomethacin can bind to PPAR␥ in vitro. On the other hand, diclofenac was a weak PPAR␥ agonist when tested alone, but antagonized rosiglitazone-induced PPAR␥ signaling [59]. PPAR␦ is a nuclear transcription factor that is activated by prostacyclin (PGI2), a COX-2-derived prostaglandin [60, 61]. PPAR␦ is often overexpressed in colon cancer cells [61] and PPAR␦ expression is negatively regulated by the APC tumor suppressor pathway through -catenin/Tcf-4-responsive elements in the PPAR␦ promoter [62]. Sulindac sulfide (200 M) and indomethacin (400 M) suppressed PPAR␦ activity in HCT116 and SW480 human colon cancer cells by disrupting DNA binding ability of the PPAR␦/RXR␣ heterodimer [62]. Furthermore, overexpression of PPAR␦ partially rescued colon cancer cells from sulindac sulfide-induced apoptosis. These findings suggest that the NSAIDs sulindac and indomethacin can mimic the effects of APC by downregulating the transcriptional activity of PPAR␦. On the other hand, studies with PPAR␦⫺/⫺ HCT116 cells revealed that PPAR␦ is not required for sulindac sulfide-induced apoptosis since there was no difference in sensitivity to sulindac sulfide between PPAR␦⫺/⫺ and PPAR␦⫹/⫺ cells. However, it appears that PPAR␦ is required for efficient tumorigenicity in nude mice since PPAR␦⫺/⫺ HCT116 cell lines were defective in establishing tumors when grown as
Soh/Weinstein
270
xenografts in nude mice [63]. Taken together, these results suggest that PPAR␦ may be an important but not essential antitumor target for certain NSAIDs. Thus, further studies are required to clarify the possible roles of PPARs ␣, ␥, and ␦ in modulating the effects of various NSAIDs on growth and apoptosis in cancer cells. Collagenase Matrix metalloproteases (MMP) such as collagenase and stromelysin play important roles in enhancing the degradation of collagen and proteoglycan in the extracellular matrix that occurs in the joints of patients with rheumatoid arthritis and osteoarthritis [64]. Nimesulide (2 M), piroxicam (14 M), meloxicam (15 M), sulindac (28 M) and sodium meclofenamate (27 M) directly inhibit collagenase type XI enzyme activity in vitro [65]. This inhibition is reversible because most of the collagenase activity is restored after dialysis of the enzyme-drug complex. These data suggest that some NSAIDs might exert their therapeutic effects in patients with arthritis, at least in part, by inhibiting enzymes involved in degradation of the extracellular matrix. This mechanism might also contribute to the in vivo antitumor effects of some of these agents since MMPs play an important role in tumor invasion and metastasis. Indirect Cellular Targets In addition to the evidence summarized above that NSAIDs and related compounds can directly target several cellular molecules that differ with respect to their specific biochemical functions, there is accumulating evidence that treatment of cells with these compounds can indirectly induce changes in the levels or activities of various protein kinases, transcription factors and cell cycle control proteins, and thereby alter pathways of signal transduction, profiles of gene expression and cell cycle control. These indirect effects are summarized in table 1 and discussed below. Protein Kinases Mitogen-activated protein kinases (MAPKs) are proline-directed, serinethreonine kinases that can be activated by various extracellular stimuli. Major substrates of MAPKs are transcription factors such as c-Fos, c-Jun, Elk-1 and ATF-2 [66]. Upon activation by upstream kinases, the three classes of MAPKs – (1) extracellular signal-regulated kinases (ERKs, p42/p44 MAPK), (2) c-Jun N-terminal kinases (JNKs) and (3) p38 MAPKs – translocate into the nucleus, phosphorylate various transcription factors and thereby modulate gene expression. Treatment of human fibroblasts with sodium salicylate (20 mM) blocked the activation of ERKs in response to TNF [67]. Aspirin (10 mM) and sodium
COX-Independent Targets of NSAIDs and Related Compounds
271
salicylate (10 mM) also inhibited ERK activity in neutrophils [68]. Aspirin (20 mM) and sodium salicylate (20 mM) caused activation of JNK in COS-1 monkey kidney cells and in HT-29 human colon cancer cells, but sodium salicylate (20 mM) inhibited TNF-induced JNK activation in FS-4 human fibroblasts [69, 70]. Activation of JNK was not observed with ibuprofen, acetaminophen or indomethacin in COS-1 cells. Sodium salicylate (20 mM) also caused activation of p38 MAPK in COS-1 cells [70, 71]. The activation of p38 MAPK by sodium salicylate correlated with its inhibitory effect on TNFinduced IB-␣ phosphorylation and degradation. As discussed above, activation of PKG by sulindac sulfone can lead indirectly to activation of JNK1 through phosphorylation of MEKK1 (see also figure 1). This activation occurs rapidly (within 30 min) and is persistent for at least 24 h [30], which is characteristic of JNK1 activation associated with apoptosis. Determination of cell survival versus cell death is known to depend on the balance between intracellular cell proliferation/survival and death signaling pathways [72]. ERK activation has been implicated mainly in growth promoting and survival signaling, whereas JNK and p38 MAPK activation have been implicated mainly in stress response signaling and apoptosis. The ability of various NSAIDs to indirectly modulate MAPK activities in cancer cells may play an important role in the cytotoxicity and apoptosis induced by various NSAIDs and related compounds. Akt. Akt or protein kinase B (Akt/PKB) is a serine/threonine kinase that can be activated by diverse stimuli such as hormones, growth factors, and extracellular matrix components [73, 74]. Akt signaling is believed to promote cell proliferation and survival, thereby contributing to cancer progression. PI3K generates phosphatidylinositol-3,4,5-trisphosphate (PIP3), a lipid second messenger necessary for the translocation of Akt to the plasma membrane (fig. 3). As discussed above, Akt is then phosphorylated and activated by PDK1 at the plasma membrane. Activated Akt then phosphorylates and regulates the function of several cellular proteins involved in metabolism (inhibition of GSK3), apoptosis (inhibition of Bad and caspase-9), and proliferation (activation of p70S6K, nitric oxide synthase, IB-␣ and CREB). Evidence has been obtained that the COX-2 inhibitor celecoxib (50 M) induces apoptosis in LNCaP and PC-3 human prostate cancer cells by blocking the phosphorylation of Akt independently of Bcl-2 [75]. Indeed, overexpression of constitutively active Akt protected PC-3 cells from celecoxib-induced apoptosis. These data are consistent with the finding that celecoxib (100 M) inhibits PDK1 activity [50], as described above. p70S6K. p70S6K is a serine/threonine kinase that phosphorylates the 40S ribosomal protein S6, in response to various factors that stimulate cell growth [76]. In addition to its involvement in regulating protein translation, p70S6K
Soh/Weinstein
272
activation has been implicated in cell cycle control and neuronal cell differentiation. Sodium salicylate (10 mM) was reported to inhibit phorbol ester- or EGF/IGF-1-induced p70S6K activation in Balb/MK cells [77]. The inhibition of p70S6K was associated with inhibition of DNA synthesis and protein synthesis. The mechanism by which sodium salicylate treatment leads to inhibition of p70S6K is not known, but could be related to above-described effects of NSAIDs on Akt. Transcription Factors NF-kB. In the above section on Direct Cellular Targets, we reviewed the NF-B pathway (see also figure 2) and the evidence that some NSAIDs and related compounds can directly inhibit the kinase activity of IKK, and thereby inhibit NF-B activity. This section reviews several additional studies in which these compounds have been reported to inhibit NF-B activity in various types of cell systems, but the direct targets were not identified. For example, aspirin (5 mM) and sodium salicylate (5 mM) inhibited the DNA binding activity of NF-B and also NF-B-dependent transcriptional activity in human Jurkat T cells and PD31 mouse pre-B cells [78, 79]. However, acetaminophen and indomethacin did not inhibit NF-B activity in these cells [79]. TNF-␣-induced IB-␣ phosphorylation and degradation were inhibited by sodium salicylate (20 mM) in PANC-1 and BxPC-3 human pancreatic cancer cell lines [80]. Sodium salicylate (20 mM) also inhibited TNF-␣-induced phosphorylation of IB-␣ and activation of NF-B in HUVEC human epithelial cells [81]. Both the R- and S-enantiomers of flurbiprofen (1 mM) inhibited NF-B activity in RAW 264.7 mouse macrophages [82]. It is of interest that the R-enantiomer of flurbiprofen does not inhibit COX activity at therapeutically relevant concentrations [82]. Ibuprofen (3 mM) inhibited the activation and nuclear translocation of NF-B by blocking the degradation of IB-␣ in U937 human peripheral mononuclear cells [83]. Ibuprofen (2 mM) also inhibited the constitutive activation of NF-B and IKK␣ in androgen-independent PC-3 and DU-145 prostate tumor cells, and blocked TNF-␣-stimulated activation of NF-B in androgen-sensitive LNCaP prostate tumor cells [84]. NF-B is constitutively activated in the PC-3 and DU-145 but not in the LNCaP cell line. Celecoxib showed no effect on nuclear concentrations of NF-B at 1–10 M, but slightly reduced the degradation of IB in the zymosan-induced hindpaw inflammation model in rats. However, in this model system celecoxib (50 M) enhanced IL-1-stimulated nuclear translocation of NF-B and led to complete degradation of cytosolic IB and upregulated expression of the COX-2, iNOS and TNF-␣ proteins [85]. It remains to be determined whether high concentrations of celecoxib or other NSAIDs can enhance rather than inhibit NF-B activity in other cell systems. Further studies are also required to determine whether NSAIDs
