Prostaglandins and their derivatives are signaling lipophilic molecules that are involved in diverse homeostatic and reactive functions, including platelet aggregation, clot formation, vasodilation, and gastric cytotoxic protection, as well as renal sodium and water balance. The first step in the synthesis of prostaglandins is the hydrolysis of membrane phospholipids to arachidonic acid by phospholipase A2. Arachidonic acid is then converted to an unstable product, prostaglandin G2, which is rapidly converted to prostaglandin H2 by the peroxidase activity of a cyclooxygenase (COX). Prostaglandin H2 is converted by tissue-specific enzymes to other prostaglandins and thromboxanes.
There are two isoforms of the COX enzyme encoded by two different genes. Cyclooxygenase-1 (COX-1) is a constitutive enzyme that is present in most normal tissues and mediates the synthesis of prostaglandins required for normal physiologic functions. The gene for COX-1 is on human chromosome 9. Cyclooxygenase-2 (COX-2) is an inducible form of the enzyme that is not normally detected in most tissues and is coded by a gene on human chromosome 1. Cyclooxygenase-2 is induced by cytokines, growth factors, tumor promoters, and carcinogens. Cyclooxygenase-2 is also induced by several oncogenes, such as v-Src, v-Ha-ras, HER2/neu and Wnt.[4-7]
Although prostaglandins are involved in many normal physiologic functions, these moleculesand the COX enzymes involved in prostaglandin productionmay also be involved in tumorigenesis. Indeed, there is evidence that prostaglandins may contribute to tumorigenesis by inhibiting the immune system, stimulating cell growth, enhancing angiogenesis, increasing mutagen production, enhancing cell invasion, or inhibiting apoptosis. In this article, we will discuss these mechanisms and will explore the role of COX-2 in tumorigenesis. We will also review clinical trials that have investigated COX-2 inhibitors as chemoprotection agents and will discuss future directions in the study of COX-2 as a therapeutic target in cancer management.
Prostaglandins have a variety of immunosuppressive effects. For example, prostaglandin E2 diminishes the cytotoxic activity of natural killer cells, inhibits T-cell and B-cell growth, and decreases the production of cytokines including tumor necrosis factor-alpha. Huang et al showed that pretreatment with a prostaglandin inhibitor prevents an increase in interleukin-10 synthesis by peripheral blood lymphocytes. Furthermore, prostaglandins may also interfere with antigen processing by dendritic cells. The ability of prostaglandins to inhibit the immune system may allow tumors to grow without surveillance and may contribute to tumorigenesis.
Enhanced prostaglandin synthesis may also contribute to tumorigenesis by direct stimulation of cell growth. Data have shown that prostaglandin E2 and prostaglandin F2-alpha can be mitogenic in BALB/c3T3 fibroblasts in the presence of epidermal growth factor. Furthermore, proliferation of mammary epithelial cells can be stimulated in the presence of epidermal growth factor by prostaglandin E1 and prostaglandin E2. In breast tissue, prostaglandins may stimulate cell growth by stimulating the aromatase gene, CYP19, and thus enhancing estrogen synthesis.[12,13] Interestingly, enhanced expression of CYP19 and COX has been found in human breast cancer specimens.
Inhibition of Apoptosis
Prolonged survival of abnormal cells favors the accumulation of genetic changes that may result in tumor formation. Therefore, inhibition of apoptosis may increase the tumorigenic potential of initiated cells. Sheng et al demonstrated that prostaglandin E2 may inhibit apoptosis by inducing bcl-2. Recently it was shown that inhibition of COX-2 by celecoxib(Drug information on celecoxib) (Celebrex), a selective COX-2 inhibitor, resulted in a decrease of production of prostaglandin E2 and TXB2, and was associated with an increase in apoptosis in vivo.
Effects on Metastatic Potential
Enhanced COX-2 expression may also contribute to tumorigenesis by increasing cell invasiveness. DuBois et al showed that COX-2 overexpression in rat intestinal epithelial cells increases cell adhesion to the extracellular matrix. Recently, Stockton and Jacobson demonstrated that COX-2 is required for NIH3T3 cell migration in a process that appears to be regulated by the extracellular signal-regulated kinase 1/2. Furthermore, the activity of enzymes responsible for digesting cellular basement membrane is enhanced by COX-2 overexpression in the breast cancer cell line Hs578T and colon cancer cells, which likely contributes to the increased ability of these cells to invade through a layer of Matrigel.[19,20]
Increased Production of Mutagens
Another mechanism by which COX-2 overexpression may play a role in carcinogenesis is an increase in the production of mutagens. One such potent mutagen is malondialdehyde, which can be produced by isomerization of prostaglandin H2. Malondialdehyde acts by forming adducts with deoxynucleotides, which cause frame-shifts and base-pair substitutions. In addition, Wiese et al showed that the peroxidase activity of cyclooxygenases can catalyze the formation of mutagens by the oxidation of aromatic amines, heterocyclic amines, and dihydrodiol derivatives of polycyclic hydrocarbons. Thus, COX-2 overexpression may lead to DNA damage that may eventually lead to carcinogenesis.