COX-Independent Targets of NSAIDs and Related Compounds
273
and related compounds modulate NF-B activity entirely through their effects on IKK or whether other molecules in this pathway provide direct targets. AP-1. The transcription factor AP-1 is a dimer composed of Jun, Fos or ATF. AP-1 is activated in response to a variety of stimuli, including mitogenic growth factors, inflammatory cytokines, growth factors of the TGF- family, UV and ionizing irradiation, cellular stress, antigen binding, and neoplastic transformation [86]. Aspirin (2 mM) and sodium salicylate (2 mM) inhibited AP-1 activity and also inhibited tumor promoter-induced cell transformation through a prostaglandin-independent pathway in JB6 mouse epidermal cells [87]. Indomethacin showed no effect on TPA-induced AP-1 activity in these cells. The R- and S-enantiomers of flurbiprofen (1 mM) inhibited AP-1 activity in RAW 264.7 mouse macrophages [82]. Aspirin (2 mM) and sodium salicylate (2 mM) inhibited UVB-induced AP-1 activity in the JB6 mouse epidermal cells [88]. These inhibitory effects on UVB-induced AP-1 activity appear to be mediated through inhibition of members of the MAP kinase family of proteins, including ERKs, JNKs and p38 MAPKs, but the direct targets have not been identified. As discussed above, sulindac sulfone and related compounds can activate AP-1 through a novel PKG-MEKK1-SEK1-JNK1 pathway (fig. 1). Gene and Protein Expression NSAIDs have been reported to induce either upregulation or downregulation of the expression of various genes during their induction of growth inhibition and apoptosis in cancer cell lines. Some of these changes in gene expression appear to be critical to the antiproliferative effects of these compounds. However, the precise mechanisms that modulate these changes in cellular levels of gene expression and their roles in antitumor effects remain to be elucidated. 15-Lox-1. The induction of apoptosis by sulindac sulfone and NS-398 in DLD-1 (COX-2-negative) colon cancer cells is associated with upregulation of cellular levels of 15-lipoxygenase-1 (15-Lox-1) [89, 90]. 15-Lox-1 is the major enzyme that metabolizes linoleic acid to 13-S-hydroxyoctadecadienoic acid (13-S-HODE). Inhibition of 15-Lox-1 activity by caffeic acid blocked the apoptosis induced by sulindac sulfone or NS-398, and apoptosis was restored by the addition of 13-S-HODE the metabolite of 15-Lox-1. These data suggest that the induction of 15-Lox-1 by certain NSAIDs and related compounds plays an important role in apoptosis, independent of the activity of COX-2, in colon cancer cells, presumably by increasing cellular levels of 13-S-HODE. The mechanism by which sulindac sulfone and NS-398 induce 15-Lox-1 is not known. Bax and Bcl-XL. Members of the Bcl-2 family of proteins are important regulators of programmed cell death. Bcl-2 and Bcl-XL suppress apoptosis
Soh/Weinstein
274
while Bax and Bad promote apoptosis [91]. The balance between Bcl-2/Bcl-XL and Bax/Bad control mitochondria-mediated pathways of apoptosis. Sulindac sulfide (120 M) and indomethacin (500 M) induced downregulation of Bcl-XL without changing the level of Bax in HCT116 cells, thus resulting in an altered ratio of Bax to Bcl-XL and enhancement of mitochondria-mediated apoptosis. However, these NSAIDs were not able to induce apoptosis in Baxnegative HCT116 cells, suggesting that Bax is required for NSAID-mediated apoptosis in colon cancer cells. In another study, indomethacin induced apoptosis in several esophageal adenocarcinoma cell lines and this effect was associated with upregulation of Bax and translocation of mitochondrial cytochrome C to the cytoplasm [3]. NAG-1. Baek et al. [92] isolated a cDNA named NAG-1 (NSAIDactivated gene) from an indomethacin-induced mRNA library of HCT-116 human colon cancer cells, which are devoid of COX enzymes. NAG-1 has proapoptotic and anti-tumorigenic activity in cell culture and tumorigenesis assays. It is induced by various NSAIDs including indomethacin (100 M), aspirin (10 mM) and ibuprofen (500 M), independent of exogenous COX overexpression or the presence of PGE2 or arachidonic acid [93]. Nur77. Indomethacin (1 mM) was reported to induce transcription of Nur77 during indomethacin-induced apoptosis in HCT-15 colon cancer cells [94]. Nur77 is an orphan member of the steroid/thyroid receptor superfamily and also an immediate-early response gene. Overexpression of a dominantnegative Nur77 or inhibition of Nur77 expression by antisense Nur77 inhibited T-cell receptor-induced apoptosis, and constitutive expression of Nur77 induced apoptosis [95, 96]. All-trans-retinoic acid blocked the induction of Nur77 and also the apoptosis induced by indomethacin. b-Catenin. -Catenin plays an important role in cell adhesion by binding to cadherins at the intercellular surface of the plasma membrane. -Catenin also plays a signaling role in the cytoplasm as the downstream mediator of the Wnt signaling pathway [97]. In colon cancer cells, -catenin degradation is blocked either by mutations in -catenin itself or in the tumor suppressor gene product APC which normally increases the degradation of -catenin. As a result, the stabilized -catenin accumulates in the nucleus and forms complexes with LEF/TCF transcription factors, thus enhancing the expression of various target genes including c-myc and cyclin D1. Aspirin decreased intracellular levels of -catenin and prevented tumor formation in a murine model of familial adenomatous polyposis [98]. Indomethacin (600 M) also induced downregulation of -catenin in either COX-2-positive HT29.Fu or COX-2-negative SW480 human colon cancer cells [5]. However, aspirin (1.5 mM) and NS-398 (600 M) did not induce downregulation of -catenin in this study. As discussed above, activation of PKG in cells treated with sulindac sulfone or related
COX-Independent Targets of NSAIDs and Related Compounds
275
compounds leads to downregulation of -catenin, apparently because PKG directly phosphorylates -catenin and this enhances its degradation [27]. Effects on the Cell Cycle As discussed above, sulindac sulfone causes decreased levels of cyclin D1 in SW480 cells [27]. The COX-1-specific inhibitor SC560 and the COX-2specific inhibitor celecoxib induce apoptosis in COX-2-deficient HCT-15 cells as well as in COX-1/2-expressing HT-29 and Caco-2 colon cancer cells [99]. Treatment with celecoxib was associated with a G0/G1 cell cycle arrest and increased levels of the cyclin-dependent kinase (CDK) inhibitors p21Cip1 and p27Kip1 in all of these cell lines. Aspirin induced cell cycle arrest in the S and G2/M phases and induced necrosis in SW620 and HT-29 human colonic tumor cells, independent of prostaglandin production [100]. In another study, sodium salicylate (10 mM) induced growth inhibition and concomitant downregulation of c-myc, cyclin D1, PCNA and cyclin A in Balb/MK cells [77]. Sodium salicylate inhibited serum-induced progression from the G1 to S phase, cellular proliferation and the expression of cyclin D1 in BxPC3 and Panc-1 human pancreatic cancer cell lines [101]. Sodium salicylate also inhibited vascular smooth-muscle cell proliferation with concomitant upregulation of p21Cip1 and p27Kip1 [102]. Sulindac (1.2 mM) and sulindac sulfide (200 M) inhibited proliferation and caused cell cycle arrest in G0/G1 in HT-29 human colon cancer cells [103]. Sulindac and sulindac sulfide inhibited CDC2 (CDK1) and CDK2 activity in these cells. Sulindac sulfide (200 M) also induced accumulation of cells in G0/G1 and apoptosis in MCF-7 human breast cancer cells [104]. Sulindac sulfide caused growth inhibition, cell cycle arrest in G0/G1 and increased levels of p21Cip1 and p27Kip1 in HL60 human promyelocytic leukemia cells [105]. Sulindac sulfone and its derivative CP461 had no significant effect on the cell cycle profile of human glioma cells, but the sulindac sulfone derivative CP248 caused marked arrest of the cells in mitosis which was caused by microtubule depolymerization [106]. Thus several but not all of the NSAIDs and related compounds that have been tested can cause cells to arrest in G0/G1, and in some cases this is associated with decreased cellular levels of cyclin D1 and increased levels of p21Cip1 and p27Kip1. However, the precise target(s) responsible for this effect are not known.
Overview and Future Directions
The above-described studies provide strong evidence that the antitumor effects of various NSAIDs, including the recently developed COX-2-specific inhibitors, and the antitumor effects of sulindac sulfone and related derivatives,
Soh/Weinstein
276
cannot be simply explained through ihibition of COX activity. The types of evidence include: (1) a lack of correlation between expression of COX-1 or COX-2 activity in a particular cell system and the antiproliferative response to the agent tested; (2) the fact that some of the compounds with antiproliferative activity completely lack COX inhibitory activity (i.e. sulindac sulfone and derivatives), and (3) the definitive demonstration of alternate direct cellular targets (i.e. PDE2/5, IKK, etc.) for some of these agents. At the same time, these studies do not exclude the likelihood that for specific compounds and cell systems COX-1 and/or COX-2 are critical targets with respect to the antitumor effects of some of these compounds. Furthermore, as summarized in table 1, the treatment of cancer cells with these agents can induce a wide variety of changes in cellular functions and gene expression, but the precise direct cellular targets and downstream signaling pathways that lead to these effects are for the most part not known. In view of these pleiotropic effects it is likely that even a single compound can target multiple cellular molecules and pathways, which may in some cases depend on drug concentration and cell context. It is important to stress that most of the studies that we have described in this review were done with cell cultures of rodent or human cancer cell lines. Therefore, their in vivo significance remains to be determined. Another important issue is the question of dosage. Thus, in these cell culture assays aspirin and sodium salicylate were used in the range of 1–5 mM, various conventional NSAIDs were used in the range of 0.1–1 mM, and the COX-2-specific inhibitors in the range of 20–100 M. Although there is limited pharmacokinetic data available on these compounds in humans, these concentrations probably often exceed maximum in vivo blood levels. On the other hand, during chronic administration to patients, some of these compounds may be concentrated in tumor tissue and also exert accumulative effects. Indeed, despite relatively low blood levels sulindac, indomethacin and sulindac sulfone do exert antitumor effects in rodent models of carcinogenesis and tumorigenesis. There is evidence that colonic epithelial cells are exposed to sulindac sulfide concentrations that are 20-fold higher than those in the serum [107, 108]. Furthermore, NSAIDs may be concentrated in the mildly acidic extracellular environment of tumor tissue [109], and possibly further concentrated within tumor cells. Consistent with this possibility is the finding that when celecoxib (1,250 mg/ kg) was administered to mice it inhibited the growth of xenografts of HCA-7 tumor cells under conditions in which the plasma concentration was only 2 M, and yet 2–3 M celecoxib did not display toxicity to cell cultures of HCA-7 cells [12]. Recent clinical studies indicated that celecoxib caused a statistically significant reduction in polyp growth in patients with FAP only when used at a relatively high dose, i.e. 800 mg/day, but not at the recommended antiinflammatory dose of 100–200 mg b.i.d. These findings raise the possibility
COX-Independent Targets of NSAIDs and Related Compounds
277
that the in vivo antiproliferative effects of celecoxib involve targets with a lower affinity than COX-2, but other explanation have not been excluded. Taken together, the above findings indicate the need for further in vivo studies on tissue concentrations and tumor uptake of various NSAIDs and related compounds. In addition, although subcellular and tissue culture studies are highly useful for identifying direct and indirect targets involved in the antiproliferative effects of these compounds, there is a major need to confirm the in vivo roles of these targets in experimental animal studies and human clinical trials. For this purpose it will be desirable to develop pathway-specific biomarkers that can be assayed in tumor tissues, before and after treatment with the compound of interest. Hopefully, further elucidation of the critical cellular targets and pathways that mediate the antitumor effects of NSAIDs and related compounds will provide a platform for developing even more effective agents for use in both cancer prevention and treatment, while minimizing potential toxicities. Acknowledgements This work was supported by NIH grant CA 26056 and awards from the National Foundation for Cancer Research, the T.J. Martell Foundation and Cell Pathways, Inc. (to I.B.W.).
References 1 2 3
4
5
6
7
8
Thun MJ, Henley SJ, Patrono C: Nonsteroidal anti-inflammatory drugs as anticancer agents: Mechanistic, pharmacologic and clinical issues. J Natl Cancer Inst 2002;94:252–266. Baron JA, Sandler RS: Nonsteroidal anti-inflammatory drugs and cancer prevention. Annu Rev Med 2000;51:511–523. Aggarwal S, Taneja N, Lin L, Orringer MB, Rehemtulla A, Beer DG: Indomethacin-induced apoptosis in esophageal adenocarcinoma cells involves upregulation of Bax and translocation of mitochondrial cytochrome C independent of COX-2 expression. Neoplasia 2000;2:346–356. Vogt T, McClelland M, Jung B, Popova S, Bogenrieder T, Becker B, Rumpler G, Landthaler M, Stolz W: Progression and NSAID-induced apoptosis in malignant melanomas are independent of cyclooxygenase II. Melanoma Res 2001;11:587–599. Smith ML, Hawcroft G, Hull MA: The effect of non-steroidal anti-inflammatory drugs on human colorectal cancer cells: Evidence of different mechanisms of action. Eur J Cancer 2000;36: 664–674. Hanif R, Pittas A, Feng Y, Koutsos MI, Qiao L, Staiano-Coico L, Shiff SI, Rigas B: Effects of nonsteroidal anti-inflammatory drugs on proliferation and on induction of apoptosis in colon cancer cells by a prostaglandin-independent pathway. Biochem Pharmacol 1996;52:237–245. Waskewich C, Blumenthal RD, Li H, Stein R, Goldenberg DM, Burton J: Celecoxib exhibits the greatest potency amongst cyclooxygenase (COX) inhibitors for growth inhibition of COX-2negative hematopoietic and epithelial cell lines. Cancer Res 2002;62:2029–2033. Elder DJ, Halton DE, Hague A, Paraskeva C: Induction of apoptotic cell death in human colorectal carcinoma cell lines by a cyclooxygenase-2 (COX-2)-selective nonsteroidal anti- inflammatory drug: independence from COX-2 protein expression. Clin Cancer Res 1997;3:1679–1683.