Cyclooxygenase-2 has also been implicated in enhanced angiogenesis, which plays a role in carcinogenesis. Tumor growth depends on increased blood supply via secretion of angiogenic promoters such as vascular endothelial growth factor. Cyclooxygenase-2 overexpression in colon cancer cells is correlated with increased production of vascular growth factors and formation of capillary networks. This angiogenic effect of COX-2 can be blocked by NS398, a selective COX-2 inhibitor. Williams et al demonstrated that tumor formation is markedly decreased in COX-2 knockout mice compared with wild-type mice. Furthermore, the pharmacologic inhibition of COX-2 leads to a decrease in vascular endothelial growth factor production that may contribute to a decrease in tumor formation. Masferrer et al showed that celecoxib, a selective COX-2 inhibitor, blocks corneal blood vessel formation in a rat model. Thus, COX-2 overexpression may increase tumor blood supply and may contribute to tumor growth.
Cyclooxygenase-2 is upregulated in multiple human premalignant and malignant conditions, including tumors of the colon, breast, stomach, lung, pancreas, cervix, prostate, bladder, liver, skin, head and neck, and esophagus (Table 1).[26-40] There are multiple lines of evidence suggesting a link between levels of COX-2 and tumorigenesis. Increased levels of COX-2 are detected in premalignant intestinal tumors in experimental animal models. The knockout of the COX-2 gene led to a marked reduction in the size and number of polyps in APCD716 mice. In addition, APCD716 mice treated with a selective COX-2 inhibitor had reduced adenoma formation. Further, COX-2 knockout mice develop fewer skin papillomas than control mice. Taken together, these results suggest that inhibition of COX-2 could be important in the prevention of a variety of epithelial tumors.
Synthetic and naturally occurring inhibitors of COX (eg, sulindac [Clinoril], ibuprofen, flurbiprofen(Drug information on flurbiprofen) [Ansaid], indomethacin [Indocin]) have also been shown to protect against mammary, colon, oral, and esophageal cancer in experimental animals.[44-50] For example, studies have shown that flurbiprofen, an inhibitor of COX-1 and COX-2, can inhibit the growth of transplanted mammary tumors and increase the mean survival duration in mice. In addition to having therapeutic activity against established mammary tumors, flurbiprofen inhibits mammary carcinogenesis induced by a low dose of N-methyl-N-nitrosourea in rats. Other data have also demonstrated that indomethacin, another inhibitor of both COX-1 and COX-2, has significant chemoprotective activity in rats when administered during either the early or late stage of mammary tumorigenesis. Importantly, epidemiologic studies have shown that chronic intake of nonsteroidal anti-inflammatory drugs (NSAIDs) reduces the incidence of various human cancers, including cancers of the colon, breast, lung, esophagus, stomach, and bladder.[53-60]
Celecoxib has been evaluated as a possible chemopreventive agent for the inhibition of tumorigenesis. In a study to evaluate the inhibitory activity of celecoxib against azoxymethane-induced aberrant crypt foci formation in the colon of rats, celecoxib significantly suppressed the total number of aberrant crypt foci in rats compared with placebo. In a second study of azoxymethane-treated rats, celecoxib reduced the incidence of colon cancer by 93% and tumor multiplicity by 97%. These animal data provide strong evidence that celecoxib has cancer activity in the prevention of cancer when tested in a defined model of tumorigenesis.
Recently, Harris et al compared the chemoprevention effects of celecoxib with ibuprofen(Drug information on ibuprofen) and placebo in the development and growth of 7,12-dimethyl-benz(a)anthracene (DMBA)-induced rat mammary tumors. Seven days prior to receiving a single dose of 15 mg of DMBA by intubation, rats were fed either a control diet or diets containing 1,500 ppm of celecoxib or 1,500 ppm of ibuprofen. Both celecoxib and ibuprofen significantly increased tumor latency, reduced tumor burden, and prevented tumor formation compared with placebo. However, celecoxib was more potent than ibuprofen. At 105 days after DMBA treatment, tumor incidence was 100% in control rats compared with 32% and 60% in rats fed either celecoxib (P < .001) or ibuprofen (P < .001), respectively. The control rats had an average of 3.2 tumors compared with 0.4 and 1.5 tumors in rats treated with celecoxib (P < .001) and ibuprofen (P < .001), respectively. Additionally, the average tumor volume was 1.5 cm³ in control rats vs 0.3 cm³ and 0.6 cm³ in rats treated with celecoxib (P < .001) or ibuprofen (P < .001), respectively. The higher potency of celecoxib suggests that this agent may have an in vivo advantage over other NSAIDs as a chemoprevention agent.
The DMBA-induction model was also used to evaluate celecoxib for efficacy against established tumors. Alshafie et al examined the effect of celecoxib, given as a daily diet (1,500 ppm), on the growth of established DMBA-induced tumors in rats over a 6-week treatment period. Tumors in the control group continued to grow; whereas tumors in the celecoxib group markedly decreased in size. The average reduction in tumor volume relative to baseline was approximately 32% (P < .04). At the end of the 6-week treatment period, average tumor volume was 1.45 cm³ and 0.13 cm³ in the control and celecoxib groups, respectively. Tumor volume regression occurred in 90% of rats treated with celecoxib. These results suggest that celecoxib has potent antitumor activity and chemoprevention activity in this rat mammary carcinoma model.