Soh/Weinstein
278
9 10
11
12
13 14 15
16
17
18
19
20
21
22
23
24
25 26 27
Richter M, Weiss M, Weinberger I, Furstenberger G, Marian B: Growth inhibition and induction of apoptosis in colorectal tumor cells by cyclooxygenase inhibitors. Carcinogenesis 2001;22:17–25. Song X, Lin HP, Johnson AJ, Tseng PH, Yang YT, Kulp SK, Chen CS: Cyclooxygenase-2, player or spectator in cyclooxygenase-2 inhibitor-induced apoptosis in prostate cancer cells. J Natl Cancer Inst 2002;94:585–591. Zhang X, Morham SG, Langenbach R, Young DA: Malignant transformation and antineoplastic actions of nonsteroidal anti-inflammatory drugs on cyclooxygenase-null embryo fibroblasts. J Exp Med 1999;190:451–459. Williams CS, Watson AJ, Sheng H, Helou R, Shao J, DuBois RN: Celecoxib prevents tumor growth in vivo without toxicity to normal gut: Lack of correlation between in vitro and in vivo models. Cancer Res 2000;60:6045–6051. Chiu CH, McEntee MF, Whelan J: Sulindac causes rapid regression of preexisting tumors in Min/⫹ mice independent of prostaglandin biosynthesis. Cancer Res 1997;57:4267–4273. Weber A, Yildirim H, Schror K: Cyclooxygenase-independent inhibition of smooth muscle cell mitogenesis by ibuprofen. Eur J Pharmacol 2000;389:67–69. McCracken JD, Wechter WJ, Liu Y, Chase RL, Kantoci D, Murray ED Jr, Quiggle DD, Mineyama Y: Antiproliferative effects of the enantiomers of flurbiprofen. J Clin Pharmacol 1996; 36:540–545. Wechter WJ, Kantoci D, Murray ED Jr, Quiggle DD, Leipold DD, Gibson KM, McCracken JD: R-flurbiprofen chemoprevention and treatment of intestinal adenomas in the APC(Min)/⫹ mouse model: Implications for prophylaxis and treatment of colon cancer. Cancer Res 1997;57: 4316–4324. Wechter WJ, Leipold DD, Murray ED Jr, Quiggle D, McCracken JD, Barrios RS, Greenberg NM: E-7869 (R-flurbiprofen) inhibits progression of prostate cancer in the TRAMP mouse. Cancer Res 2000;60:2203–2208. Bak AW, McKnight W, Li P, Del Soldato P, Calignano A, Cirino G, Wallace JL: Cyclooxygenaseindependent chemoprevention with an aspirin derivative in a rat model of colonic adenocarcinoma. Life Sci 1998;62:PL367–373. Lim JT, Piazza GA, Han EK, Delohery TM, Li H, Finn TS, Buttyan R, Yamamoto H, Sperl GJ, Brendel K, Gross PH, Pamukcu R, Weinstein IB: Sulindac derivatives inhibit growth and induce apoptosis in human prostate cancer cell lines. Biochem Pharmacol 1999;58:1097–1107. Goluboff ET, Shabsigh A, Saidi JA, Weinstein IB, Mitra N, Heitjan D, Piazza GA, Pamukcu R, Buttyan R, Olsson CA: Exisulind (sulindac sulfone) suppresses growth of human prostate cancer in a nude mouse xenograft model by increasing apoptosis. Urology 1999;53:440–445. Rahman MA, Dhar DK, Masunaga R, Yamanoi A, Kohno H, Nagasue N: Sulindac and exisulind exhibit a significant antiproliferative effect and induce apoptosis in human hepatocellular carcinoma cell lines. Cancer Res 2000;60:2085–2089. Piazza GA, Alberts DS, Hixson LJ, Paranka NS, Li H, Finn T, Bogert C, Guillen JM, Brendel K, Gross PH, Sperl G, Ritchie J, Burt RW, Ellsworth L, Ahnen DJ, Pamukcu R: Sulindac sulfone inhibits azoxymethane-induced colon carcinogenesis in rats without reducing prostaglandin levels. Cancer Res 1997;57:2909–2915. Piazza GA, Rahm AK, Finn TS, Fryer BH, Li H, Stoumen AL, Pamukcu R, Ahnen DJ: Apoptosis primarily accounts for the growth-inhibitory properties of sulindac metabolites and involves a mechanism that is independent of cyclooxygenase inhibition, cell cycle arrest and p53 induction. Cancer Res 1997;57:2452–2459. Piazza GA, Rahm AL, Krutzsch M, Sperl G, Paranka NS, Gross PH, Brendel K, Burt RW, Alberts DS, Pamukcu R, et al: Antineoplastic drugs sulindac sulfide and sulfone inhibit cell growth by inducing apoptosis. Cancer Res 1995;55:3110–3116. Rybalkin SD, Beavo JA: Multiplicity within cyclic nucleotide phosphodiesterases. Biochem Soc Trans 1996;24:1005–1009. Perry MJ, Higgs GA: Chemotherapeutic potential of phosphodiesterase inhibitors. Curr Opin Chem Biol 1998;2:472–481. Thompson WJ, Piazza GA, Li H, Liu L, Fetter J, Zhu B, Sperl G, Ahnen D, Pamukcu R: Exisulind induction of apoptosis involves guanosine 3⬘,5⬘-cyclic monophosphate phosphodiesterase inhibition, protein kinase G activation, and attenuated -catenin. Cancer Res 2000;60:3338–3342.
COX-Independent Targets of NSAIDs and Related Compounds
279
28
29
30
31 32
33
34
35
36
37
38
39
40 41 42 43 44
45 46 47 48 49
Park K, Moreland RB, Goldstein I, Atala A, Traish A: Sildenafil inhibits phosphodiesterase type 5 in human clitoral corpus cavernosum smooth muscle. Biochem Biophys Res Commun 1998;249:612–617. Saeki T, Adachi H, Takase Y, Yoshitake S, Souda S, Saito I: A selective type V phosphodiesterase inhibitor, E4021, dilates porcine large coronary artery. J Pharmacol Exp Ther 1995; 272:825–831. Soh JW, Mao Y, Kim MG, Pamukcu R, Li H, Piazza GA, Thompson WJ, Weinstein IB: Cyclic GMP mediates apoptosis induced by sulindac derivatives via activation of c-Jun NH2-terminal kinase 1. Clin Cancer Res 2000;6:4136–4141. Soh JW, Mao Y, Liu L, Thompson WJ, Pamukcu R, Weinstein IB: Protein kinase G activates the JNK1 pathway via phosphorylation of MEKK1. J Biol Chem 2001;276:16406–16410. Maundrell K, Antonsson B, Magnenat E, Camps M, Muda M, Chabert C, Gillieron C, Boschert U, Vial-Knecht E, Martinou JC, Arkinstall S: Bcl-2 undergoes phosphorylation by c-Jun N-terminal kinase/stress-activated protein kinases in the presence of the constitutively active GTP-binding protein Rac1. J Biol Chem 1997;272:25238–25242. Yamamoto K, Ichijo H, Korsmeyer SJ: BCL-2 is phosphorylated and inactivated by an ASK1/Jun N-terminal protein kinase pathway normally activated at G(2)/M. Mol Cell Biol 1999;19: 8469–8478. Loweth AC, Williams GT, Scarpello JH, Morgan NG: Evidence for the involvement of cGMP and protein kinase G in nitric oxide-induced apoptosis in the pancreatic B-cell line, HIT-T15. FEBS Lett 1997;400:285–288. Butt E, Abel K, Krieger M, Palm D, Hoppe V, Hoppe J, Walter U: cAMP- and cGMP-dependent protein kinase phosphorylation sites of the focal adhesion vasodilator-stimulated phosphoprotein in vitro and in intact human platelets. J Biol Chem 1994;269:14509–14517. Deguchi A, Soh JW, Li H, Pamukcu R, Thompson WJ, Weinstein IB: VASP phosphorylation provides a biomarker for the action of exisulind and related agents that activate protein kinase G. Mol Cancer Ther 2002;1 (in press). Butt E, Immler D, Meyer HE, Kotlyarov A, Laass K, Gaestel M: Heat-shock protein 27 is a substrate of cGMP-dependent protein kinase in intact human platelets: Phosphorylation-induced actin polymerization caused by HSP27 mutants. J Biol Chem 2001;276:7108–7113. Ammendola A, Geiselhoringer A, Hofmann F, Schlossmann J: Molecular determinants of the interaction between the inositol 1,4,5-trisphosphate receptor-associated cGMP kinase substrate and cGMP kinase I. J Biol Chem 2001;276:24153–24159. Huber A, Neuhuber WL, Klugbauer N, Ruth P, Allescher HD: Cysteine-rich protein 2, a novel substrate for cGMP kinase I in enteric neurons and intestinal smooth muscle. J Biol Chem 2000; 275:5504–5511. Thomas MK, Francis SH, Corbin JD: Substrate- and kinase-directed regulation of phosphorylation of a cGMP-binding phosphodiesterase by cGMP. J Biol Chem 1990;265:14971–14978. Murthy KS: Activation of phosphodiesterase 5 and inhibition of guanylate cyclase by cGMPdependent protein kinase in smooth muscle. Biochem J 2001;360:199–208. Ghosh S, Karin M: Missing pieces in the NF-B puzzle. Cell 2002;109(suppl):81–96. Karin M, Cao Y, Greten FR, Li ZW: NF-B in cancer: From innocent bystander to major culprit. Nat Rev Cancer 2002;2:301–310. Chen Z, Hagler J, Palombella VJ, Melandri F, Scherer D, Ballard D, Maniatis T: Signal-induced site-specific phosphorylation targets I B-␣ to the ubiquitin-proteasome pathway. Genes Dev 1995;9:1586–1597. Brown K, Gerstberger S, Carlson L, Franzoso G, Siebenlist U: Control of I B-␣ proteolysis by site-specific, signal-induced phosphorylation. Science 1995;267:1485–1488. Baeuerle PA, Baltimore D: NF-B: Ten years after. Cell 1996;87:13–20. Yamamoto Y, Yin MJ, Lin KM, Gaynor RB: Sulindac inhibits activation of the NF-B pathway. J Biol Chem 1999;274:27307–27314. Yin MJ, Yamamoto Y, Gaynor RB: The anti-inflammatory agents aspirin and salicylate inhibit the activity of I()B kinase-. Nature 1998;396:77–80. Belham C, Wu S, Avruch J: Intracellular signalling: PDK1 – a kinase at the hub of things. Curr Biol 1999;9:R93–R96.
Soh/Weinstein
280
50
51 52 53
54 55 56 57 58
59 60 61
62 63 64
65
66 67 68
69 70
71
72
Arico S, Pattingre S, Bauvy C, Gane P, Barbat A, Codogno P, Ogier-Denis E: Celecoxib induces apoptosis by inhibiting 3-phosphoinositide-dependent protein kinase-1 activity in the human colon cancer cell line HT-29. J Biol Chem 2002;277:27613–27621. Zhao Y, Bjorbaek C, Moller DE: Regulation and interaction of pp90(rsk) isoforms with mitogenactivated protein kinases. J Biol Chem 1996;271:29773–29779. Bjorbaek C, Zhao Y, Moller DE: Divergent functional roles for p90rsk kinase domains. J Biol Chem 1995;270:18848–18852. Stevenson MA, Zhao MJ, Asea A, Coleman CN, Calderwood SK: Salicylic acid and aspirin inhibit the activity of RSK2 kinase and repress RSK2-dependent transcription of cyclic AMP response element binding protein- and NF-B-responsive genes. J Immunol 1999;163: 5608–5616. Boguski MS, McCormick F: Proteins regulating Ras and its relatives. Nature 1993;366:643–654. Herrmann C, Block C, Geisen C, Haas K, Weber C, Winde G, Moroy T, Muller O: Sulindac sulfide inhibits Ras signaling. Oncogene 1998;17:1769–1776. Houseknecht KL, Cole BM, Steele PJ: Peroxisome proliferator-activated receptor ␥ (PPAR␥) and its ligands: A review. Domest Anim Endocrinol 2002;22:1–23. Sporn MB, Suh N, Mangelsdorf DJ: Prospects for prevention and treatment of cancer with selective PPAR␥ modulators (SPARMs). Trends Mol Med 2001;7:395–400. Lehmann JM, Lenhard JM, Oliver BB, Ringold GM, Kliewer SA: Peroxisome proliferatoractivated receptors ␣ and ␥ are activated by indomethacin and other non-steroidal antiinflammatory drugs. J Biol Chem 1997;272:3406–3410. Adamson DJ, Frew D, Tatoud R, Wolf CR, Palmer CN: Diclofenac antagonizes peroxisome proliferator-activated receptor-␥ signaling. Mol Pharmacol 2002;61:7–12. Hatae T, Wada M, Yokoyama C, Shimonishi M, Tanabe T: Prostacyclin-dependent apoptosis mediated by PPAR␦. J Biol Chem 2001;276:46260–46267. Gupta RA, Tan J, Krause WF, Geraci MW, Willson TM, Dey SK, DuBois RN: Prostacyclinmediated activation of peroxisome proliferator-activated receptor-␦ in colorectal cancer. Proc Natl Acad Sci USA 2000;97:13275–13280. He TC, Chan TA, Vogelstein B, Kinzler KW: PPAR␦ is an APC-regulated target of nonsteroidal anti-inflammatory drugs. Cell 1999;99:335–345. Park BH, Vogelstein B, Kinzler KW: Genetic disruption of PPAR␦ decreases the tumorigenicity of human colon cancer cells. Proc Natl Acad Sci USA 2001;98:2598–2603. Walakovits LA, Moore VL, Bhardwaj N, Gallick GS, Lark MW: Detection of stromelysin and collagenase in synovial fluid from patients with rheumatoid arthritis and posttraumatic knee injury. Arthritis Rheum 1992;35:35–42. Barracchini A, Franceschini N, Amicosante G, Oratore A, Minisola G, Pantaleoni G, Di Giulio A: Can non-steroidal anti-inflammatory drugs act as metalloproteinase modulators? An in-vitro study of inhibition of collagenase activity. J Pharm Pharmacol 1998;50:1417–1423. Seger R, Krebs EG: The MAPK signaling cascade. Faseb J 1995;9:726–735. Schwenger P, Skolnik EY, Vilcek J: Inhibition of tumor necrosis factor-induced p42/p44 mitogenactivated protein kinase activation by sodium salicylate. J Biol Chem 1996;271:8089–8094. Pillinger MH, Capodici C, Rosenthal P, Kheterpal N, Hanft S, Philips MR, Weissmann G: Modes of action of aspirin-like drugs: Salicylates inhibit erk activation and integrin-dependent neutrophil adhesion. Proc Natl Acad Sci USA 1998;95:14540–14545. Schwenger P, Alpert D, Skolnik EY, Vilcek J: Cell-type-specific activation of c-Jun N-terminal kinase by salicylates. J Cell Physiol 1999;179:109–114. Schwenger P, Bellosta P, Vietor I, Basilico C, Skolnik EY, Vilcek J: Sodium salicylate induces apoptosis via p38 mitogen-activated protein kinase but inhibits tumor necrosis factor-induced c-Jun N-terminal kinase/stress-activated protein kinase activation. Proc Natl Acad Sci USA 1997; 94:2869–2873. Schwenger P, Alpert D, Skolnik EY, Vilcek J: Activation of p38 mitogen-activated protein kinase by sodium salicylate leads to inhibition of tumor necrosis factor-induced IB-␣ phosphorylation and degradation. Mol Cell Biol 1998;18:78–84. Cross TG, Scheel-Toellner D, Henriquez NV, Deacon E, Salmon M, Lord JM: Serine/threonine protein kinases and apoptosis. Exp Cell Res 2000;256:34–41.
COX-Independent Targets of NSAIDs and Related Compounds
281
73 74 75
76 77
78 79 80
81
82 83 84
85
86 87 88
89
90
91 92
93
94
Hemmings BA: Akt signaling: Linking membrane events to life and death decisions. Science 1997;275:628–630. Nicholson KM, Anderson NG: The protein kinase B/Akt signalling pathway in human malignancy. Cell Signal 2002;14:381–395. Hsu AL, Ching TT, Wang DS, Song X, Rangnekar VM, Chen CS: The cyclooxygenase-2 inhibitor celecoxib induces apoptosis by blocking Akt activation in human prostate cancer cells independently of Bcl-2. J Biol Chem 2000;275:11397–11403. Grammer TC, Cheatham L, Chou MM, Blenis J: The p70S6K signalling pathway: A novel signalling system involved in growth regulation. Cancer Surv 1996;27:271–292. Law BK, Waltner-Law ME, Entingh AJ, Chytil A, Aakre ME, Norgaard P, Moses HL: Salicylateinduced growth arrest is associated with inhibition of p70s6k and down-regulation of c-myc, cyclin D1, cyclin A, and proliferating cell nuclear antigen. J Biol Chem 2000;275:38261–38267. Frantz B, O’Neill EA: The effect of sodium salicylate and aspirin on NF-B. Science 1995;270: 2017–2019. Kopp E, Ghosh S: Inhibition of NF-B by sodium salicylate and aspirin. Science 1994;265: 956–959. McDade TP, Perugini RA, Vittimberga FJ Jr, Carrigan RC, Callery MP: Salicylates inhibit NF-B activation and enhance TNF-␣-induced apoptosis in human pancreatic cancer cells. J Surg Res 1999;83:56–61. Pierce JW, Read MA, Ding H, Luscinskas FW, Collins T: Salicylates inhibit I B-␣ phosphorylation, endothelial-leukocyte adhesion molecule expression, and neutrophil transmigration. J Immunol 1996;156:3961–3969. Tegeder I, Niederberger E, Israr E, Guhring H, Brune K, Euchenhofer C, Grosch S, Geisslinger G: Inhibition of NF-B and AP-1 activation by R- and S-flurbiprofen. Faseb J 2001;15:2–4. Stuhlmeier KM, Li H, Kao JJ: Ibuprofen: New explanation for an old phenomenon. Biochem Pharmacol 1999;57:313–320. Palayoor ST, Youmell MY, Calderwood SK, Coleman CN, Price BD: Constitutive activation of IB kinase-␣ and NF-B in prostate cancer cells is inhibited by ibuprofen. Oncogene 1999;18: 7389–7394. Niederberger E, Tegeder I, Vetter G, Schmidtko A, Schmidt H, Euchenhofer C, Brautigam L, Grosch S, Geisslinger G: Celecoxib loses its anti-inflammatory efficacy at high doses through activation of NF-B. Faseb J 2001;15:1622–1624. Shaulian E, Karin M: AP-1 in cell proliferation and survival. Oncogene 2001;20:2390–2400. Dong Z, Huang C, Brown RE, Ma WY: Inhibition of activator protein 1 activity and neoplastic transformation by aspirin. J Biol Chem 1997;272:9962–9970. Huang C, Ma WY, Hanenberger D, Cleary MP, Bowden GT, Dong Z: Inhibition of ultraviolet B-induced activator protein-1 (AP-1) activity by aspirin in AP-1-luciferase transgenic mice. J Biol Chem 1997;272:26325–26331. Shureiqi I, Chen D, Lotan R, Yang P, Newman RA, Fischer SM, Lippman SM: 15-Lipoxygenase-1 mediates nonsteroidal anti-inflammatory drug-induced apoptosis independently of cyclooxygenase-2 in colon cancer cells. Cancer Res 2000;60:6846–6850. Shureiqi I, Chen D, Lee JJ, Yang P, Newman RA, Brenner DE, Lotan R, Fischer SM, Lippman SM: 15-LOX-1: A novel molecular target of nonsteroidal anti-inflammatory drug-induced apoptosis in colorectal cancer cells. J Natl Cancer Inst 2000;92:1136–1142. Chao DT, Korsmeyer SJ: BCL-2 family: Regulators of cell death. Annu Rev Immunol 1998;16: 395–419. Baek SJ, Kim KS, Nixon JB, Wilson LC, Eling TE: Cyclooxygenase inhibitors regulate the expression of a TGF- superfamily member that has proapoptotic and antitumorigenic activities. Mol Pharmacol 2001;59:901–908. Baek SJ, Wilson LC, Lee CH, Eling TE: Dual function of nonsteroidal anti-inflammatory drugs (NSAIDs): Inhibition of cyclooxygenase and induction of NSAID-activated gene. J Pharmacol Exp Ther 2002;301:1126–1131. Kang HJ, Song MJ, Choung SY, Kim SJ, Le MO: Transcriptional induction of Nur77 by indomethacin that results in apoptosis of colon cancer cells. Biol Pharm Bull 2000;23: 815–819.
Soh/Weinstein
282
95 Liu ZG, Smith SW, McLaughlin KA, Schwartz LM, Osborne BA: Apoptotic signals delivered through the T-cell receptor of a T-cell hybrid require the immediate-early gene nur77. Nature 1994;367:281–284. 96 Woronicz JD, Calnan B, Ngo V, Winoto A: Requirement for the orphan steroid receptor Nur77 in apoptosis of T-cell hybridomas. Nature 1994;367:277–281. 97 Polakis P, Hart M, Rubinfeld B: Defects in the regulation of -catenin in colorectal cancer. Adv Exp Med Biol 1999;470:23–32. 98 Mahmoud NN, Dannenberg AJ, Mestre J, Bilinski RT, Churchill MR, Martucci C, Newmark H, Bertagnolli MM: Aspirin prevents tumors in a murine model of familial adenomatous polyposis. Surgery 1998;124:225–231. 99 Grosch S, Tegeder I, Niederberger E, Brautigam L, Geisslinger G: COX-2-independent induction of cell cycle arrest and apoptosis in colon cancer cells by the selective COX-2 inhibitor celecoxib. Faseb J 2001;15:2742–2744. 100 Subbegowda R, Frommel TO: Aspirin toxicity for human colonic tumor cells results from necrosis and is accompanied by cell cycle arrest. Cancer Res 1998;58:2772–2776. 101 Perugini RA, McDade TP, Vittimberga FJ Jr, Duffy AJ, Callery MP: Sodium salicylate inhibits proliferation and induces G1 cell cycle arrest in human pancreatic cancer cell lines. J Gastrointest Surg 2000;4:24–33. 102 Marra DE, Simoncini T, Liao JK: Inhibition of vascular smooth muscle cell proliferation by sodium salicylate mediated by upregulation of p21(Waf1) and p27(Kip1). Circulation 2000;102: 2124–2130. 103 Shiff SJ, Qiao L, Tsai LL, Rigas B: Sulindac sulfide, an aspirin-like compound, inhibits proliferation, causes cell cycle quiescence, and induces apoptosis in HT-29 colon adenocarcinoma cells. J Clin Invest 1995;96:491–503. 104 Han EK, Arber N, Yamamoto H, Lim JT, Delohery T, Pamukcu R, Piazza GA, Xing WQ, Weinstein IB: Effects of sulindac and its metabolites on growth and apoptosis in human mammary epithelial and breast carcinoma cell lines. Breast Cancer Res Treat 1998;48:195–203. 105 Ikui AE, Yao Y, Zhou P, Weinstein IB: Induction of apoptosis by sulindac sulfide in HL60 cells is enhanced by p21CiP1 or p27KiP1. Anticancer Res 2001;21:2297–2303. 106 Yoon JT, Palazzo AF, Xiao D, Delohery TM, Warburton PE, Bruce JN, Thompson WJ, Sperl G, Whitehead C, Fetter J, Pamukcu R, Gundersen GG, Weinstein IB: CP248, a derivative of exisulind, causes growth inhibition, mitotic arrest, and abnormalities in microtubule polymerization in glioma cells. Mol Cancer Ther 2002;1:393–404. 107 Duggan DE, Hooke KF, Hwang SS: Kinetics of the tissue distributions of sulindac and metabolites. Relevance to sites and rates of bioactivation. Drug Metab Dispos 1980;8:241–246. 108 Waddell WR, Ganser GF, Cerise EJ, Loughry RW: Sulindac for polyposis of the colon. Am J Surg 1989;157:175–179. 109 Bhujwalla ZM, Artemov D, Aboagye E, Ackerstaff E, Gillies RJ, Natarajan K, Solaiyappan M: The physiological environment in cancer vascularization, invasion and metastasis. Novartis Found Symp 2001;240:23–38. 110 Zhang L, Yu J, Park BH, Kinzler KW, Vogelstein B: Role of BAX in the apoptotic response to anticancer agents. Science 2000;290:989–992.
Jae-Won Soh Herbert Irving Comprehensive Cancer Center College of Physicians & Surgeons, Columbia University HHSC-1509, 701 West, 168th Street, New York, NY 10032 (USA) Tel. ⫹1 212 3056921, Fax ⫹1 212 3056889, E-Mail
[email protected]
COX-Independent Targets of NSAIDs and Related Compounds
283
Author Index
Altorki, N.K. 107
Gately, S. 179
Müller-Decker, K. 72
Baron, J.A. 1 Blanke, C.D. 243
Hawk, E.T. 210 Howe, L.R. 90 Huang, M. 138
Pold, M. 138
Kerbel, R. 179
Sabichi, A.L. 163 Saha, D. 193 Sharma, S. 138 Soh, J.-W. 261 Subbaramaiah, K. 107
Lippman, S.M. 163
Umar, A. 210
Mao, J.T. 138 Marks, F. 72 Masferrer, J.L. 243
Viner, J.L. 210
Choy, H. 193 Isakson, P.C. 25 Dannenberg, A.J. 90, 107 Dixon, D.A. 52 Dohadwala, M. 138 Dubinett, S.M. 138 DuBois, R.N. 124 Fürstenberger, G. 72
Weinstein, I.B. 261
284
Subject Index
o-(Acetoxypheny)hept-2-ynyl sulfide, suicide inhibition of cyclooxygenase-2 125 Actinic keratoses, diclofenac management 222 Akt non-steroidal anti-inflammatory drug targeting and antiproliferative effects 272 signaling and cyclooxygenase-2 expression 63, 64 Angiogenesis cancer role 180 cyclooxygenase-2 apoptosis inhibition 184 inhibition rationale 180, 185, 186 microvasculature expression 181, 182 role in angiogenesis 112, 113, 180–184, 195 stromal cell expression 182 tumor expression 181 eicosanoid product modulation 183 non-steroidal anti-inflammatory drug inhibition, cyclooxygenaseindependent mechanisms 185, 186 overview 179, 180 prostaglandins angiogenesis promotion 184 vascular endothelial growth factor induction 182, 183 thromboxane A2 contribution 183, 184 AP-1, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 274
Apoptosis cyclooxygenase-2 effects cell cycle 148, 149 lung cancer 147–149 upper respiratory tract cancers 111 cyclic GMP-dependent protein kinase modulation and targeting 266 non-steroidal anti-inflammatory drugs and apoptosis induction cell cycle modulation 276 inhibition and angiogenesis role 184 lung cancer 147–149 mechanisms 149 radiation-induced apoptosis enhancement 148 Aspirin chemoprevention advantages over other agents 235 ongoing chemoprevention trials 224, 225 wide-spread use rationale 234, 235 heart disease prevention 2, 234 AUF1, cyclooxygenase-2 AU-rich RNA element binding 60 Bax, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 274, 275 Bcl-2, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 274, 275
285
Bladder cancer cyclooxygenase-2 expression 165, 244 epidemiology 163 inflammation in etiology 164, 165 non-steroidal anti-inflammatory drug chemoprevention and treatment animal model studies 165–167, 173 cell line studies 168, 169 epidemiology studies 13, 164, 165 prospects for study 173, 174 rationale 163 types and management 164 Brain tumor cyclooxygenase-2 expression 255 non-steroidal anti-inflammatory drug trials 255 Breast cancer cyclooxygenase-2 expression clinicopathological correlates 92, 93, 101 immunoreactivity studies 91, 92 rodent models 94 survival outcomes 92 transgenic mouse studies 91, 94 tumorigenesis roles 98, 99 non-steroidal anti-inflammatory drugs chemotherapy combination trials 250 epidemiology studies 10–13, 94 HER2-positive cancer studies 96, 97, 101, 250 rodent model studies 94–97 Cardiovascular safety, cyclooxygenase-2specific inhibitors 40–42 -Catenin, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 275, 276 C/EBP, cyclooxygenase-2 regulation 55, 78, 79 Celecoxib angiogenesis inhibition 202 assays of cyclooxygenase-2 inhibition 31, 32 familial adenomatous polyposis clinical trials 5, 6, 131, 132 indications and dosing 33–35 kinetics of inhibition 30, 31
Subject Index
ongoing chemoprevention trials 224, 225 opioid-sparing effects 35, 36 pharmacokinetics absorption 32 drug interactions 33 metabolism and excretion 32, 33 plasma protein binding 32 safety cardiovascular safety 40–42 CLASS trial 229, 230 gastrointestinal safety 26, 36–38 platelet safety 39, 40 renal safety 42–44 structure and cyclooxygenase-2 active site binding 27–29 Cervical cancer cyclooxygenase-2 expression 99, 100, 244, 254 menstrual cycle stage and surgery 100 non-steroidal anti-inflammatory drugs chemoprevention 100, 101 chemotherapy combination trials 254, 255 Chemotherapy cellular response 244, 245 non-steroidal anti-inflammatory drug combination trials, see also specific cancers cisplatin 246 irinotecan 246, 247 lung cancer 150, 206, 245, 246 metronomic chemotherapy 256 taxanes 245, 246 toxicity reduction 247, 248, 256, 257 Collagenase, non-steroidal antiinflammatory drug targeting and antiproliferative effects 271 Colorectal cancer cyclooxygenase-2 animal model studies of inhibition 125, 126 cancer growth role 126, 127 expression 53, 54, 64, 65, 244 knockout mouse studies 90 tumorigenesis role 128–130 mortality 124
286
non-steroidal anti-inflammatory drug epidemiology and treatment studies aspirin benefits 4, 5, 90 chemoprevention trials 223, 226, 227 chemotherapy combination trials 250–253 cost-effectiveness analysis vs. screening 233, 234 cyclooxygenase-independent effects 127, 128, 262, 263 duration of use and risks 5 familial adenomatous polyposis clinical trials 5, 6, 131, 132, 214–217, 222, 223, 226, 248 observational studies 131 ongoing clinical trials 132, 133 prospects for study 132, 133 relapse chemoprevention trials 252, 253 sporadic colorectal neoplasia clinical trials 6, 7, 217, 219 ulcerative colitis patients 5 prostaglandin receptors 130 CRE, cyclooxygenase-2 gene regulation 54, 55 Cyclic GMP-dependent protein kinase apoptosis role 264, 266 non-steroidal anti-inflammatory drug targeting and antiproliferative effects 264, 266 Cyclooxygenase active site and inhibitors 27–29 assays 31 bifunctional activity 139 carcinogen metabolism 142 early tumor development role 195, 196 functions 25, 26, 72, 73, 90 gene regulation, see Expression regulation, cyclooxygenase-2 inflammation role 26 isoforms 25, 52, 109, 124, 125 therapeutic targeting rationale in cancer 212, 213 tissue distribution 25, 26, 52, 73 Dendritic cell, cyclooxygenase-2 regulation in lung cancer 144
Subject Index
Diclofenac, actinic keratoses management 222 Esophageal cancer cyclooxygenase-2 and carcinogenesis angiogenesis role 112, 113 apoptosis inhibition 111 inflammation and immunosuppression 112 invasion and metastasis 113 preclinical evidence 109, 110 prospects for study 116–118 smoking and gene induction 109, 111 therapeutic targeting 113–116 xenobiotic metabolism 110, 111 cyclooxygenase-2 expression and survival 116, 244 epidemiology 107 non-steroidal anti-inflammatory drugs epidemiology studies 7, 19, 108 esophageal squamous dysplasia studies 223 Expression regulation, cyclooxygenase-2 AU-rich RNA element in posttranscriptional regulation binding proteins AUF1 60 HuR 58, 59 TIA-1/TIAR 60 tristraprolin 59, 60 discovery 56, 57 tumorigenesis 57, 58 keratinocytes 78, 79 promoter CRE 54, 55 NF-IL-6 55, 78 nuclear factor-B 56 PEA3 55 TATA box 54 regulatory elements 52, 53 signal transduction pathways and regulation Akt 63, 64 ERK 63 JNK 62, 63 p38 63 Ras 62
287
Expression regulation, cyclooxygenase-2 (continued) signal transduction pathways and regulation (continued) Rho 62 Wnt1 and APC signaling 61, 62 smoking and gene induction 109, 111 tumor expression 53, 54, 64, 65 Extracellular signal-regulated kinase, signaling and cyclooxygenase-2 expression 63, 78 Familial adenomatous polyposis, see Colorectal cancer Gastric cancer, see Stomach cancer Gastrointestinal safety cyclooxygenase-2-specific inhibitors 37, 38 non-selective non-steroidal antiinflammatory drugs 26, 36, 37, 228, 229 Gene regulation, see Expression regulation, cyclooxygenase-2 Head and neck cancer cyclooxygenase-2 and carcinogenesis angiogenesis role 112, 113 apoptosis inhibition 111 inflammation and immunosuppression 112 invasion and metastasis 113 preclinical evidence 109, 110 prospects for study 116–118 therapeutic targeting 113–116 xenobiotic metabolism 110, 111 epidemiology 107 smoking and cyclooxygenase-2 induction 109, 111 HuR, cyclooxygenase-2 AU-rich RNA element binding 58, 59 Interleukins, cyclooxygenase-2 regulation in lung cancer 143, 144 Intraepithelial neoplasia, clinical trial endpoint 212 Invasion, cyclooxygenase-2 role
Subject Index
lung cancer 146, 147 upper respiratory tract cancers 113 JTE-522, upper respiratory tract cancer response 113–115 Jun N-terminal kinase non-steroidal anti-inflammatory drug targeting and antiproliferative effects 271, 272 signaling and cyclooxygenase-2 expression 62, 63 Leukemia, non-steroidal anti-inflammatory drug epidemiology studies 18 Lipoxygenase non-steroidal anti-inflammatory drug targeting and antiproliferative effects 274 prostate cancer role 172, 173 Lung cancer cyclooxygenase-2 apoptosis regulation 147–149 carcinogen metabolism 142 expression 139–141, 195–197, 244 growth factor induction 142 immunity regulation antigen-presenting cell 144 cytokine balance 143, 144 invasion role 146, 147 tumorigenesis role 139, 142 mortality 139 non-steroidal anti-inflammatory drug chemoprevention and treatment chemoprevention trials 151–153 chemoradiotherapy combination trials 205, 206 chemotherapy combination trials 150, 206 epidemiology studies 18, 139 non-small cell lung cancer trials 249, 250 radiation sensitization 150, 151, 196, 197 Melanoma, see Skin cancer Metastasis, cyclooxygenase-2 role 113
288
Mitogen-activated protein kinases, see Extracellular signal-regulated kinase, Jun N-terminal kinase, p38 NAG-1, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 275 Nimesulide, apoptosis induction 148 Non-steroidal anti-inflammatory drugs cyclooxygenase-independent mechanisms Akt 272 angiogenesis inhibition 185, 186 AP-1 274 Bax 274, 275 Bcl-2 274, 275 -catenin 275, 276 cell culture studies of cyclooxygenase expression and antiproliferative effects 261, 262 cell cycle modulation 276 collagenase 271 colorectal cancer 127, 128 direct cellular targets 263–265 dose-response 277, 278 enantiomer studies of antiproliferative effects 263 indirect cellular targets 271–276 lipoxygenase 274 mitogen-activated protein kinases 271, 272 mouse model of intestinal tumorigenesis 262, 263 NAG-1 275 nuclear factor-B 267, 268, 273, 274 Nur77 275 p70S6K 272, 273 peroxisome proliferator-activated receptors 270, 271 phosphodiesterases 264, 266 3-phosphoinositide-dependent kinase 1 268 prospects for study 276–278 Ras 269, 270 RSK2 268, 269 cyclooxygenase-2 inhibitors, see specific drugs
Subject Index
epidemiology of cancer prevention, see also specific cancers bladder 13 breast 10–13 colorectum 4–7, 20 confounding factors in analysis 1–3 esophagus 7, 19 indirect evidence using rheumatoid arthritis patients 3 leukemia 18 lung 18 ovar 13, 14 overview 213, 214 pancreas 10, 19 prostate 15, 18 renal cells 13, 15–17, 19 skin 79–81 stomach 7, 10, 19 risk-benefit ratio in chemoprevention 232, 233 safety of non-specific vs. cyclooxygenase-2-specific inhibitors 26, 36–40, 42–44, 228–230 types 211 Nuclear factor-B cyclooxygenase-2 gene regulation 56 IB regulation 267, 268 non-steroidal anti-inflammatory drug targeting and antiproliferative effects 267, 268, 273, 274 Nur77, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 275 Ovarian cancer, non-steroidal antiinflammatory drug epidemiology studies 13, 14 p38 non-steroidal anti-inflammatory drug targeting and antiproliferative effects 271, 272 signaling and cyclooxygenase-2 expression 63, 78 p70S6K, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 272, 273
289
Pancreatic cancer cyclooxygenase-2 expression 253 non-steroidal anti-inflammatory drugs chemotherapy combination trials 253, 254 epidemiology studies 10, 19 PEA3, cyclooxygenase-2 gene regulation 55 Peroxisome proliferator-activated receptors, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 270, 271 Phosphodiesterases, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 264, 266 Phosphoinositide-dependent kinase-1, non-steroidal anti-inflammatory drug targeting and antiproliferative effects 268 Platelet safety, cyclooxygenase-2-specific inhibitors 39, 40 Prostaglandins angiogenesis promotion 184 vascular endothelial growth factor induction 182, 183 breast cancer tumorigenesis role 98, 99 carcinogenesis role evidence 107, 108 lung cancer invasion role 146 receptors colorectal cancer receptors 130 knockout studies 231, 232 types 231, 232 synthesis 108, 109, 124 Prostate cancer cyclooxygenase-2 expression 170 epidemiology 169 inflammation and fatty acids in pathogenesis 169, 170, 172, 173 non-steroidal anti-inflammatory drug chemoprevention and treatment animal model studies 171–173 clinical trials 255 epidemiology studies 15, 18, 170 prospects for study 173, 174 rationale 163, 255
Subject Index
Radiation therapy cyclooxygenase-2 expression and outcomes 197 mechanism of action 193 molecular and biological responses 194 non-steroidal anti-inflammatory drug combination therapy anti-angiogenic activity 202–204, 206, 207 apoptosis enhancement 148, 207 chemoradiotherapy combination 204–206 human cancer cell line studies 201, 202 lung cancer sensitization 150, 151, 196, 197 mouse tumor model studies 197, 199, 200 rationale 248 Ras non-steroidal anti-inflammatory drug targeting and antiproliferative effects 269, 270 signaling and cyclooxygenase-2 expression 62, 79 Renal cancer, non-steroidal anti-inflammatory drug epidemiology studies 13, 15–17, 19 Renal safety cyclooxygenase-2-specific inhibitors 43, 44 non-selective non-steroidal antiinflammatory drugs 42, 43 Retinoic acid, cyclooxygenase-2 inhibition 117 Rheumatoid arthritis, cancer incidence studies 3 Rho, signaling and cyclooxygenase-2 expression 62 Rofecoxib angiogenesis inhibition 203, 204 assays of cyclooxygenase-2 inhibition 31, 32 indications and dosing 33–35 kinetics of inhibition 30, 31 ongoing chemoprevention trials 225 opioid-sparing effects 35, 36 pharmacokinetics
290
absorption 32 drug interactions 33 metabolism and excretion 32, 33 plasma protein binding 32 safety cardiovascular safety 40–42 gastrointestinal safety 26, 36–38 platelet safety 39, 40 renal safety 42–44 VIGOR trial 229, 230 structure and cyclooxygenase-2 active site binding 27–29 SC-58125, angiogenesis inhibition 202 Skin cancer cyclooxygenase-2 expression basal vs. squamous cell carcinoma 77, 78 constitutive expression 74, 75 growth factor induction 74 keratinocytes 78, 79 melanoma 78 transgenic and knockout mouse studies 72, 81, 82, 85 epidemiology of types 76–78 inflammation and cyclooxygenase isoform response 74–76 morbidity and mortality 77, 78 non-steroidal anti-inflammatory drug prevention studies in animal models 79–81, 219, 222 skin architecture 73, 74 Smoking bladder cancer risks 163 carcinogen metabolism by cyclooxygenase-2 142 cyclooxygenase-2 induction 109, 111 Stomach cancer cyclooxygenase-2 and carcinogenesis angiogenesis role 112, 113 apoptosis inhibition 111 inflammation and immunosuppression 112 invasion and metastasis 113 preclinical evidence 109, 110 prospects for study 116–118 therapeutic targeting 113–116
Subject Index
xenobiotic metabolism 110, 111 epidemiology 107 non-steroidal anti-inflammatory drug epidemiology studies 7, 10, 19, 108 smoking and cyclooxygenase-2 induction 109, 111 Sulindac apoptosis induction 148 ongoing chemoprevention trials 224 prostate cancer management 173 TATA box, cyclooxygenase-2 gene regulation 54 Therapeutic index, non-steroidal antiinflammatory drug improvements 231 Thromboxane A2, angiogenesis contribution 183, 184 TIA-1/TIAR, cyclooxygenase-2 AU-rich RNA element binding 60 Tristraprolin, cyclooxygenase-2 AU-rich RNA element binding 59, 60 Valdecoxib assays of cyclooxygenase-2 inhibition 31, 32 indications and dosing 33–35 kinetics of inhibition 30, 31 opioid-sparing effects 35, 36 pharmacokinetics absorption 32 drug interactions 33 metabolism and excretion 32, 33 plasma protein binding 32 safety cardiovascular 40–42 gastrointestinal 26, 36–38 platelet 39, 40 renal 42–44 structure and cyclooxygenase-2 active site binding 27–29 Vascular endothelial growth factor non-steroidal anti-inflammatory drug inhibition 202 prostaglandin induction 182, 183 Wnt1, APC signaling and cyclooxygenase-2 expression 61, 62
291
